UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Test Gear and Workshop Equipment

Notices

Vintage Test Gear and Workshop Equipment For discussions about vintage test gear and workshop equipment such as coil winders.

Closed Thread
 
Thread Tools
Old 5th Sep 2021, 3:38 pm   #1
6SN7WGTB
Hexode
 
Join Date: Apr 2021
Location: Crawley, West Sussex, UK.
Posts: 442
Default VTVM - appears to under-read by ca. 10% on AC and DC

Hello All,

I have a Taylor 172A VTVM, schematic below.

I'm running into an issue that I'm struggling to resolve, so really would appreciate some help. I've done my best to diagnose but am reluctant to chop things around without a good reason/understanding.

Before I restored the device I made a 1M DC probe - as this was absent and is required.

I've now restored the device, and capacitors renewed (except 3300pf grid caps - grey hollow cylinders with values printed on), and all components checked for within tolerance.

On the latter, the panclimatic/ladder resistors were spot on, and any other resistors appears within 20% max. Both grid caps measured ca. 2400pF. It's engine out to replace so although I have them, I haven't swapped them out yet.

The sitrep is:
1. On DC, it under-reads by 10% with the probe! But is as good as spot on without. It calibrates OK at the spec of 50V DC, on both V- and V+ without probe, but with 1M probe the calibrations (V+ and V-) run out of travel.
2. On AC it under-reads by 10% without a 1M probe - calibration again out of travel. This was with a 50Hz source.
3. On ohms it is as good as spot on on all ranges
4. Meter Zero and Ohms Adjust (FSD) are both about centre of rotation.
5. Mech zero OK.

This all smells like some-one had badly modded it to embody the DC 1M inside.

However, I cannot see any sign at all of any modification. There is certainly no 1M resistor anywhere and all the other components are original and match the parts list.

I have done the following:

CHECKED and/or CHANGED
(a) Been through every component to see if it within spec (as noted above)
(b) The only ones I question are the two 3300pF grid caps which I did not change - as noted above, but reluctant to take it all to bits without some more knowledgeable suggestions/confirmation
(c) Checked triode heater balance (there's an adjustment) - and I'm getting 5.95 V and 6.05 V so close enough I'd say. Adjusting it doesn't fix the problem.
(d) Tried a new triode (ECC82/12AU7) - no change (didn't think it would help)
(e) Tried two new rectifier tubes (6AL5) - no change (ditto). Note I used CV4025s as replacements temporarily as they won't clip in as the originals are EAA91 so shorties.

CALIBRATION
(a) Calibrated it per manufacturer instructions.
(b) As stated above, DC is fine without 1M probe - calibrates on V+ and V- in central area of slider. Note the amount of travel on the DC adjustment pots suggests they only allow a small adjustment - not anything like 10%
(c) AC: Now this has not been perfect as the AC calibration (R21 2M pot) appears (just) not to allow a needle zero when the leads are shorted on the lowest AC range (1.5V) - it's almost out of range which suggests something wrong. Then when you switch AC ranges the meter zero moves considerably off of zero which it is not supposed to. Then you go to the normal Zero and adjust it, but then it's now a loop of AC zero negating meter Zero.

UNVERIFIED/NOTED
The one variation between schematic and meter is around the R4/R5, R7/8 and R9/10 area, where the actual meter has a slightly different arrangement:
(1) The line from R14 splits into two ca. 4k resistors, one feeding R5 and R10 (the V+ and V- calibration pots), and one feeding R8 (the AC calib pot).
(2) So, in the meter R4,7,9 which are on the schematic do not exist.
(3) Oddly, or maybe commensurately, they are NOT shown in the parts list and...
(4) The ca. 4k resistors are not shown either. The calib pots are ca. 2k

What have I missed?
(i) Is it the grid caps, if so, I'm a fool for not swapping them...
(ii) Could the movement be somehow to blame? Feels unlikely as a 10% 'mechanical offset' is hard to understand as it mechanically and electrically appears to zero OK.
(iii) Is there an issue in the calibration network?
(iV) I haven't touched the AC freq calibration trimmer - surely can't be this? (Calibration is done at 100kHz)

ATTACHMENTS
i. The schematic
ii. The calibration board - showing the wirewound VRs and what look like w/w 4k resistors as explained. VRs are 2k spec and appear OK.

I'm flummoxed. The meter is mint and so so very nearly working correctly.

Any assistance very much appreciated.
Attached Thumbnails
Click image for larger version

Name:	IMG_8459D.jpg
Views:	140
Size:	55.3 KB
ID:	240725  
Attached Files
File Type: pdf Taylor 172A schematic.pdf (632.1 KB, 124 views)
6SN7WGTB is offline  
Old 5th Sep 2021, 5:46 pm   #2
Chris55000
Nonode
 
Join Date: Aug 2007
Location: Walsall Wood, Aldridge, Walsall, UK.
Posts: 2,853
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Hi!

Can you scan/post the rest of the manual for this if you have it please?

I can certainly offer some further advice on fault–finding it, but an anonymous circuit diagram with no component– values doesn't help much I'm afraid!

Chris Williams
__________________
It's an enigma, that's what it is! This thing's not fixed because it doesn't want to be fixed!
Chris55000 is offline  
Old 5th Sep 2021, 6:59 pm   #3
6SN7WGTB
Hexode
 
Join Date: Apr 2021
Location: Crawley, West Sussex, UK.
Posts: 442
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Chris,

Yes, I appreciate that. I tried to give the component values that seemed most relevant.

Thank you very much for your offer.

The manual is unfortunately protected as I had to pay for it, and so wholesale uploading isn't something I feel comfortable about doing.

I have attached the parts list.

Calibration is as follows:

[MY ANNOTATIONS IN BOLD]


4/2. VALVE REPLACEMENT
This is unlikely to be necessary for avery long time due to the conditions under which the valves operate.

Before leaving the works every instrument is given athorough soak test to ensure that it is in athoroughly reliable condition. This also serves to “age” the valves and other components and results in amore reliable and consistent performance. It is strongly recommended that any valve replacements which may become necessary are thoroughly “aged” before use, either by running them continuously for at least 24 hours in the instrument or by operating them under full load conditions for at least 12 hours before fitting to the instrument.

V.2, the double triode (type 12AU7), may need replacement if it becomes impossible to set up the instrument correctly or to adjust zero on all ranges.

V.1 (type 6AL5) may need replacement if the instrument fails to indicate on the A.C. ranges.

V.3 (type 6AL5) may need replacing if the instrument fails to function on any range, as one half of this double diode is used for the H.T. rectifier. The other half of this diode is used to counteract the voltage which appears at the grid of the 12AU7 due to contact potential within V.1 when used as an AC rectifier.

4/3. CALIBRATION
We recommend that all major recalibrating and repairs be carried out by our service department.

In the event of this being impracticable, the following procedure should be adopted. If any components or valves are replaced, the instrument should be allowed to “age” by leaving it operating under normal conditions for a period of time, approximately 24 hours.

D.C.-CALIBRATION
Set the Function switch to “-V” and the Range switch to 50V. Adjust zero with the “Zero Adjust” control and connect the valve voltmeter terminals to the D.C. supply source (observing the correct polarity), adjust the supply to exactly 50 volts output. Then adjust the slider R10 (located on the resistor panel of the 172A) until the pointer reads exactly full scale 50 volts.
[WORKS WITHOUT 1M DC PROBE WITH ADJ CENTRAL ON SLIDER, BUT ADJ OUT OF RANGE WITH PROBE]

D.C. + CALIBRATION
Set the Function switch to "+V” and the Range switch to 50V, reverse the connections to the 172A and repeat the above procedure by adjusting slider R5, until the pointer reads exactly full scale 50 volts.
[DITTO]

A.C. CALIBRATION
Set the Function switch to “A.C.” and the Range switch to 1.5V. Short circuit the input terminals of the 172A and adjust R21 (located on the printed circuit panel) until the pointer reads zero [NOT ENOUGH RANGE ON POT], then rotate the Range switch through the remaining ranges.

The pointer should remain exactly on zero on all A.C. ranges. If the pointer cannot be adjusted to zero [IT CANNOT] or if it rnoves from zero when the Range switch is rotated [IT DOES], the two 6AL5 valves should be interchanged; if this fails to correct the fault, one or both of the 6AL5 valves should be changed. [TRIED - DOES NOT CURE]

Set the Range switch to 150V and apply to 150 volts A.C. 50 c.p.s. to the 172A and adjust R8 (located on the resistor panel), until the pointer reads exactly full scale 150 volts. [NOT DONE AS DON'T HAVE THAT HIGH AC AVAIL. BUT AT LOWER VOLTAGES 10-40V 10% LOW]

COMPENSATION ADJUSTMENT [NOT DONE]
Set the Function switch to A.C. and the range switch to 500V. Adjust meter to zero. Apply exactly 150 volts at a frequency of 100 kc/s to the 172A and adjust Cl until the pointer reads exactly 150 volts.
NOTE: The accuracy of the instrument if re-calibrated can only be as good as the equipment used to monitor the voltages.
Attached Files
File Type: pdf Parts List.pdf (204.8 KB, 78 views)
6SN7WGTB is offline  
Old 5th Sep 2021, 9:31 pm   #4
ronbryan
Octode
 
Join Date: Jan 2003
Location: Guildford, Surrey, UK.
Posts: 1,958
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Is it possible that the paxolin circuit board mounting the 12AU7 has absorbed some moisture and there is a leakage path in the valve grid circuit? If you think this is likely, you could put the instrument in the airing cupboard or near the boiler for a week or so to see if an improvement is obtained.

Ron
ronbryan is offline  
Old 6th Sep 2021, 7:13 am   #5
6SN7WGTB
Hexode
 
Join Date: Apr 2021
Location: Crawley, West Sussex, UK.
Posts: 442
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Quote:
Originally Posted by ronbryan View Post
Is it possible that the paxolin circuit board mounting the 12AU7 has absorbed some moisture and there is a leakage path in the valve grid circuit? If you think this is likely, you could put the instrument in the airing cupboard or near the boiler for a week or so to see if an improvement is obtained.

Ron
It's a thought but the device has been kept in a warm well-ventilated area for several weeks. I can test the board for conductivity to be sure, so will do that.

What I'm wondering is is the grid to ground caps are in fact leaky. As I said I didn't replace those - could this explain the consistently low readings if they're passing the boas to ground?
6SN7WGTB is offline  
Old 6th Sep 2021, 9:03 am   #6
6SN7WGTB
Hexode
 
Join Date: Apr 2021
Location: Crawley, West Sussex, UK.
Posts: 442
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Further information:

Just checked AC secondaries: htr 6.7 V, HT 157 V.

The VTVM is being fed with an isolation transformer so I have ca. 245 V at the primary.

Across HT filter cap B+ 172 V.
6SN7WGTB is offline  
Old 6th Sep 2021, 10:47 am   #7
WME_bill
Octode
 
Join Date: Oct 2007
Location: Exeter, Devon, UK.
Posts: 1,553
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Taylor VTVM.
These voltmeters, and the similar Avo and Heathkit models with the balanced double triode circuit
behave very oddly sometimes.
Following up RonBryan's sensible comments about damp, thoroughly clean all round the valve base, the valve holder and range switches (isopropyl alcohol or even meths).
Stray grid current in these high impedance circuits can be the problem.
So try another valve. This I have found will often solve it.
Some versions of this meter use the Brimar 13D8 which is a selected variant of the 12AU7 / ECC82. Any will do.
wme_bill

Last edited by WME_bill; 6th Sep 2021 at 10:55 am.
WME_bill is online now  
Old 6th Sep 2021, 11:36 am   #8
6SN7WGTB
Hexode
 
Join Date: Apr 2021
Location: Crawley, West Sussex, UK.
Posts: 442
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Bill,

Have cleaned it all carefully - it was dry and almost mint. Contacts done carefully so no spraying of potentially conductive stuff about the place.

Have tried a different 12AU7. Identical behaviour.

Next step is to test the grid to ground caps for resistance (i.e. a surrogate for leakage) in circuit. As I said, engine out to replace, but looks like I may have to.

Can anyone confirm/comment on:
(a) Are the tubular caps prone to leakage?
(b) If they ARE in fact leaking this would appear to explain why I can't get 'enough' range on the calibration VRs on the triode anode?

Thanks all. Appreciate the guidance and I equally appreciate very hard to diagnose at arm's length...!
6SN7WGTB is offline  
Old 6th Sep 2021, 1:04 pm   #9
Chris55000
Nonode
 
Join Date: Aug 2007
Location: Walsall Wood, Aldridge, Walsall, UK.
Posts: 2,853
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Hi!

Thank you for posting the parts list – that and the diagram is enough for me to look at it!

The first thing to try, if replacement EB91/6AL5 and ECC82/12AU7 valves do not help, is to check whether the meter retains the correct sensitivity, as loss of magnetic field strength in the pole–pieces will cause a moving–coil meter to read low!

To do this, switch the instrument to the lowest d.c. voltage and apply the full scale d.c. voltage to the instrument (it will be marked on the range switch what this is) – if the meter reads low on this test, and all the anode and cathode resistors for the ECC82 are all within tolerance, check R9 and the setting of "DC Cal" R10 again to make sure, if still no luck the meter movement will need checking or remagnetising by an instrument repairer – I think there are firms that can do this in the UK!

As a guide, a 200μΑ movement, as used in Heathkit instruments will read two–thirds full scale when tested using a 10k resistor and a 1.5V dry cell in series.

If you look at the meter scale–plate carefully, you might find it has the f.s.d. in microamperes marked on it, in which case it's easy to test it with a second known good meter or D.M.M. and a suitable resistor in series with a 1.5V cell or variable PSU!

If this test shows the meter sensitivity is within the original maker's tolerance, then there must be a changed value resistor somewhere, and you'll need to dig into it carefully against the parts list!

Chris Williams
__________________
It's an enigma, that's what it is! This thing's not fixed because it doesn't want to be fixed!
Chris55000 is offline  
Old 6th Sep 2021, 1:24 pm   #10
Nickthedentist
Dekatron
 
Nickthedentist's Avatar
 
Join Date: Jan 2005
Location: Oxford, UK.
Posts: 17,820
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Following on from C's excellent advice, I suppose you could put a multimeter set to measure uA in series with the movement, drive the VTVM and see what the VTVM's own meter's FSD actually is.

Then compare this figure with what its FSD is meant to be (assuming the data are available).

Or would the multimeter's internal resistance not being zero mess things up?

N.
Nickthedentist is online now  
Old 6th Sep 2021, 1:51 pm   #11
6SN7WGTB
Hexode
 
Join Date: Apr 2021
Location: Crawley, West Sussex, UK.
Posts: 442
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Parts list shows meter as 200µA FSD. Scale isn't marked.

Testing with 1.5V cell plus 10K resistor gives a current through meter of 149µA, with meter reading 37V on 50V scale (not that the scale matters as it's linear).

So, current is 74.5% FSD, reading is 74%. Checks out.

Onwards we go...
6SN7WGTB is offline  
Old 6th Sep 2021, 1:54 pm   #12
6SN7WGTB
Hexode
 
Join Date: Apr 2021
Location: Crawley, West Sussex, UK.
Posts: 442
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Quote:
Originally Posted by Nickthedentist View Post
Following on from C's excellent advice, I suppose you could put a multimeter set to measure uA in series with the movement, drive the VTVM and see what the VTVM's own meter's FSD actually is.

Then compare this figure with what its FSD is meant to be (assuming the data are available).

Or would the multimeter's internal resistance not being zero mess things up?

N.
Fair point, but 1.5V cell plus 10k = 155µA, so by my reckoning meter resistance (Vm) is ca. 67 ohms.

1.5 = 149E-6 x (10k + Vm)
6SN7WGTB is offline  
Old 6th Sep 2021, 9:56 pm   #13
Chris55000
Nonode
 
Join Date: Aug 2007
Location: Walsall Wood, Aldridge, Walsall, UK.
Posts: 2,853
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Hi!

Fair enough, we've established that the meter movement is in order with no significant loss of pole–piece field strength!

You can repeat the test I suggested for the lowest d.c. voltage range, but this time, supply a d.c. voltage of the amount needed to produce exactly 100% f.s.d.

If this voltage is substantially greater than the nominal voltage marked on the range switch, either there is some residual leakage to earth across the grid circuit of the input valve, or there's still a component value change that's not come to light yet!

A modern type of D.M.M. will be of assistance here, if you have one those lab–quality ones that will read into the GΩ range, (or can borrow one) this will be useful for testing for leakage paths!

If you don't have access to a lab–type meter, then you need to make sure the rotary wafer switches are scrupulously clean with no signs of leakage, dirt or grot across any of the wafers or contacts!

If the worst comes to the worst, you might have to remove the range–switch assembly from the instrument and dry in it an oven set to about 110°C or as low as you can reasonably set it for a few hours, checking the switch periodically to make sure it's not getting too hot to touch!

A fault like this is difficult to diagnose over a forum without having the instrument beside me, but I'll try my best!

Have you double–checked the "set zero" control between the anodes of the ECC82?

Chris Williams
__________________
It's an enigma, that's what it is! This thing's not fixed because it doesn't want to be fixed!
Chris55000 is offline  
Old 7th Sep 2021, 8:05 am   #14
6SN7WGTB
Hexode
 
Join Date: Apr 2021
Location: Crawley, West Sussex, UK.
Posts: 442
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

A new day, a new development.

When initially tested before work started, the 3M3 grid resistor was about 10% high. Wrong direction obviously, but shouldn't have been the reason for the large skew.

However, retesting all the resistors again after having the unit on several times revealed this was now 4M2. So we may have a culprit.

I will parallel it with ca. 15M to bring it into line and see where that gets me before replacing.

The meter innards are pretty much mint, although that won't negate ageing of course. I have carefully cleaned contacts.

Chris - yes I have a Fluke 87, which whilst not lab quality, does do nS so will measure very high resistance indeed at least directionally (using good leads).

The 0µ1 and 0µ047 HUNT caps I replaced were all measuring 50-100% high on value, but with resistances in the 300+ M range. However, they were in the AC circuit as you know.

The 16µ electrolytic was about 22µ and 16M - however the ESR was poor (but by modern standards) and I'd be daft to leave a 50-year old electrolytic in place which could burn out the unobtanium transformer.

I'm yet to be wholly convinced by LV cap testers - but that's another story, and can of worms!

EDIT: Zero Adjust remains centre range, resistance is good and it's smooth mechanically and electrically.

Last edited by 6SN7WGTB; 7th Sep 2021 at 8:16 am.
6SN7WGTB is offline  
Old 7th Sep 2021, 9:59 am   #15
6SN7WGTB
Hexode
 
Join Date: Apr 2021
Location: Crawley, West Sussex, UK.
Posts: 442
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

UPDATE: with a parallel resistor on the out-of-spec 3M3 I achieved 3M265.

However, makes no noticeable difference to the under-reading.

To answer Chris' question, a 1.55V D cell on the 1.5 V FSD range WITH the 1M probe measures 1.4V. WITHOUT probe just off of FSD so close.

As noted, WITHOUT the 1M probe and using a 50V calibration on the 50V FSD range, I cannot get the calibration VR to get the reading to 50V - maybe 48/49V. Obv with probe it's at least 10% low and way beyond adj.

Looks like I will have to look further. I suppose I'm now bound to replace the grid to ground caps...
6SN7WGTB is offline  
Old 7th Sep 2021, 4:21 pm   #16
G4BZI Roger
Pentode
 
Join Date: Nov 2005
Location: Crewe, Cheshire, UK.
Posts: 136
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Hi,

I too have a 172A vtvm that is waiting in the refurbishment queue.
I checked and mine does not have R4/7/9 as per schematic, but rather two ~4k8 resistors one for AC cal and the other for DC (+ & -) cal in series with the respective slider preset resistors.
Some of the Hi-stab resistors in the voltage divider chain are out of tolerance and awaiting replacement so I can't do meaningful calibration checks at present but if you need any other help with comparison (e.g. supply voltages, etc) just let me know.
Roger
G4BZI
G4BZI Roger is offline  
Old 7th Sep 2021, 4:38 pm   #17
6SN7WGTB
Hexode
 
Join Date: Apr 2021
Location: Crawley, West Sussex, UK.
Posts: 442
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Quote:
Originally Posted by G4BZI Roger View Post
Hi,

I too have a 172A vtvm that is waiting in the refurbishment queue.
I checked and mine does not have R4/7/9 as per schematic, but rather two ~4k8 resistors one for AC cal and the other for DC (+ & -) cal in series with the respective slider preset resistors.
Some of the Hi-stab resistors in the voltage divider chain are out of tolerance and awaiting replacement so I can't do meaningful calibration checks at present but if you need any other help with comparison (e.g. supply voltages, etc) just let me know.
Roger
G4BZI
Roger - all help most welcome!

Mine has the same arrgt for the calibration as yours - so we both differ from the schematic, albeit in a minor way.

If I were to ask two things, to check some basics:
(a) What DC voltage do you have across the C6 PSU cap? I'm at 176V.
(b) What voltage do you have at the common of those calibration VRs (i.e. the anode of V2b)? I have 162V.

Many thanks.
6SN7WGTB is offline  
Old 7th Sep 2021, 5:18 pm   #18
G4BZI Roger
Pentode
 
Join Date: Nov 2005
Location: Crewe, Cheshire, UK.
Posts: 136
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Measurements as requested:
Instrument set to DC V+ 1500V range
AC input: 241V
Across C6: 171V
V2b anode (to -ve end of C6): 156V
G4BZI Roger is offline  
Old 7th Sep 2021, 6:08 pm   #19
6SN7WGTB
Hexode
 
Join Date: Apr 2021
Location: Crawley, West Sussex, UK.
Posts: 442
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

Quote:
Originally Posted by G4BZI Roger View Post
Measurements as requested:
Instrument set to DC V+ 1500V range
AC input: 241V
Across C6: 171V
V2b anode (to -ve end of C6): 156V
Super, thanks. That rules out any major issues in that area. I'm running on an isotran so probably have closer to 250V given the load so that supports the deltas.
6SN7WGTB is offline  
Old 7th Sep 2021, 7:26 pm   #20
frsimen
Heptode
 
Join Date: Jan 2011
Location: Croydon, London, UK.
Posts: 767
Default Re: VTVM - appears to under-read by ca. 10% on AC and DC

A closer look at the circuit diagram has, I think, allowed the area of trouble to be narrowed down. The error is similar on both AC and DC volts. These ranges use entirely separate switches and resistors, so it is unlikely to be a problem in that area.

I don’t think there is a leakage issue as the error was unchanged when you corrected the value of R15.

The gain of the amplifier valve, V2 is determined (to a first approximation) by the ratio of the anode resistors/cathode resistors. In your checks you said that the divider resistors were within spec and all the others were within 20% of the correct value. In the case of the resistors around V2, particularly R16 and R19, 20% is rather a wide tolerance and could result in the gain and balance problems that you are observing.

Paula
frsimen is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 12:14 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.