I’ve started restoring my Triplett 850 VTVM.
The Triplett 850 Type 2 manual is available in my Liberated Manuals section. Didier (KO4BB) also has an excellent scan of a somewhat newer manual in his repository, along with an annotated schematic and parts list.
A friend from the Antique Radio Forum kindly sent me a knob for the range switch. I still need the handle. Perhaps I can make a reproduction, if some other 850 owner would be kind enough to take measurements of the handle and send them to me.
The Cornell Dubilier Tiny Chief capacitors are molded paper caps, so they have to be replaced, along with the electrolytic filter capacitor in the power supply. I’ve already cut out the filter capacitor and some drifted resistors in this photo.
I checked all the resistors. Like Sol over on the Antique Radio Forum, my 850 deviated a bit from the manual. R16 is 1 KΩ rather than the 510 Ω specified in the manual, and R21 is 30 KΩ rather than the 27 KΩ called out in the manual. Fred Scoles says those are the values used prior to production run #20 of the 850’s.
The DC range voltage divider is in pretty good shape. Only R3 needs to be replaced. In the AC range divider, R25 and R26 need to be replaced, but thankfully R27, the big 900KΩ precision resistor, is close enough to being in spec.
Like the similar RCA WV-98A, the ohmmeter reference resistors are 5% carbon composition and most have drifted out of tolerance. This part of the range switch is going to need to be totally rebuilt.
Of the remaining carbon composition resistors, most of them have drifted out of tolerance and thus will need to be replaced. Of note are R21 and R22, which used to be attached to the right-hand terminal strip in the above photo. R22 had drifted from 30 KΩ up to 38.4 KΩ, and R21, part of the power supply voltage divider, was essentially open-circuit (around 1.2 MΩ, depending on how much I wiggled the leads). This failure of R21 would have prevented the meter from functioning. I guess that’s why the cosmetics are so good - no scratches on the meter face and the front panel in good condition.
Note the 1N1693 silicon diode. I have to remember to check the diode the next time I’m at the bench.
The filter capacitor was a 12 µF 250 volt electrolytic. I don’t stock that value, and I was reluctant to bump it up to 22 µF, so I placed a Mouser order. There were also a few drifted resistors whose values I didn’t have in stock, so I added them to the order.
It’s a little tight working in this area. Aside from the capacitors, the 5.1K and 3.6K resistors (R23 and R20) were out of tolerance, so most of the parts here have to be removed and replaced with new components.
The 15 KΩ resistor at the top of the photo (R17) was also out of tolerance.
I took enough photos to ensure I could get the replacement wired up correctly.
Since a number of resistors need to be replaced on the range switch, I pulled it from the front panel. Even so, there’s not a lot of room to work. I’ll probably break off a wire or two. Maybe I should bite the bullet and disconnect the range switch from the rest of the circuitry?
Here I’ve replaced R41 on the function switch.
The capacitor in the lower center of the photo is C3, the 0.1 µF 1000V coupling capacitor for the AC input voltage divider. I should have ordered a replacement in my Mouser order but I forgot. I’m hoping it’ll test good on my GR 1864B and I won’t have to replace it, but that’s probably wishful thinking.
Another view of the range switch. You can see some of the ohms reference resistors on the left-hand half of the rear wafer. They’re all wired up to a common ground bus that’s just hanging in space. I bet the factory used a jig to wire up the ground bus before installing the set of resistors on the range switch.
Here I’ve replaced the Tiny Chief capacitors with modern polyester film caps. I’ve replaced R23, but I’ll have to wait for my Mouser order to arrive to replace R20, R21, and R22, and the filter capacitor.
When I’ve been working on the bottom of the chassis, I’ve been propping up the meter on a screwdriver. This is really a recipe for disaster - the meter could take a tumble off the workbench and land on the concrete floor. I should clamp a wooden leg to the chassis.
Update
My Mouser order arrived and I’ve finished the range switch and the chassis. C3, the 0.1 µF 1000 volt capacitor tested at 110 MΩ on my GR 1864-B, so I’m calling it good.
I uploaded my spreadsheet for the resistors in the Triplett 850, both the original MacOS Numbers version and an exported Excel version. The Numbers version looks better, because my fancy number formatting for resistances didn’t export to Excel.
The bottom of the chassis is finished.
Here I've replaced the electrolytic with a new Sprague Atom 12 µF 250 volt capacitor, and installed replacements for R21 and R22.
Here I’ve installed the new R20, fresh from Mouser.
Here’s a partial view of the resistance string on the range switch.
Here’s the rest of the resistance string.
Rather than risk damage to the solder lugs on the range switch, I opted to cut out the resistance range resistors, then clean up the lugs separately.
Neither Mouser nor DigiKey stocked a 8.45 MΩ 1/2 watt or larger resistor to replace R39 (8.49 MΩ), so I replaced it with an 8.06 MΩ in series with a 475 KΩ.
R25 and R26 in the AC voltage divider have been replaced.
To install the new resistance range resistors, I first soldered all of them in to the lugs on the range switch.
Then I went along and connected the other ends into a common bus. I re-used the old varnished cambric insulated tubing where the lead passes by the spacer on the switch bolt.
The 10 MΩ resistor is a Vishay BC VR37 series resistor, while the rest are Vishay Dale CCF60 series resistors except for the 9.53 Ω resistor. I had to settle for a small-bodied 1/2 watt resistor for that one, as the CCF60’s aren’t made smaller than 10 Ω.
I still have to make and install a battery eliminator, and then test and calibrate the VTVM.
Update: I’ve built and installed the battery eliminator.
The Triplett 850 takes a C cell for its ohmmeter circuit. I used a 1 3/4 inch length of 3/4” copper pipe as the enclosure. It fits the C cell battery holder quite well.
It was a bit of a challenge fitting everything in.
Here’s the wiring side.
The circuit obviously needs to be insulated from the enclosure.
A piece of 1” heat-shrink tubing would have been great, but I didn’t have any. I ended up just wrapping the board with several layers of electrical tape.
The yellow wire is the 6.3 VAC feed to the battery eliminator. I tapped in to the filament circuit at V2.
I wired the ground and the 1.55 VDC output to the terminals on the battery holder.
This is something of a cross between a schematic and a layout drawing for the battery eliminator. It’s a minimal LM317 circuit. I did use a 60 volt 1 amp Schottky diode as the rectifier. I bought it surplus from All Electronics, it doesn’t have a JEDEC number on it, just “SB160” and a date code of 0212G. I don’t recognize the manufacturer’s logo.
I fired up the meter and it responds to the “zero” and “ohms” controls. The zero shifts a bit between the DC voltage ranges. I think that means there’s a bit of grid current perhaps due to gas. I’ll let the tubes age a while and hopefully the getter will absorb the gas.
I’ll need to make a probe. After that, I’ll be ready to calibrate the meter.
Update:
I found my stock of 0.1 µF 1000 volt capacitors, so I went back and replaced C3.
I re-used the varnished cambric spaghetti from the old capacitor.
Another update:
I’ve completed a set of probes for the 850, using my design for a switchable VTVM probe.
The Triplett 850 has a separate pin jack for the ground lead. I used a modern-style pin plug.
It’s warming up in preparation for calibration.
Calibration
I raise the instrument to eye-level when calibrating it. This makes it easy to read the meter without parallax error, without having to stoop down or squat.
Next time I calibrate a meter, I’ll build a table rather than improvise one from storage boxes.
I record the checks in a spreadsheet as I make them. I need a better mount for my computer, too, or else get an iPad.
I set the calibration pots per the manual, but half-way through verifying the AC ranges, the calibration shifted. It took a bit of thinking to get to the root of the problem. At first I suspected the AC calibration control of drifting, as it had needed a good cleaning when I first set it.
After a while I thought to check the resistance of my homebrew probe. It turns out that the switch in my homebrew switchable probe had failed, and stopped shorting out the 1 MΩ resistor in the probe.
While my probe wasn’t designed to be serviced, I managed to slice it up neatly with a razor saw and miter box and expose the switch and its mounting board. I’m not sure why the switch failed. One possibility is that I may have overheated it when soldering the circuit together. The switch body is a molded thermoplastic.
At any rate, I replaced the old switch and circuit board with a new one, and solvent-welded the probe back together. That fixed the immediate calibration problem.
Click the above thumbnail for a readable image.
The meter is well within its specified accuracy except for the 500 volt and 1500 volt AC ranges. The 1500 volt AC range is just within spec, but it’s far worse than the lower ranges. The 500 volt AC range is far out of its specified accuracy.
The Triplett 850 AC measurement circuit, unlike the Heathkit VTVMs and the RCA WV-97A, has both the standard pre-diode divider (to limit the maximum voltage applied to the 6AL5 diode) and separate AC range voltage divider. (Many VTVMs use the same range voltage divider for both DC and AC measurements.) So, the fault could be in either AC divider.
I’ll have to go make some measurements to isolate the problem. I suspect I missed changing out an out-of-tolerance resistor in the AC circuitry, perhaps R25 or R26. I’m a little nervous about troubleshooting this, though. The Triplett 850 circuitry is pretty compact, and I’ll have to be probing the crowded area next to the range switch decks with 900 volts AC applied. This calls for caution!
Another update:
ARFer stevef asked about the wiring of the power line ground on the Triplett 850. I noticed I hadn’t photographed that area of my 850, so I took some photos for him. The wiring is a bit unusual.
The earth ground on the power line connects only to the shield on the probe cable, connector, and front panel, and to C8. This grounds the exposed metal on the connectors for the probe and the aluminum front panel, but allows the copper chassis and measurement common to float above earth ground.
The ceramic capacitor in this photo is C8, and it connects between the green power cord ground wire and the two black wires that connect to the copper chassis and the measurement common input jack.
The green power cord ground wire connects to the aluminum front panel in two ways. First, in the photo above, you can see it is soldered to a lug attached to one of the tie bolts on the function switch. These screws connect to the front plate of the switch, which connects to the front panel via its mounting collar.
Second, a black wire leads from the lug where the green power line ground wire terminates to the solder lug for the Amphenol 75-MC1M probe connector.