Ballantine 323 True RMS Voltmeter
Posted: Thu Aug 03, 2023 2:47 pm
This came from a swapmeet a few weeks back. It was in the closing stages when prices were being marked down. It was rather dirty. I'd heard the name Ballantine and True RMS meters are always interesting. I offered the seller £2 and he accepted.
I thought it would have matched heated thermocouples, but it doesn't. It uses the logarithmic response of two matched backwards diodes and a very unusual logarithmic meter movement to give a true RMS response. It has a 1st stage attenuator, preamp and then second stage attenuator. The signal passes into an RMS detector arrangement with back diodes and is then amplified and brought out to the meter with a mechanically chopped chopper amplifier.
I believe this item of TE came out in the 60s and is still produced today with refinements, such as the mechanical chopper being long gone.
It was in a filthy state and looked as if it had been kept in the damp. There's a sticker on it saying the guarantee expires 25th June 1975. Inside was a festering NiCd battery, which I removed. I cleaned up the covers with surface cleaner and removed the cal stickers. Looking at the circuit diagram the battery is essential to provide the 12V supply for the true RMS meter and 6V for a multivibrator which drives the mechanical chopper. Some of the copper on the PCBs was discoloured. I cleaned them up with diluted white vinegar. Nothing seemed eaten through, it just looked horrible.
With the battery removed, and I had no intention of replacing it, I investigated the power supply. I powered it up and it showed some signs of life. After it had been powered on for a few minutes, the mains transformer gave off a wisp of smoke and the mains fuse blew. Maybe I should have replaced the reservoir cap, which is only 100µF as this instrument uses very little power. I decided to redo the power supply arrangement. The rectifier and reservoir capacitor, plus what was essentially a NiCd charger circuit, were on a board at the rear. I found a small mains transformer which produced about 13V O/C and which conveniently mounted using the holes for the battery mounting arrangement. I replaced the reservoir capacitor with a 470µF one, and removed the original PSU/charger components. The 12 and 6V supplies were provided by 7812 and 7806 regulators on a strip board, which was mounted via nylon standoffs using two holes already existing in the rear circuit board. The mains wiring was redone and wiring for the battery was removed. I cleaned up the wafer switch contacts, which is a common problem with old TE.
When it was powered on I checked it using a function generator and it seemed to work but was very insensitive. I checked the 12 and 6V supply lines at the PSU and a few other points in the circuit then I checked for ripple. The 6V supply line had ripple at 6MHz. It could have been a troublesome 6V regulator. I had put ceramic caps across the output and earth. I put together another regulator board and this time added 47µF capacitors as well as the ceramics. No oscillations this time. There was still a problem with sensitivity. It took 15mV to produce an FSD on the 1mV range. I investigated with a scope. I find an analogue scope is more pleasant to use for this sort of troubleshooting. It seemed to be a problem with the preamp after the FET. They put expected voltages on the circuit diagram. Sure enough the base of Q3 was at 3.69V (should be 1.0V), the emitter was at 0.04V (should be 0.25V), so no expected 0.7V B-E drop, and the collector was at 10.8V (not 3.2V). Looked like a dead transistor. It could have been spontaneous, or it could have been caused by the electrolytic C15 leaking. I replaced C15.
The transistor is a 2N3563 (specially selected by Ballantine and only to be replaced by an approved spare from Ballantine with the correct colour code). It's a transistor in a TO106 package, which has been obsolete since the 70s. It has a fT of 800MHz. I had a look through the collection and about the closest I could find was a ZTX 313 with an fT of 500MHz. I suspect any small signal NPN transistor with a respectable fT would do, such as 2N3906. The original transistor is O/C across all pins.
With the transistor replaced I tried it again. The voltages around the transistor were about as expected and the instrument worked very well, agreeing with a newish Siglent function generator to within 1 or 2 %. I checked it using the adjustment procedure in the manual but very little needed adjusting.
It does seem to be a reliable instrument. It's very fast responding and Ballantine claimed it was accurate across the scale, unlike inferior rivals using thermocouples. The meter seems different to most moving coil meters and was obviously designed to put up with some punishment. I'd be tempted by a less tatty one, if I came across it for the right price. I'd go for the mains only version rather than one with the inevitably decaying NiCd battery with its mess and hazards.
I thought it would have matched heated thermocouples, but it doesn't. It uses the logarithmic response of two matched backwards diodes and a very unusual logarithmic meter movement to give a true RMS response. It has a 1st stage attenuator, preamp and then second stage attenuator. The signal passes into an RMS detector arrangement with back diodes and is then amplified and brought out to the meter with a mechanically chopped chopper amplifier.
I believe this item of TE came out in the 60s and is still produced today with refinements, such as the mechanical chopper being long gone.
It was in a filthy state and looked as if it had been kept in the damp. There's a sticker on it saying the guarantee expires 25th June 1975. Inside was a festering NiCd battery, which I removed. I cleaned up the covers with surface cleaner and removed the cal stickers. Looking at the circuit diagram the battery is essential to provide the 12V supply for the true RMS meter and 6V for a multivibrator which drives the mechanical chopper. Some of the copper on the PCBs was discoloured. I cleaned them up with diluted white vinegar. Nothing seemed eaten through, it just looked horrible.
With the battery removed, and I had no intention of replacing it, I investigated the power supply. I powered it up and it showed some signs of life. After it had been powered on for a few minutes, the mains transformer gave off a wisp of smoke and the mains fuse blew. Maybe I should have replaced the reservoir cap, which is only 100µF as this instrument uses very little power. I decided to redo the power supply arrangement. The rectifier and reservoir capacitor, plus what was essentially a NiCd charger circuit, were on a board at the rear. I found a small mains transformer which produced about 13V O/C and which conveniently mounted using the holes for the battery mounting arrangement. I replaced the reservoir capacitor with a 470µF one, and removed the original PSU/charger components. The 12 and 6V supplies were provided by 7812 and 7806 regulators on a strip board, which was mounted via nylon standoffs using two holes already existing in the rear circuit board. The mains wiring was redone and wiring for the battery was removed. I cleaned up the wafer switch contacts, which is a common problem with old TE.
When it was powered on I checked it using a function generator and it seemed to work but was very insensitive. I checked the 12 and 6V supply lines at the PSU and a few other points in the circuit then I checked for ripple. The 6V supply line had ripple at 6MHz. It could have been a troublesome 6V regulator. I had put ceramic caps across the output and earth. I put together another regulator board and this time added 47µF capacitors as well as the ceramics. No oscillations this time. There was still a problem with sensitivity. It took 15mV to produce an FSD on the 1mV range. I investigated with a scope. I find an analogue scope is more pleasant to use for this sort of troubleshooting. It seemed to be a problem with the preamp after the FET. They put expected voltages on the circuit diagram. Sure enough the base of Q3 was at 3.69V (should be 1.0V), the emitter was at 0.04V (should be 0.25V), so no expected 0.7V B-E drop, and the collector was at 10.8V (not 3.2V). Looked like a dead transistor. It could have been spontaneous, or it could have been caused by the electrolytic C15 leaking. I replaced C15.
The transistor is a 2N3563 (specially selected by Ballantine and only to be replaced by an approved spare from Ballantine with the correct colour code). It's a transistor in a TO106 package, which has been obsolete since the 70s. It has a fT of 800MHz. I had a look through the collection and about the closest I could find was a ZTX 313 with an fT of 500MHz. I suspect any small signal NPN transistor with a respectable fT would do, such as 2N3906. The original transistor is O/C across all pins.
With the transistor replaced I tried it again. The voltages around the transistor were about as expected and the instrument worked very well, agreeing with a newish Siglent function generator to within 1 or 2 %. I checked it using the adjustment procedure in the manual but very little needed adjusting.
It does seem to be a reliable instrument. It's very fast responding and Ballantine claimed it was accurate across the scale, unlike inferior rivals using thermocouples. The meter seems different to most moving coil meters and was obviously designed to put up with some punishment. I'd be tempted by a less tatty one, if I came across it for the right price. I'd go for the mains only version rather than one with the inevitably decaying NiCd battery with its mess and hazards.