|
Hints, Tips and Solutions (Do NOT post requests for help here) If you have any useful general hints and tips for vintage technology repair and restoration, please share them here. PLEASE DO NOT POST REQUESTS FOR HELP HERE! |
|
Thread Tools |
16th Feb 2010, 12:19 pm | #1 |
Nonode
Join Date: Mar 2008
Location: Virginia Water, Surrey, UK.
Posts: 2,880
|
10p fix for low-reading Avo Valve Tester meters
10p worth of carefully selected resistors can sort out an aging VCM meter which is reading low. These meters are "unobtainium" and at 30uA sensitivity cannot easily be replaced. You might fiddle with magnetic shunts but I think this is more elegant.
(If you don't want the workings, skip to the attached spreadsheet which does it all for you!) In my case (see other threads), all calibration of Vg and gm worked fine, but the end result was low readings on everything, in my case by almost exactly 10%. (You check by comparing indicated to actual anode currents using an Avo on current range across an opened A1 shorting link. The Avo should read exactly 50% of indicated current on a MK4) So I had a hard look at the circuit and realised that, unlike a normal Avo, which has varying shunts, a VCM (Mk3, Mk4 or CT160) always has a fixed 10K shunt across the meter. And Avo always have a small swamp resistor within the meter housing to allow for variation and help with temperature compensation. The meter is nominally 30uA, in my case having a resistance of about 2707 ohms in series with a 543 ohm swamp resistor (actually a small coil). With a 10K shunt this gives an fsd of 39.75uA and a resistance of 2453 ohms. I measured the total meter resistance very carefully with a DMM and then equally carefully took the meter faceplate off and measured the swamp resistance. Then I checked the fsd with an Avo in series with a couple of other (digital) meters as checks and got a reading of just about 33uA. If I could persuade the current in the meter leg of the pair to increase to the actual meter sensitivity, with a corresponding diminution in the shunt current, that would sort out the sensitivity problem. BUT, I must also make sure that the total resistance of the meter plus shunt remains the same, or the external circuitry (multipliers etc) will be upset. This took me back to school! We now have a classic two variable, two constraint problem: So now came the "back to school" bit - simultaneous equations with two variables - new swamp resistor, new shunt resistor. 1) (Meter resistance + new swamp)/New shunt must = meter fsd / remaining shunted current (eg 39.75 - meter fsd) [to get correct new current ratio] 2) (Meter resistance + new swamp) paralleled by new shunt must still = 2450 ohms [to keep total resistance constant] Suffice it to say that there was a lot of head scratching and pencilled sheets of workings before I got a solution. And I had to try to remember the way of solving the quadratic equation you get at the end. Anyway, I realised it was eminently workable, and set about making a spreadsheet that would do the work for my in any situation, compete with working out suitable parallel resistor combinations to make up the inevitably weird values I needed. I would point out that the dis-assembly of the meter housing must be done very delicately. Use a very clean space, don't let magnetised tools clang against the meter magnet, don't let anything touch the meter pointer. And I strongly suggest leaving the swamp in place and just shunting it with a suitable resistor. Keep the meter terminals shorted whenever possible. And remember on re-installation to re-tighten the retaining bolts just enough to keep the meter secure - no more, or you can crack the meter housing. In a CT160 the shunt resistor is mounted on the rear of the paxolin meter bracket so is easily accessible. For VCMs like my Mk4, to locate the correct resistor, look for the wires coming back from the meter to extra switch wafers on the circuit selector switch. On My Mk4 they are blue and green. The easiest method is to cut one lead of the existing shunt resistor near the tag and solder in an additional resistor in series. The end result is pleasingly exactly as predicted. The meter reads correctly now, which means that both Ia and gm are reading correctly. The only minor change is that the meter is slightly less well damped as a result of decreased swamp and increased shunt. But overall a good, cheap and to my mind elegant solution, which brings a classic piece of test equipment back into original spec. (Spreadsheet attached in next post) Regards
__________________
Jeremy, G8MLK, BVWTVM Friend, VMARS, BVWS Secretary. www.pamphonic.co.uk www.bttt.org.uk |
16th Feb 2010, 12:21 pm | #2 |
Nonode
Join Date: Mar 2008
Location: Virginia Water, Surrey, UK.
Posts: 2,880
|
Re: 10p fix for low-reading Avo Valve Tester meters
spreadsheet....
and by the way, this will cope with meters that are down by up to 25% or so. It is not a magic bullet to allow 50uA meters to be used! You still need a total fsd of just under 40uA.
__________________
Jeremy, G8MLK, BVWTVM Friend, VMARS, BVWS Secretary. www.pamphonic.co.uk www.bttt.org.uk |
16th Feb 2010, 12:54 pm | #3 |
Nonode
Join Date: Mar 2008
Location: Virginia Water, Surrey, UK.
Posts: 2,880
|
Re: 10p fix for low-reading Avo Valve Tester meters
If it helps, here are the main solutions based on my meter. They should be reasonably universal:
R6 is shunted across the meter swamp resistor R7 is inserted in series with the meter shunt (10K) fsd: 31, 32, 33, 34, 35, 36 uA R6: 2k7//15k, 1k//10k, 470//15k, 270//1k5, 100//1k2, s/c (zero). R7: 1k2//27k, 2k7//68k, 4k7//82k, 8k2//47k, 12k//100k, 18k//150k There are no other solutions - if your meter is worse than 36uA fsd you need a new meter, or use the op-amp solution described on other threads. At higher meter fsds the swamping will be very loose and the meter will take a while to settle. Get the most accurate lower-value resistor you can. Use a DMM. The higher resistor of the pair is of course less sensitive.
__________________
Jeremy, G8MLK, BVWTVM Friend, VMARS, BVWS Secretary. www.pamphonic.co.uk www.bttt.org.uk |