OD LED Power Level vs resistor packs

Swapping out the resistor packs on the ADC board used to measure OD allows you to tune the range of measurable ODs and the amount of noise in the data. The choice is evaluated by doing an OD calibration, to map eVOLVER scattering measurements to real OD600 values.

The power level & resistor choices are determined empirically, and it would be great for someone to do an in-depth study into how resistor choice and power level affect the mapping between actual culture density and eVOLVER scattering measurement.

For the Wong et al 2018 smart sleeves, we described two options:

Low density, Narrow range, Low Noise:
Optimal for experiments in the 0.05-0.4 range
1M ohm resistor pack
Power level setting: 3700

Moderate density, Moderate range, Moderate Noise:
Optimal for experiments in the 0.1-0.8 range
1M ohm resistor pack
Power level setting: 2125

For the Fynch Bio smart sleeves, different power levels are used, as the geometry is a bit different.

What resistor pack / power levels have worked best for your experiments?

1 Like

We are still testing the new systems out, but on the Fynch Sleeves, there are two modes we can set up the system. To check which mode you are on, please check if the photodiode is in slot Spare A or in the normal photodiode slot.

Mode 1: HIGH OD - Photodiode in Normal PD slot (90 Degree offset from IR LED)
Resistor Pack: 1MOhm
PWR Level: 4095

Preliminary tests showed that if you did more of a 90 degree angle, you can hit significantly higher OD readings (~3.0 OD600) but is less accurate in very low (0 to .2) OD ranges.

image

Mode 2: LOWER OD - Photodiode in Spare A B (135 180 Degree offset from IR LED)
Resistor Pack: 1MOhm
PWR Level: 2500

This shows a better linear response in the lower OD ranges, leading to better differentiation between the lower and higher OD.

image

NOTE: Data from Zack Heins

@bgwong any thoughts on why the voltage decreases in Mode 1, but increases in Mode 2? I haven’t gone and looked at the schematics yet, but I imagine there must be something different or missing in the Spare A circuit? Is there a chance I flipped something in installation that would cause this?

Im not quite sure yet. They should both decrease. With LONG end going to + for the Photodiode. Even if the code is slightly different, I’m not quite sure why it would display the opposite format. When you get a chance, just double check the leads.

I definitely did LONG -> +

The PDs would be in a different bias mode and not sense if they were flipped, you’d just be seeing the voltage drop over the diode always and not detect anything.

OK following up on this. Zack and I just chatted and realized that he put it in a different slot that what we expected, making it 180 degrees offset from the LED, instead of 135 degrees. The values would increase (meaning the signal is decreasing, i know its confusing) because it’s measuring absorbance instead of scattering. This actually results in a really good calibration curve, might move to do this instead of 135 degree offset

2 Likes

Sorry for the confusion - I misunderstood you when I was making that mod. It’s really cool that it ended up working out so well. I’m glad I did that broad range power sweep/calibration now with that setup.

How difficult is it to switch between slot spare A and spare B in between experiments?

If the hardware is installed in both, and you’ve got calibrations for both arrangements, then it should be relatively simple to make code to switch back and forth between experiments (though this isn’t formalized yet).

If instead, you’re hoping to swap out hardware (LEDs or photodiodes) and/or move them around between experiments, then you’d want to recalibrate after each swap, which will get very annoying.

I’m choosing a resistor pack now and I don’t see any part numbers in the documentation. Does anyone know what these should be? It looks like they should be the following based on the # of pins:
RES ARRAY 9 RES 1M OHM 10SIP part # 770101105P or 4610X-101-104LF
RES ARRAY 9 RES 100K OHM 10SIP part # 770101104P or 4610X-101-105LF

Thanks for the question. Sorry if anything is unclear. The original post is a bit outdated now.

EDIT: I got it mixed up (and 104 and 105 was mixed up above):

4610X-101-105LF -
(1M ohm) Should be used for the OD90

4610X-101-104LF -
(100 kohm) Should be used for OD135

Either way, the resistor packs are really cheap and it is helpful to have several resistances lying around.

3 Likes

Hey Ben,

Sorry I mixed it up a bit. Edited the comment above to correct it.

Brandon

Got it! That was my original mistake. Thanks for the clarification.
The values are in the supplemental data, I just couldn’t find documented part numbers.

While we are on the topic. It’s part # 4610X-101-103LF (10kohm) for the thermister ADC?

Yea the thermistor readings get maximized when the resistor is the same value as the thermistor itself (10k ohm)

Brandon

I’ve tried looking through the code, and I may have missed this, but where do we specify the power level settings for the different detectors?

You can change these values in the server configuration file conf.yml, in the following line:

For this,

  1. Log in to your eVOLVER machine using ssh:
    ssh pi@<evolver_ip>
  2. Backup your configuration file, just in case:
    cp evolver/evolver/conf.yml evolver/evolver/conf.yml.bak
  3. Open and edit the configuration:
    nano evolver/evolver/conf.yml

I should have a script to make this safe and more automatic, so I’ll share it when I have some time to document it properly.

Hope it helps!

4 Likes

Found it! Thanks so much for the help!

1 Like

@danigarcia that’s a good idea. Currently the only way to modify these are through the DPU or manually as you described. We can maybe use the server test script as a base for this.

The other option is a way to modify the config through the GUI - that’s slightly more work though and there are a lot of un-released features that need to be pushed out first.

I uploaded the script we use to change the LED power here:

And you can find the instructions here.

Hope it is useful!

3 Likes