20-channel DIY Scanner card for Keithley DMMs and DAQs

[cozdas]

I implemented this some time ago and mentioned that in the DMM6500 topic

Few people showed interest in this project thus I wanted to opensource it. You can find the details along with project files and sources in this github project page.

Feel free to ask questions and post comments.

Happy holidays.

[Alexcnc]

Thank you for bringing the project open source!

[thm_w]

Thanks for the notification, this is really good documentation  . Hope to build it.

Some photoMOS, first is one OP used:

Code: [Select]

    PN     Voltage  Resistance   Leakage
TLP3558A -  160V   2R    1uA    $1.5-2.50
AQY212EH -  60V    2.5R  1uA    $0.25+
AQY210EH -  350V   25R  1uA     $0.25+
AQV253A  -  250V   8R   1uA     $0.8
LH1540   -  350V   22R  1uA     $0.8
PVA1354N -  100V   5R   1nA?    $2+  (limited life)
PVA3324   - 300V   24R  1nA?    $2

- Equivalents for AQY: TLP227G, TLP222G, G3VM-351D, LCA110L (6pin, lower R typ) $0.5-3

Reed relay
- D1A050000 - 200V 0.1R 1nA - $0.60

Not sure how much resistance matters? Shame if 25R is too high as there are a ton of surplus ones, probably from telecom gear.
Voltage will depend on what circuits you plan to measure. Original spec was ~200V but I'm sure the relays were rated for slightly higher.

[cozdas]

Thanks for the notification, this is really good documentation  . Hope to build it.

Some photoMOS, first is one OP used:
- TLP3558A - 160 V 2 R - $1.5-2.50
- AQY212EH - 60V 2.5R - $0.25+
- AQY210EH - 350 V 25 R - $0.25+
- Equivalents: TLP227G, TLP222G, G3VM-351D, LCA110L (6pin, lower R typ) $0.5-3

Not sure how much resistance matters? Shame if 25R is too high as there are a ton of surplus ones, probably from telecom gear.
Voltage will depend on what circuits you plan to measure. Original spec was ~200V but I'm sure the relays were rated for slightly higher.

If you implement it let me know the outcome.

My thinking regarding the on-resistance was to keep the loading error less than 1 digit (1ppm). With 10MOhm input impedence, TLP3558A results in ~0.2ppm which is 3.2 LSB of the 24bit ADC.

[Gregg]

Thank you codzas for your hard work.  Now I have another reason to want a DMM6500.      

[Kleinstein]

I would expect the 25 Ohms or even 100 Ohms to be still OK. It may effect the 10 MOhms input impedance ranges a little, but this would be higher than 10 V and usually the lower accuracy ranges anyway.

I would be more worried with leakage between the channels. So it may be a good idea to not use the very low resistance ones. At least unused channels should not contribute much to leakage.  So the scanner may not be a good choice for high impedance source.

[thm_w]

I would expect the 25 Ohms or even 100 Ohms to be still OK. It may effect the 10 MOhms input impedance ranges a little, but this would be higher than 10 V and usually the lower accuracy ranges anyway.

I would be more worried with leakage between the channels. So it may be a good idea to not use the very low resistance ones. At least unused channels should not contribute much to leakage.  So the scanner may not be a good choice for high impedance source.

I added leakage current spec to my post. Seems the mos ones above at least are similar, for their max spec.

LCSC has some reed relays for ~$0.60, with better leakage (nA). But of course life is limited (10^8 cycles, about 3 years switching every second). There are also normal relays, and spec for the cheapies is really bad (~5uA, 10^5 life only!). Maybe some value in having a few reed and few photomos or similar.

One thing I'm not sure of, what is the purpose of channel 10 on the original scan card? It uses K111 relay and U105 3301 photomos in series, with some 100R resistors. The other channels use a single relay.

[cozdas]

One thing I'm not sure of, what is the purpose of channel 10 on the original scan card? It uses K111 relay and U105 3301 photomos in series, with some 100R resistors. The other channels use a single relay.

SCAN-2000 card that I bought from eBay (probably an older revision) doesn't have solid-state switches, just mechanical ones. Although the SCAN-2000 manual doesn't explain the why there are Solid-State Relays (apparently in the newer revisions), SCAN-2001 manual does explain it. I think with a special mode enabled in the DMM, the mechanical relays of both channel 5 and 10 are closed and the fast switching between ch5 & ch10 is achieved by the photo-mosfets.

[thm_w]

I went to test the card and it looks like I've got the pinout back to front 
Time for Rev 2

[voltsandjolts]

I have a K2000 DMM (2006 vintage, latest firmware) which probably won't support a 20 channel scan card.
I say that because there is a specific version (K2000-20) which does support a 20 channel card.
Does anyone know what the differences are between K2000 and K2000-20?
Is it just firmware?

Cross-reference to 10-channel card which works with K2000 DMM:
https://www.eevblog.com/forum/circuit-studio/example-project-relay-scan-card-for-k2000-dmm/

[voltsandjolts]

Another option for 20 channel SSR scan card here:
https://www.eevblog.com/forum/circuit-studio/example-project-20-channel-solid-state-scan-card-for-k2000-dmm/

[Anders Petersson]

I built one (slightly modified). It works with Keithley DMM6500 but is not recognized by Keithley 2010.
My original 2000-SCAN card is recognized by Keithley 2010, so it's likely the 2010 doesn't support 20-ch cards, only 10-ch. (Neither the manual of DMM6500 nor 2010 mentions support for 20-ch cards.)

EDIT:
Looking at voltsandjolts version (https://www.eevblog.com/forum/circuit-studio/example-project-20-channel-solid-state-scan-card-for-k2000-dmm/), I see a suspicious section in the source code:

    //The SPI clock line from K2000 idles high, whilst from DMM6500 it idles low.
    //Determine idle state of SPI clock by waiting for 50 samples of the same value

Possibly the Arduino code just needs to do something similar to support K2010.

[Neuromodulator]

It would be interesting to fabricate the piece that keeps the PCB in place.