Keithley 2000 Firmware Update

[Inverted18650]

Greetings All,

Can someone teach me how to update the firmware on my K2000 DMM's?

Background:
I found version A19 of the firmware here but I have no idea how to install it. I have never flashed anything except an Arduino. I would like to run version 19 of the firmware so I can use LabView. I believe A20 is the newest update but requires new calibration.

I installed the K I/O layer, then GPIB 9.0 (for the plain KUSA-488) from Kiethley and then my instrument was detected by my PC and showed up in NI-Max. I then installed the "ke2000" packages through the LabView/NI installer and kept getting an error when trying to run the examples.

Right now I am running A06 on both meters, Windows 7 Ult (SP1), and tried to get the meter to talk to LabView (17.0, 32-bit) with no luck. After spending a full day playing with it I re-read the app note from Keithley on the xDevs forum and realized I need to update the firmware in order to use my KUSA-488 ( no prefix) GPIB cable.

If anyone has the time to help me out I would greatly appreciate it. I will sit down and do a complete video on the topic and post it to YT once I know what I am doing. I'd like to update the firmware and then document how to get it working with LabView. To my knowledge, there is no resource for it yet. I am not after any credit for making the video, just want the resource to be available for those who have the same issues. It would be awesome if "TiN" was able to make the video, but I understand he's got shit to do as well.

[wictor]

I believe you will need new EEPROM/Flash chip and programmer to program the chip with new firmware image. Do you have such tools? Something like TL866 could be used to program the new chip.

When right tool, new chip and FW image are available, then the programming itself is quite easy task.

Wictor

[Inverted18650]

I can order it..no worries. I was also thinking that I may just be able to swap out the old EEPROM with a chip that has already been flashed to Rev A19? Maybe someone who already has the resources would flash a couple chips for me and allow me to buy them. If I can just desolder and replace them, as I am not sure it works that way.

Per "https://xdevs.com/fix/kei2000/ " I will order the SST39SF020A-70 FLASH ROMs and the TL866 rpogrammer and install the software if my above idea is not feasible.

[wictor]

The chip seems to be PLCC32. If you can get the programmed chip, you just remove the old one and push the new one to the socket. No soldering needed.

This kind of tool is used to get the old chip out:
https://www.amazon.com/uxcell-Professional-Extractor-WTS-610-Assisted/dp/B007Q8J2HG

[wictor]

I can order it..no worries. I was also thinking that I may just be able to swap out the old EEPROM with a chip that has already been flashed to Rev A19? Maybe someone who already has the resources would flash a couple chips for me and allow me to buy them. If I can just desolder and replace them, as I am not sure it works that way.

Per "https://xdevs.com/fix/kei2000/ " I will order the SST39SF020A-70 FLASH ROMs and the TL866 rpogrammer and install the software if my above idea is not feasible.

The good thing with Flash chip is that you can reprogram it.  And calibrations are normally stored in separate chip so you don't loose them. Only when calibration scheme changes, you maybe need to redo the calibrations. I think this is the case with newer A20 FW.

I have K2000, but I haven't upgraded the FW in that unit. But I have updated one Agilent PSU FW, which is similar process.

[Inverted18650]

Thank you brotha! I feel progress for the first time in more than 24 hours...

So I can just flash the old chips (pictured) and not have to buy new ones. (Though I think I will buy a few anyway just to see how it reacts with the Rev A20 vs A19 or maybe flash a bunch and be ready to help others, if it works). I saved the images from the every Rev I could find from the Keithley website. I can play with them...

I bought a Fluke 335A so I could eventually recal them myself anyway and I am going through the steps to build up to it. First I want to log the output from the 335A for a while and see how much its drifts and what issues it may have, then get IT calibrated so it can be a transfer standard. 


Thank you again and I will get it the programmer ordered right now.

[wictor]

So I can just flash the old chips (pictured) and not have to buy new ones. (Though I think I will buy a few anyway just to see how it reacts with the Rev A20 vs A19 or maybe flash a bunch and be ready to help others, if it works). I saved the images from the every Rev I could find from the Keithley website. I can play with them...

No, I'm 99% sure that those that are now installed are EEPROMs, that cannot be reprogrammed. Order at least 2 new Flash chips and then you can program them as many times as you wish(~at least 10000 times anyway...) SST39SF020A-70 chips should work according following post:
https://www.eevblog.com/forum/testgear/keithley-2000-rs-232-wtf/msg669225/#msg669225

SST39SF020A-70 is the chip type and last digits, -70 in this case is the speed rating. Don't buy chips with bigger speed rating number , lower is faster and should work fine. Well, according quick googling, there is only 55ns and 70ns versions, so both should be fine.

[Inverted18650]

What do you think about this package? It supplies the TL866, the tweezers and a PLCC 32 Pin breakout. I will get the chips direct from Digikey, they have the -70 in stock, but there tools are a bit pricey.

https://www.ebay.com/itm/High-Speed-MiniPro-USB-TL866CS-Programmer-EPROM-EEPROM-FLASH-BIOS-AVR-AL-PIC/232489072939?hash=item36216db92b:m:mpRcGphYNBRc0sOiuFXmYDw

[wictor]

What do you think about this package? It supplies the TL866, the tweezers and a PLCC 32 Pin breakout. I will get the chips direct from Digikey, they have the -70 in stock, but there tools are a bit pricey.

https://www.ebay.com/itm/High-Speed-MiniPro-USB-TL866CS-Programmer-EPROM-EEPROM-FLASH-BIOS-AVR-AL-PIC/232489072939?hash=item36216db92b:m:mpRcGphYNBRc0sOiuFXmYDw

That can do the job. I have the TL866A version, which has the in-circuit programming cable also. It's not needed for this job and I think some have converted also the CS version to A version.

[Samogon]

I have spare already programmed laying around, PM me.

[Inverted18650]

I have spare already programmed laying around, PM me.

Thank you so much for the offer, but I already placed the eBay order. The offer is extremely kind of you.

edit: I JUST realized you are saying you have a spare chip that is already programmed, not a "spare programmer"...I will PM you in a moment. I read it too fast and misread what you were saying..

[Samogon]

Np
Just be prepared to see 306.2 and 306.4 errors on self test, thise introfuced lately. And old 2000 will fail no matter what

[Inverted18650]

Thank you for the heads up on the errors. I just read a bit about it and the leakage current errors it throws. Seems that the error has no adverse effect on the actual measurements so I will have to live with it (if I want to use LabView).

[Macbeth]

TL866 is a no brainer and a very useful and cheap tool. Definitely get the extra adapters with it for a few dollars more, because it will just cost 10x as much otherwise.

The first thing you want to do is make a backup of the little 24LC16B EEPROM that holds the cal data. You can do this with an 8 pin SOIC clip, but I found the electrolytics surrounding the EEPROM would interfere and it was easier to use 3 pin grabbers instead. You do this while the instrument is powered up, but not running. What I found was forcing RESET was not good enough, which works with K2001/2002. But pulling the 2 firmware OTP EPROMs and powering up allows you to backup and program the EEPROM no problem.

Now it is well known that moving up to A20 requires a new DC calibration, indeed if upgrading the firmware if you are quick you will see the message

"+514   DC calibration data lost"

However, I also found upgrading to A19 messed with the AC calibration, I got the message

"+513   AC calibration data lost"

This didn't bother me as the meter I was fixing had a totally fecked up AC calibration and I had to replace a voltage divider and so recal the AC anyway.

The main thing is make sure you make a backup of your original A06 calibration. Perhaps try stepping up firmwares until A19 and watch out for the AC cal lost message knowing you can reflash the cal and downgrade happily.

Also with old hardware and new ROMs you will no doubt have the pleasure of the 306.2 and 306.4 "errors" when running the built in self test.

[Inverted18650]

Thank you MacBeth. I will be sure to back up the Cal chip. I am re-watching this now.

https://www.youtube.com/watch?time_continue=23&v=FLG03f_ua5g

[Inverted18650]

UPDATE: I was able to get the LabView working on my Keithley 2000 (Rev06) with the OEM KUSB-488 without updating to Rev19. After sleeping on it, I decide to remove all the softawre I installed a few days ago and try again. What I needed:

Keithley GPIB_488_v9 Driver
https://www.tek.com/accessory/kpci-488a-software


Keithley I/O Layer (IVI Drivers)
https://www.tek.com/software/KIOL/850C07#

NI's : ke_2000_driver package (with examples)
http://sine.ni.com/apps/utf8/niid_web_display.download_page?p_id_guid=E3B19B3E960B659CE034080020E74861
(for this file you can just use the NIMAX->Search, as well)

After I reinstalled the "I/O layer" (which I installed last), NIMAX recognized the not just the Keithley KUSB-488, but finally assigned my VISA address "GPIB0::1::INSTR" so I could run the VI example.

I will try to film it this week and get the footage up for anyone else who finds themselves stuck. 

[Inverted18650]

The IC's and TL866 have arrived. I am ready to install the TL866 software and prepare to flash my first chip.

[Inverted18650]

I just want to make sure I have the steps in order prior to starting this project:

1: Using a 3 (or Pin grabber connected to 24LC16B, I need to pull both ODD and EVEN EEPROMs and then power up the meter so I can save the CAL data from the 24LC16B. I do this by having the TL866 connected to the other end of the grabber and running the "Read From Chip" program in MiniPro V.6.60?

2: Using "https://xdevs.com/fix/kei2000/#firmware", start with A07 and work my way up to A19.

3: Once I verify that the CAL data is saved I can power down the meter. Then I install the newly flashed SST39SF020A-70 chips and reboot. Run B.I.T. and record any errors along the way. (take a few measurements on the different ranges and settings?)

I did notice the "Erase Before" box is checked by default in ProMiniV6.60 and I need to be damn sure I un-check that box prior to attempting to save the CAL info from the 24LC16B. Are there any "gotcha's" that I need to worry about? I can just upload the files directly into MiniPro V.6.60 as 'BINARY, NORMAL' and then use "Program Chip" button to send the data to the new chip?

[Macbeth]

1: Using a 3 (or 8 ) Pin grabber connected to 24LC16B, I need to pull both ODD and EVEN EEPROMs and then power up the meter so I can save the CAL data from the 24LC16B. I do this by having the TL866 connected to the other end of the grabber and running the "Read From Chip" program in MiniPro V.6.60?

Yes. If using the 8 pin clip do not connect Vcc as the MiniPro complains.

2: Using "https://xdevs.com/fix/kei2000/#firmware", start with A07 and work my way up to A19.

I don't think you have to go through each revision. IIRC I got the AC cal lost message with A19, then restored my calibration and went to A15 for a while then A19 and never got the AC Calibration lost message again. However I did have to make a repair to the AC on my meter and a DIY cal adjustment.

Perhaps you can go straight to A19 without issue? Nobody else has reported it.

Worst case - you get the AC Calibration lost message and just restore 24LC16B backup if needed. Then flash A15 then A19.

3: Once I verify that the CAL data is saved I can power down the meter. Then I install the newly flashed SST39SF020A-70 chips and reboot. Run B.I.T. and record any errors along the way. (take a few measurements on the different ranges and settings?)

Keep a sharp eye (and your camera phone recording) on the meter when you switch it on after upgrading the PROMs as the upgrade messages run by quickly if you have any. You can do self tests and measurements after. You may get error messages about GPIB address lost and invalid system communication. You just reset the GPIB and serial settings with the front panel.

I did notice the "Erase Before" box is checked by default in ProMiniV6.60 and I need to be damn sure I un-check that box prior to attempting to save the CAL info from the 24LC16B. Are there any "gotcha's" that I need to worry about? I can just upload the files directly into MiniPro V.6.60 as 'BINARY, NORMAL' and then use "Program Chip" button to send the data to the new chip?

I don't think Erase Before is a valid option when reading. It's only used when flashing. Just leave Erase Before and Verify After ticked.

Yes, just use binary files non of that Intel HEX or Motorola rubbish.  Good luck!

[Inverted18650]

No luck getting the cal data out of the 24LC16B. I am using just pins 5,6, & 7 and the TL866 says the read is successful but it does not appear to be actually capturing  anything. Once the 'read' is complete, the table still has nothing but FF in every position. Any ideas?

I ended up just soldiering 3 thin wires directly to the legs of the 24LC16B. (I thought the kit i bought had them but was wrong). I only used SDA(5), SLC(6), and W-R(7) and made sure to put the wires into the programmer in the correct footprint location per the diagram. When I removed the EVEN and ODD EEPROM, I get "No Comm Link" error. Meter works fine again with them re-installed.
 

[Samogon]

No Comm Link means central CPU is not working, sure have no FW to run on when  FW chips are out.

[TiN]

Did you connect ground? You dont need mess with WR pin.

[Inverted18650]

Did you connect ground? You dont need mess with WR pin.

No sir. As much as I wanted to findle about, I decided to wait for guidance. I will try SCL, SDA, & GND tomorrow. Thank you.

[Inverted18650]

Using GND(4), instead of WR(7) worked. Thank you all.

[Inverted18650]

Is there a repository for the 24LC16B data? Both of my K2K have been reset to factory specs and I have saved both sets of files. After I get through tinkering with the firmware and I going to attempt the calibration with the Fluke 335A (not AC). If I am successful, I will pull the data from the 24LC16's again.

This has been very rewarding and opened my mind a bit. I assume if I mess up one of the 24LC16B's, I would have to buy a new chip, flash and replace it, but does that mean I could swap cal data from another unit? Of course the data is specific to each meter as they age and settle differently but the possibilities seem pretty cool. In theory, I could flash a chip and send it across the world and anyone with a K2K could solder it in. It would not be spot on but if you had a skeleton K2K and could only rely on 3-4 digits, it may work