Finally 7.5 Digit - My "new" Keithley DMM7510

[KungFuJosh]

Why pay $2000?? You can buy them on the used market for less than that. If there's a problem with the meter you receive, you can return it.

[BillyO]

Ooohhhh....
Unfortunately, I need every penny for my move at the moment, otherwise I would definitely give in again.

Actually, you should buy two, then wait until the “glut” from China subsides.
Then offer one of them for sale, and you're sure to recoup the price for both.

Just a note Martin.  You were 100% correct.  I dragged my feet on this until they were all gone and no such deal has appeared again.

Sigh.  I still seek....

[HalFoster]

Why pay $2000?? You can buy them on the used market for less than that. If there's a problem with the meter you receive, you can return it.

Well, it did include a cal with data which was due - that would be another $800 spent on anything else that I bought.  The repair itself consisted on replacing both the analog and digital boards with new ones so I wound up with what was, in effect, a brand new meter. 

[KungFuJosh]

Why pay $2000?? You can buy them on the used market for less than that. If there's a problem with the meter you receive, you can return it.

Well, it did include a cal with data which was due - that would be another $800 spent on anything else that I bought.  The repair itself consisted on replacing both the analog and digital boards with new ones so I wound up with what was, in effect, a brand new meter.

You need to find some better options. I have two local cal labs (both known/respected), and both quoted around $200 to cal a DMM7510.

Anyway, give me your cursed DMM7510.

[HalFoster]

Why pay $2000?? You can buy them on the used market for less than that. If there's a problem with the meter you receive, you can return it.

Well, it did include a cal with data which was due - that would be another $800 spent on anything else that I bought.  The repair itself consisted on replacing both the analog and digital boards with new ones so I wound up with what was, in effect, a brand new meter.

You need to find some better options. I have two local cal labs (both known/respected), and both quoted around $200 to cal a DMM7510.

Anyway, give me your cursed DMM7510.

A full calibration on the DMM7510 checks 179 separate items but a cheap cal will typically only cover a small subset of that - I know, I shopped around on previous calibrations and requested samples of the report that would be returned... heck, with a full calibration there are 40 items checked just on DC voltage.  That being said, a limited - and much cheaper - calibration is often all that most users actually need, myself included, really, but I just like having a complete OEM calibration on that meter.  To clarify, the DMM7510 is a very nicely designed and functional meter that, in the majority of cases, works very reliably. My issues are with the Keithley/Tektronix service department who can couldn't find their a$$ with with both hands and a flashlight much less an obvious and repeatable equipment fault.  Hell, this last go round it took me a month to even convince them that the unit *had* an issue even after the unit wouldn't boot and had I send screenshots of the error codes it produced the few times that it did.

[KungFuJosh]

A full calibration on the DMM7510 checks 179 separate items but a cheap cal will typically only cover a small subset of that - I know, I shopped around on previous calibrations and requested samples of the report that would be returned... heck, with a full calibration there are 40 items checked just on DC voltage.  That being said, a limited - and much cheaper - calibration is often all that most users actually need, myself included, really, but I just like having a complete OEM calibration on that meter.  To clarify, the DMM7510 is a very nicely designed and functional meter that, in the majority of cases, works very reliably. My issues are with the Keithley/Tektronix service department who can couldn't find their a$$ with with both hands and a flashlight much less an obvious and repeatable equipment fault.  Hell, this last go round it took me a month to even convince them that the unit *had* an issue even after the unit wouldn't boot and had I send screenshots of the error codes it produced the few times that it did.

That's a good question. I should ask what their protocols are.

So is your meter working fine now??

[HalFoster]

So far, so good - I've used it a *lot* since it's been back.  As I said, they are dependable meters... until they aren't

Please let me know if you find anything out on the cal labs - while not for the Keithley, I would love to find somewhere more affordable to send some other equipment that I have.  And again, limited cals are a great way to save money if you can find a lab that offers them - at work we have a couple of Fluke 5720A calibrators that we only use for certain things so I send them out for limited cals and it costs about half of what a full one would.

[KungFuJosh]

These are the measurement points from the two meters I asked them about (shared with permission). What do y'all think?

Thanks,
Josh

[HalFoster]

That's nice and, as far as I am concerned, quite complete.  Would you be willing to share the company?  I have several things that I wouldn't mind having quoted and I also have a few things at work that are due next month including three 3458A meters so that might also be something to discuss with them.

For reference, attached is the calibration done after the repair by Tektronix.

Also, I have used CustomCal quite a bit and while not exactly inexpensive they are cheaper than an OEM and are extremely nice people to work with.

[KungFuJosh]

They're a local Transcat affiliated lab. I'm waiting for more information from a different lab also. I might share contact info, but you should be able to find somebody local to you that's similar.

[rusoaie]

Datasheet updated in March 2026.
Attached revised and previous versions 

[johnbaker26]

So I decided to upgrade to one of these and put the word out with my sources in China. I go there every month for work. I was quickly connected up to a guy who had a back room full of these. Also had 2450s but the price on those were still 5500 dollars. I offered 11000 rmb (1600 USD) on the nicest one he had and he immediately accepted. I regret not offering less but the unit is very nice not a scratch on it. The plastic film had never been removed from the faceplate. It's been about a month now using it and I couldn't be happier. A noisy fan is my only complaint.
If anyone wants to go this route his name is Tan Huiyan. Nice guy but no English. WeChat phone number +8613145876435. He only accepts WeChat pay (you can link Xoom to Paypal) and not sure if he ships. Probably will.

[exe]

It's been a while since I turned mine on. So I decided to check if it still works. And it does)

Comparing to https://www.eevblog.com/forum/testgear/finally-7-5-digit-my-new-keithley-dmm7510/msg6083319/#msg6083319 , the value seems to stable (taking into account much less averaging and different leads). Unfortunately, I lost my records of voltage and humidity, so cannot check if the system (reference+DMM) drifted anywhere. I guess I should start logging).

[exe]

So, I made a script (okay, AI did that for me  ). It's a plot of deviation in PPM from average value of my reference (but not a ppm of the range, so it's around 0.7uV). The reference is ADR1399, and I have no idea how stable it is. I read value from dmm every 10s (settings: impedance auto, range 10V, nplc 5, line sync).

There is not much data yet, but I was rushing to share first data with you guys. How measurement works: every 10 seconds I take a sample from DMM, I also take internal temperature, and a measurement from external thermometer. The gap on the plot at the end is because external thermometer wasn't available and script crashed 

Let me know if you have any suggestions what/how should i measure. I already added humidity and atmospheric pressure to plots.

[BillyO]

There doesn't seem to be a lot of correlation.  Seems the voltage reference oven is doing it's job very well.

[KungFuJosh]

So, I made a script (okay, AI did that for me  ). It's a plot of deviation in PPM from average value of my reference (but not a ppm of the range, so it's around 0.7uV). The reference is ADR1399, and I have no idea how stable it is. I read value from dmm every 10s (settings: impedance auto, range 10V, nplc 5, line sync).

There is not much data yet, but I was rushing to share first data with you guys. How measurement works: every 10 seconds I take a sample from DMM, I also take internal temperature, and a measurement from external thermometer. The gap on the plot at the end is because external thermometer wasn't available and script crashed 

Let me know if you have any suggestions what/how should i measure. I already added humidity and atmospheric pressure to plots.

Why don't you use TestController to do all of that?

[exe]

Why don't you use TestController to do all of that?

I didn't know about it. Though it seems TestController requires at least java and UI, and I have a headless setup, without desktop environment. If it can read temperature from home assistant, then it's a viable option.

[Kleinstein]

So, I made a script (okay, AI did that for me  ). It's a plot of deviation in PPM from average value of my reference (but not a ppm of the range, so it's around 0.7uV). The reference is ADR1399, and I have no idea how stable it is. I read value from dmm every 10s (settings: impedance auto, range 10V, nplc 5, line sync).

If noise is relevant like in this measurement one would want more data / longer conversions. So more like read nearly as fast as possible, or use long integration like 100 PLC or even more.

A good part of the noise may still be from the ADR1399. The short 5 PLC intergration also gives a relatively large noise BW (5 Hz for 100 ms integration) for the source.

[exe]

I monitored my reference for about a week, here is what I have. Too bad I don't know if it's a reference ageing, or DMM drifting 

[Kleinstein]

AFAIK the DMM7510, afer some burn in should not show much drift in the 10 V range. The internal reference is supposedly rather stable. The test during warm up show gain corrections from time to time and this suggests and internal kind of ACAL step for the 10 V range. So for just the 10 V range there may be no need for an explicite ACAL. ACAL would effect the other ranges, but may have essentially no effect on the 10 V range.

A ADR1399 reference is expected to have some drift in the first few weeks. So I would expect the drift to be largely from the external reference. Also the temperature effect could be from there.

[KungFuJosh]

The voltage drift follows the temperature drift. You have to eliminate the environmental inconsistencies to profile the voltage drift better.

[JimKnopf]

Last year I tried to build this project https://github.com/envido32/rfid-zapper. It's a RFID-Zapper. There was something wrong with the design. It didn't work. NE555 didn't swing. I could not find the fault.

Yesterday i let my friend Claude-Code inspect the project files and find the fault. After fixing, the device works as expected. During my tests i observed the Big-Cap charging with the DMM7510. A very strange fault occured when switching back from Full-Graph-View back to the previous Half-Graph/Half-Digit-View using the Exit Button.

A diagonal line from the lower left corner to the actual curve point appeared on screen in Half-Graph mode. Switching to Full-Graph, the line disappieard. Hitting again Exit Button, the line was present again at the new curve point in the Half-Graph-View. After some minutes, the curve was flat at the top, the diagonal line disappeard and did not come back after switching to Full-Graph-View and back to Half-Graph-View.

[KungFuJosh]

Did you try clearing the buffer?

[exe]

A diagonal line from the lower left corner to the actual curve point appeared on screen in Half-Graph mode.

I've seen artifacts like that too. Clearly it's a firmware bug.

[JimKnopf]

Today i cannot reproduce the strange line. Graph settings are the same, Scale=Show all. But this time, no diagonal line.