Even though it is long-in-the-tooth, I like my Keithley 2000 and a scan card is a useful add-on.
CircuitStudio project attached, should import to newer versions of Altium OK.
PDFs/Gerbers also included.
Here is a crude Python script to log scan card channels to CSV using the K2000 serial interface.
Tested with above card.
I recommend pyvisapy backend, works, small size.
That looks great.
Also, I wouldn't describe a Keithley 2000 as long in the tooth. Now, my HP 3455A is long in the tooth.
Hello,
This card should also work with DMM6500 ?
Thank you !
Ok. Thank you very much !
Maybe I will do also the 20 channel card in the future, but for my needs I have choose to build this model.
In fact I have already made it m but I have realized that the model 2001-TCSCAN is a little different and need to be sure .
Thank you very much for your work !
P.S.There are 4 PCB's left for everyone need it.
Do you have a 3D model for the clamp ? To print it ?
I see you populated R145 and R146 - these are optional, check original manual.
I also have 4 leftover board if somebody wants them :-)
Yes you are right, I did not pay attention to them.
I picked them from BOM.
Thank you for sharing!
I took a differnet approach and "saved" the terminal blocks. Since the original 20-channel card could do with a 44-pin DSUB, I found I could use a good quality 25-pin. I have very limited space behind the instrument, so I moved it inside of the case. A pluggable connection is better for me, and it seems to be OK for my use.
I built one too! Thanks for the great plans, this was a fun and rewarding project.
I only ran into one issue, the MIC5841 was out of stock everywhere EXCEPT in a weird 20PLCC package, so I made a small adapter board to compensate.
If anyone is interested I can clean up and post the kicad and gerber files!
I also have 4 leftover boards for the main PCB if anyone wants them in the US/Canada. I only have 2 extra PLCC ones so not enough to finish the project.
I built one too! Thanks for the great plans, this was a fun and rewarding project.
Glad you had fun building it
I only ran into one issue, the MIC5841 was out of stock everywhere EXCEPT in a weird 20PLCC package, so I made a small adapter board to compensate.
Nice solution. This component shortage is a PITA in so many ways.
I built one of these. Works very well and fund project!
Thanks to voltsandjolts for sharing!!
I have now been using this card for some time. Generally it works very well.
Once issue though: when measuring some very precis voltages that are not too well buffered, when the card scan over these voltages it loads down with around 50 to 300 micro volt. I have another 6.5 digit multimeter where i can see the voltage jump up and down with the scanning. When using the front input of the DMM6500 and tap the measurement point with test leads this do not happen. So its not the DMM6500 that load down, but switching from the card.
Is this a general thing that can happen with scan cards, or could there be some problem with mine?
I also build voltsandjolts's SSR 20 chanel card where i dont have this problem. Sometimes the last digit on the 6.5 meter i use to check voltages can move a bit more than if no scanning, but thats all ok, and just some noise.
Some load transients when switching between channels is a general problem. The inputs have a certain capacitance and the protection part may have a limit in the slew rate for some internal signals (e.g guards, bootstrapping for protection). The scan card should not add much to this (a little capacitance and maybe a little capacitive coupling to a control signal). The VOR1142 SSR has some 39 pf and this adds up with 20 channels. So a version with fewer channels or additional seprating groups (e.g. like the mode for 4 wire ohms) can help.
Scanning with high impedance signals will always be tricky.
Just taping with the leads by hand may be misleading as the DMM inputs tends to stay charged to the previous signal and this may be the same signal as before when doing it by hand.
Perhaps the fact that you see the problem on the armature relay card but not on the SSR card is a clue.
Could it be the impulse applied to the relay coils has this slight affect on the high-impedance signal, passing through the contacts close to the coil?
Do you see the effect permanently (e.g. when scanning just one channel) or only temporarily when changing channels?
Hello, iam interested in this project. Does anyone have spare PCB for sale ? I dont want to order 5 or more pieces from PCB manufacturer. Thanks !
Even if somebody has boards, you still have to pay shipping.
New boards are so cheap you won't be saving much.
Here is the gerber zip I uploaded to JLCPCB, check out the quote for 5 boards delivered:
Indeed. It's cheaper to order.
I just checking, and sending from Romania to Czech Republic is almost 20 euro.
If you find a cheaper way tell me and I will send a board. I have 3 available.
Thank you for reply both guys. I´m aware of shipping price, shipping from JLCPCB is 24 eur to CZE, given the fact i have to pay customs its even more. And ordering 5psc if i need only one its kind of wasting of material for me
Another problem i realized is availability of Micrel MIC5841, mouser states availability in 10/2023
For the Micrel chip there could be alternatives, like TPIC6C595 and similar ones from ST or others. Depending on the relays even just 74AC595 or HV595 + AC541 may work too.
This may be a reason to make your own, slightly modified design instead of getting a PCB for an out of stock chip.
For some countries (there can be restrictions) it may be possible to use a letter for shipping a signel PCB at a lower price.
Yeh, those AliExpress MIC5841BN would probably do the job, assuming there is some silicon inside the DIP package
Hi, I made an alternative layout that supports the SOIC-18 versions of MIC5841YWM, I have not made a board yet, so there might be errors, I don have much layout experience...
I'm not sure when I will test a prototype as I want to check first if I can replicate the 2001-TCSCAN and manage to use thermocouples, the AD590MH Temperature Sensor is quite expensive so I want to look for alternatives. has anyone thought about doing that?
anyway for now I leave my export here if anyone wants to try
Cool, and in KiCAD, thanks for posting.
The 2001-TCSCAN is a bit of a nuisance because you (ideally) need the isothermal block for the cold junction, but surely it's doable.
I just use 4-wire RTDs with a standard scan card.
It seems difficult to match the performance of the original 2001-TCSCAN without spending almost as much as a used card
but I think I will try to make a layout with an AD590KRZ (lower specs SOIC-8 version of the sensors) that goes for around 20€ and place it under the binding posts on the bottom of the board on a nice ground plane full of stitching vias, I would think that would be enough to use at least the two neighboring posts.