Marshall Valvestate 8100 low heaters and high 15v rails

[Indyryder02]

This amp came from another shop in town that had supposedly just changed the tube but this did not fix the original concern of low output of the boost ch.
I found broken solder pads at R112 & R113. It appears all the capacitors were removed and hot glued back in place…(some fuckery has been in action)
I repaired the solder connections.. and replaced the 12AX7 as it had failed (still had low output after solder repairs, changing preamp tube regained volume).. I now have a loud and proud boost channel.
but a basic check or voltages shows around 9.5v across pins 4&5 of the tube. Shouldn’t this be around 12.6? (Is this what caused the tube to fail?)
the 15v rails at Zd1 and ZD2 are around 16.9v and it’s fitted with 16v zeners. And the adjacent caps are only rated at 16v.. so that can’t be right.. Schematic calls for 15v 3w. I assume these were changed?

all the components in this part of the circuit appear to test within spec. So haven’t figured out why my heaters seem low and my rails seem high.

anyone run into this before? Am I missing something basic here?
thanks.

[g1]

I would change back to the stock 15V zeners. C40 and C41he caps are only rated for 16V.
I guess it's possible someone thought higher value zeners might put less load on the heater lines? Are R112 and R113 for heaters ok? I suppose you could tweak those values to bring up the heater voltage.

[Indyryder02]

I would change back to the stock 15V zeners. C40 and C41he caps are only rated for 16V.
I guess it's possible someone thought higher value zeners might put less load on the heater lines? Are R112 and R113 for heaters ok? I suppose you could tweak those values to bring up the heater voltage.

I forgot to mention.. those resistors were smoking. Literally. Not sure if it’s residual hot glue. Can’t tell. I think it has finally stopped smoking. They both test within spec.
I did consider changing their value to increase the heater voltage.. but then I shouldn’t need to. Surely it wasn’t designed this way. And I don’t want to miss another issue that’s causing this. I expect some more of the previous shop’s shenanigans, but I haven’t been able to make sense of it yet.. and there’s not even much to it.

[g1]

Probably worth trying another tube in there to be sure.

[Indyryder02]

Probably worth trying another tube in there to be sure.

Forgot to mention. I did switch to some 15v zeners which brought the rails back to around 15.5v those caps should really be more like 25v not 16. Seems awfully close. Anyway. Didn’t change the heaters at all.

I will try another preamp tube. Just in case.

[Indyryder02]

Tried another tube. No change. Which would have be wild if it did change it really.
so far, I lifted the leg of R114 to Eliminate the auto tx. And the rest of that HV line. No change. Changed ZD1& Zd2 to 15v zeners. Which brought the 15v rails to around 15.41 and 15.24.
replaced c40 & c41 as well as c61&c62. No change to the heater voltage. Which remains around 9.3-9.5
R112&R113 read 184r so pretty close still.
r114 is 48.3r
About 30.1 vac from transformer going to r112&r113.
gonna have to put this to the side for a minute.
should have had this figured out already.

[Pedro Vecino]

I've had countless Valvestate but have never measured the filament voltage at the tube. I see very likely that those 9.5V are the original voltage.

[Indyryder02]

I've had countless Valvestate but have never measured the filament voltage at the tube. I see very likely that those 9.5V are the original voltage.

Just seemed awfully low… but maybe you’re right. Could be trying to fix a non issue.. it does work now after all. That was the initial goal

[g1]

Yes, it could be normal. As you said in the first post, the tube is working 'loud and proud'.

[Indyryder02]

Yes, it could be normal. As you said in the first post, the tube is working 'loud and proud'.

Think I was just worried it caused the tube in it to fail.. but I suppose that would take a really long time. Last shop supposedly changed the tube and it didn’t help the initial low boost channel concern. (Course the heaters weren’t connected due to the bad solder joints at the dropping resistors. )
it’s a newer Psvane tube also. Just thought it was strange after fixing the solder joints. I still had to swap the tube to a new JJ to get the sound to come up to a proper level. Could just be a fluke or they could have put the original tube back in. Which still wouldn’t have been original to the amp as the Psvane tubes are fairly new to the market still. Oh well I guess! I’ll let it burn in with some pink noise into a load for a few hours and make sure it holds up without further issues. Thanks

[J M Fahey]

Seen that problem at DIY Audio and somebody was intrigued enough to do a lot of measuring.

Yours is a real problem, and of course Valvestates left Factory working properly.

What this guy found (by **measuring** none less) is that although all 12A*7 work with 12.6V across pins 4 & 5 THEY DO NOT HAVE THE SAME COLD DCR depending on brand.

Russian/Sovtek ones (what Marshall used) have significantly higher resistance, at least when cold.

IF you feed filaments from a constant voltage source: meaning a transformer winding OR a regulator, say a 7812, no problem, when hot all pull 150mA

IF fed from a constant current source,same thing, fixed 150mA will force them to heat up , in which case they will drop 12.6 Volts.

BUT if designer is cheap, and it´s an SS amp, some avoid the extra 6.3V or 12.6V winding by feeding filament from existing power transformer mid-voltage winding (here 30+30VAC) through dropping resistors.
Cheap ass solution.

Problem is resistors are neither a constant voltage source nor a constant current one, but a voltage divider (resistor + filament resistance which is highly nonlinear and PTC), so it will work properly ONLY with the same type and brand of tube the original designer used, here a Russian 12AX7 ¨Period.

Any other brand/origin tube plugged there will give unpredictable results.

This guy measured many brands, and Russian/Sovtek *always* had a much higher DC resistance than any other.

So others will never reach proper 12.6V needed, filaments will be cold by a significant amount, and dropping resistors will heat up more ... a perfect storm.

I suggest you try a Sovtek/Rusian tube there.

[Pedro Vecino]

The model used on the Valvestate was the12AX7 China labeled Marshall. The typical model from the 90s with a getter similar to a ravioli.

[Helmholtz]

What this guy found (by **measuring** none less) is that although all 12A*7 work with 12.6V across pins 4 & 5 THEY DO NOT HAVE THE SAME COLD DCR depending on brand.

Good info!

I also considered differences in heater resistance, but lacked the background info.

To add, if the series resistors drop more voltage than "normal" with some tubes, the heater current must be higher than "normal" or nominal even though the resulting heater voltage is low.
Actual heater current can be calculated from the voltage drop across one of the resistors.

The 7W limit of the 180R resistors is reached at 200mA. If the current is higher than the nominal 150mA (corresponding to a voltage drop of 27V) they will get very hot.

[g1]

Further to what JM said, I was initially a bit surprised that the replacement tube was working fine with less than 10V on the heaters.
I had figured drop out would occur around the 10V mark but that may be faulty recollection.

Did you by chance try the 'bad' psvane tube in another amp with proper heater voltage?

[Indyryder02]

This is all good info!
once I get a chance. I’ll definitely try the suggested tube as well as trying the Psvane in another amp. Likely both will prove to be true. And quite interesting!