I recently acquired and have been working on an AVO Test Bridge. Its an interesting instrument and does work but it does have some problems. I found a thread on UKVR about the same instrument and asked some questions there, but I have reached a point where I am probably hijacking the thread so figured that I ought to start my own separate discussion.
This is the thread on UKVR:
https://www.vintage-radio.net/forum/showthread.php?t=181388
There is also one thread on this forum:
https://www.radios-tv.co.uk/community/general-discussion/avo-test-bridge-repair/
I haven't been on here in a very long time, so long in fact that the site URL seems to have changed and I have had to re-register! Anyway, I thought I would start a thread on here for a change.
This unit had no paper waxies present, so it seems they had been replaced previously. The C5 replacement was the wrong value so the 20nF part was replaced with a 220nF one rated at 400V. A number of resistors that had gone high or were borderline on tolerance (R7, R9, R11) have also been replaced. R8 also reads high (105kΩ) is due to be replaced using a 120kΩ with a 470kΩ in series which equates to 95.5Ω, however, I would like to solve the other problems first.
The selenium rectifier (W1) is still in place and seems to be working just fine.
The instrument measures fine on the C2 and C3 ranges as well as R1 through R3. The 'L' (capacitor leakage) mode also seems to work. So far this has been tested only with resistors up to 22MΩ in value at which point it just barely lights up. R7 had gone high and has been replaced. I haven't yet tested the Ext mode.
The problems still to be solved are these:
1. Readings on the C1 range are a long way off being correct. A 100pF mica capacitor reads '2' or 200pF. On my DE-5000 it reads 100.01. I have checked other caps in the pF range and all readings are way off the mark. I disconnected the common end of C1/C8 and checked on the DE-5000. The reading was 110pF. I adjusted C8 until it read 100.2pF, however, the 100pF capacitor still reads '2'. Might the reading be skewed by item 2 ?
2. In voltmeter mode, the movement is deflected just over halfway without any input being present. Shorting 'Ext' to the earth terminal drops this to zero. There is approximately 0.5Vdc on the grid (point marked TC) and 13V on pin 5, the output to R8, the series resistor to the movement. With the valve removed those voltages go away so might there be some leakage on the 6B8G?
3. Like the author of the second thread above, since the case is metal, I would like to replace the power cord with a three core lead and earth the case. Should "chassis earth" point be bonded to this mains earth or should it be left floating? Since the "chassis earth" point is the transformer core and this is mounted on the bakelite front panel, there is currently no electrical connection between transformer iron and case. I think I know the answer, but would like to be sure before doing anything that might impact safety.
This unit also has an interesting mod. The main dial potentiometer (R6) has a 1Ω "preset" either side and the potentiometer itself has been modified. You can see the gap and additional wire connected on the LHS of the winding in the photo. I have been told on UKVR that my unit is quite a late one (June 47) but this mod has not been seen in other units. if anyone has any ideas why it might have been done I would be interested to know.
Hi Wavey, and a warm welcome back, though wish it were under better circumstances. I'm afraid you will find this forum a pale shadow of its former self. Circumstances now are that we have slightly less than two handfuls of regular members visiting and posting these days.
Personally, it is appreciated that you took the time to rejoin. However, if you're seeking advice to resolve problems with this device, you will find that persevering with that thread on UKVRRR, or perhaps starting a new one if you're concerned about hijacking.
I only mention this, so you do not think the lack of response to your query here is due to no interest or undue rudeness. Sadly, these are the circumstances VRAT finds itself in these days.
CrustyTV Television Shop: Take a virtual tour
Crusty's TV/VCR Collection: View my collection
Crustys Youtube Channel: My stuff
Crusty's 70s Lounge: Take a peek
Hi Wavey, Looking at the schematic, I was wondering if both earth connections on the 50V winding via R5 - 350Ω should be there. Comparing with the original AVO circuit neither input connections are grounded.
The later AVO Universal Bridge has a three core mains lead with earth connected to chassis. Topology is similar but a twin triode is used instead. Not clear if the core is earthed and it may be for reasons of dealing with stray capacitance etc which may affect the balance when measuring low capacitance.
Rich
Let's assume just for a moment that the valve is faulty. If it is gassy or leaky in any way, then the more sensitive ranges will be more likely to be indicated wrongly? If it were me, I would like to re-check with a known good valve.
As Rich pointed out above, I can't see any way it could work with both the points grounded. I dare say, with due respect to the person who drew the schematic, that this is an error.
With sensitive capacitance readings, as you may well know, the layout of wiring and indeed the internal capacitance of the valve will have a marked effect.
I must confess to having no experience with this model, so am simply offering a personal opinion based on what we have. A better mind than mine may be required to explore it if problems still exist after replacing the valve, and the circuit is all correctly made. 😉
To understand the black art of electronics is to understand witchcraft. Andrew.
Not sure if it helps as there are mods to this unit, but the original AVO manual is in the library.
CrustyTV Television Shop: Take a virtual tour
Crusty's TV/VCR Collection: View my collection
Crustys Youtube Channel: My stuff
Crusty's 70s Lounge: Take a peek
Thank you for the comments! There seem to be only two conditions when one of those terminals is connected to chassis ground:
- main dial set to the 'V' position, in which case the terminal marked Earth on the Ext pair is earthed.
- main dial set to the 'L' position, in which case the terminal marked Earth on the pair for connecting the component under test is earthed. The opposite side of the component is then connected in series with the neon bulb and a limiting resistor to V+.
In any other position (i.e. reading the dial value), I agree with the above. neither terminal could be earthed to chassis ground, and certainly not both at the same time! The terminals in this case become bridge connection points.
Its a clever arrangement to provide multiple functions and probably explains why those terminals are marked with an Earth symbol, but somewhat confusing because they have a dual function depending on the mode selected.
Marconi_MPT4, thanks for the update on the later version mains lead arrangement. On this unit, the chassis earth rail on the voltmeter board is connected via a mounting screw to the metal transformer mounting bracket. On another part of the mounting bracket there is another common "chassis earth" tapping point that connects several wires running to different parts of the instrument. To all intents and purposes, the transformer mounting bracket and core are the "chassis". With the instrument having a twin cord, and the "chassis earth" being on the transformer secondary side, it is currently a "floating" common point connected to one end of the secondary winding. The metal case has no connection to this "chassis earth" point so is effectively isolated.
So the question is, if the twin core cord were to be replaced with a three core one, should the currently floating "chassis earth" be connected to mains Earth? Should the metal case be connected to mains Earth as well, thereby making the case common with "chassis earth" as well? Or is it best to leave the metal case isolated as at present, and retain the original twin core cable arrangement? Generally any metal casing on a mains appliance is earthed for safety reasons so I just want to confirm what the Earth arrangement should be on this unit in order to meet modern safety standards.
The replcement valve arrived today, a military grade 6B8G (CV1893). I plugged it in and powered on, but unfortunately, the meter is exhibiting the same behaviour. It is still showing over halfway deflection on the voltmeter mode setting. 🙁
I don't mind having a spare valve, but am very puzzled as to where the voltage in question is coming from.
Possible leakage in and around the bridge circuit?
Might be worth temporarily disconnecting C6 to confirm if meter indication drops to near zero.
Interesting to note that when in voltmeter mode R6 still has AC current flowing through it which appears in series with chassis ground and the input socket “earth”. Maybe to use voltmeter mode R6 wiper should be set to end marked ‘V’ as shown in the diagram.
If R6 has already been set to 'V' but the errant indication has returned when reconnecting C6, either it is just pickup of noise coupled through wiring or possibly leakage across the selector switch. Sometimes if switch cleaner has been used in the past it might lead to insulation leakage. This might not be of consequence when connected to an actual circuit.
Rich
So the question is, if the twin core cord were to be replaced with a three core one, should the currently floating "chassis earth" be connected to mains Earth? Should the metal case be connected to mains Earth as well, thereby making the case common with "chassis earth" as well? Or is it best to leave the metal case isolated as at present, and retain the original twin core cable arrangement? Generally any metal casing on a mains appliance is earthed for safety reasons so I just want to confirm what the Earth arrangement should be on this unit in order to meet modern safety standards.
I decided to retrieve the AVO Universal Bridge and check which parts are earthed. For this model the metal case, transformer core and power supply secondaries are all bonded to mains earth as shown in attachment. Confirmed by measurement.
However on browsing through the user manual a paragraph pertaining to measurement error and polarity of the mains input appears to be relevant due to both models using a similar topology. Extract from AVO manual.
"It is essential that this instrument is correctly connected to the mains, in so far as phase and neutral are concerned, and to this end, the mains wire is colour-coded according to the standard convention, the yellow or green lead being earthed or connected to the earth socket of the mains plug. Users should ensure that they affix the mains plug so that correct connections are maintained at their output plug.
The effect of transfer capacity in the transformer is to give a minimum reading (approximately 12/15 pF) on the lowest range (see capacity testing). A reversal of the mains connections may cause this figure to be considerably increased."
In the light of this information I would proceed carefully in earthing the test bridge and evaluate different configurations. There will always be some extra transfer capacity to cause bridge imbalance despite the 50V bridge winding being isolated on a separate former.
Rich
Rich, some interesting observations. Your unit seems to be different internally to mine although the operating principle is no doubt the same. I appreciate you taking the time and trouble to test the earthing arrangements. Unfortunately I have been ill and in bed for the last couple of days so in no state to do anything further at present. It will have to wait until I get better but I will have a look at the points you mention.
Your unit seems to be different internally to mine although the operating principle is no doubt the same.
It's a completely different model: the one illustrated is the successor to yours in a larger grey wrinkle-finish case, with switchable voltages to test leakage in caps &c. The case style matches the valve voltmeter and signal generator that AVO built in the early 1950s.
Mod edit:
Added a photo to help illustrate Catho's post
Now that I am finally feeling a bit better, I tried reversing the mains wires to see whether it would make any difference. Unfortunately it did not. I have also now replaced the smoothing capacitor and selenium rectifier as well as swapping the 100k resistor for a 120k to offset the reduced voltage drop of the replacement silicon diode. Unfortunately none of these changes have made the slightest bit of difference.
Disconnecting C6 from S2 as well as S1 common does indeed cause the meter reading to drop to zero. Connecting S2 causes some deflection. Adding S1 increases the deflection further so both are contributing to the deflection in some way.
Glad you are feeling better.
Disconnecting C6 illustrates the vacuum tube voltmeter section of the bridge is capable of reading zero with no signal and for now can be assumed to work correctly. The HT supply only feeds the VTVM and the capacitor leakage test neon.
With C6 reconnected and the meter deflection still present, does it disappear when R7 in series with the neon is temporarily disconnected with R6 turned to one end of its travel to ground R5,R6 junction, shown as position 'V' on the diagram?
Rich
Temporarily disconnecting R7 does not seem to make any difference. With R6 turned to R5/R6 junction in position 'V' there is still exactly the same deflection as previously. I also checked resistance between chassis ground and jct R5/R6 and got a reading of approximately 0.21Ω of which the meter leads account for 0.17Ω so negligible.
I have now also replaced the two-core lead with a 3-core one and connected earth to chassis-ground. That reduces the deflection from about 1V to around 0.5V. Isn't the input to the valve Hi-Z? If so, then could the deflection be simply showing a residual voltage, similar to what one gets on a DMM?
I did some tests with a signal generator, generating a sine wave at 100hz. I tried in 1Vrms (1.41Vp) steps up to 6Vrms (8.46Vp) and the meter read each step correctly up to 6Vrms. I can therefore confirm that aside from the 0.5V deflection, the voltmeter measures correctly and seems to be quite accurate.
The bridge also works quite accurately on all C2 and C3 capacitance and R1 - R3 resistance ranges, but the C1 range is rather inaccurate and can be up to 20% off the mark at the top end (e.g. 10nF reads 8nF). It seems to vary a little depending on whether the metal case is present or not. Overall, its not bad for an instrument that is around 75 years old!
