Author Topic: Restoration glory of Keithley 2001 DMM  (Read 203987 times)

0 Members and 1 Guest are viewing this topic.

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4052
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: Restoration glory of Keithley 2001 DMM
« Reply #250 on: April 07, 2015, 07:12:28 pm »
Published draft article for repair Keithley 2001. Currently listed only items for unit #2 , will update with rest of info later this week.
Would like to consolidate all information we have about these instruments in one spot for easy newbie access!
Hires photos of newer units (manufactured later than 1994) are very welcome!
YouTube | Chat room | Live-cam | Have documentation to share? Upload here! No size limit, firmware dumps, photos.
 

Offline ManateeMafia

  • Frequent Contributor
  • **
  • Posts: 719
  • Country: us
Re: Restoration glory of Keithley 2001 DMM
« Reply #251 on: April 07, 2015, 07:41:52 pm »
TiN,

I believe that I sent you some pics of my 2001 back in February. It was manufactured in 2001. Let me know if you need any additional pictures. I have it sitting in the corner waiting for the new front panel.
It looks like Keithley might be making them in batches because it only took them a week to send me the new 2400 display board and keypad. I have been waiting almost a month now...
 

Offline analogNewbie

  • Contributor
  • Posts: 46
  • Country: cn
Re: Restoration glory of Keithley 2001 DMM
« Reply #252 on: April 07, 2015, 11:43:12 pm »
My 2001 seems much noiser than 34401 at 6.5 digits mode. Is it measuring the real noise or it's a noisy DMM?
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4052
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: Restoration glory of Keithley 2001 DMM
« Reply #253 on: April 08, 2015, 12:48:05 am »
Did you set max NPLC, Line-sync, autozero to max? Measure short, or some signal?
Any data to reference? I can help to capture with same settings and post for you later on.
YouTube | Chat room | Live-cam | Have documentation to share? Upload here! No size limit, firmware dumps, photos.
 

Online macboy

  • Super Contributor
  • ***
  • Posts: 1970
  • Country: ca
Re: Restoration glory of Keithley 2001 DMM
« Reply #254 on: April 08, 2015, 02:10:36 am »
TiN, my four units' analog boards have the following revisions: Rev H, K, N, and P.
 

Offline Galaxyrise

  • Frequent Contributor
  • **
  • Posts: 526
  • Country: us
Re: Restoration glory of Keithley 2001 DMM
« Reply #255 on: April 08, 2015, 02:17:16 am »
My 2001 seems much noiser than 34401 at 6.5 digits mode. Is it measuring the real noise or it's a noisy DMM?
Interesting to see this on someone else's.  I believe mine is a noisy DMM as it's almost 10x noisier than my K2K both with a 4-wire short and with a very stable 7V reference.  I don't have a gpib adapter though, so I'm not able to log it.
I am but an egg
 

Offline ManateeMafia

  • Frequent Contributor
  • **
  • Posts: 719
  • Country: us
Re: Restoration glory of Keithley 2001 DMM
« Reply #256 on: April 08, 2015, 03:20:26 am »
@Galaxyrise

Does it pass all self-tests?

I have yet to compare both my functioning 2K1's against my 8846A's, but maybe we can all agree on very specific settings like TiN suggested (NPLC, etc..) to try and compare results. I think all the ADC boards are probably running the same FW.

I am hoping Tektronix ships my last couple of parts so I can reassemble one of them. The other is ready for use once I do some performance tests.
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4052
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: Restoration glory of Keithley 2001 DMM
« Reply #257 on: April 08, 2015, 04:22:22 am »
Ok, here's idea. What if we have a page where performance data can be stored and updated?
I think I'm good enough with D3.js now to have charts for various meters displayed on page.
Participants can upload their CSV-files to FTP, and I'll add those data into page as well.
Would be nice to see what meter is quietest. Can add other measurements as well, such as zero ohm short test, own-reference measurement test, etc.
YouTube | Chat room | Live-cam | Have documentation to share? Upload here! No size limit, firmware dumps, photos.
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4052
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: Restoration glory of Keithley 2001 DMM
« Reply #258 on: April 08, 2015, 04:51:33 am »
TiN, my four units' analog boards have the following revisions: Rev H, K, N, and P.
Can you take some good photos of these puppies? I will add revision data into article. I saw later revisions have better input divider, using Caddock ceramic hybrid resistors instead of blue single substrate ones.
« Last Edit: April 08, 2015, 04:57:07 am by TiN »
YouTube | Chat room | Live-cam | Have documentation to share? Upload here! No size limit, firmware dumps, photos.
 

Online zucca

  • Supporter
  • ****
  • Posts: 2078
  • Country: it
  • EE meid in Itali
Re: Restoration glory of Keithley 2001 DMM
« Reply #259 on: April 08, 2015, 08:55:07 am »
Those are the days I think we can conquer the world guys! Still can't believe the FULL schematics are now in our hands  :-+.

Nice job TiN and macboy.

From my side I am still hunting the 1801 preamp, but I think it is official and extinct animal. Surely somewhere it is just sleeping in some dusty big company lab corners (:palm:).
Worth to try with the Keithley forum brothers (as it happen here), and why not ask there for the B16 FW? I can do it if nobody already tried that.

My 2001 unit is still sailing the ocean finding his way from Canada to Germany....

Quote
'Grandpa, were you a hero in the war?' Grandpa said 'No... but I served in a company of heroes.'
« Last Edit: April 08, 2015, 09:00:05 am by zucca »
Can't know what you don't love. St. Augustine
Can't love what you don't know. Zucca
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4052
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: Restoration glory of Keithley 2001 DMM
« Reply #260 on: April 10, 2015, 08:03:40 pm »
My 2001 seems much noiser than 34401 at 6.5 digits mode. Is it measuring the real noise or it's a noisy DMM?
Interesting to see this on someone else's.  I believe mine is a noisy DMM as it's almost 10x noisier than my K2K both with a 4-wire short and with a very stable 7V reference.  I don't have a gpib adapter though, so I'm not able to log it.

Cross-post to new thread, I think would be nice to compare other meters too, not just 2001.
YouTube | Chat room | Live-cam | Have documentation to share? Upload here! No size limit, firmware dumps, photos.
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4052
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: Restoration glory of Keithley 2001 DMM
« Reply #261 on: April 11, 2015, 10:43:41 am »
Someone forgot to replace capacitors in their Keithley 2001? Well, it's time to do so just now.
Everything was looking fine but analog board throwing lots of 4xx.x errors.. Remove all caps, and now look at this, ugh..  :palm:



Time to fix this bad boy..
YouTube | Chat room | Live-cam | Have documentation to share? Upload here! No size limit, firmware dumps, photos.
 

Offline Galaxyrise

  • Frequent Contributor
  • **
  • Posts: 526
  • Country: us
Re: Restoration glory of Keithley 2001 DMM
« Reply #262 on: April 11, 2015, 06:25:45 pm »
Cross-post to new thread, I think would be nice to compare other meters too, not just 2001.
Ooo, nice to see what a functioning 7.5 digit DMM should look like!  I'm going to run again with the SCPI commands you listed, but here's what I'm seeing after configuring it from the front panel.  You can see it's basically 10x as noisy as the k2k I posted in your DMM noise thread, and like 50x as noisy as yours! If your steps change the output, I'll update here.

I am but an egg
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4052
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: Restoration glory of Keithley 2001 DMM
« Reply #263 on: April 11, 2015, 09:27:44 pm »
Today's update..

Took very first unit I bought in 2011, to try fix it with new knowledge.
Run self-test: gotcha 309.1,401.1,402.1,403.1,405.2,405.4,405.6,405.8,406.6,407.1,407.2,407.3,409.6,410.1,411.1,411.2.

Sometimes there was 304.7, but that got fixed by cleaning area around ohms sources.

First step:

Recap analog board and repair damages:



Fix:







What is the outcome? :)

Anybody have extra plastic covers for analog board for sale? Two of my remaining 2001's lack plastic covers (top and bottom) for analog board.
Or I can trade to other 2001 parts maybe. Have extra digital board and analog donor board, case, transformer, front panel (no VFD).
YouTube | Chat room | Live-cam | Have documentation to share? Upload here! No size limit, firmware dumps, photos.
 

Offline Smith

  • Frequent Contributor
  • **
  • Posts: 349
  • Country: 00
Re: Restoration glory of Keithley 2001 DMM
« Reply #264 on: April 12, 2015, 03:26:07 pm »
Do you mean the plastic standoffs?

I just noticed there is no metal cover as far as I can see around the ADC board? The 6517 electrometer  uses the same ADC board, but is has a metal cover covering the whole ADC board like the one in the upper left corner:
Trying is the first step towards failure
 

Offline ManateeMafia

  • Frequent Contributor
  • **
  • Posts: 719
  • Country: us
Re: Restoration glory of Keithley 2001 DMM
« Reply #265 on: April 12, 2015, 03:36:11 pm »
I think TiN means the plastic covers that sandwich most of the board. They are black and the interior is painted with a conductive coating.
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4052
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: Restoration glory of Keithley 2001 DMM
« Reply #266 on: April 12, 2015, 04:05:39 pm »
Seems like everyone having nice weekend, and only weird voltnuts are messing with fixing some 25 year old instruments :)
So I will drop the suspense and show rest of progress on very first meter I got. It is a reason of all this Keithley madness, root of all evils.  :box:



As above, this 2001 had analog board revision L, with avalance diodes on front side.
Obviously capacitors leaked after all those years and caused shorts in bootstrap regulator, allowing high current to do some PCB damage around C117.
Pad and connection to negative terminal of cap were completely burnt.

Lucky enough, no low-signal tracing on inner layers was affected. I had quick bodge with some noname fung-hwa-chi capacitors, and had this meter running for
few years in 2012-2013, as it was working for DCV,ohms,current ranges. After that it was collecting dust most of time, as I got other 2001's to play with which were in better shape.

Another issue - was open pin connection on U531 DAC, which was throwing lots of 4xx errors. Previous owner probably was not careful removing analog board, you can see DAC pins bent, like someone pushed poor IC off the PCB.
Fixed by reflowing pins with soldering iron. Easy enough.



As all Rev.L board I saw, this one have more stable Vishay R356,R358,R365 for current ranges, and ULT H5185A 1Meg 1% in little metal can (R360).
Cleaning around them fixed 304.7 issue for nA ohm current source.



Had issue with 309.1, Amps. It was wrong Q305 on board. Replacement to IRFR020 from donor 2001 board fixed this problem.



Replaced fan to ADDA



No reworks for TRMS area.



No reworks on input side either. It's interesting to note, I had 6 units (2001) and none of them had any blown parts on input. Is this a case when sensitive electronics does job to protect 1 dollar protection circuit from failing well? :)



VREF is alright too, LMx99 with SL40057 IC-527 with datecode 22 week 1993. It have a crack on top cover, since I was curious to see what's inside 5 years ago when got this meter. :)



Here's recapped board with replaced avalance diodes, CR514,CR513 zeners.
I desoldered U505, U507, U508, U509, cleaned all PCB area, cleaned ICs themselves with KONTAKT60 (not best cleaner, but works OK and not aggressive as acetone)
Then soldered all except U508 back. U508 (74HC00) was replaced to new one 74F00 (just don't have HC ones in drawer). It's jellybean logic part, why not replace it to new? :)

Notice that I used 1000uF x 50V capacitors for C117,C116. Same diameter, just taller, but still fit nicely. This should reduce wear&tear forcaps, as they are near hot LDO.
I'm not sure why Keithley had 35V rated caps while output on bootstrap circut is +38..+44 and -38 to -44 VDC?

Also this time used general purpose Chemicon KMG 470uF x 63V instead of fattier Low-ESR KZE series Chemicon's.

Here replaced parts marked blue on schematics:



After you gave some capacitor love to analog board, don't forget digital too! There are three lonely 1000uF x 16V, I bet they want to leak and sneeze after all those decades too.
There are no high-power path there, so problem less evident, but it's worth to replace those too. Especially footprint is designed to fit wider 1000uF x 35V.
On my board C630 and C624 had some electrolyte escape already.



Replaced to all three to 1000uF x 35V Chemicon KY's.

Overall photos of digital and analog boards:



How let's slap A/D board on and check this beast..
I used one of A/D boards bought few years ago from our fellow member, free_electron.



Meter power's on. -39 and +39 after bootstrap there, all voltages ok, no smoke or smell elsewhere. Time to run selftest.



And pass, no errors!



Run in continuous loops few dozen times, no problems.



Time to check voltages?



Sourced 1.00028VDC from 2400, measured 0.99029 VDC. No good.



Sourced 10V from EDC MV106, measured 9.89829 VDC. No good, definately need calibration, but should be no issues with that.

Bottom 2001 - one fixed last weekend, calibrated using VPG 19K and 1Meg resistors and 2V and 20V from EDC MV106 measured by reference 2001 (it's meter on top right, calibrated by Tektronix in February 2014)

Like always, enlarged photos on clicks ;)

Bonus



A/D board with probe port to Mictor (which in turn runs into Mictor-P6860 adapter and my TLA714. :) What's going on here?
Hope you enjoy reading..
YouTube | Chat room | Live-cam | Have documentation to share? Upload here! No size limit, firmware dumps, photos.
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4052
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: Restoration glory of Keithley 2001 DMM
« Reply #267 on: April 12, 2015, 04:10:21 pm »
Hi Smith,

No, i mean black cover which covers analog board from top and bottom.



It's unique to Keithley 2001 and 2002 units only.

I would LOVE to see more high-resolution photos of your 6517 internals. Information about electrometer insides are scarce. Maybe even A/D board if it's not same revision as one's I have (J and H).
You can upload you photos on my FTP and I would rehost on website for people repairs. Size unlimited, everything is welcome :)
Also firmware would be nice to dump, if you have tool to read ROMs. We have only A10, A12 and A13 and it's only for 6517B.

Also maybe someone have front panel board with VFD? Since I had salvaged VFD to fix my 2400 long time ago, and now I need working front panel to one of 2001's.
If nobody have - then I will sacrifice my 7001 and desolder VFD from it's board and solder back on 2001 front panel PCB (still have it, just no glass).
« Last Edit: April 12, 2015, 04:17:42 pm by TiN »
YouTube | Chat room | Live-cam | Have documentation to share? Upload here! No size limit, firmware dumps, photos.
 

Offline Smith

  • Frequent Contributor
  • **
  • Posts: 349
  • Country: 00
Re: Restoration glory of Keithley 2001 DMM
« Reply #268 on: April 12, 2015, 05:23:34 pm »
I will take a look tommorow, I just started calibrating and checking the 6517A. The main electrometer input circuit is pure electronics porn, so I will definetly take some pics. Too bad I don't have a EEPROM reader. I still have C03, C05 is newer but has to be flashed directly in the two EEPROMs.

The VFD is quite bad, it had been up and running for about 8 years, so some characters  can only be read from nearby. Maybe I'll try to get some donor, a new one is too expensive.
« Last Edit: April 12, 2015, 05:26:35 pm by Smith »
Trying is the first step towards failure
 

Online macboy

  • Super Contributor
  • ***
  • Posts: 1970
  • Country: ca
Re: Restoration glory of Keithley 2001 DMM
« Reply #269 on: April 13, 2015, 06:23:37 pm »
Here's a neat trick to increase the "resolution" when viewing signals much less than full scale, such as measuring a few microvolts or a null (like in the DMM Noise thread). Apply a mX+b math function with m=1000 (for example). Then the 200 mV full scale becomes 200 uV full scale... and you still get 7.5 digits on screen.  :-DMM  Enable a big long digital filter to tame some of those extra digits. Clearly one must understand what they are and are not achieving with such a trick, or just don't bother.
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4052
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: Restoration glory of Keithley 2001 DMM
« Reply #270 on: April 13, 2015, 06:55:29 pm »
Yes, math flexibility is one reasons why I like 2001's.

Today no more meters get fixed, but got prepared frontpanel.
I have 2 units left to with, that's not counting donor unit, which is unrepairable.

One of units I bought had no front panel board. And VFD from donor front panel board was moved to my 2400, as original 2400's glass was broken.. So I end up having 2001 front panel PCB, but no VFD glass.

Later few years I bought pair of 7001 for 100$. One had broken glass, second one fully functional. Since I don't have any scan cards, 7001 goes to destruction derby now. Already took DC-AC module from it before, so it's decided...



Same glass, different PCBs.

This job is not for nervous people. It took me about an 1.5 hours and extensive use of censored language , together with solder wick and flux to "free" glass pins in thruhole pads.
I did not want to heat up whole board and glass, so did it this way, one by one. These glass mounted pins usually pretty gentle and easy to peel off if apply even little force.



I was gentle..

Then clean pin holes in 2001's board. To do so, just heat pad and push toothpick thru, to open hole.



After this everything is trivial. Pop glass in, it drops in like a charm with zero force, and solder it pin by pin with 1-1.5mm tip.



End result - it works :) Front cover need some cleaning, that will do later.

Any one know how to make yellow plastic look like new?
« Last Edit: April 13, 2015, 06:57:21 pm by TiN »
YouTube | Chat room | Live-cam | Have documentation to share? Upload here! No size limit, firmware dumps, photos.
 

Online macboy

  • Super Contributor
  • ***
  • Posts: 1970
  • Country: ca
Re: Restoration glory of Keithley 2001 DMM
« Reply #271 on: April 13, 2015, 07:13:56 pm »
Any one know how to make yellow plastic look like new?
Look up retr0bright.
It is a combination of very strong hydrogen peroxide (hair bleach) and a little bit of laundry oxygen booster. The other essential ingredient is strong UV light which is free in the form of sunshine. Now that spring has sprung here and the outdoor temperature is neither too cold nor hot I will be restoring a few bits of my own (one K2001 in particular is very yellow).
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4052
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: Restoration glory of Keithley 2001 DMM
« Reply #272 on: April 13, 2015, 07:18:09 pm »
How about labels, could H2O2 destroy them? I remember strong one is pretty agressive.
YouTube | Chat room | Live-cam | Have documentation to share? Upload here! No size limit, firmware dumps, photos.
 

Online macboy

  • Super Contributor
  • ***
  • Posts: 1970
  • Country: ca
Re: Restoration glory of Keithley 2001 DMM
« Reply #273 on: April 13, 2015, 07:30:36 pm »
How about labels, could H2O2 destroy them? I remember strong one is pretty agressive.
People use this stuff on old computers, gameboys, and other things made of ABS all the time, with little fear over painted / silkscreened labels. Paper labels and stickers are another story. This stuff is not a solvent; it is specifically reacting with bromine (fire retardant) in the plastic which is the cause of the yellowing.

I had to search for the right peroxide, I found what I wanted at a professional beauty supply store: 30 or 40 "vol" peroxide gel. The gel is convenient because you can just paint the mixture onto the plastic rather than needing to submerge the entire piece in a liquid. You can also add thickener xanthan gum (get that from a health food store; it is used as a binder in gluten-free baking).
 

Offline McBryce

  • Super Contributor
  • ***
  • Posts: 1363
  • Country: de
Re: Restoration glory of Keithley 2001 DMM
« Reply #274 on: April 15, 2015, 11:35:49 am »
The first Retrobrite attempts involved buying and mixing lots of hard to get chemicals, but the guys from the retro computer scene have moved to using this product: http://www.amazon.co.uk/Jerome-Russell-Bblonde-Maximum-Peroxide/dp/B00DG8RTHK/ref=pd_cp_beauty_3/280-1602579-8962721
Not sure if it's available where you are, but this can be used "as is" without having to add anything further. You just smear it on, cover the part in clear plastic and leave it out in the sun. Just rinse off afterwards to reveal a like new colour.

McBryce.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf