[12-OCT-16] Our Subcutaneous Transmitters (SCTs) are small, battery-powered sensors designed to be implanted beneath the skin of laboratory animals. Each SCT measures one or more biopotentials. It samples the biopotentials, usually with sixteen-bit precision, and transmits the digital samples through the skin of the animal with radio waves. The primary purpose of the system is to record biopotentials continuously for weeks at a time, and to analyze the recordings reliably and automatically. When recording EEG in epileptic animals, for example, we will want to count seizures automatically in tens of thousands of hours of recording.
Our Subcutaneous Transmitter System consists of implanted SCTs, antennas to pick up their radio-frequency transmissions, amplifiers and demodulators to detect and decode these transmissions, communication hardware to permit data acquisition computer over ethernet, and software to record and analyze the data. Each system can monitor up to fourteen single-channel transmitters. The animals in which these fourteen transmitters are implanted can all live in the same cage, or in separate cages. Almost all users SCT systems make use of faraday enclosures to keep local radio-frequency interference away from the receiving antennas, to isolate SCT systems from another so they can operate without interfering with one another, to eliminate low-frequency electrical noise that might otherwise corrupt the biopotential recordings, and to block ultrasonic noise that would otherwise disturb the animals.
The SCT system consists of a data acquisition and power supply assembly called the LWDAQ Driver such as the A2071E, a radio frequency data receiver such as the A3027E, one or more faraday enclosures, up to sixteen antennas such as the A3015C, and up to fourteen subcutaneous transmitters implanted in animals. In addition, we may have location tracking platforms beneath some or all of the animal cages to monitor the location and movements of the animals. These location trackers, such as the A3032A, use signals from the data receiver and communications via their own connection to the LWDAQ driver. The schematic below attempts to clarify the arrangement of the system components. If we want to operate two or more systems in the same faraday enclosure without data from one system being recorded by the other, we use data receivers and transmitters that operate with different set numbers. If we want to record all signals from a hundred devices with one data receiver, we can do that as well. The system supports 196 independend signal channels.
The subcutaneous transmitters will monitor electroencephalograph (EEG), local field potential (LFP), electrocardiogram (ECG) or electromyography (EMG). We describe the sources of EEG and LFP in The Sources of EEG. For recordings from live animals, see our Example Recordings page. Our A3028 series of subcutaneous transmitters supports sample rates up to 4048 SPS. For recording EEG, most of our customers choose a bandwidth of 0.3-160 Hz with a sample rate of 512 SPS.
The A3028A provides two input channels, allowing EEG to be monitored in two locations, or EEG and EMG to be monitored at the same time. Its volume is 1.4 ml, which fits easily into an adult mouse. It samples its X and Y inputs at 512 SPS each, and its amplifiers provide bandwidth 0.3-160 Hz. It runs for 360 hrs. The X and Y inputs share the same common terminal, so there are three leads emerging from the device, plus the antenna. The A3028B is the same size as the A3028A, but provides only the X input and runs for 600 hrs. The A3028D is electrically identical to the A3028A, with a larger battery. It's volume is 2.8 ml and it runs for 1900 hrs. The A3028E is also 2.8 ml, but provides only the X input and runs for 3200 hrs.
Reception of radio-frequency signals from a freely-moving object is unreliable when you use only one antenna to pick up the signal. Within one of our faraday enclosures, a single antenna will receive the signal from an implanted transmitter over 95% of the time. Two independent antennas will pick up the signal 99.5% of the time. Our Octal Data Receiver (A3027) provides eight independent antenna inputs. Each input has its own amplifier, demodulator, and decoder. By distributing these antennas between our animal cages, the probability that at least one of them will receive any particular message approaches 100%. With eight antennas distributed through an IVC rack, for example, we obtained 100% reception everywhere within the rack. If we want to record from more than fourteen animals in the same IVC rack, we run several SCT systems that operate with different set numbers.
In our conclusion, we give examples of subcutaneous transmitter systems and their performance. You will find the prices of subcutaneous transmitter system components in our price list. In the body of this report, we describe the performance and operation of our Subcutaneous Transmitter hardware. The descriptions are mingled with the history of our development, which makes them long-winded. For that we apologize. We have done our best to extract as much of the history as possible and place it in our history section. Our Seizure Detection page gives a history of our automatic event detection software.
If you can't understand our radio-frequency terminology consult our terminology page, where we define the terms RF, IF, LO, mixer, attenuator, dBm, and dB. We explain the significance of 50 Ω in our circuits, and we provide a table that relates power in 50-Ω circuits like our to signal amplitude.
[21-MAY-14] Our Subcutaneous Transmitter (A3028) provides one or two, depending upon how it is configured. Our Data Receiver (A3037) comes in a 30 cm × 22 cm × 11 cm metal box. The box connects to our LWDAQ data acquisition system through an RJ-45 connector. It receives power through the same connector. Eight BNC sockets on the back of the box provide connections for a coaxial cables to the antennas. There are thirty indicator lights on the front of the box. There is one white light for each transmitter channel, one white light for each antenna input, and power, reset, upload, and empty lights.
The Octal Data Receiver comes with eight Dampled Loop Antenna (A3015C). These attach to the antenna inputs with 240-cm coaxial cables. Each antenna can stand on its own, or we can take the mounting brackets off and lay it on a table or wedge it between the cages in an IVC rack.
When first setting up a system, we recommend you do so with the help of a faraday enclosure such as the front-loading, stackable FE2C, which is large enough for two rat cages. Within an faraday enclosure you are sure to get perfect reception, and by opening the enclosure door, you can determine how severe your local interference is. If you want to operate in a larger space, such as an IVC rack, you can set up a Radio-Frequency Isolation Chamber. These chambers are straightforward to construct our of microwave absorbing foam and steel mesh fabric. We present a prototype here. Gaps between the walls are well-tolerated by the system because the isolation chamber needs to attenuate ambient interference only by a factor of four to guarantee adequate reception with the octal data receiver, even in urban laboratories.
and Manual Link
|A3028||Subcutaneous Transmitter (Single or Dual Channel)||Active|
|FE2B||Faraday Enclosure (Top Loading)||Active|
|FE2C||Faraday Enclosure (Front Loading)||Active|
|A3027||Data Receiver (Eight Independent Antennas)||Active|
|A3014MT||Modulating Transmitter||Test Equipment|
|A3014SO||SAW Oscillator||Test Equipment|
|A3018||Data Receiver (One Independent Antenna)||Obsolete|
|A3019||Subcutaneous Transmitter (Single Channel)||Obsolete|
|A3021||Active Antenna Combiner (Four Inputs)||Obsolete|
Radio frequency isolation chambers are a new development in 2014, and we believe laboratories can put these together themselves with our advice. So we do not have a part number for a specific isolation chamber design.
[27-NOV-15] Here is a list of parts you need to set up a new recording system. All parts are supplied by Open Source Instruments, except for the first one: the power cable. The power cable is a computer power cable left-over from a retired PC or monitor. We don't ship these because they have to have your local power plug on the end, and we know that everyone has plenty of them stashed away in a draw in their lab.
Recommendation: Set up your SCT recording system several weeks before you plan to perform your first animal experiments. During these weeks, you can install the LWDAQ software and learn how to acquire and analyze transmitter signals. In particular, we recommend that you place one or two transmitters within a faraday enclosure and record their signals continuously for at least a week, in order to make sure that your computer, local area network, and hardware is functioning correctly. Recording might be interrupted every few hours, and you will be glad to figure out what causes these interruptions before you have animals with transmitters implanted. Reception from the transmitters may cease all-together every few hours, and you will be glad to figure out what is interfering with reception long before you have begun an experiment with live animals.
Assuming you have the above parts in hand, follow these steps. They will take you through the set-up of the hardware and software, so that by the end you will be able to record and analyze transmitter signals.
In passing through all the above steps, you will have learned how to use the system in detail, and you will be ready for your experiments. If you do not follow these steps before you begin your experiments, you will be in a rush to solve any hardware or software problems you encounter. We do our best to send you working equipment and bug-free software, but we make mistakes. We do not want our mistakes to cause you unnecessary inconvenience.
The Data Recorder (A3018) plugs into a Long-Wire Data Acquisition (LWDAQ) system. It acts as a LWDAQ device, although it violates the current-consumption limits of the LWDAQ.
We use the Data Recorder with the LWDAQ Driver with Ethernet Interface (A2037E). The A2037E connects to the Internet, your Local Area Network, or directly to your computer via an RJ-45 Ethernet socket. You communicate with the A2037E, and therefore the Data Recorder, via TCP/IP.
On the computer you use for data acquisition, you run the LWDAQ software, which you can download from here.
Download the latest version of the LWDAQ software from here. To help you with installation and use of the LWDAQ software, consult the User Manual. With the Subcutaneous Transmitters and Data Receiver, you will use the Recorder Instrument and Neuroarchiver Tool.
The Recorder Instrument is a set of routines that run inside the LWDAQ program. You can use the routines by opening the Recorder Instrument window from the Instrument menu. You can call the routines from the LWDAQ's console. The Recorder Instrument software downloads blocks of binary data from the Data Recorder hardware and divides them into blocks of fixed time-duration. It displays transmitter signals as it receives them, each transmitter trace in a different color, and prints a summary of the received signals to the screen.
The Neuroarchiver Tool uses the Recorder Instrument to download signals from the Data Receiver. The Neuroarchiver downloads, filters, displays, and stores to disk selected signals from the Data Recorder hardware, and does so using the Recorder Instrument as an intermediary. The Neuroarchiver calculate and display the Fourier Transform of the incoming signals. It stores data and transforms to disk.
[04-MAR-16] The Subcutaneous Transmitter (A3019) provides one analog input channels by means of two flexible, insulated stainless steel flexible wires. In the A3019A, this input is low-pass filtered by a three-pole active filter with roll-off of at 160 Hz and high-pass filtered by a single-pole filter with roll-off at 0.2 Hz. Its dynamic range is 20 mV. The A3019A can detect a 5-μV sinusoid of any frequency between 5 Hz and 160 Hz, and measure its amplitude with 1-μV accuracy. The A3028 is a two-channel transmitter with similar performance, depending upon how it is configured.
All transmitter models digitized at with sixteen-bit precision. The A3019A and A3028A digitize each input channel at 512 SPS. Each sample is transmitted in its own, isolated, 7-μs radio-frequency message. Transmission takes place through a dedicated loop antenna. The A3019A and A3028A transmit in the 902-928 MHz ISM band. It uses two frequencies to represent two logic levels. The messages propagate through space as radio waves. See How Antennas Work in our Technical Proposal for an explanation. The power transmitted by the antenna during the 7-μs burst is roughly 300 μW (−5 dBm).
The transmitter's analog circuits operate continuously to amplify and filter its analog signals. But the logic and RF circuits are inactive almost all the time. In the case of our A3019A, the logic and RF circuits wake up for 7 μs every 1953 μs, which is 0.36% of the time. During this 7-μs burst of activity, the transmitter converts its single analog input into a sixteen-bit number, and transmits this number, together with some synchronizing bits, the transmitter's four-bit ID, and a checksum. We call this transmission a message. The A3019A's message rate is 512 per second. Within each message, the transmitter sends bits at the bit rate. The A3019 and A3028 bit rate is 5 MBPS, or 200 ns per bit. For more details of the message encoding, see below.
The radio-frequency (RF) messages are received by our Loop Antenna (A3015). The messages are joined in the antenna by RF signals from other sources of similar frequency. The messages and the interference propagate along a coaxial cable to the Data Recorder (A3018) or Octal Data Receiver (A3027). The signal power arriving at the antenna must be at least four times (12 dB) greater than the interference power from other radio-frequency sources, and at least 25 pW (−76 dBm) if there is no interference. Interference power of 10 nW (−50 dBm) in the 902-928 MHz band is common in urban areas, at which level, the receive antenna must receive at least 40 nW, which we cannot rely upon when the transmitter is implanted in a moving animal, even if the receive antenna is no more than 30 cm from the animal. In order for reception to be reliable in the presence of 10 nW of interference, we must place the transmit and receive antennas inside a faraday enclosure to reduce the interference power below 1 nW. Faraday enclosures such as our FE2F provide at least a factor of ×1000 (30 dB) isolation from external interference, and so make it possible to obtain reliable recordings with one antenna from up to fourteen cohabiting animals. Larger, less restrictive faraday enclosures offer isolation of 10 dBm, and with the help of multiple, independent antennas connected to the same data recorder, we obtain reliable reception despite prominent interference.
The radio signal enters the metal enclosure of the Data Recorder and connects to the RF input of a demodulating receiver (such as the A3017). The RF messages and interference are amplified by 20 dB and passed through a bandpass filter that rejects signals outside the 902−928 MHz. What remains are the messages and ISM-band interference. These are amplified by another 20 dB and enter a mixer, which downshifts the 902-928 MHz RF signal to a 38-64 MHz intermediate-frequency (IF) signal.
The IF signal passes through three limiting amplifiers. Each provide 22 dB of gain, and limit their output. When we disconnect the antenna, the signal on IFL is random, but not quite saturated. When we connect the antenna, the RF interference in our laboratory causes IFL to saturate almost all the time. Because IFL is saturating, its amplitude is constant. Only its frequency changes. The fixed-amplitude IFL passes into a tuned circuit called the discriminator. The discriminator attenuates the lower IF frequencies, so that its output amplitude depends upon the frequency of IFL.
The discriminator output passes into a full-wave demodulator. The demodulator turns the AC signal into a signal proportional to the AC amplitude. The demodulator output is D in the S3017_2 schematic. The alternating radio frequencies of the transmitter messages appear in D as a square wave.
The D signal is mildly band-pass filtered by two RC networks to form the signal S in the S3007_1 schematic. The transmitter messages appear in S as a square wave centered about the 0-V potential.
A comparator transforms S into a logic HI when S is above 0 V, and logic LO when it is below 0 V. The comparator output is called C. The logic levels of the original transmitter message now re-appear as logic levels on C. Outside the messages, C is a random or pseudo-random sequence of logic levels. It is random when the dominant source of RF power at the antenna input is electronic noise, and pseudo-random when the dominant source is interference in the ISM band.
The C logic signal passes into the Data Recorders's logic chip, where it is synchronized with a 40-MHz message clock. The synchronized version of C is SC. It changes only on the rising edges of the message clock. Pulses on C that take place between these rising edges do not appear in SC. The Data Recorder monitors the stream of bits on SC, looking for messages embedded within the stream. These messages could arrive at any time. Messages can also appear in the SC bit sequence by chance, as a result of thermal noise or interference. We call these bad messages.
The Data Recorder stores any messages it detects in its 512 KByte first-in first-out buffer as thirty-two bit records, as we defined by the SCT message encoding. The last eight bits are a timestamp, which counts cycles of the Data Recorder's 32.768 kHz clock.
The Data Recorder stores clock messages 128 times per second in among the SCT messages. The clock messages take the form of a transmitter message from a transmitter with ID zero. The sixteen-bit data of a clock message is a sixteen-bit counter that increments by one with every clock message stored. The final eight bits are a timestamp, which is always zero, because the clock messages are stored whenever the Data Recorder's 32.768 kHz cycle counter wraps around to zero. You can think of the clock messages as being the product of a virtual transmitter with ID zero, transmitting 128 messages per second, and whose data is a counter that increments from one message to the next.
The Data Recorder connects to the LWDAQ. The LWDAQ in turn connects to an Ethernet. Your data acquisition computer runs the LWDAQ software and communicates with the LWDAQ by TCPIP.
The LWDAQ software on your data acquisition computer uses its Recorder Instrument to download blocks of messages from the Data Recorder's first-in first-out buffer. It plots these messages on the screen, including the clock messages. The Recorder provides you with blocks of data that cover a time interval you specify. As it acquires data, the recorder makes no effort to eliminate bad messages or insert substitute messages. Its data blocks are therefore of variable memory-size, but fixed time-duration.
The Neuroarchiver Tool calls the Recorder to supply fixed time-duration data blocks. The Neuroarchiver records all data downloaded from the Data Receiver and stores it to a file. The file contains the raw values downloaded, with no processing or alterations made, and also provides a substantial metadata field for text data describing the recording. The file is arranged in our general-purpose NDF file format.
Meanwhile, the Player section of the Neuroarchiver will read, process, and display recordings from any NDF file, including the one that is receiving the freshly-downloaded data. We tell the Neuroarchiver which channel numbers we want to reconstruct and display. Each channel corresponds to a transmitter recording. A two-channel transmitter will have two consecutive channel numbers for its two signals. Our subcutaneous transmitters have channel numbers between 1 and 14. Channel number zero is reserved for the Data Receiver's clock messages. Channel number 15 is reserved for slow communication, such as command acknowledgement in our Implantable Sensor with Lamp. The reconstruction of each channel from the raw data involves removing bad messages and inserting substitute messages where necessary. The Neuroarchiver displays the reconstructed channels and their discrete Fourier transforms. When playing back an archive, the Neuroarchiver will perform processing on the signals, so as to calculate metrics for event classification, export the signal to another recording format, or determine the total power in various bands of the signal spectra.
The bit rate is the rate at which the transmitter sends bits during one of its message transmissions. The A3019A, for example, transmits a total of 40 bit-values in 8 μs. The bit rate is 5 MBPS (megabits per second). We describe the function of these bits in the Message Encoding section here.
With the exception of a few temporary, slower versions we made for the following tests, all Stage Four and Stage Three circuits run at 5 MBPS. Here we compare the performance of the Stage Four system at 5 MBPS and 2.5 MBPS and show that the gain in operating range we make by dropping the bit rate is slight, while the loss in operating life is significant.
We set up a Demodulating Receiver (A3017) with a SAW Oscillator (A3016SO) and Data Recorder (A3007C). We connected a Dipole Antenna (A3015B) on a 240-cm (96") coaxial cable (RG58C/U). We recorded messages from a Subcutaneous Transmitter (A3013). We configure the transmitter and Data Recorder for message transmission at 5 MBPS (200 ns per bit), then for 2.5 MBPS (400 ns per bit). The switch between bit rates requires the following changes.
We measure a variety of performance parameters for each bit rate. We present our measurements in the following table.
|2.5 MBPS||Parameter||5 MBPS|
|918±4 MHz||Frequency Modulation||918±4 MHz|
|13 MHz||Bandwidth (90% Power)||18 MHz|
|12 μs||Message Duration||8 μs|
|108 μA||Active Current||75 μA|
|18 μA||Inactive Current||18 μA|
|15 m||Maximum Range (favorable orientation)||10 m|
|<0.1/s||Bad Message Rate|
|<0.1/s||Bad Message Rate|
|<0.2%||Blocked Message Rate|
(Transmitter Adjacent to Antenna)
|≈3%||Missing Message Rate|
(Transmitter Moving Randomly 1 m from Antenna)
|≈15%||Missing Message Rate|
(Transmitter Moving Randomly 1 m from Antenna
and Enclosed Between Two Hands)
The 5 MBPS and 2.5 MBPS transmissions perform best with a modulation depth of ±4 MHz. The 90% power bandwidth of the 5 MBPS messages is 18 MHz, which leaves us with at least 8 MHz of extra room in the 902-928 MHz band. The power bandwidth of the 2.5 MBPS messages is 13 MHz. In theory, the narrower the bandwidth, the less likely a hole in the reception caused by interference (such as the this) will lie within the transmission bandwidth, and therefore the less likely that the transmission will be corrupted by interference.
The biggest benefit of the lower bit rate is a 50% increase in operating range. The biggest cost of decreasing the bit rate are the increase in active current. The current rises from 75 μA to 108 μA. The operating life drops from nine weeks to six weeks. Another cost of decreasing the bit rate is the increased transmission time, which means a greater probability of collisions between multiple transmitters sharing the same receiver. The probability of collision between any two transmitters on any given message transmission is 0.7% at 5 MBPS and 1.4 % at 2.5 MBPS. With ten transmitters, each would lose 7% of its samples at 5 MBPS and 14% at 2.5 MBPS. Our system can handle a 30% loss of messages. At 2.5 MBPS we would use up most of this missing message budget in the normal operation of ten transmitters.
We decided to stay with a 5 MBPS bit rate. Our collaborators at ION stated that battery life and bandwidth were more important than operating range, because in either case the operating range appears to be greater than the size of a standard rat cage.
[24-JAN-17] The subcutaneous transmitters encode ones and zeros in a frequency-modulated radio signal. The signal consists of two frequencies separated by roughly 8 MHz, and both within the 902-928 MHz band. The transmitter messages are a sequence of ones and zeros sent with Manchester Encoding. We first described the SCT message structure here. Each message begins with eleven bits of value one (1). These bits serve two purposes. They synchronize the receiver clock with the message clock, and their exact period serves to distinguish the start of a transmitter message from random noise. The message body begins with a start bit of value zero (0). After the start bit comes the lower four bits of the transmitter's eight-bit channel number. The most significant bit comes first. The sixteen-bit sample comes next, most significant bit first. After the sample comes a four-bit completion code. The completion code is a function of the channel number, and allows us to deduce the upper four bits of the channel number from the lower four bits.
The message contains no checksum. The primary means by which the data receiver picks out genuine messages from interference and noise is through close examination of the timing of the signal. The encoding of the message produces a frequency transition every 195-215 ns. If a transition fails to occur, the receiver will abort decoding and look for another sequence of sychronizing bits. In our experience, each bit that satisfies the timing constraint makes it three times less likely that the message will turn out to be a bad message arising from interference or noise.
Where C is the completion code, N is the lower four bits of the channel number, and S is the upper four bits of the channel number, also called the set number. It would have been simpler to replace the completion code with the set number in the message encoding, but the completion code was othe one's complement of a channel number in the original four-bit channel number system. The completion code allows us to retain backward compatibility with the original system. The original system is equivalent to a system that uses only set zero. The following table gives examples of set numbers, channel numbers, and completion codes.
We reserve channel number 0 for clock messages, and call it the clock channel. Clock messages are generated by the data receiver only. Clock messages are stored among the transmitter data messages. We reserve all channel numbers for which the lower four bits are 0 for other future internal use. We reserve all channel numbers for which the lower four bits are 15 for auxiliary data. We call these auxiliary channels. Each set has its own auxiliary channel. We reserve channel numbers 224-225 for future applications.
In the Recorder Instrument and the Neuroarchiver Tool, we refer to signals by their channel numbers. The Recorder Instrument allows us to configure data receivers to select only one set, which we specify with a set number. By default, the Recorder Instrument attempts to configure its data receiver to record from all sets.
We introduced channel numbers 16-223 in January 2017. A firmware upgrade will permit any data receiver to record from the higher channel numbers. An Octal Data Receiver (A3027E) with firmware version ≥11 can be configured to receive from any single set zero (0) through thirteen (13), or from all sets zero through thirteen (0-13) simultaneously. We configure the data receiver with the daq_set_num parameter in the Recorder Instrument. When this parameter is 0-13, the data receiver accepts messages only from the corresponding set. When this parameter is "*", the data receiver accepts messages from all sets.
Note: There are some advantages to recording from only one set. Suppose we have a large faraday enclosure around an IVC rack containing forty animals each with a transmitter implanted. We want to know which transmitters are turned on. Our A3027 data receiver has channel activity lights numbered 1-15. All channel numbers with remainder 1 when divided by 16 will illuminate the first activity light. When we select only one set, we can be sure of which channel number corresponds to which light. Furthermore, we reduce the amount of data we must download. We could have four independent data receivers recording from four sets of ten transmitters. The auxiliary channel number light is blue on the A3027, while the other lights are white. There is no activity light for the reserved channels or for the clock channel.
Data receivers store transmitter messages as four-byte records. The first byte is the transmitter channel number. The next two bytes are the sixteen-bit data and the last byte is a timestamp. The timestamp counts periods of the Data Recorder's 32.768 kHz clock. The timestamp returns to zero 32768 ÷ 256 = 128 times per second. Whenever the timestamp returns to zero, the data receiver inserts a clock message into the data. The clock message has zero for its channel number, this being the clock channel. Its sixteen data bits are a counter that increments by one from one clock message to the next. Its final byte contains the data receiver firmware version number.
The auxiliary channels allow variants of the SCT, such as the Implantable Sensor with Lamp (A3030C), to transmit acknowledgments and other slow data. An SCT with an on-board accelerometer or thermometer would be able to use the auxiliary channel to send slow data without interfering with EEG transmission. An auxiliary message begins with the four bits of the number fifteen, followed by sixteen content bits, and then the completion code, calculated as above. The sixteen content bits begin with four bits for the identifying number of the transmitter that sent the auxiliary message. Because the completion code implies the set number, and we have an auxiliary channel number set aside for every set of channel numbers, we need only specify the identifying number to specify which transmitter in the set is using the auxiliary channel. This will be a number between one (1) and fourteen (14). Now come four field address bits, which indicate what the final eight bits, the data bits, contain. We reserve field addresses zero (0) and fifteen (15) for future use. Field address one (1) means this auxiliary message is an acknowledgement, and the byte following is the acknowledgement key. Field address two (2) means this message provides synchronization: the byte following is the SCT's internal timestamp at the time of transmission.
The Recorder Instrument supports auxiliary channels by diverting their contents into an auxiliary message list. Each element in the list is itself a list of four numbers: the channel number of the signal source that generated the auxiliary message, the field address, the data bits, and a timestamp. The timestamp is the value of the data receiver's 24-bit clock, which counts up at 32.768 kHz, so that the precision of the clock is ±15 μs and the clock cycle is 512 s.
Much of our discussion concerns itself with robust radio-frequency reception. We define reception robustness as the fraction of the time during which we receive 80% or more transmitted messages while the transmitter rotates and moves randomly at a particular range. Our sample rate and filtering is designed to tolerate the lost of 20% of the samples without loss of signal quality. We define robust reception as reception with robustness 95% or higher. In other words: signal quality is degraded by loss of samples for less than 5% of the time. A transmitter's operating range is the maximum range from the receiving antenna at which reception is robust. We use faraday enclosures to guarantee robust reception at ranges up to 300 cm in all laboratories. A transmitter's operating life is the maximum time for which the transmitter can provide robust reception in the absence of ambient interference. Its collision tolerance is the maximum number of transmitters of its type that can share the same receiver and still obtain robust reception in the absence of ambient interference.
The Data Recorder can fail to identify the message in its incoming bit stream (see Signal Path), even though they are received clearly by the Demodulating Receiver. The Data Recorder might be occupied with something that looks like a message, and when the real message arrives, all it knows is that the previous message was invalid. The recorder goes back to looking for a new message, but is too late to identify the real message. We measure the rate at which the Data Recorder loses messages by putting a single transmitter next to the antenna and watching for missed messages. Our signal is strong and there are no other transmitters to collide with it. We assume that any missing messages have been lost in the Data Recorder. It turns out that the number of synchronizing bits that the Data Recorder requires at the beginning of a message is what dictates the missing messages rate.
|Number of Synchronizing Bits||Missing Message Rate|
Only after it has received its required number of synchronizing bits does the Data Recorder proceed to message recording, and it will lose an incoming message only if it is already recording a false message. By insisting upon five synchronizing bits, we ensure that the Data Recorder loses less than one in a five hundred messages. As it turned out, we settled upon 11 synchronizing bits because this improved the performance of the system when the signal was weak.
We placed a single transmitter within a faraday enclosure, on an upturned paper cup 10 cm from the antenna within a faraday enclosure. We recorded continuously for an hour. Average reception was 99.7%. When we turn off the transmitter, the bad message rate is less than 0.1%. We conclude that roughly 0.3% of message are being blocked.
Each transmitter, when active, transmits messages in a continuous stream. The A3019A transmits 512 sixteen-bit data samples per second. A message that the Data Recorder fails to receive is a missing message.
The Neuroarchiver Tool handles missing messages by creating a substitute message with equal data sample value to the previous message received from the same transmitter. It does this only for the channels you are recording.
Despite the Neuroarchiver's message substitution, missing messages degrade the effective bandwidth and quality of our received signals. But they do not cause catastrophic problems, nor do they confuse our data acquisition system, which can rely upon a guaranteed stream of messages from each transmitter it monitors.
The primary cause of missing messages is lack of signal strength at the antenna when compared to ambient interference power, as we discuss in Ambient Interference. The primary cause of loss of signal strength is cancellation of the RF signal by its own reflections arriving at the antenna, as we discuss in Operating Range. The primary source of ambient interference is cordless phones operating in the 902−928 MHz band, and mobile phones operating in the frequency bands immediately above and below the 902−928 MHz band. Even if such mobile phones transmit only 1% of their power outside their designated frequency bands, they will still combine to produce substantial interference in neighboring bands. Another source of interference is other subcutaneous transmitters, in which case we refer to the interference as a collision.
Mobile phone interference in the ION laboratory in London, which is on the eighth floor of a building, was so severe that reception from an implanted transmitter at range 50 cm would sometimes drop as low as 20%, and was rarely above 80%. With the help of one of our faraday enclosures, reception jumped up to an average of 99%, with a minimum of 98% during any four-second period.
Noise and interference can generate messages on their own. We call these bad messages. Bad signals can raise false alarms, spoil the scale of self-adjusting displays, and wreak havoc with your Fourier Spectrums. We have made every effort to avoid them.
Bad messages have a valid transmitter ID number, but their data is invalid. The Recorder Instrument rejects them when they occur with ID numbers you have not asked it to record.
The bad message rate in our laboratory is less than one per ten minutes when there are no transmitters active. In London, the bad message rate was roughly one per second, sometimes coming in bursts of ten in a second. With transmitters active, we find that some messages get corrupted, as we describe below, and these message become bad messages.
We spend the rest of this section showing how bad message can, in theory, arise from noise and interference, and how our message encoding makes such bad messages very unlikely.
As we describe in Signal Path, the Data Recorder monitors a logic signal called SC, looking for messages. These messages could arrive at any time. If the stream of bits is random, there is a chance it will generate a message by chance. If the stream of bits has some pattern to it because of radio interference, the chance of a message appearing in it at random might be much higher.
When we disconnect the Data Recorder's antenna, and replace it with a 50-Ω terminator, we stop radio signals from reaching our receiver. All that remains at the input to our RF amplifier is white noise, whose power in our radio-frequency passband is around −87 dBm. The receiver amplifies this noise until it is large enough to generate an active pattern of zeros and ones on SC. This alternation is random, because the white noise is random. The output of the Demodulating Receiver (A3017) is an analog signal, S. This signal is sharply bandwidth-limited by the Demodulating Receiver's radio-frequency passband, and mildly high-pass filtered when it enters the Data Recorder. The Data Recorder uses a comparator to generate the logic level SC from S. When we examine SC on an oscilloscope with the antenna disconnected, the bandwidth of the bit stream is around 10 MHz, which we determine with the help of various low-pass filters.
Our messages represent individual bits values as edges in SC, not as levels of SC. An edge is a transition from logic LO to HI or from HI to LO. A rising edge is a one, and a falling edge is a zero. A sequence of synchronizing one-bits at the beginning of the message tells the Data Recorder when to look for data-carrying edges, because it knows the rising edges in the synchronizing sequence are the data edges in these initial one-bits. Every subsequent data-carrying edge allows the Data Recorder to adjust its expectations for when the next data-carrying edge will arrive. With a 5 MBPS bit rate, the data-carrying edges are separated by 200 ns. Each bit takes roughly 8 periods of our 40-MHz message clock. We say "roughly" because the bit rate is not exact. The transmitter's message clock is generated by a ring oscillator, and is accurate to only ±10%.
If the Data Recorder detects an edge within one or two clock periods of another, it rejects the entire message and goes back to looking for synchronizing bits. It rejects synchronizing bits using the same constraint. The Data Recorder tests the level of SC for premature edges twice during each bit period. In order for a random sequence of bits to create a valid message, it must satisfy both tests for all bits of the message. There are around 25 bits in the message, so the random stream must take on the correct value 50 times. Furthermore, the message must have the correct checksum at the end, which is the same as saying that it must take the correct value for another five bits.
The likelihood of a random sequence on SC matching our message protocol is roughly 0.530. To the first approximation, there are 105 opportunities for the match to occur every second, because each message occupies roughly 10 μs. We expect our bad message rate from noise and random interference to be of order one per hour. And indeed this is what we observe: after running for ten minutes we saw no bad messages with the antenna disconnected.
Most interference, however, is not random. It is generated by communication devices and its purpose is the transfer of information in a regular fashion. Such interference can have a far greater probability of creating a bad message than random interference. In a basement laboratory with −68 dBm interference power in the 902-928 MHz band, we recorded the number of bad messages in consecutive one-second intervals with a Data Recorder (A3018C). We used no faraday enclosure, and no transmitters were running nearby.
The average number of bad messages per second was 1.1 and the standard deviation was 2.5. Given that we expect to receive of order 500 messages per second from active transmitters, this bad message rate is not significant.
A corrupted message is one that has been interfered with, but which still passes through our error-checking. A corrupted message has the same effect upon data acquisition as would a bad message with the same ID.
We can reduce the likelihood of message corruption by checking for errors in the content of the message, as opposed to only by comparing the final four bits to the first four bits. A four-bit cyclic redundancy check at the end of all transmitted messages would give us better rejection of corrupted messages. Instead of ending up as bad messages, most corrupted messages would end up as missing messages. But the cyclic redundancy check would make our Stage Four transmitters incompatible with those of Stage Three, so we decided to tolerate a higher than necessary corrupted message rate for the sake of backward-compatibility.
The text below is a list of messages between one clock message and the next, taken from data recorded from a system with six active A3013A transmitters (starting time 71.0625 s in archive M1288538199.ndf). Each line represents one message, either a clock (ID 0) or a sample (IDs 1 to 14). We give the message index, the channel number, the sample value, the timestamp, and the hexadecimal representation of the four message bytes.
24 0 7050 5 $001B8A05 25 8 42595 0 $08A66300 26 12 43431 26 $0CA9A71A 27 7 43084 31 $07A84C1F 28 10 40959 43 $0A9FFF2B 29 8 42613 53 $08A67535 30 12 405 83 $0C019553 31 7 43100 90 $07A85C5A 32 6 42185 92 $06A4C95C 33 4 180 106 $0400B46A 34 10 40987 115 $0AA01B73 35 8 42615 126 $08A6777E 36 12 43416 160 $0CA998A0 37 6 42111 160 $06A47FA0 38 7 43116 162 $07A86CA2 39 5 42234 169 $05A4FAA9 40 10 40988 177 $0AA01CB1 41 8 42661 191 $08A6A5BF 42 7 43197 218 $07A8BDDA 43 12 43330 221 $0CA942DD 44 6 42310 235 $06A546EB 45 10 41052 242 $0AA05CF2 46 8 42689 246 $08A6C1F6 47 0 7051 5 $001B8B05
These messages were recorded inside faraday enclosures, so interference from non-SCT sources is minimal. Message 30 is a corrupted message from No12 in which the ID has remained the same. This corruption gives rise to a glitch in the No12 signal. Message 33 has ID 4, but there is no No4 transmitter in the system. We suspect that a message from No5 was corrupted in such a way as to appear as a message from No4.
The above errors occur more often when the signal from one or more transmitters is exceptionally weak. The errors manifest themselves as spikes in the signal. Because they are rare, we can remove them with a glitch filter. More of a problem than the spikes is the poor reception that goes with the weak signals that cause spikes.
[27-JAN-17] When two transmitters send a message at the same time, we say that they collide. When two transmitters collide, we can lose one or both of their messages. Consider message detection by a single antenna amplifier, as is the case with our original Data Receiver (A3018). If the first message is much more powerful than the second message arriving slightly later, the second will be lost. The second message will become a missing message. But the first message will be received correctly. If the second message is more powerful than the first, it will interfere successfully with the first, preventing its reception. The second message may also be lost, depending upon how fast the message detector can recognize and recover from the corruption of the first message. If the two signals are of equal power, neither will be received. If we have two or more independent receiving antennas, we can hope to detect both messages. The Octal Data Receiver (A3027) provides eight receiving antennas, each with its own amplifier, demodulator, and message detetor. Collisions occur at each of the antennas, but the outcome of the collision differs from one antenna to the next, so that the rate at which we lose messages to collisions is reduced.
Collisions occur because two transmitters are transmitting at the same time. If the transmission period were exactly regular, and two clocks drifted into exactly coincidence, collisions could occur systematically at every transmission instant. But the transmission period is not regular. The average period is exact, but the moment of each individual transmission is displaced by the transmitter by a small, random, amount of time. In the A3019A, the transmission period is 64 cycles of its 32.768 kHz clock, or 1952 μs. The A3019A delays its moment of transmission by 0 to 15 clock cycles, so that the actual moment of transmission can be delayed by up to 456 μs.
We call the displacement of the transmission instant transmission scatter. We describe how the data acquisition system handles transmission scatter in our Recorder Manual. The transmitter uses the lower four bits of its sixteen-bit ADC conversion as the source of a random number. We describe in more detail how the hardware implements the scatter in the Transmission section of our A3013 Manual. The figure below shows transmitter scatter on the oscilloscope screen. As you can see, the scatter is not perfectly random.
Faraday enclosures provide isolation of transmitters from ambient interference, and also from transmitters that do not share the same receiver. If we have eight transmitters in four separate faraday enclosures, each faraday enclosure will have its own antenna. We combine the signals from the six antennas and feed the combination into our Data Receiver (A3018) using an Antenna Combiner (AC4A). Transmitters in these six enclosures will interfere with one another whenever their transmissions overlap.
At 512 SPS, the transmission period is 1952 μs and the transmission itself lasts for only 7 μs. Each transmitter transmits for 0.35% of the time. The chance of one 7 μs transmission overlapping another is 0.7%. The average loss due to collisions when we have n transmitters sharing the same receiver is (n-1) × 0.7%. If we have If we have twelve transmitters sharing a receiver, we will lose roughly 8% of messages to collisions when averaged over a long time period. We say the average collision loss is 8%.
But the collision loss varies with time, as shown in the following graph. Its average value may be only 0.7% for two transmitters, but it's peak value can be ten times higher.
In the above graph, we see cyclic variation in reception caused by a slight difference in the transmitter clocks. Over the course of eighteen minutes, the cycles grow in amplitude from 1% to around 8% and shrink again. Another twenty minutes goes by with no evidence of collisions, and the collision cycles begin again.
The period of message transmission at 512 SPS is roughly 2 ms (64 cycles of the transmitter's 32.768 kHz oscillator). An entire collision sequence, such as the thirty-six minute sequence captured in the figure above, takes place as the two clocks drift with respect to one another by 2 ms. The ASH7KW clock we use on the A3028 is accurate to ±20 ppm over its entire temperature range. We expect differences between the clocks of 5 ppm to be common. In the example above, it takes 2200 s for the two clocks to drift apart by 2 ms. The two clocks differ by 0.9 ppm. This combination of clocks is unusual. More often we see collision cycles of period several hundred seconds, like this one taken from transmitters in live animals at ION in London.
The individual cycles in the above sequence have period thirty seconds, as we can see in the following figure.
Each cycle corresponds to a drift of 30.5 μs between the two clocks (one period 32.768 kHz). Because our clocks are only 0.9 ppm apart, we get a 34-s cycle. We see the structure of the cycle clearly. There is a 18-s period with no collisions in each cycle, and a 16-s period where the collisions take place.
The first period of collisions in the entire collision sequence begins when the earliest transmission window of one transmitter coincides with the last transmission window of the other. In our case, the two 7 μs windows overlap for 16 s (that's twice 7 μs divided by 0.9 ppm). Collisions in this first period should be rare, because only one window overlaps. The chance of a collision should be 1/16 × 1/16 = 0.4%.
As we see in the oscilloscope trace above, the four-bit values the transmitters use for random numbers are not uniformly-distributed across their sixteen possible values. Furthermore, one transmitter might be so much stronger than another that the interference is only one-way. In our example sequence, we see the first cycle has a depth 6% for No4 and 0% for No11. The peak cycles have depth 8% for both transmitters.
The second cycle of collisions occur 34 s after the first, when the clocks have drifted 30.5 μs to the next window overlap. Now the earliest window of one transmitter coincides with the second-to-last last window of the other. But now we have the second-earliest window coinciding with the last window as well. Our chance of collision is, in theory, 0.8%. The overlap lasts for 16 s, and 18 s later comes the next overlap. The chance of a collision is 1.2%. On the sixteenth cycle, all sixteen windows overlap, and the collision probability is 6%. The cycles shown in detail in the figure above are the largest in the collision sequence. Both transmitters experience a 6% drop in reception.
Sixteen cycles after the peak, none of the windows overlap. The clocks drift another 1 ms apart over the next 1200 s and there are no collisions. Now the entire sequence starts again.
If we place n transmitters in the same faraday enclosure, it is inevitable that their sixteen transmission windows will coincide at some point in time. At that time, the collision rate between transmitters will be at its greatest, and reception will be at a minimum. If we consider a message occurring in one of the windows, the probability of no other message occurring in the same window is 0.94(n−1). If we assume near-perfect reception in the absence of collisions, which appears to be the case in faraday enclosures, this probability is the minimum reception rate for the n transmitters.
|Number||Average (%)||Minimum (%)|
For robust reception we need to receive more than 80% of messages for 95% of the time (robustness of 95% or higher). It's not obvious from the minimum and average reception values whether reception will be robust. To obtain a good estimate of robustness, we simulated the collision cycles of n transmitters. Our simulation program, collisions_1.pas, simulates sets of n transmitters working together with randomly-distributed clock periods over a period of two thousand seconds. The figure below shows reception from the first four transmitters in a set of fourteen.
When the simulation begins, all clocks are synchronous, and we obtain the minimum reception. After that, the interactions between the clocks become complex. For a 100-s detail, see here. For each simulation of n transmitters, we obtained n values for minimum reception, average reception, and robustness. These values were consistent from one transmitter to the next to within a few percent, so the values we give in the table below are the average values taken over the n transmitters.
|Number||Average (%)||Minimum (%)||Robustness (%)|
Robustness is greater than 95% all the way up to n = 12, so the collision tolerance of a set of 512 SPS transmitters is 12. This is why we talk about 12 transmitters sharing a single receiver in other sections. Robustness for n ≤ 5 is 100%.
In our simulation, we assume that any collision between two transmitters will result in the loss of both messages. This is not true in practice. If the power received from one transmitter is 12 dB greater (16 times greater) than the power received from another, the more powerful signal will dominate. If reception of the more powerful signal is taking place, the weaker signal will be ignored. If the weaker signal is being received, there is a good chance that the receiver will have time to abandon the weaker message and receive the stronger message. In our recording of two stationary transmitters, shown here, we see that fewer messages are lost from No7 than No4. Transmitter No7 is in air with an 80 mm antenna. Transmitter No4 is in water with a 50-mm antenna. We assume the signal from No7 is stronger than that from No4.
The following graph shows measured transmission over the course of half an hour with nine active transmitters.
Average reception over the half-hour period was 92% for No3 and 99% for No1, with the others in between. The minimum reception observed in any four-second period was 71% from No2 and 94% for No1, with the others in between. Reception is almost always greater than 80% for all transmitters, so we conclude that we obtain robust reception from nine transmitters despite collisions.
All these calculations assume that we have only one antenna amplifier and demodulator with which to detect all transmitted messages. But the Data Receiver (A3027) provides eight independent antennas. When two transmitters collide but are in different locations, both messages are likely to be received correctly because at each location one transmitter dominates over the other. We placed nine transmitters in a faraday enclosure, emitting a total of 8192 SPS, and received their messages with three independent antennas. Our simulation suggests a loss of around 12% of messages due to collisions at each of the antennas, assuming reception would be perfect otherwise. In practice, we observe an average loss of 4%.
We define operating range as the greatest range from the pick-up antenna at which we obtain robust reception. An A3019A transmits 512 messages per second, so the operating range is the greatest range at which we receive 410 or more messages per second in 95% of orientations and positions at the operating range. In practice, we operate the transmitters inside a faraday enclosure, so their operating range must be adequate to cover the range from the antenna in the center of the enclosure to the farthest corner of the enclosure volume.
When we hold a transmitter between two fingers in our basement laboratory, and move it around randomly at range 100 cm for a minute (see movie), we receive 98% of the messages it transmits. The operating range of the A3019 is greater than 100 cm in our basement laboratory. In our our office in the center of Waltham, operating range is closer to 70 cm. Operating range in the ION laboratory in London appears to be around 30 cm. The decreasing operating range is the result of ambient interference, which dominates transmitter messages whenever the transmitter signal is attenuated by reflections and unfavorable orientation.
Regardless of the operating range of transmitters in a laboratory, we recommend the use of enclosures with conducting walls to block out ambient interference. Such enclosures are called faraday cages. We call them faraday enclosures so we don't get them confused with animal cages, which are used to contain animals. We discuss ambient interference below, and we describe the performance and construction of faraday enclosures in our Faraday Enclosures report.
Because faraday enclosures give us at least a 30-dB (one thousand-fold) reduction in ambient interference, they increase the operating range of our transmitters by a factor of 30 (square root of one thousand). Even if the operating range is only 10 cm without a faraday enclosure, it will be 300 cm within a faraday enclosure. We expect robust reception from transmitters operating within an enclosure 500 cm × 500 cm × 200 cm, provided the antenna is in the center of the floor of the enclosure. Our FE2A enclosure is far smaller than this.
Robust reception within a faraday enclosure appears to be guaranteed. Nevertheless, we always benefit from increased signal strength compared to ambient interference, and we have worked hard to increase the operating range of our transmitters in the absence of faraday enclosures. We devote the remainder of this section to a discussion of what phenomena limit the operating range of our transmitters. The discussion explains why we use a poorly-tuned loop antenna to receive our transmitter signals, and why we use a poorly-tuned bent antenna to transmit them.
When we implant a transmitter in an animal, the 50-mm antenna of an A3013A-E will resonate more efficiently, and so transmit more power (see Antenna Length). We may lose some or all of that additional power by absorption in the animal's body (see Transmitter in Baby Rat Corpse).
We compared the operating range of the Data Receiver (A3010B), which is the Stage Three receiver upgraded with new firmware and an antenna socket, and the Data Receiver (A3018A), which is the original Stage Four receiver. (The original Stage Three receiver, the A3010A, suffered from a firmware bug that reduced its performance even farther with weak signals.) We used the same Subcutaneous Transmitter (A3013A) for both receivers. We used the same Loop Antenna (A3015A) for both receivers. We unplugged it from one receiver and plugging it into the other.
It appears that our Stage Four development has doubled the effective range of the transmitters. We claim that we are nearing the physical limits of message detection, as we shall now explain.
The noise in our receiver is the thermal and amplifier noise at the antenna input. The interference is RF power in the receiver's pass-band arriving at the antenna from transmitters other than those the receiver is intended to receive. The effective noise power at the input of our demodulating receiver is −90 dBm (see Noise and Interference). But interference power, even in our basement laboratory, is over a hundred times more powerful than this noise, at −68 dBm (see Noise and Interference).
Our signal must have power 12 dB greater than the interference in order to avoid corruption (see Foreign Interference). When we transmit power across 1 m of space from a quarter-wave antenna to a loop antenna, we lose at least 32 dB compared to connecting the power directly to the receiver circuit with a cable (see Reception). Even with our omni-directional antennas, an unfavorable relative orientation of the transmitting and receiving antennas causes a 17 dB drop in received power (see Omnidirectional Antennas and Transmit Antenna). We must also deal with reflections, or multi-path interference (see Multi-Path Interference and Radiated Power). Reflections of the main signal interfere at the receiver. We can easily get a 10-dB loss due to destructive interference. In other words: we can lose 90% of our power easily
Our transmitters produce roughly −4 dBm (see Radiated Power and Modulating Transmitter). Transmission across 1 m of air with can drop our received power to −53 dBm. This −53 dBm is still more than 12 dB above our interference power of −68 dBm.
But we must contend with multi-path interference as well, which give rise to reception dead spots. With the antenna in an unfavorable orientation, in which we receive only −53 dB from line-of-sight transmission, we have power radiating more effectively in other directions. This power can reflect off nearby conducting surfaces and arrive at the receiving antenna with as much strength as the line-of-sight signal. The reflection adds to the line-of-sight signal. It might reinforce the line-of-sight signal, or cancel it, depending upon their relative phase.
The phase difference between a reflected and line-of-sight wave is a strong function of frequency. If the reflected path length is three meters long, this is a hundred wavelengths. A 1% change in frequency will cause a 2π change in phase. A 0.25% change in frequency will cause a π/2 change in phase. The figure below shows two holes in the response of our our Demodulating Receiver (A3017). The antennas are 1 m apart, on either side of our body.
We see that each hole is about 1 MHz wide. A 0.1% change in frequency causes the cancellation to stop. We take such sharp holes in the frequency response to be evidence of multi-path interference. The width of the holes is consistent with a reflected path length of several meters. Longer path lengths would give holes that were even more sharp.
Aside: The two holes are separated by 24 MHz, and you will notice yet another hole, on the left side, which is 24 MHz below the middle hole. As the frequency changes by 24 MHz, the phase difference between the line-of-sight and reflected waves changes by 2π. The number of wavelengths in the path difference changes by one. Because the frequency changes by 2.5% to bring about this one-wavelength change, the path difference must be 40 wavelengths, or 12 m. This suggest to us that the wave is bouncing off one of the metal shelves five or six meters from our transmitting antenna. If that's the case, then the reflected signal will be about 25 dB weaker than the line-of-sight signal, except for the fact that our body is in the line of sight. Human tissue attenuates 900 MHz by approximately 1 dB/cm, so we expect a loss of around 25 dB through a human torso. Both signals are each strong enough for reception, but they cancel one another at particular frequencies.
With 10 dB loss due to cancellation by reflections occurring at the same time as 17 dB loss due to poor antenna orientation and 32 dB loss due to transmission across 1 m of air, our signal drops to −63 dBm, which is only 5 dB above our −68 dBm interference. Reception will fail.
With a favorable orientation of the transmitter, however, we can expect −36 dBm at 1 m, and − 56 dBm at 10 m. Even at 10 m, we can receive a signal that is 12 dB above our interference. In fact, we find we can get reliable reception at up to 15 m in our lab by orienting the transmitter properly.
Radio frequency power from sources other than our subcutaneous transmitters is what we call ambient interference. If it's large enough, ambient interference can dominate the transmitter signals and cause missing messages. In early 2009, we determined that our operating range in the ION animal laboratory in London was only 25 cm, after observing a 50-cm operating range for most of 2008 (details in this e-mail). A range of 25 cm is not adequate to cover even a single animal cage, let alone four at once. We assumed that ambient interference was to blame for the the short operating range, and began to look at ways to block ambient interference from reaching our receiving antenna.
We can block out ambient interference with a faraday enclosures, which is a box with electrically conducting walls. There can be holes in the conducting walls, but the holes must be less than 1% of the wavelength of the radiation if we want to block 99% of the power. In our case, the wavelength in air is around 300 mm, so holes less than 3 mm will be fine. We describe our experiments with home-made faraday enclosures in our Faraday Enclosures report.
Another way to block ambient interference is to turn your entire laboratory into a faraday enclosure. You can paint the walls, ceiling, and floors with shielding paint. You could cover the windows with aluminum mosquito mesh, or make curtains out of shielding fabric. Alternatively way to block ambient interference is to move your animal laboratory into a basement with a concrete ceiling.
Our FE2B faraday enclosure sells for around a thousand dollars US. It is large enough to contain two rat cages. Four litter-mates could live together in each cage, each with an implanted transmitter. We believe this is the most cost-effective way of performing long-term studies.
Ambient interference can penetrate the recording system through the cable that joins the data receiver to the LWDAQ Driver. To cut down on such penetration, we must use a shielded LWDAQ cable for this connection, as we describe in Shielding.
Our experiments with the Subcutaneous Transmitter (A3009) showed us we needed to increase the gain of the analog input from ×30 to ×300 if we were to see normal EEG activity in a mouse brain. We also detected a couple of sources of noise in our analog-to-digital conversion process. Apart from the increase in amplification, and the addition of an auxiliary input, the Subcutaneous Transmitter (A3013) is much the same as its predecessor.
The A3013, with its greater gain, has roughly 12 μV of cross-talk and thermal noise on its input. It can detect a 5-μV 160-Hz sinusoid with no difficulty, once you take the Fourier Transform of its output in the Neuroarchiver.
See the Analog Inputs section of our A3013 Manual for a detailed discussion of the analog inputs. See our Mains Hum report for more on the origins of mains hum. Our transmitters pick up no measurable mains hum when implanted in animals. The electrodes measure the difference in potential between two parts of the brain. Mains hum tends to operate upon the entire animal body. Furthermore, with the introduction of Faraday Enclosures to isolate transmitters from radio-frequency interference, we find that the transmitters are isolated from mains hum as well.
We encapsulate transmitters in epoxy and coat them in silicone. The result is shown here. For an account of our work on rugged encapsulation that is resistant to water, fatigue, and vacuum, see our Encapsulation report.
We use stainless steel coiled wire with silicone insulation for the analog leads, and a stranded steel cable with silicone insulation for the antenna. We describe our selection of wires suitable for long-term implanting in lab animals in our Flexible_Wires report.
[15-OCT-13] The operating life of a transmitter is how long it takes to consume its battery capacity in its active state. The A3019A provides 1 voltage-sensitive input (for EEG), 1 magnetic field input (for turning the transmitter on and off), and one radio-frequency output (for transmitting messages to the receiver). During radio-frequency transmission, the A3019A consumes around 10 mA, which would run our battery down in one day. But transmission takes place only 1% of the time, so current consumption is reduced by a factor of one hundred. The transmitter trusts that the receiver will be listening all the time, and with such vigor as to pick out messages from a dozen different from a continuous, random message background. The transmitter consumes 100 uA, the receiver consumes 20 W.
We have moved the complexity and power consumption of the communication into the receiver, out of the transmitter, and this is why our transmitters last for hundreds of times longer than any other radio-frequency telemetry devices on the market. These other radio-frequency systems use transceiver chips that set up continuous reception and transmission between the implanted device and its recorder.
The shelf life of a transmitter is how long it takes for the transmitter to consume its battery capacity in its inactive state. When a transmitter is inactive, its magnetic switch consumes some current. In the case of the A3019A, the switch consumes around 5 μA. In the A3028A, it consumes 2.5 μA. Both transmitters have a BR1225, 48 mA-hr primary lithium cell as a power source. These cells last for ten years or more without losing charge. At 5 μA, the battery will run down in about a year. At 2.5 μA, the battery will last two years.
With the introduction of faraday enclosures, robust reception from implanted transmitters became assured. We describe the development of our faraday enclosures in our Faraday Enclosures report. Our faraday enclosures reduce ambient interference by at least 30 dB, and most likely 40 dB. The result is an increase in our operating range of 30-fold to 100-fold. Not only is interference kept out of the enclosure, but the transmitter signals are kept in the enclosure. Transmitters in separate enclosures do not interfere with one another unless they share the same receiver. Nor will they interfere with other users of the same radio-frequency band.
Here we discuss the implantation procedure. For recordings of EEG and other biopotentials from live animals see our Example Recordings page.
[15-OCT-13] The volume of a transmitter must be small enough to be accommodated by the target animal comfortably. We can get an idea of what is acceptable in the field by looking at the size of implantable mini-pumps such as those made by Alzet. According to Alzet, a 0.5-ml pump is fine in mice 10 g and larger, a 1.0-ml is fine in mice 20 g and larger, and the 6.5-ml is fine in rats of 150 g and larger (see here). Our veterinary collaborators tell us that two such pumps can be implanted in one animal without excessive discomfort.
From these observations, we arrive at our rule of thumb for implant size in mice and rats. These animals can tolerate a single implant up to 0.07 ml/g or two implants each up to 0.05 ml/g. In this paper on rat growth, we see that rats are roughly 80 g when they are a month old. They can tolerate up to 5.6 ml in one implant. Our 2.8-ml A3028E is well within this limit. For mice, their weight is a strong function of mouse strain. By and large, however, they get to 10 g after 3 weeks and 20 g after 6 weeks. A 20-g mouse can tolerate our 1.2-ml A3019A, but a 10-g mouse would find such an implant uncomfortable.
The choice of electrodes at the tip of the leads depends upon what we want to measure, as well as the size of the animal. We present our work on electrodes here. In mice, the skull is only 200 μm thick near the bregma. For EEG monitoring, we recommend our 0.6-mm long, 0.5-mm diameter threaded electrode. In rats, the skull is around 500 μm thick, and over the course of many weeks, will grow thicker. We recommend our 3.2-mm long, 1.5-mm diameter screws. We show the various screws we have used as electrodes here. For monitoring EEG near an optical cannula, we can insert the bare end of the steel lead.
Rats are strong animals, and aggressive scratchers of incisions or points of irritation. The analog pick-up leads of the transmitter, running from the back to the skull beneath the skin of the neck, will irritate a rat unless the wires are routed correctly. This diagram by Rob Wykes of ION shows how best to route the wires. When the wires were routed directly up the top side of the neck, half the rats scratched the leads out of their skin, which is an indication of intense discomfort on their part. When the wires were routed along the side of the neck, with adequate slack, none of them scratched the leads out, suggesting that this routine is far more comfortable for them.
[21-MAY-14] Pishan Chang (Neuroscience, Physiology & Pharmacology, UCL, London) implanted four A3019A transmitters in P28 rats (post-natal 28 days). The animals weighed 55-77 g at the time of implantation. The volume of the transmitter is 1.2 ml and its mass is around 2.6 g. The following table shows her measurement of their wight over the next seven days, compared to one control animal.
The weight gain suggests that the animals are healthy. They are cohabiting in the same cage, and we are able to record continuously from all transmitters. This video shows the rats moving around in their cage soon after implantation. Note the shaved areas around the incisions, and the stitches closing the incisions on the head and back. The transmitter is beneath the incision on the back. We run the leads under the skin to the incision in the scalp, where M0.5 screws act as electrodes through holes in the skull
[15-MAY-14] Implant the SCT with the battery facing in, so that the circuit board holding the magnetic sensor is closest to the skin. Use a large magnet. A small magnet, even if it is at close ranges, still has a short range. Its field will not penetrate easily through the skin. If you have the battery facing out, it is between the magnet and the magnetic sensor, and you will have trouble turning on the transmitter with a small magnet. With a large and moderately strong magnet, you should have no trouble turning transmitters on and off while implanted.
When implanting, you may be working on steel tables with steel implements. These can become magnetic and they can turn on and off the transmitter while you are working. One way a table or a tool can become magnetic is by storing your on-off magnet on the table or the tool handle. Turning off and on the transmitter during implantation causes no harm to the transmitter, but can cause concern that the transmitter has failed, when it has not.
[01-NOV-16] In our Technical Proposal, we laid out our plan to develop a transmitter small enough to be implanted in the body of a rat, fast enough to transmit four hundred data samples per second, powerful enough to be detected at a range of three meters, and efficient enough to operate for three months on a lithium battery. We divided our development into four stages.
The electronics of Stage Four were ready in March of 2007. It was then that we confronted the problem of robust, water-proof encapsulation for the transmitters. At the time we wrote our Technical Proposal, we believed that two coats of silicone dispersion would be adequate to protect the transmitter from water. Given the simplicity of the procedure, there would be no point in coating the transmitters ourselves. We proposed to send transmitters to our users without batteries, wires, or coating. Our user could solder their preferred wires to the transmitters, install the battery, and apply the coating.
Archive Photograph: See here for photograph of the prototype electronics for Stage Four in 2007. The Data Receiver is the aluminum box with its lid off. Alice is holding a transmitter (prototype version with external battery and programming connector). The loop antenna on a cable sits at one end of the table. A dipole antenna (disconnected) is next to the loop. Two Modulating Transmitters are in the background. The black box is the LWDAQ Driver, connected to our computer. All circuits shown in the picture are now obsolete, but their replacements look much the same. The LWDAQ Driver is a black box, the Data Receiver remains a silver box.
Water-proof encapsulation is not straightforward. Capillary action makes water a relentless invader of any opening or crack. During encapsulation, air bubbles trapped beneath components emerge into the curing encapsulation. In the low pressure of an aircraft cargo hold, the bubble beneath the battery pushes outwards, and will burst an encapsulation made of silicone. Our work on encapsulation is an appendix to the work we committed to in our Technical Proposal. After nine months of effort, we arrived at an encapsulation process that uses both epoxy and silicone, which we called Process X. We made our first transmitters with Process X in December 2007. We describe our work on encapsulation and encapsulation methods in our Encapsulation report. We later improved Process X and arrived at Process B. We made our first transmitters with Process B in June 2009.
Once our transmitters were water-proof in Process X, they endured for long enough in live animals for the wires to break from repetitive stress. Wires broke at the neck of a rat, at their solder joints, and where they emerged from epoxy encapsulation. Our work on flexible wires is an appendix to the work we committed to in our Technical Proposal, and was funded in part by ION's purchase of ten prototype transmitters. We describe our work on wire fatigue in our Flexible Wires report. In Spring 2009 we began trials using a variety of steel wires, including steel springs, as we describe in the Trials section of Flexible Wires. We eventually arrived at a stranded wire for the antenna and helical steel wires for the input leads. We insulate both with silicone, and this insulation forms part of the outer cover of the transmitter body.
Interference with our subcutaneous transmitter signals from outside sources proved to be a problem in ION's London laboratory. We solved the problem of interference with faraday enclosures. Our work on faraday enclosures is an appendix to the work we committed to in our Technical Proposal, and was funded in part by ION's purchase of ten more prototype transmitters. We describe the development of a practical faraday enclosure, and document its success, in faraday enclosures. Faraday enclosures are now an integral part of the subcutaneous transmitter system. Not only do they provide immunity to outside interference, but they allow many transmitters to operate simultaneously in the same space without interfering with one another.
At the end of October 2013, we had prototypes of our new Dual-Channel EEG Monitor (A3028). We were working on an automated thermo-plastic inner encapsulation for these devices, followed by silicone dipping. We were building the first of our Octal Data Receivers (A3027).
At the end of January 2017, the Octal Data Receiver (A3027E) with firmware version ≥12 supports channel numbers 1-222. We can, in principle, operate 196 transmitters in the same faraday enclosure and record data from all of them with a single A3027E. The Animal Location Tracker (A3032A) allows us to track the movements of animals over a 16 cm × 32 cm platform. Our Implantable Sensor with Lamp (A3030D) is receiving commands and flashing its optogenetic stimulator while implanted in rats. We have manufactured over two thousand transmitters to date, and are currently shipping fifty of them per month.
[01-NOV-16] The A3028E is a single-channel EEG monitor with volume 2.8 ml. It is designed for use in rats. Its EEG amplifier is protected from condensation by an epoxy encapsulation, and made water-proof by a silicone coating. Its analog leads are steel springs insulated with silicone. Its antenna is a stranded steel wire also insulated with silicone. The encapsulation and wires will resist the fatigue and corrosive fluids within an animal's body for at least one thousand days. When active, the A3028E transmits 512 samples per second for at least 18 weeks when it starts with a fresh battery. Each sample is sixteen bits, with a dynamic range of −13 mV to +9 mV. The electronic input noise is 12 μV rms. Its frequency response is limited to 160 Hz by a three-pole low-pass filter on the transmitter. Its near-ideal differential input picks up less than 1 μV of mains hum when implanted in a live animal. The A3028A is a dual-channel 512 SPS EEG/EMG monitor designed for implantaion in mice. Its volume is 1.4 ml. Its operating life is fifteen days. The A3028L is a dual-channel 1024 SPS EEG monitor with volume 6.5 ml, which is small enough to be tolerated by an adult rat. Each channel has bandwidth 0.3-320 Hz. Its operating life is 5.5 months. Many other versions of the same transmitter circuit exist.
In certain rural locations, and in certain basements, it is possible to operate a fourteen-channel SCT system with no faraday enclosure. But in most laboratories, a faraday enclosure is necessary to ensure reliable reception from the implanted devices. We have cooperated with customers to develop and build faraday enclosures to suit their laboratories. We have SCT systems operating in IVC racks, with an enclosure of steel mesh fabric around the rack. We have stacks of smaller enclosures on wheels that can be moved from one laboratory room to another. Most of our customers are able to build their own enclosures after receiving a few examples from us. We describe the enclosure components and performance here.
Within a larger enclosure, we can operate several SCT systems simultaneously with the help of transmitter set numbers, or we can record from a hundred transmitters with a single system. With the addition of the Animal Location Tracker (A3032A), a single SCT system can track the locations of up to fourteen animals in one or two cages. The A3032A platform is 16 cm × 32 cm, and thin enough to fit beneath a cage in an IVC rack. Its accuracy in measuring animal location is of order 20 mm, and its precision in detecting movement is of order 2 mm. The tracker makes as many measurements of position as it receives samples from the implanted transmitter. If the transmitter is running at 512 SPS, the tracker can monitor animal movements at up to 1000 mm/s.
All our hardware and software is open-source, distributed under the GNU Public License. Our LWDAQ Software is free and distributed under the GNU Public License. It runs on Linux, MacOS, and Windows. The software downloads transmitter signals from the Data Receiver via a LWDAQ Driver. The connection between the computer and the LWDAQ Driver is TCPIP. Eight A3028E transmitters, each producing 512 sixteen-bit samples per second, generate only 16 kBytes/s, so a local area network, or an office wireless network, has no difficulty transferring the data for recording.
Once the data has been recorded on disk, our Neuroarchiver Tool, bundled with our LWDAQ Software, provides an Event Classifier that allows us to find EEG events automatically in tens of thousands of hours of recordings. We describe our work on automatic event detection in Seizure Detection.