Repair and cleanup of a vintage Keithley Instruments 600A Electrometer.
After searching online for a little over a year I came across this Keithley meter on an Ebay auction. I had been holding out for a repairable digital display Keithley model, but with the starting bid and shipping on this vintage model I thought it would be worth a try. Luckily I was the only bidder on this unit and picked it up for under $40 USD.
Shipping took 10 days from the West coast, so I had some time to prepare for it’s arrival.
One item that would need to be replaced from viewing the auction photos was the dual D‑cell battery holder. The photos indicated lots of corrosion and electrolyte leakage in the past, so I ordered a Keystone Electronics 176 battery holder which arrived before the meter. I also ordered a full set of batteries for the unit.
From the angle of the auction photos, I did not notice that the meter bezel was pushed in past the front panel. Other than that the meter arrived in good shape in the condition that I expected.
Next order of business after it arrived was to open it up and check the condition of all the internal components, and for any obvious problems.
The first item I found was a broken wire to the “Zero Check” mid support terminal. The next item was a terminal block shorted to the outer case of the “Set Zero Center” potentiometer which is connected to case ground.
Battery electrolyte on top of resistors and old flux on solder connections.
Overall the inside of the unit had a lot of residual battery electrolyte that had leaked and then flaked off onto the sides and components. It took nearly a day and a half of cleaning and checking before I was satisfied.
The problem with cleaning Electrometers is that many of the connections between components and their physical mounting points need an extremely high resistance either between components or ground for accurate measurements. The Keithley 600A Electrometer uses PTFE bushings and standoffs which provide a very high electrical insulation of > 1018 ohms. To preserve this high resistance all surfaces of the PTFE must be contamination free.
Other areas that must be contamination free are the input connector and the glass envelope of the vacuum tubes where the leads enter the glass.
My weapons of choice for cleaning PTFE and gold contacts are 99.9 % Isopropyl alcohol, and then DeoxIT Gold on the contacts. I usually start with a physical cleaning using the Isopropyl alcohol and a soft bristle toothbrush, and then multiple rinses with the alcohol and drying with compressed air between rinses.
After performing other various repairs including the meter housing, it was time to install the batteries.
The operations manual called for 1.34 V mercury D‑cell batteries and 8.4 V mercury batteries which are non-existent now days. I ended up using some Tenergy Centura low self-discharge Ni-MH D‑cell and 9 volt batteries. Both had similar voltage specifications to the mercury batteries and are rechargeable. I went with the low self-discharge type to reduce the need to open up the case every few months.
The other hard to find battery is the No. 413 type 30 volt units. The 600A requires two of these (B4-B5) which are expensive, but available online.
I did end up using some DeoxIT D5 on the front and back switch contacts, and followed up with several more rinses with Isopropyl alcohol on the PTFE mounts.
The main board is vibration isolated at four points to reduce microphonics in the vacuum tubes mounted on the board. The copper wires from the input connector and range selector are made from fine braided wires and are looped to also reduce vibration conducted into the main board.
For shipping there is a metal angle bracket that supports the main board and is removed for operation. The bracket is attached thru the hole in the back panel above the “Normal-Fast” selector switch.
Keithley 600A Electrometer Right side after repair and cleanup. 5886 electrometer tubes under shield
After warming up for 30 minutes I performed a calibration on the Volts setting, then several current and resistance checks on the meter. I currently do not have any equipment or standards to check the upper ranges of this meter, but all mid-range checks were accurate.
I did get a reading of over 100 Teraohms ( 1014 ohms ) with no leads attached.
Not bad for a meter over 55 years old made in the early 1960’s.
The operations manual is available from the Keithley website.
Hi, I just bought one of these, but the manual talks about using mercury batteries, which are no longer available (I found the 30V one, though). Are you sure that it’s supposed to use the 9Vs and D cells, because the manual that I found online says otherwise. Thanks
Hi Mike,
I am sure that is NOT supposed to use standard 9 volt and alkaline D‑cell batteries.
In my post I stated, “I ended up using some Tenergy Centura low self-discharge Ni-MH D‑cell and 9 volt batteries. Both had similar voltage specifications to the mercury batteries and are rechargeable”.
I have been using this electrometer for the past 20 months with no issues and have recharged the batteries twice in that time period.
Greg (Barbouri)
One of my manuals came with a penciled-in extra resistor, in series with each of the 2 D‑cells. If you add a well chosen resistor, you should be able to use standard 1.54 V alkaline D‑cells. The peanut tubes each draw 10 mA of filament current, and each cell drives 2 filaments, so a 10 Ohm resistor should work well there.
I have not tried this, but it might be a good solution for those D‑cells. I got sticker shock last time I priced those in NiCad or NiMH. How well this works would depend on how flat the discharge curve is on the alkaline D‑cells.
I have a couple of these that I’m just in the process of cleaning up and getting going again. Yours appears to be an even earlier model, with sockets for the peanut tube leads. Later versions are all soldered, to avoid any chance of varying contact resistance.
You can match your own pairs of 5886 electrometer tubes by setting up a test jig with a NiCad cell filament supply, a 9 V battery, and a 620 kOhm cathode bias resistor. This is 2x the bias resistor in the 600A, because you’re only testing one tube at a time. Tubes that test with the same V across that resistor are a matched pair, but different Keithley models used pairs with different Vs. You could also change the bias resistor in your 600A, within limits, if you can’t find a matched a pair that works well with your instrument.
I see that the clamp for your 9 V batteries is held down by thumb nuts. None of mind came to me this way, and my units don’t have enough headroom to clear. If the knurled nuts short out against the case/cover, that shorts out the floating ground of the 600A. In that case, just use normal hex nuts.
My 600As have a “tapered” cover that overhangs the meter if it installed correctly. If it’s installed backwards, the overhang is in the rear. I don’t see any front overhang in your 3rd photo, but it’s possible that the early units did not have this feature.
I see an extra carbon comp 10 MOhm resistor tacked onto the top of the coarse balance switch (rear panel.) The resistors at each end of that switch should be 430–450 kOhms, but people often fudged that to get mismatched 5886 tubes to work. It’s not the best workaround.
The spring that adds extra contact force to the battery holders is a really good idea. They made a big difference in connection reliability in mine. I had to make another spring to replace one that was missing out of my 2 units.
Also, in the battery holders, even if you replaced the holders, it’s a good idea to SOLDER the rivet to the lug in each battery terminal. Those tend to go resistive over time if you don’t.
Your holders show something wrapped around the upper ends of the side plates. What is that? It looks like a rubber band, but I can’t see well enough to be sure.
The thumb nuts on my unit have good clearance from the case side, so no issues with shorting.
You are correct with the tapered cover. The photo shows how the meter was shipped to me with the cover reversed. If it had been installed correctly it might have prevented some of the damage to the meter bezel in shipping.
Good to know about that 10 MOhm resistor on the coarse balance switch. It looks to be original to the unit, but will check into it.
There is a rubber band on the D‑cell holder, for a bit more holding force, as the spring would not work for the new style holder.
I am getting ready to recharge the batteries this month so will also solder the rivets to the lugs of the battery holders.
Will also be replacing the NEDA 210 batteries with some 3D printed holders with CR2354 cells.
Thanks,
Greg (Barbouri)
I like your replacement for the #210s. Would you consider selling a couple of them?
I’d recommend adding a rounded gusset at each end of each side rail, to strengthen and stiffen that corner. Sharp inside corners that are under tension tend to be places where cracks start.
I bought a couple of the #210 replacements. They work, but they are somewhat expensive. We’ll see how long they last. Your solution has the advantage of being able to replace individual cells easily.
BTW, one of the 600As that I just finished going thru, seems to work well, except that it lacks the large central knob. If anyone has a parts unit that they’d be willing to sell me a knob from, please try to contact me thru this web page. Greg has my email and should be able to pass your message on to me.
Hello,
I just finished repairing my 600A (also from eBay, for about 40€): one of the 450k resistors on the coarse grain zero adjust had lost internal contact (gone to infinity).
I am using also NiMH rechargeable batteries for the heating supply (C‑cells (baby) since I didn’t have D‑cells (mono) around — plenty for the 20mA heating current). I am using 8.4V rechargeable NiMH 9V-block batteries for B2a and B2b, and 9V alkaline blocks for B6 and B7. For the 30V batteries, B4 and B5, the previous owner had already prepared holders for alkaline 23A batteries (12V), two of them in series. But the resulting 24V were too little, so I wedged 4 x AG12 coin cells (1.5V each) into the holder as well, increasing the voltage to exactly 30V. I think this is a relatively cheap battery solution, which works well.
The 600A works like a charm now! I am so happy — I had wanted a high impedance meter for many years!
Would you make your 3D-print model for the coin cell holder available for download? Maybe on thingieverse or TinkerCAD or just here?
Bye
Helmut
Hi Helmut,
Here is the link for the battery holder I used on Thingiverse.
Battery Holder for ten CR2345 batteries
Greg (Barbouri)
Thanks for sharing the coin cell holder file, but I just noticed that there’s a typo in the name. It says CR2345 instead of CR2354.
I’ve downloaded the file and corrected the name.
Jim
i have also just recieved one of these. its got a little bit of rust and some weird white particulate within the case but other than that it seem to work ok. ive put in full sized alkaline d cells into the unit (i am a bit worried if the current is too much after reading some comments) and ive taken apart some 23A batteries for their inside cells and shrink wrapped 20 of them in series to make the proper 30v. seems to work ok though they have about 1/4rd of the capacity of the 413 cells. my zero check knob snapped off and need to be glued back together. I have also been dealing with the issue of noise on the lower sensitivity ranges. a fluctuation of about 4ish millivolts. Im not sure if its normal but it takes a long while for my unit to settle down as well. it takes about 15–20 minutes for the needle to stop drifting to less than 1 millivolt/ minute. ill add in a series resistor to the alkaline batteries and hope that helps. anyways thanks for the article. it really helped me out