FAQs for MINT uLED-12-32
John Seymour
Frequently Asked Questions
Q: I am seeing a stimulation artifact whenever I pulse the LEDs. Can this be avoided?
Stimulation artifacts can occur in the recordings at the beginning (stimulation onset) and the end (stimulation end) of the voltage / current pulse to the micro-LEDs in a shape of high-frequency, high-amplitude swing of the signal. Stimulation artifact is typically observed on the recording channels on the same shank of the LED being stimulated, but it is possible that it occurs on all the recording channels. The artifact amplitude and pulsewidth varies depending on position relative to the LED and the impedance of the given electrode.
The cause is mostly dominated by the photovoltaic effect for which we hope to have a solution available in 2019. Whether artifacts can ever be completely eliminated is yet to be proven but we have reason to be hopeful. Fortunately, there are ways to deal with it when it occurs, namely pulse shaping and subtraction methods.
Pulse shaping, i.e. slowing the slew rate (or rate of change) of the pulse signal provided to the LEDs, will reduce the artifact amplitude. Driving your LED with a sine wave is an extreme form of pulse shaping. In standard CMOS drivers, the turn-on period is in nanoseconds. It has been observed that sinusoidal or gaussian off-to-on and on-to-off transitions exhibit stimulation artifacts of smaller amplitudes than that of (high-speed) ramp. It has also been observed that ramp transitions with rise (and fall) times as long as a few milliseconds can also greatly reduce the stimulation artifact.
Another pulse shaping method to reduce the artifact is to reduce the total step change in the voltage required to drive the LED. Choose a low voltage where current is effectively zero, usually 2V works well, and program that to be off state for the LED. The ON state would be as it was before but in this situation the dV/dt is reduced and so is the artifact. Similar argument works for a current driver but this requires very high resolution, e.g. choose 100 nA.
The artifact can also be subtracted out before your spike
sorting/PCA analysis. Given your timestamps for each ON and OFF event, remove
the signal recorded for a 1 ms period at the stimulation onset. This should be
repeated at the stimulation OFF event. The missing signal is then replaced with
a linear interpolation. This has been done in English,
McKenzie, et al., “Pyramidal cell-interneuron circuit architecture and dynamics
in hippocampal networks”, Neuron, 2017.
Q: I am wondering if my LEDs are still working properly. What is the best way to test them?
Micro-LEDs can be damaged due to extended exposure to high current. It is recommended that the micro-LEDs are not exposed to current higher than 100 μA for multiple stimulation cycles. If suspicious about proper operation, you can measure the current-to-voltage (I vs. V) characteristics of the potentially damaged micro-LED and compare that to the original characteristics as that can be used as a good indicator for the operation of the micro-LED. You can also measure the optical power output using an optical power meter (although measurement using optical power meters is not as accurate as the integrating sphere we use internally). This allows you to compare the radiant flux-to-voltage (E vs. V) or the radiant flux-to-current (E vs. I).
For current measurement, you can use a sourcemeter or a combination of a DC voltage source and a multimeter (with microampere resolution). For optical power measurement, you can use an optical power meter which uses calibrated silicon-based sensor and set the wavelength at 470 nm.
Q: Are the micro-LEDs ESD protected?
ESD can permanently damage the micro-LEDs. Currently, there is no ESD protection circuitry integrated to the PCBs to protect the micro-LEDs on the optoelectrode. It is important that you discharge yourself before handling the micro-LED optoelectrode PCBs, especially when handling them in dry environment. It is also recommended that you use ESD protection equipment (e.g. ESD-safe mats and wrist straps) and ESD-safe (dissipative) tools for handling if available.
Q: Can I drive the LEDs with voltage driver instead of a current driver?
Current drivers are generally a safer way to use your µLEDs but if you choose to use a voltage driver follow the I-V curve carefully and consider placing a high precision resistor in series and monitor current as well. Any oscilloscope or voltmeter would work well across your resistor in this case and ensure you know the current. This is also a simple way to make your own I-V curve if you want to evaluate the µLEDs on your own.
Q: Can you recommend commercial current drivers for the uLEDs?
We do not want to endorse particular commercial products but we will gladly share our experience with products as we learn more. Plexon, for example, has begun testing our µLEDs with their Plexon PlexStim system. We will share that information in Google Group forum and encourage EVERYONE to share their own experience so the community can learn.
Q: How do I interpret the I-V curves for the micro-LEDs?
A turn-on voltage between 2.8V and 7V is considered usable, although near 3V is typical. If the I-V curve is flat, i.e. there is no current at any voltage, then the LED is open. Please do not use. If the turn-on is higher than 7V, it is also considered damaged. If the I-V curve is linear (I=a*V) then there is a short and that too is faulty.
Q: What is a normal working range for impedance values at 1 kHz frequency?
A normal working range is 100 kΩ to 1.5 MΩ. Outside of that, you are not likely to see spiking activity.
New Questions?
Please post to this forum! If you need to communicate directly with someone in the NSF Funded MINT group, please email Conta...@umich.edu.
--UM Yoon Group
Nathan Slager
Summer 2021 microLED Optoelectrode Virtual Training Workshop Q&A
Q: For the microdrives, what is the distance per turn of the screw?
A: 270 um per turn
Q: Do you use any specific micromanipulators for holding the probe or just helping hands?
A: Helping hands typically suffice.
Q: Would the base and cap be the same for a rat?
A: Different base caps should be used for mice versus rats.
Q: How many uLED can be held in a cap? In other words, how many uLED can be inserted into the rat simultaneously?
A: Typically two can be implanted without adjusting the current design. Of course this depends on your target regions as well. We would generally recommend using the traditional base with copper mesh for your first surgery to ensure you can fit both drives in an optimized spacing prior to using the newer cap.
Q: What do you recommend to clean the skull well?
A: Start with a skull scraper then finish with a wipe using 2% hydrogen peroxide.
Q: Is the reference and ground bound together on the same screw? And do you connect electrode's ground to animal's ground, Faraday cage or both?
A: The Intan system shorts out the ground and ref, so we use just one screw, which is also connected to the Faraday cage.
Q: Is the Faraday cage necessary to minimize noise or can it be eliminated in some cases?
A: They may not be necessary with digital preamplifiers, such as the Intan headstage, but we have not fully tested it out without the caging yet.
Q: For the microdrive screws, are they all made of metals? Do we have any sense on how radio-opaque they are?
A: We haven't observed any issues with electrical noise from these screws, but they are not radio-opaque.
Q: I combine high speed X-ray with e-phys, so wondering if the screws can be made of non-metal or other materials so that X ray images of screws are not projected on the receivers.
A: I am not currently aware of any non-metal replacements for these screws. However they are standard #00-90 screws, so they can be replaced with non-metal screws, if you have a source for these. There are also two metal bars in the drive that are used to stabilize the drive motion, so you would have to somehow replace those as well.
Q: How do you clean the base of the microdrive?
A: It's usually clean because it is shielded with the detachable base which is left on the skull post-surgery.
Q: Do you have any tips for habituating the animal to attaching the cables?
A: Use a pulley system to balance the weight, and give froot loops and snacks to the animal while connecting and disconnecting.
Q: Would there be any possibility to use this system in aquatic environments? or is it too exposed?
A: At this point, probably too exposed. But it may be possible with the rat headcap system if it is testing for water-sealing.
Q: Thank you for the great presentation! Can you say again what/if plans are in the works for a hands on workshop?
A: Plans are still the development stage. We look forward to feedback from the virtual workshop.
Q: Is there any commutator solution for integrating both cables?
A: There is a commutator solution from Plexon. https://plexon.com/plexon-systems/neurolight-stimulator/
https://plexon.com/products/carousel-commutator-neurolight/
Q: How long can the headstages last after surgery before falling-off and have good signal at the same time?
A: Up to 6 months!
Q: Any suggestion on what to use to seal the craniotomy? Silicone vs bone wax
A: We have had the best experiences with bone wax. But the Dowsil works well too, including in combination with bone wax. Details of the bone wax are provided in protocol (roughly equal mix of paraffin and mineral oil)
Q: Do you get a lot of movement artifacts?
A: Not with good grounding
Q: What is the difference between acute and chronic probes, apart from the flexible cable?
A: Only the flexible cable. The acute probe could still be used in chronic experiments however.