christarak 7/12/2018 1:53 AM
Marshall Artiste with low P/T secondary and low output
A pre-emptive thanks to my learned colleagues.

This is a problem with a Marshall Artiste 50w from about 1978.

It came in with several problems which I have worked through (V1 plate resistor, and Bias rectifier diode). The amp is now going but with a very, very low output (4.8VRMS across the speaker).

The B+ voltage was very low (360V where I was expecting 420V). I replaced all the original filter caps and the main rectifier diodes. B+ is still 360V). Next I measured the P/T secondary (with rectifiers out of circuit) and it was reading 570VRMS (Fluke 87V) where the specs on the Classic tone site had me expect 690V http://www.classictone.net/40-18023.pdf . Thatís down about 20%.

Question #1: I thought Rectified DC = 0.9xVRMS. Clearly I am wrong. I donít understand why a secondary voltage of 690VAC would result in a B+ voltage of only 420VDC according to Marshall. Removing the HT fuse didnít see any notable increase in B+.

Here are relevant readings with all tubes installed:
[ATTACH=CONFIG]49684[/ATTACH]

(and in text for, just in case picture doesn't load)

Measured location Expected Value Actual Value
P/T Secondary Vrms 690 570
V @ rectifier VDC 360
V5plate 420 358
V5 screen 420 349
V5 grid -30 -25 (maxed)
A4A plate 175 153
A4B plate 170 141
A3A plate 210 171
A3B plate 125 98
A2A plate 140 113
A2B plate 100 88
A1A plate 140 67
A1B plate 105 92

Question #2: Could having a V1 plate as low 67V be the cause of the low output?. Looking at the 1KHz sine wave at the speaker output shows a slight distortion on the top half of the sine wave.

Question #3: It looks like the P/T is a couple of windings short of a picnic. Is there anything you recommend I check before breaking the bad news to the customer.
 
dstrat 7/12/2018 6:29 AM
I wouldn't be too alarmed there were quite a few Marshalls from 74 that had lower then normal voltage, I in fact have one.
came from the factory this way sounds great but not at all a 50 clean watts amp.
I can bias it colder and get the plates up to 390 but sounds better biased hot.

Yeah the low output is definitely a problem.
 
Helmholtz 7/12/2018 6:33 AM
Heater voltage? Line voltage? Mains selector position?
 
Enzo 7/12/2018 6:41 AM
!. You have a cap input filter, so rectified is 1.414 times RMS. (0.9 x RMS is for choke input filters) 570vAC, which is 285-0-285. 285 x 1.414 is 402v, about 5% low from the print. Any chance the amp is set for 240v and your mains are 220v?

The Classic Tone transformer looks too high a voltage spec. It would result in almost 500v B+. Is that the transformer actually in your unit? Or are you just assuming the specs from Classic Tone are correct?

Your B+ is 60v low, so is there a ton of ripple or is it smooth but low? And you measured the transformer with no rectifiers, but what does teh transformer measure with the circuits connected?
 
drewl 7/12/2018 1:09 PM
An Artiste'? Wow!
never saw one in person, can you post some pictures?

Love to see the insides as well as the outside.
 
christarak 7/12/2018 4:04 PM
Quote Originally Posted by Helmholtz View Post
Heater voltage? Line voltage? Mains selector position?
Hi Helmhotz. The heater is 6.29-6.30VAC. Line voltage is currently 241.3VAC (at 7:30am) Mains selector is on 240.
 
christarak 7/12/2018 4:39 PM
Quote Originally Posted by Enzo View Post
!. You have a cap input filter, so rectified is 1.414 times RMS. (0.9 x RMS is for choke input filters) 570vAC, which is 285-0-285. 285 x 1.414 is 402v, about 5% low from the print. Any chance the amp is set for 240v and your mains are 220v?

Thanks for rectification multiplier. I need to hit the books everytime you help me Enzo. I must be getting old The amp is set for 240V, the mains fuse has 241.3v sitting on it and the filament is 6.29-6.30 VAC


Quote Originally Posted by Enzo View Post
The Classic Tone transformer looks too high a voltage spec. It would result in almost 500v B+. Is that the transformer actually in your unit? Or are you just assuming the specs from Classic Tone are correct?
I just assumed that Classic Tone know what they're talking about compared to me.

Quote Originally Posted by Enzo View Post
Your B+ is 60v low, so is there a ton of ripple or is it smooth but low? And you measured the transformer with no rectifiers, but what does the transformer measure with the circuits connected?
When connected to the circuit and all tubes in place

P/T secondary: 275-0-275VAC
DC side of rectifier:
HT Fuse IN: 358V with 11.4 v p-p ripple .http://music-electronics-forum.com/a...0&d=1531434075
HT fuse OUT: 393VDC with 138mV p-p ripple http://music-electronics-forum.com/a...9&d=1531433994

Is a 35v drop excessive with a load on the power supply? I expect the ripple to increase, but is 11.4V too much?
 
christarak 7/12/2018 4:47 PM
Quote Originally Posted by drewl View Post
An Artiste'? Wow!
never saw one in person, can you post some pictures?

Love to see the insides as well as the outside.
I can't wait to hear this much-maligned baby sing again. Photos as requested[ATTACH=CONFIG]49691[/ATTACH][ATTACH=CONFIG]49692[/ATTACH][ATTACH=CONFIG]49693[/ATTACH][ATTACH=CONFIG]49694[/ATTACH]
 
nickb 7/12/2018 4:53 PM
What is happening on the grids on the output tubes when it starts to clip on the output?

BTW, I had the 100W version of one of these here recently.
 
christarak 7/13/2018 4:06 AM
Quote Originally Posted by nickb View Post
What is happening on the grids fn the output tubes when it starts to clip on the output?

BTW, I had the 100W version of one of these here recently.
Hi Nick, I'm just looking at the Channel 2's V2 with a 100mV 1Khz signal at the input. The signal on the grid of V2 is 106mV and its plate is 1.27VAC. On V4 the pin 2 Grid is 1.08V and the pin plate is 19.2v. On V5 (EL34) the signal on the grid is 18.2v (I can't get any more negative than -24VDC bias).

HERE IS THE WEIRD THING: With the reverb pot on minimum, the signal across the speakers is 8.17Vrms ( 8 watts). When I turn the reverb pot to maximum the signal across the speakers goes to 19.4vrms (47watts). The B+ and Plate voltages are still all low. I have NO reverb tank connected to the reverb leads. At maximum volume with maximum reverb the output signal is clipping as one would expect to see. If I turn down the reverb the signal cleans up as it reduces. Its like the reverb control is acting as a Master Volume.

Is it normal to have the terminals on the reverb drive transformer exposed like that? Probing fingers can touch the supply voltage line.
 
Helmholtz 7/13/2018 7:23 AM
Quote Originally Posted by christarak View Post
Hi Nick, I'm just looking at the Channel 2's V2 with a 100mV 1Khz signal at the input. The signal on the grid of V2 is 106mV and its plate is 1.27VAC. On V4 the pin 2 Grid is 1.08V and the pin plate is 19.2v. On V5 (EL34) the signal on the grid is 18.2v (I can't get any more negative than -24VDC bias).

HERE IS THE WEIRD THING: With the reverb pot on minimum, the signal across the speakers is 8.17Vrms ( 8 watts). When I turn the reverb pot to maximum the signal across the speakers goes to 19.4vrms (47watts). The B+ and Plate voltages are still all low. I have NO reverb tank connected to the reverb leads. At maximum volume with maximum reverb the output signal is clipping as one would expect to see. If I turn down the reverb the signal cleans up as it reduces. Its like the reverb control is acting as a Master Volume.

Is it normal to have the terminals on the reverb drive transformer exposed like that? Probing fingers can touch the supply voltage line.
As DC voltages look reasonable for an open loop PT secondary voltage of 570VRMS and heater voltage is spot-on, it is most probable that the PT is O.K. (having had lower than typical voltage from the first day). Shorted secondary turns would eat power, heat up the PT and cause heater voltage to drop as well. The same with shorted primary turns. In case of doubt you may do R.G. Keen's transformer short test:
https://www.premierguitar.com/articl...ormer-tester-1

There is also no sign that the low voltages are caused by some fault in the amp. (Not saying that there is no other problem.) In other words I see no chance to noteably increase voltages with this PT - except when using the 220V primary tap. This would raise heater voltage to around 6.9V (can be dropped by a suitable small value resistor if you care) and increase supply and bias voltages by 9%.

The reverb transformer should isolate from supply voltage, I guess what you feel when touching the secondary terminals is a high frequency oscillation in V3a (scope) caused by the open transformer secondary (similar to a power amp without speaker load).
How is the performance with reverb tank connected?
How do you measure grid voltages? These are high impedance circuits (around 1M) and the meter/scope should have at least 10M input impedance..
 
nickb 7/13/2018 2:44 PM
Quote Originally Posted by christarak View Post
Hi Nick, I'm just looking at the Channel 2's V2 with a 100mV 1Khz signal at the input. The signal on the grid of V2 is 106mV and its plate is 1.27VAC. On V4 the pin 2 Grid is 1.08V and the pin plate is 19.2v. On V5 (EL34) the signal on the grid is 18.2v (I can't get any more negative than -24VDC bias).
106mVAC in and 1.27VAC out is only a gain of 11. That is way too low. Can you retest but use a 10nF 630V ( or so) capacitor in series with your meter lead? That said this is very much a clean amp. You can always mod it to make it more JCM800 like.

HERE IS THE WEIRD THING: With the reverb pot on minimum, the signal across the speakers is 8.17Vrms ( 8 watts). When I turn the reverb pot to maximum the signal across the speakers goes to 19.4vrms (47watts). The B+ and Plate voltages are still all low. I have NO reverb tank connected to the reverb leads. At maximum volume with maximum reverb the output signal is clipping as one would expect to see. If I turn down the reverb the signal cleans up as it reduces. Its like the reverb control is acting as a Master Volume.
The reverb mixer is rather, er, lacking. You can try on old trick and that is to swap the wires to the reverb level pot wiper and clock wise pins. It will help to stem the gain loss with no reverb.

I notice you can get 47 watts out, that's good. Therefore your issue is not the power supply or output stage. It's just gain and these don't have a lot of that. You can add cathode bypass caps to V1 on CH1 to give a bit more gain.

Is it normal to have the terminals on the reverb drive transformer exposed like that? Probing fingers can touch the supply voltage line.
Safety standards have evolved over the years. At the time that was OK. Feel free to protect more to your own satisfaction.

Here is a gut shot of the 100W version. Notice some mods around the input stages.

[ATTACH=CONFIG]49705[/ATTACH]
 
g1 7/13/2018 4:44 PM
If you can't get more negative than -24VDC at power tube grids, I think you should pull the power tubes and fix that before doing anything else.
The change in output level with the reverb tank disconnected makes no sense to me. There is no signal path through the reverb then. Makes me think it is just oscillating.
 
drewl 7/13/2018 8:10 PM
Thanks for the pix, these are rare.

I actually have an Artiste' 2x12 cab I found on the early days of Ebay.

I built a Marshall 18w chassis for it back in the day and still use it alot.
 
nickb 7/14/2018 1:10 AM
Quote Originally Posted by g1 View Post
If you can't get more negative than -24VDC at power tube grids, I think you should pull the power tubes and fix that before doing anything else.
The change in output level with the reverb tank disconnected makes no sense to me. There is no signal path through the reverb then. Makes me think it is just oscillating.
I missed the -24VDC. g1 is right that is a problem. I'm surprised the EL34's aren't red plating. Might also explain you low volts.

The gain drops as you increase the reverb level due to interaction in the reverb / dry mixer.
 
christarak 7/14/2018 1:27 AM
Quote Originally Posted by g1 View Post
If you can't get more negative than -24VDC at power tube grids, I think you should pull the power tubes and fix that before doing anything else.
Thanks for the tips G1. I put a 680K across the 220K supplying the Bias rectifier. That gave me enough negative volts to have -32v on pin 5 in the middle position of the bias pot. -32 v gave me 40ma at 370v on the plate.

Quote Originally Posted by g1 View Post
The change in output level with the reverb tank disconnected makes no sense to me. There is no signal path through the reverb then. Makes me think it is just oscillating.
With the reverb tank connected, I can hear the anaemic reverb when I have the pot at around 3 o'clock. Now here's the brain scrambler: when I disconnect the reverb tank the reverb pot acts like another gain stage. I can hear dramatically this with either my signal generator or guitar. I agree with you G1, I can't see a signal path to V3b. The actual amp seems to match the schematic. Its got me stumped!
 
christarak 7/14/2018 1:32 AM
Very interesting that you have a low voltage Marshall, too. I had to jiggy some values to get the bias voltage (-32v) to the point where the EL34 was drawing 40mA at plate voltage of 370v. I can't get it any higher than 370.
 
christarak 7/14/2018 1:45 AM
Thanks Nick. Weirdest thing. V1 had a cathode bypass cap tacked on, but I'll be damned if I could find the bypass cap on V2. I attached a 25uF/25v across the cathode resistor and boom ...the gain doubled. I haven't swapped the reverb leads yet but I have bigger issues to solve first. The main one being the Reverb pot acting like an extra gain stage when there is no reverb tank but then acting like a reverb mix pot when the tank is connected.

Channel 1 was oscillating when I removed the 270K between grid and plate that a previous tech had installled....so I reattached it. Channel 1 sounds like crap, as if the treble pot is on 0 when its on 10.
 
christarak 7/14/2018 2:03 AM
Yes, modding the bias circuit to get a -32v bias certainly helped a lot. The was no red plating, but I don't think I was too far away. I saw 70mA through one of the tubes before I dived for the off switch. Its now at 40mA.

Nick, what would you expect when you disconnect the reverb tank (pulling out RCA connectors at amp end) and you dial up the reverb? I'm with G1 and expect nothing, and yet my ears don't lie. Its like another gain stage has kicked in.
 
dstrat 7/14/2018 3:24 AM
Hi Chris, from the looks of the wiring, some of it does not look stock to me. Its possible someone has modded the reverb to act as a gain stage. I can not see a 1M resistor anywhere around the 3 470K mixers.
also make sure your bias is feeding from the switch side of the stand by switch, and not diode side as that could cause blown fuses and no bias for a few seconds while it charges to its desired voltage.
maybe its already been taken care of, hard to tell from the photos.

Nick I get a error when clicking your attachment?
 
nickb 7/14/2018 8:02 AM
Quote Originally Posted by christarak View Post
Yes, modding the bias circuit to get a -32v bias certainly helped a lot. The was no red plating, but I don't think I was too far away. I saw 70mA through one of the tubes before I dived for the off switch. Its now at 40mA.

Nick, what would you expect when you disconnect the reverb tank (pulling out RCA connectors at amp end) and you dial up the reverb? I'm with G1 and expect nothing, and yet my ears don't lie. Its like another gain stage has kicked in.
Look at the way the dry and wet mix are done. The reverb level pot when set to zero shunts the dry path with 470K to ground so attenuating the signal by half. When the level pot is about halfway you'll get less attenuation. The difference is around 1.5dB if everything is as it should be. If it isn't then all bets are off.
 
nickb 7/14/2018 8:08 AM
Quote Originally Posted by dstrat View Post
.....hard to tell from the photos.

Nick I get a error when clicking your attachment?
Thx - I've re-uploaded the pic. It seems to be there now http://music-electronics-forum.com/a...2&d=1531577029 . Ignore the "Attachment" text. It's not real and doesn't show up in advanced edit. Weird...

Anyway, so now you have a second image to compare the reverb wiring with.
 
Chuck H 7/14/2018 8:17 AM
70mA through ONE of the tubes... What about the other?

Obviously you'll replace power tubes if one is drawing substantially more current than the other. I only mention it because I think that (after the signal chain is working and) once you get the bias sorted out with known good tubes you'll probably have around 380Vp. And that's likely normal. I've had two 2204 amps. One from the late 70's and the other from the mid 80's I think. The 70's amp had 380Vp and the 80's amp had 420Vp. This is a pretty well known anomaly for some Marshall 50 watt models. Though I don't have an encyclopedic knowledge of the models, years and details this is probably what I would chalk it up to and I wouldn't sweat it.
 
Helmholtz 7/14/2018 8:40 AM
With the reverb tank connected, I can hear the anaemic reverb when I have the pot at around 3 o'clock.
Did you verify the oscillation without tank on your scope? Oscillation can produce very high voltages across the reverb transformer primary and may eventually damage the transformer, cause arcing in the tube and give false meter readings
I recommend to short the reverb transformer secondary while the tank is disconnected.
 
christarak 7/14/2018 3:50 PM
Quote Originally Posted by Helmholtz View Post
Did you verify the oscillation without tank on your scope? Oscillation can produce very high voltages across the reverb transformer primary and may eventually damage the transformer, cause arcing in the tube and give false meter readings
I recommend to short the reverb transformer secondary while the tank is disconnected.
Thank you Helmholtz. I checked for oscillation with my scope, but I couldn't see or hear any. With the tank connected, I can hear reverb. Both the In and the OUT of the tank measure 166R. Is it OK to short the reverb secondary terminals or is it kinder to the transformer to connect a 100R resistor between them? I get this (nice sounding) gain boost when increasing the Reverb pot with the tank removed and the reverb transformer secondary left open (which I only do very briefly).

I have just emailed the customer to see if he is aware of a mod in his amp to convert the reverb to provide more gain. I doubt it as then I would not expect the reverb that I can hear. I just fail to see an electrical connection to pin 7 of V3 (the recovery half) other than the reverb tank.
 
christarak 7/14/2018 3:52 PM
Quote Originally Posted by nickb View Post
Thx - I've re-uploaded the pic. It seems to be there now http://music-electronics-forum.com/a...2&d=1531577029 . Ignore the "Attachment" text. It's not real and doesn't show up in advanced edit. Weird...

Anyway, so now you have a second image to compare the reverb wiring with.
Thank you Nick. I will be studying your photos when I get home on the bigger monitor to pick up any discrepancies. Thank you for taking that trouble.
 
g1 7/14/2018 4:15 PM
See nicks post #21 about the gain boost. Or you might think of it as a 'cut' when reverb is turned down.

I think a 100ohm resistor to load the reverb TX should be fine.
 
christarak 7/14/2018 4:26 PM
Quote Originally Posted by nickb View Post
Look at the way the dry and wet mix are done. The reverb level pot when set to zero shunts the dry path with 470K to ground so attenuating the signal by half. When the level pot is about halfway you'll get less attenuation. The difference is around 1.5dB if everything is as it should be. If it isn't then all bets are off.
You are bloody brilliant Nick!! Does my little sketch match what you're telling me? If so, then there is no mystical voodoo happening at all. [ATTACH=CONFIG]49723[/ATTACH]
 
christarak 7/14/2018 4:31 PM
Quote Originally Posted by Chuck H View Post
70mA through ONE of the tubes... What about the other?

Obviously you'll replace power tubes if one is drawing substantially more current than the other. I only mention it because I think that (after the signal chain is working and) once you get the bias sorted out with known good tubes you'll probably have around 380Vp. And that's likely normal. I've had two 2204 amps. One from the late 70's and the other from the mid 80's I think. The 70's amp had 380Vp and the 80's amp had 420Vp. This is a pretty well known anomaly for some Marshall 50 watt models. Though I don't have an encyclopedic knowledge of the models, years and details this is probably what I would chalk it up to and I wouldn't sweat it.
Sorry Chuck, I should have been more specific. Both tubes were drawing large currents, but I only put one value down, The lettering on the tubes have browned from the overheating. I have now corrected the bias to give me good levels of current. My B+ is still low at 370, but I feel much better about it from what you have said. Thank you so much for sharing your experience and knowledge with me.
 
Helmholtz 7/14/2018 4:32 PM
Quote Originally Posted by christarak View Post
Thank you Helmholtz. I checked for oscillation with my scope, but I couldn't see or hear any. With the tank connected, I can hear reverb. Both the In and the OUT of the tank measure 166R. Is it OK to short the reverb secondary terminals or is it kinder to the transformer to connect a 100R resistor between them? I get this (nice sounding) gain boost when increasing the Reverb pot with the tank removed and the reverb transformer secondary left open (which I only do very briefly).

I have just emailed the customer to see if he is aware of a mod in his amp to convert the reverb to provide more gain. I doubt it as then I would not expect the reverb that I can hear. I just fail to see an electrical connection to pin 7 of V3 (the recovery half) other than the reverb tank.
Shorting the secondary or terminating with 100R shouldn't be a problem. Perfect would be a load resistor corresponding to the tank's input impedance (might be found from the model number). Does this take care of the electrifying/shock when touching?

What is the bandwidth of your scope? Oscillation may be in the MHz range.
 
nickb 7/14/2018 4:53 PM
Quote Originally Posted by christarak View Post
You are bloody brilliant Nick!! Does my little sketch match what you're telling me? If so, then there is no mystical voodoo happening at all. [ATTACH=CONFIG]49723[/ATTACH]
You are on the right lines. When reverb is at max it will be loaded by the reverb recovery tube so you'll see about a 510K load rather than 1.47meg.

So this is the big question in my mind. If the reverb works normally then the gain variation should not be too bad. If on the other hand it's been modded or the recovery stage is not working as it should be (so that it's output impedance is higher) then the gain variation will be more.
 
christarak 7/15/2018 2:37 AM
Thanks for the suggestion Helmholtz. My scope is 25MHz. I haven't checked yet, but I have no idea of what amplitude of oscillation is significant. I guess the answer is ANY amplitude is bad.
 
christarak 7/16/2018 5:27 AM
Loading the secondary of the reverb transformer with 100 ohm (The resistance on the reverb tank was 166 ohm) seems to have sorted the massive gain and the increase in gain is now quite small. I am now getting 38 watts across the speaker on Channel 2 which I should accept as OK when you consider my P/T was down. Channel 1 is putting out a little less but sounds very muddy. Then I need to lift that circuit board without causing more damage to properly solder in my repair work.
 
pdf64 7/16/2018 6:30 AM
FYI another report of Marshall with lowish HT http://music-electronics-forum.com/s...ad.php?t=21173
I remember plenty others over the years.
And the Marshall voltage chart http://www.freeinfosociety.com/elect...iew.php?id=495
As the HT will sag substantially under load, I'd only expect 30-40 watts from these amps.
 
Helmholtz 7/16/2018 6:34 AM
Quote Originally Posted by christarak View Post
Loading the secondary of the reverb transformer with 100 ohm (The resistance on the reverb tank was 166 ohm) seems to have sorted the massive gain and the increase in gain is now quite small. I am now getting 38 watts across the speaker on Channel 2 which I should accept as OK when you consider my P/T was down. Channel 1 is putting out a little less but sounds very muddy. Then I need to lift that circuit board without causing more damage to properly solder in my repair work.
If the reverb input DCR is 166 Ohm, its impedance will be 1k to 2K. So you could increase the (dummy) load resistor up to 2K without a problem, if you like.
Somewhere above you mentioned that the reverb wiring carries "supply voltage". What made you think so?

In your post#10 you mentioned 47W output (before clipping?). This means that the power stage voltages are sufficient for this kind of output power and it must be possible to get the same power again.
 
nickb 7/16/2018 11:18 AM
Quote Originally Posted by Helmholtz View Post
If the reverb input DCR is 166 Ohm, its impedance will be 1k to 2K. So you could increase the (dummy) load resistor up to 2K without a problem, if you like.
Somewhere above you mentioned that the reverb wiring carries "supply voltage". What made you think so?

In your post#10 you mentioned 47W output (before clipping?). This means that the power stage voltages are sufficient for this kind of output power and it must be possible to get the same power again.
He was referring to the to the B supply to the reverb transformer primary, I believe.

Chris -sounds like you had the reverb primary signal crosstalking into the reverb recovery circuit.

We still need to look into the low gain reported above in post 10. You had a gain of 11 where is should be 50-ish...
 
christarak 7/16/2018 1:55 PM
Quote Originally Posted by pdf64 View Post
FYI another report of Marshall with lowish HT http://music-electronics-forum.com/s...ad.php?t=21173
I remember plenty others over the years.
And the Marshall voltage chart http://www.freeinfosociety.com/elect...iew.php?id=495
As the HT will sag substantially under load, I'd only expect 30-40 watts from these amps.
Thank you pdf64. I missed that series of posts from 8 years ago. I'll have a read through that this morning. I'll print out and keep that Marshall voltage chart, too. Many thanks.