Some readers may be aware of my strange interest in weird old test equipment, and this latest one goes perhaps further than any before!
I bought two of these units from the VMARS stand at the NVCF several years ago for the princely sum of £8. One instrument had a smashed Bakelite case but looked fairly clean inside, whilst the other was in better shape but had suffered badly from battery electrolyte corrosion. As I was looking for a meter capable of measuring resistance down to very low values, the scale of 0 to 0.1 ohm interested me and I hoped that I would be able to make a working instrument from the two wrecks.
When I got round to dismantling them, the full extent of the damage was revealed. On the plus side, I had (just) enough unbroken case parts to assemble a whole instrument. On the minus side, however, one movement was severely corroded, and the other had a broken ligament. I could also find out nothing about these meters at all on the internet so, after some fruitless research and starting to regret my purchase, I put the various parts into a box towards the bottom of the ‘to-do’ pile.
Recently, the chance acquisition and successful repair of a small
Megger circuit testing ohmmeter reminded me of the old Record bond testers, and I resolved to have another look at them. More research revealed that my units were similar to an Evershed and Vignoles bonding tester, and some very helpful information including a circuit diagram was available on the excellent
Richard’s Radios website. The movements are of the dual-moving-coil ‘Megger’ patented true ohmmeter type having a deflection coil and a control or restoring coil set at an angle to each other, but no hairsprings in the conventional sense. The coils are fed by two pairs of fine wire ligaments, which wind and unwind onto a barrel and so the pointer tends to float until the meter is energised.
Despite being badly corroded, the older movement (serial number 4382 dated 1944) still worked. I eventually fathomed out the circuit and with the aid of a multimeter identified the movement's connections and polarity, from which point I was quickly able to sketch out most of the rest of the circuit apart from the probes, which were missing. I started by removing the corrosion from the working movement, after taking advice from the Forum, using a de-scaling solution of equal parts of acetic acid and lemon juice, which was moderately successful. However despite taking great care and rinsing all the descaling solution away with boiled water, as I was reassembling the instrument I noticed that one of the control coil’s ligaments had become detached at its outer end. This wire is so fine that it is barely visible, and after some abortive attempts to solder it, I almost gave up on the whole project!
The other movement, serial number 5103 and bearing a ‘repaired’ date of 1957, was not corroded at all, but had a broken deflection coil ligament. The deflection coil carries more current than the control coil, and its ligaments are correspondingly thicker, being flat in cross-section rather than circular wire. Part of the ligament had disappeared, so I extended the fixed contact using a stiff copper wire to the point where I thought I could solder it onto the remaining section of ligament. After several unsuccessful attempts to ‘tin’ the tarnished end of the ligament, I took advice from the Forum again and it was suggested that I use a more aggressive flux. Eventually, and after much struggle and some profanity, I managed to make a good soldered joint. The movement had been brought back to life!
I replaced all the corroded internal wiring, fitted a paralleled pair of 1.2 volt sub-C size NiMH cells in place of the original alkaline NiFE accumulator that had caused all the damage, and lashed up a couple of test leads. The instrument then worked, but was reading far too high. A bit more research suggested that the original probes had internal resistances in them, although the values weren’t known. I tried experimenting with various values in series with the positive voltage probe feeding the deflection coil, and eventually settled on a value of 1.89 ohms (1.5 + 0.39 ohms in series) which gave very good accuracy across the whole scale range. I have some very low value 3% tolerance resistors, and on test the meter reads correctly and well within the 3%.
I read somewhere that four-terminal high current milliohmmeters of this type were designed for checking electrical cross-bonding links in aircraft, and originally came with heavy probes which rotated before closing a switch in the battery positive lead, thus ensuring that good solid contact was made with the resistance under test before the battery was switched on. In the absence of any details, I had to experiment and make something up, and settled for 0.75 sq mm two-core mains flex with simple crocodile clips as the voltage detector probes, plus a third ‘prod’ acting as the current injector. Accuracy is virtually independent of battery voltage or test lead resistance, within reason. The test leads are fitted with standard BS546 5-amp round-pin plugs. The earth pins aren't connected, and serve only to ensure correct polarity. The meter works with the leads plugged in to either socket.
I quickly discovered that it is vital for the voltage prods to make good contact with the resistance under test, as failure to do so causes a roughly 13-times overload which results in the meter movement being flung violently with a bang against the end stop when current is applied. If I’d had any worries about the integrity of my soldered repair to the ligament, it was dispelled by the time I’d done this on several occasions, and the movement survived... these things certainly are robust!
In use, the deflection coil behaves as a low-resistance millivoltmeter and the prods are clamped tightly across the resistance under test, before touching the positive end with the current prod. Battery current of anything between 1.7 and 2.0 Amps flows via the shunted control coil through the resistance under test, and the pointer indicates the resistance. For example, with 0.1 ohms across the prods, around 170mV appears across them which drives the deflection coil towards FSD. A proportion of the battery current also flows through the control coil, creating an opposing torque, and when equilibrium is reached the pointer indicates the ratio of voltage/current, hence behaving as a ‘true ohmmeter’ with its basic accuracy determined by its physical construction. I calculate that at FSD measuring 0.1 ohm, the 1 ohm deflection coil is passing 58mA with 58mV across it, and the 60 ohm control coil is passing 15mA with 950mV across it. Looking at the coils shows that they are wound with substantially heavier gauge wire than that seen in more conventional moving coil meters. There is, incidentally, no ‘zero’ adjustment on these movements other than bending the pointer where it is fixed to the moving coil assembly.
The final steps were cosmetic. The case is made up of very substantial Bakelite mouldings which responded well to polishing with Greygate No 5, and the scale window was removed, cleaned and refitted. All brass contacts were also polished. The instrument is quite heavy and was originally fitted with a leather strap as a carrying handle, fitting the two brass studs a bit like those on an Avometer. The studs are positioned towards the front of the case to counterbalance the weight of the original nickel-iron wet accumulator. I made a carrying handle from a scrap leather dog lead, but with the relatively lightweight NiMH cells I’d fitted, the instrument hung out of balance. I cured this by adding a small plastic bag filled with gravel in the large battery compartment! The meter including its leads and prods fitted neatly into a spare Avo leather carrying case.
If you’ve stayed with it thus far, you’ll probably be wondering why I need a meter that can measure resistances as low as 0.1 ohm. I do a lot of Avometer repairs. The 10 Amp range shunt in the Model 7 is 0.0125 ohms and that in the Model 8 is 0.05 ohms. I can now measure these accurately, as well as checking the resistance of the closed cut-out contacts, test leads and so on.
The Record Bond Tester is described in Air Publication A.P. 19744, which I've been unable to trace, and have an RAF reference number 5G/2126, so they were probably only ever sold to the Military and the vast majority were probably scrapped decades ago. Most of them were probably dropped and smashed, or ruined by the effects of spilled alkaline potassium hydroxide battery electrolyte. Judging from the complete absence of any pictures or information about these meters on the internet, I'm tempted to wonder whether I own the only working example in existence. Could this be a Record...?
A few ‘before’ photos are attached, plus the circuit diagram. Thanks for reading!
Phil