Re-calibrating a AVO Model 8 Mk5

[Specmaster]

Hello I'm trying to recalibrate another meter for my collection, its An Avo Model 8 Mk5 and the voltage ranges are reading which means that I'm looking for a source of replacement resistors for it. The manual just lists the values and the Avo part numbers and thats it,

I suspect that these would be wire wound, maybe something like 5w each and 1% to 0.3% accuracy. They are the ones located on the top board in the photo, blue in colour and the only current info I have is taken from the service manual.

Resistor Value and circuit ref.    Avo Part number
8M (R23)                                6120-281
6M (R22)                                6120-280
400K (R18)                             6120-275
4M (R21)                                6120-279
800K (R20)                             6120-278

Does anyone know where I can obtain these from or have any that they are willing to part with?

[Edwin G. Pettis]

These are metal film resistors not wire wound and of various wattages.  Have you tried to measure them as it is rather unlikely that all of them have gone bad (usually some signs of over heating for example).  Most of them are not off the shelf values but could be built up from available values.  That appears to be a 40M in the photo (unless the decimal point is not visible).  Check around eBay and surplus parts on line, you might be able to find similar parts.  I would eliminate other possibilities of failure before replacing these resistors, as dividers they are in series with the input, put the meter on the lowest voltage input range, if there is no voltage reading on the meter, look for a component in series with the input that has failed.  You can check with another VOM for voltage and follow it through towards the meter circuit.  The biggest resistor (physically) is probably 4 to 5 watts, the smaller ones progressively less wattage, such as 3, 2 or 1 watt for the smallest.

[Jay_Diddy_B]

Hi,

Have you checked the 50uA and 100mV ranges?

These will tell you the condition of the meter movement.

From the service manual:




On the other ranges there should be a voltage proportional to the meter deflection on the 100mV test points:




For example if you put in 5V on the 10V range, there should be 50mV on these test points.

I would check this before changing resistors.

Regards,
Jay_Diddy_B

[Specmaster]

This is only applicable to the later version which is what I have, but I cannot see where the test point is located, I presume that this is accessible somewhere on the circuit board without having to dismantle anything?

[Jay_Diddy_B]

Hi,

I would start by checking the AVO 8 on the 50uA DC range. This is pretty much a direct connection to the meter movement. If the 50uA range is bad it is could be the movement.

Jay_Diddy_B

[tautech]

Have you checked the trimmer on switch position 12 as all the R values you've listed are series trimmed with it.
Depending on meter version the trimmer is 15 or 22K.

Manual attached.

[Specmaster]

Have you checked the trimmer on switch position 12 as all the R values you've listed are series trimmed with it.
Depending on meter version the trimmer is 15 or 22K.

Manual attached.

The only trimmer on switch position 12 is the zero ohms adjuster pot for the x1 resistance range, the other 2 22K trimmers are the low and high resistance ranges which all measure on known value test resistors just fine.

[tautech]

Have you checked the trimmer on switch position 12 as all the R values you've listed are series trimmed with it.
Depending on meter version the trimmer is 15 or 22K.

Manual attached.

The only trimmer on switch position 12 is the zero ohms adjuster pot for the x1 resistance range, the other 2 22K trimmers are the low and high resistance ranges which all measure on known value test resistors just fine.

Yeah OK sorry I got that way wrong. 

About the resistor ratings, there's a key top left on the schematic page with value/tolerance/Watts and how to interpret the values next to each resistor in the schematic.
Eg R24 40M 0.5% 2W

I have a AVO 8 MkII with a nackered scale that you can have the guts out of if they're any good to you.
Dunno until I open it if it has the same values but somehow I doubt it as they only went to 2.5KV not 3KV like the Mk5.

[Specmaster]

Hi,

I would start by checking the AVO 8 on the 50uA DC range. This is pretty much a direct connection to the meter movement. If the 50uA range is bad it is could be the movement.

Jay_Diddy_B

Checked the meter with a 50uA supply, gives full scale deflection, so movement is OK. Problem has to somewhere external to the movement.

[Specmaster]

Have you checked the trimmer on switch position 12 as all the R values you've listed are series trimmed with it.
Depending on meter version the trimmer is 15 or 22K.

Manual attached.

The only trimmer on switch position 12 is the zero ohms adjuster pot for the x1 resistance range, the other 2 22K trimmers are the low and high resistance ranges which all measure on known value test resistors just fine.

Yeah OK sorry I got that way wrong. 

About the resistor ratings, there's a key top left on the schematic page with value/tolerance/Watts and how to interpret the values next to each resistor in the schematic.
Eg R24 40M 0.5% 2W

I have a AVO 8 MkII with a nackered scale that you can have the guts out of if they're any good to you.
Dunno until I open it if it has the same values but somehow I doubt it as they only went to 2.5KV not 3KV like the Mk5.

Thanks for the kind offer if I need it but the shipping costs would be more than the meters really worth TBH. I just concocted myself a quick 50uA supply and the meter reads OK so that clears the movement.

[Jay_Diddy_B]

Hi,
Since the meter movement is okay, how far off are the readings on the ranges that are bad?

Have you checked the values the resistors with a DMM?


The AVO is 20k per volt. If you start on the 3V range and measure the input resistance with a DMM it should read 60k .

The 10V range should read 200k

30V range 600k .

You should be able to identify the bad resistor or resistors using this method.

Regards,
Jay_Diddy_B

[Specmaster]

I had already done this and came up with some odd values, these values also changed depending on if you were switching up the range or down. I also compared them against another Avo with the same ranges but was a Model 9 Mk2 whose values were as predicted and the meter reads correctly.

I have now gone and pulled the resistor board from the meter, I don't like those ribbon cables, horrid things, and just hoping that though go back again with no problems 

All the resistors tested and all seem to be in spec, the only one to was out was the 800K which was 810K so I'll replace that one.

[drjazz]

If all DC voltages are reading high , on 10V range  5vdc measures about 7v then one of the current shunt resistors are open circuit.

[Specmaster]

DC volts on 3, 10 and 30V ranges are OK, 100v and above are affected and also AC on same ranges. I have ordered a NOS identical resistor, even the same red colour as a replacement for the 800K so will see what happens when that is fitted.

[Specmaster]

OK then, my NOS 800K identical resistors to the original ones arrived today, marked as 0.3%, looked as if they had just left the factory TBH. Tested them before installing only to be disappointed, 1 was @ 3.2% and the other was 1.2%, far worse the one already fitted.

So now I still need the following resistors in order to bring the meter back into specification.

6Meg 0.3% 0.5W,
8Meg, 0.3% 0.5W,
4Meg 0.3% 0.5W
140K 0.3% 0.3W
800K 0.3% 0.5W

Or of course better % would be great, so if anyone knows where these can be obtained from, or has any of them, please let me know.

[Specmaster]

Quick update, traced the original manufacturers of these resistors and had a chat with them about replacements. Nothing doing, some were not being made any more and the 800K was a complete special and made in lots of 1,000 to order, err no thanks.

They however suggest a stockist that may have some other required values and the only one that they could with was the 140K but with a MOQ of 5 and after the P&P and VAT was added it would have come to £42, so thanks but no thanks FFS, I only paid £15 for meter 

Yesterday I revisited the project and did some calculations and the worst 2 resistors, 140K and 800K were also still in play on the other 3 upper voltage ranges so any errors on those first 2 would be multiplied on the higher voltage readings. These 2 were mission critical in that case, so set about trying to resolve them and, seeing as the values had drifted upwards, it meant that I could tack on some smaller wattage resistors in parallel to "trim them back into spec" and it works out that I need a 20M .25W 1% resistor to be shunted across the 141.2K to pull it back to 140K and likewise a 50M .25W 1% resistor shunted across the 811K to trim it to 800K again. These have been ordered and, will hopefully be here this week then we can fit them and see if the results are acceptable. Watch this space for further updates.

[George Edmonds]

Hi

I don't want to rain on you, but, the resistor rift with time and voltage will continue, always upward.  Please take note of the maximum applied voltage rating for any resistors that you use.

George G6HIG in Dover

[Specmaster]

Yes, I understand this will continue now that it has started, but I hopeful of finding another meter with a stuffed movement etc in the future to become a potential donor of those resistors, fingers crossed. I'm aware that they can be found on eBay but currently they are going for stupid money, even as parts only. So for in a year I have gathered 3 Avo's for approx £15 each, a Model 8 MK4 and a Nato one that are perfect but have always wanted a MKV, so, when I got this one I was disappointed to note that the values on certain resistors had drifted that much.

[George Edmonds]

Hi

I have a scrap Mk5, don't know just what condition it is in.  It is ex BT and it rattles with the case halves taped together.  I know a number ex AVO/Megger employees, they would not give a Mk5 house room.

George G6HIG in Dover

[Specmaster]

Hi

I have a scrap Mk5, don't know just what condition it is in.  It is ex BT and it rattles with the case halves taped together.  I know a number ex AVO/Megger employees, they would not give a Mk5 house room.

George G6HIG in Dover

George, I can fully understand their feelings. It is in real terms a step backwards for Avo but people were calling out for a redesign, something a bit modern and this was the result, it does look more upto date than previous models, but quality took a heavy hit in my view. It is both lighter overall and you see and feel where cost-cutting has taken place. Just the feel of the switches alone give the game away, in some positions it is possible to leave the switch midway between ranges which leaves both ranges connected at the same time and this might be a possible cause of the resistors drifting?

How much would you be looking for to part with your Mk5, the resistors on it may be to spec?

[George Edmonds]

Hi

I wanted to have time to take a look at my scrap AVO 8 MK5 before replying to you.  It is ex PO TEL, it has a burn out track on one of the printed ribbon cables and has been partial dismantled (got at) internally, it appears to be all there apart from the battery compartment cover.  Movement appears good.

I have measured only the 800K resistor with a Fluke 2000 DMM (6.5 Digits) and found it to be some 813K, some of the other resistors I cannot get easy access to so as to measure them.  It looks like the “special” 800K resistors made for AVO all have a problem with age related drift, even when not used.  If you really want 800K look on eBay, no shortage of 400K at 0.1% and in a voltage divider it is better to use two not one as this reduces the applied voltage across them, one of the main reasons resistors go high.

Now if only the 800K is in error by +13K the meter should on the 100VDC range read 0.65% low at full scale, you will be hard pushed to see this given that the best accuracy for a brand new AVO 8 Mk5 is 1% at full scale.  Calculated using +13K as a percentage of 2Meg, the total divider chain resistance on the 100VDC range.

Your postings leave a number of unanswered questions, for example what is the accuracy of your applied test voltages, and what are/were the actual meter readings obtained from them?

In one of your postings you say that when measuring the total external resistance of the AVO 8 Mk5, on say the 100VDC range when it should be 2Meg, the resistance varies by which direction you rotate the switches, this points to a problem with the switches or the internal wiring, certainly not the fixed resistors.

The 100mv test points are on the polarity reverse switch, checking the accuracy of the meter movement is the first step, before anything else, all moving coil meters lose sensitivity with time due to the degradation of their magnetic field.

One of your photographs appears to show that the resistor PCB has been subjected to long term heat around the area of the 800K resistor, caused by the AC current transformer?

George G6HIG Dover in the UK.

[Specmaster]

Hi

Your 800K is worse than mine then which is 811K and also measured on DMM @ 6.5 digits and also cross-checked with 6 others to make sure I was not being lied to by the meter, they all agreed at 811K.

I have checked all the resistors in the meter and the others, not already identified in the photo are within their specs. When in the 100V range, which is when the 800K is switched, it is in series with the 600K (600Vac), 400K (30Vdc), 140K (10Vdc) and 58K (3Vdc) making 1.998M which is switched into the meter circuit when in the 100VDC position.

I have pulled apart the rest of the meter as far as I can get it, without risking breaking it and as far as I can see everything seems to be fine. I have double-checked all solder joints and reflowed them, given all the switch contacts a good cleaning with switch cleaner etc and so far these resistors are the only smoking gun I've come across.

There has been no excessive heat within the meter as this new photo of the PCB shows, any discolouration you saw on the previous photo is purely down to the light reflections and shadows on the board. The board will be given a good scrub all over with IPA before being put back in the meter to remove any grease and other contaminates.

The 3V and 10VDC ranges read OK, as does the resistance range, the manual also says something about errors on particular voltage ranges are caused by the preceding resistors.

[Specmaster]

Update time, trimmed the first resistors (800K and 140K), still reading too high. If I give it a 100Vdc it goes off the scale and with a DMM (Hi-Z) across the test point and 100mV selected manually, a 100V input gives 111mV. If I use 100V on the 300Vdc range it gives me around 42mV  So the problem is definitely within the voltage divider chain somewhere. It's as if resistors are altering their values dependent on the voltage applied  .

The meter is perfectly linear I have checked it using my decade box on resistance and on each range the meter responds 100% correctly, 10's, 100's, 1,000's, 10,000's and Megohms, as I add another on the decade box, the meter reads it 100% dead on.

Stripped the switching PCB out and checked the switch contacts, on both sides, not an issue there, checked the bottom switch contacts, on the largest of the 2 ribbon cables, all OK, checked the ribbon cable, all OK. I have to go out for a while and when I get back I'm going to give the ribbon cable a good clean with IPA.

The printed shunt resistor does not read as per drawing but its with in an ohm. The printing does have a few cuts in it here and there but I think they are intentional where they fine tune it on final calibration. Apart from that I'm stumped 

[Specmaster]

Ok then, after more playing around I have come to the conclusion that all the resistors on the resistor board need to be evicted and replaced. I'm not too bothered about the 4M for the 3,000Vac or the 40M for 3,000Vdc as I'm not very likely to need to measure anything that high with it anyway. So its either going to be a donor meter or new resistors, which has already been discovered is a no-go.

[George Edmonds]

Hi

Now we have some real numbers to work from what I can tell you is that you appear to be barking up the wrong tree with the series resistors.

If for example the 800K resistor goes high the AVO 8 will read low not high on the 100VDC range.  In my opinion the problem is NOT in the voltage divider chain.  Yes resistors do change resistance with applied voltage across them, but. in some sixty years I have only ever seen an increase in resistance, never a decrease, an increase in resistance will have exactly the opposite effect to what you are seeing.

All of what I am saying is based on the assumption that this AVO 8 Mk5 has not been contaminated internally with a conductive substance.

Given that the AVO 8 is 20K Ohm per volt its meter movement must be 50uA full scale, no ifs no buts.  The basic meter movement is 37.5uA full scale in addition 12.5uA is passed in the parallel printed shunt resistor array making a total of 50uA.  This is may not very clear from the circuit diagram, possibly there is something wrong with either the basic meter movement sensitivity (unlikely as moving coil meters lose sensitivity with time not gain it as in your case) or the printed shunt resistors and its switching has a problem.

I believe that you have a known good AVO, put both in series on their 50uA ranges and pass 50uA through them both.  Simply done with a 1.5VDC battery and some appropriate resistors totalling around approximately 33K or slightly more for a new battery.

My guess is that when the reference AVO shows 50uA full scale your faulty one will be off scale.

Now 0.75 of the 50uA passes through the moving coil meter and 0.25 of the 50uA passes though the printed shunt.  The meter is trimmed to be exactly 2667R by R27 (a small white object) which is mounted on the Reverse Polarity switch (the moving coil meter and R27 MUST be kept together as a calibrated pair).  The total resistance of the printed shunt is 8001R which allows 0.25 of the 50uA to pass through it.  The entire printed shunt is put in parallel with the moving coil meter via switch contacts 1 plus 4 and 2 plus 3.

Yes the printed shunt resistor is trimmed by what appear to be cuts in it.  If you do try to make resistance measurements around the moving coil meter plus printed shunt do preferably not do it with an auto ranging meter as this can cause the moving coil meter to pulse and negate the readings.  Just to help you I measured the DC resistance between the plus and minus terminals on my AVO 8 Mk5 set for 50uA DC full scale. I got a measured 2012.5 Ohms against a calculated 2002 Ohms, this measurement was done using a 2 wire resistance setup without any great care being taken to null out the lead and contact resistance.

I also checked it on the 30VDC range and found it to be reading about 3% low, not surprised at this, like wise it was about 3% low on the 100VDC range, again not surprised

UNTIL YOU HAVE THE AVO 8 Mk5 50uA DC BASIC RANGE CORRECT YOU ARE WASTING YOUR TIME TRYING TO CORRECT ANY OTHER POSSIBLE FAULT AND POSSIBLY DOING DAMAGE TO THE REST OF THE METER IN THE PROCESS.

Good luck with it.  Use a lump hammer only as a last resort.

George G6HIG Dover in the UK.