Risson Marvel 12E bias setting cold

[Pavedog]

Hi,

I recently came across a Marvel 12E, an interesting Fender type circuit with 2 X 6AQ5s(EL90) and 2 X 12 AX7s.

I recently swapped out some of the bad caps(drifted) the two for the bias circuit had drifted the most.

There are two pots, one is a hum pot, the other is for the bias, but I can't get it anywhere near where my calculations say it should be. Plate dissipation is 12, 250V plate current which should be 33ma but the best I can do is 12-13. The amp sounds fine but I can't figure out what is causing the restriction. I took out the pots (hum and bias) and both tested fine, no dead spots.

I added some pictures for reference

[sluckey]

To increase plate current you need to reduce the voltage between grid and cathode. I can be more specific if looking at a schematic. Big pics (4X the size of the little ones you posted) would be helpful too.

[Pavedog]

Thanks for the help, I posted new original size pics if you click on them. 

[WimWalther]

What's the DC voltage on the 6AQ5s G1? Should be around -15V with 250V on the plate, giving ~35mA for class-AB operation.

[Pavedog]

The plate voltage is 260V on both tubes, the G1 is .09 and .65, no negative voltage. These are different than my initial readings as I swapped out the tubes that came with them for a set of newer EL90s.

[mresistor]

If you have no negative bias voltage on the grids then its best to shut the amp off pull the 6AQ5s until you find out why there is no negative bias voltage and fix it before proceeding.

[Pavedog]

To increase plate current you need to reduce the voltage between grid and cathode. I can be more specific if looking at a schematic. Big pics (4X the size of the little ones you posted) would be helpful too.
/quote]

I have been trying for a month to locate a schematic but have not had success. I even spoke with the new owner of Risson amps and he did not think he had one for the 12E, but did think he had one for the 18E and would send me it when he got a chance, that was 3 weeks ago.

[mresistor]

If Ben doesn't have the schematic probably nobody does ..

[Pavedog]

If Ben doesn't have the schematic probably nobody does ..

Yes, hopefully he will reply back at some point. However, watching videos that Bob Rissi made when released he talked about it being built like the orginal early 60's Fenders. The only thing I am not clear on is that it has two output transformers and I have yet to work on these 7pin EL90 tube amps.

[Pavedog]

To increase plate current you need to reduce the voltage between grid and cathode. I can be more specific if looking at a schematic. Big pics (4X the size of the little ones you posted) would be helpful too.

At about 1:32 in the video Bob Rissi shows the original circuit and talks about how and why he designed it that way. Not sure I understand the reasoning behind the two output transformers

[WimWalther]

The plate voltage is 260V on both tubes, the G1 is .09 and .65, no negative voltage. These are different than my initial readings as I swapped out the tubes that came with them for a set of newer EL90s.

Let me get this straight: You measure ~12mA Ip, with 260V Ep and ~0V Eg1. Is that right?

What's the voltage on the screen grids - Eg2?

Because this doesn't add up. Something is fundamentally wrong, either with the power supplies, the power tubes or your measurements. Those tubes should be roasting, glowing red hot, not limping along at 12mA.

Again, check the voltage on the screen grids..

[Pavedog]

The plate voltage is 260V on both tubes, the G1 is .09 and .65, no negative voltage. These are different than my initial readings as I swapped out the tubes that came with them for a set of newer EL90s.

Let me get this straight: You measure ~12mA Ip, with 260V Ep and ~0V Eg1. Is that right?

What's the voltage on the screen grids - Eg2?

Because this doesn't add up. Something is fundamentally wrong, either with the power supplies, the power tubes or your measurements. Those tubes should be roasting, glowing red hot, not limping along at 12mA.

Again, check the voltage on the screen grids..

Wim,  I agree with you that something is wrong. I have several sets of these 6AQ5/EL90 tubes but none have red plated, and have tested them after use and they test fine. I swapped out the 12AX7s as I had two new ones and here is what I got on each EL90 tube.

                                  V3           V4
Pin 1 - Grid -              .19V         .41V
Pin 2 - Cathode -       10.3V         15V
Pin 3 - Heater -           3.3V        3.3V
Pin 4 - Heater -           3.3V        3.3V 
Pin 5 - Plate -            296V        300V
Pin 6 - Grid 2            296V        296V
Pin 7 - Grid 1*          .19V         .41V

*pin 1 & 7 are jumpered

Of course the tubes can handle more than spec plate voltage. 

I appreciate your insight, happy to provide more numbers if you need them.

[sluckey]

Ain't nothing wrong!

[Pavedog]

Ain't nothing wrong!

The grid voltages I am listing are positive and with the bias cranked down as far as it goes. If I got the other direction both will go over 1volt. Additionally, the plate voltages should be a max of 250volts.

[WimWalther]

With cathodes 10-15V+ and g1 near ground, this is more what I'd expect from a self-biased design, not a fixed bias design.

Can you describe or draw the cathode circuits of the power tubes? Is there a resistance in there, say around 130R or so? A schematic would be really helpful right now.

[sluckey]

The grid voltages I am listing are positive and with the bias cranked down as far as it goes. If I got the other direction both will go over 1volt.

So what? Bias is the voltage between the grid and cathode.

[WimWalther]

The grid voltages I am listing are positive and with the bias cranked down as far as it goes. If I got the other direction both will go over 1volt.

So what? Bias is the voltage between the grid and cathode.

Sure, but why is one at 10V while the other is at 15V? Yet neither tube is pulling the correct Ik?

Is it possible that this is a self-bias design, and the "bias" pot he's adjusting is really a balance control?

Also, it seems that the B+ is running high, as you'd expect for a light loading.

[sluckey]

I don't know what this amp is and neither does pavedog. Without a schematic everything is pure speculation. I doubt we'll ever see one.

[PRR]

Solder looks very dubious.

[Pavedog]

With cathodes 10-15V+ and g1 near ground, this is more what I'd expect from a self-biased design, not a fixed bias design.

Can you describe or draw the cathode circuits of the power tubes? Is there a resistance in there, say around 130R or so? A schematic would be really helpful right now.

I can do the schematic, I went to school for electronics 40 years ago and worked digital motor controls for a few years.  I think as was mentioned it might be a balance and not a traditional bias adjustment. I say this because I can get them both to balance out the cathode voltage by adjusting it. As far as the cathode circuit it is straight forward, from Pin 2 both tubes trace to the pot, from the pot to the two 100uf 63V caps and in parallel there is a 7W 84.5 ohm resistor and then to ground. (all visible in the pics) Additionally there is a single wire that traces from the negative side of both bias caps and the 84.5 ohm resistor to the board. The first stop is the two 470K, 1/2W resistors and each end of the resistor to a separate .01uf 500V capacitors. I have attached a picture to show it. clearer.

I have never worked on an amp with this bias balancer, so trying to correct my knowledge gap.

I will do a full schematic but likely going to take some time with my full time job. Happy to answer other questions to help.

When I do balance out the cathode (pin2) it balances at 13.3 V across both tubes and the Plate(5) goes up to 301V.

[Pavedog]

Solder looks very dubious.

I will blame it on trying to use the solder without lead and I checked the circuit paths prior to installing....... ;)

If you see an issue on the connections let me know the position and I will check them. My solder work is the two boards, but I did make sure there was a good path to ground on all the grounds to chassis.

[Pavedog]

I don't know what this amp is and neither does pavedog. Without a schematic everything is pure speculation. I doubt we'll ever see one.

Thankfully Ben came through and I have the schematics. The output is bit more but tracing through the schematic the component values including the bias balance circuit are the same. I just need to figure out how to upload PDFs......

[Pavedog]

Schematic for the Marvel II attached

[WimWalther]

@Pavedog

Yep, nailed it! That's a DC balance control you're adjusting, not a bias control. This is a self-biased amp, the large cathode resistance (bypassed by electrolytic caps) is the giveaway.

You are correct to set the control for equal Ik between the two power tubes. It's there to reduce or eliminate the need for matched pairs.

The Ep is high for maybe two reasons: - those tubes are statistically "cool", and/or that's just how they designed it.

ETA: Now that I've seen the schematic, it's definitely a self-bias design with a DC balance pot.

[sluckey]

This is also a single-ended amp, not push pull. I've never seen that power amp before. Different but definitely very cool. 

[WimWalther]

This is also a single-ended amp, not push pull. I've never seen that power amp before. Different but definitely very cool. 

No doubt, can't say I've ever seen this design before.

It's a pair of SE outputs, driven in parallel, with the OT secs wired in series. You wonder why.. My guess is that they had plenty of stock on the SE OT, and found a way to use them in a higher-power design.

But.. now I wonder what is the target impedance? If that OT was built for 8R, the series pair will tend to prefer 16R.. unless they run a bunch of additional NFB.

[WimWalther]

Solder looks very dubious.

I will blame it on trying to use the solder without lead (...)

Oh, god.. no.

My suggestion is that you avoid the stuff like the plague that it is. Use a rosin-core solder with either a 60/40 (SN60) or, preferably, a 63/37 (SN63) alloy.

The former was the standard for 100 years, and generally works well. It's also very easy to rework. The latter is a superior material, physically stronger and far less prone to cold or weak joints - but can be much more difficult to rework.

[PRR]

Someone had too many 84.5 ohm 7 Watt parts. (NO reason the heater reference needs 7W resistors.)

It's overall a "unique" affair. So be it. Thanks for sharing.

> If you see an issue on the connections let me know the position and I will check them. My solder work

To my eyes, in the pictures, 90+% of joints look un-wetted or actually un-soldered. It can't be as bad as it looks from here.

[WimWalther]

Someone had too many 84.5 ohm 7 Watt parts. (NO reason the heater reference needs 7W resistors.)

Right. They're only doing 250mW between the pair. That's like 55X overbuilt.

It does add support to my guess that they were happily using the parts they had on-hand.

[mresistor]

Maybe Rissi was picking up surplus military parts at a DRMO facility.

[Pavedog]

Wim,

You are spot on. In the video I attached from 2012 Bob Rissi talks about finding a surplus of old amp parts in his storage unit from when he closed down his original operations in early 2000s.  I spent the night removing the old solder(pretty satisfying) and replacing it with some Kester solder 63/37 that I had on hand.

So how can I get those plates down to 250 on the tubes and what about the negative voltages that should be on the grid?  PRR posted a spec sheet in another thread showing they should be -12.

I appreciate the expertise you guys offer on the forum......

[PRR]

So what is the actual problem? Doesn't play like Eric or Jimi? Howls? Smokes?

It's not going to be your bread-winning every-night professional amplifier-- too odd and strange. EL90/6AQ5 are $15 which in today's tube market is not that bad. Yes they may stand 300V for years especially as a hobby/toy amp.

[Pavedog]

PR,

It seems like the lack of negative voltage on the grid and the high plate voltage will make it unreliable and likely change the tone of the amp too. Are you saying this is just how the circuit is designed?  Rissi worked for Fender as an engineer, it seems counter intuitive that his circuit would have these characteristics?

I like to make these reliable, and I have friend that expressed interest in using it for a lap steel guitar, so making it reliable is an added bonus.

[WimWalther]

So how can I get those plates down to 250 on the tubes and what about the negative voltages that should be on the grid?  PRR posted a spec sheet in another thread showing they should be -12.

Your grid voltages are fine.. it's just that in a self-biased amp, the bias voltage is obtained by holding the g1 at ground, while *raising* the k voltage above ground. So if Eg1 = 0V and Ek = 12V, that's the equivalent of saying g1 is -12V wrt k.

This is exactly how the 12AX7 stages are biased.. the current through Rk produces the necessary bias voltage drop. Now this technique also eats up gain, particularly in the small triode gain stages with Rk in the 1-3K range.. so the Rk is usually bypassed with a cap to restore the gain. You can see these bypass caps on the schematic and on the board.

As for the 300V Ea, I'm not so sure.

[Pavedog]

Wim,

Thanks for the explanation, I was just going through the circuit that feeds the grid and understand the bias.  Regarding the Plates being at 300V could I not us a resistor to drop the voltage down?

[WimWalther]

Why are you so sure you'd want to? Assuming everything else is in order, it's safe to say that this is how the amp was designed.

It's not at all unusual to see audio amps operating the tubes well above datasheet specs. For instance,  look at just about any EL84 design.

[mresistor]

If you are dead set on dropping the plate HV one method is to use a 50w zener diode on the center tap of the PT HV secondary. Example https://www.ebay.com/itm/383376554075. It has to be mounted in an area that can take a little heat and be securely fastened to a heat sink such as the chassis. It's not the greatest thing in the world to do but it works. The diode has to have a B at the end of the part number.
You would have to consider the resulting voltages on the rest of the power supply nodes as well. 


I tend to agree with Wim that this amp may have been designed to run the tubes a little hotter with Plate V than the design center max voltages in the specs. Maybe just run it like it is and see how it does over time. If it starts eating power tubes then it might be time to consider dropped the HV some.

[WimWalther]

I'll agree that it's still a bit of a puzzler. It is by all appearances a SE / Class-A design, and we normally expect LOWER plate volts and HIGHER plate currents in such case. Yet as it sits, it has the E & I typical of an AB design..

300V x 12mA is only 3.6W, plus a little more for screen, just call it ~4W. That's nothing, the tubes should live very long lives.

Really, the thing to do is a power test..  that should be instructive.

[Pavedog]

Wim, MR,

Thanks, I will take your advice and leave it as is for now.  Are the two output transformers something seen often in older amps?  As for the power, it was advertised as 12 watts back in the day but it sounds like that was just advertising........

I appreciate the education!

[Pavedog]

I'll agree that it's still a bit of a puzzler. It is by all appearances a SE / Class-A design, and we normally expect LOWER plate volts and HIGHER plate currents in such case. Yet as it sits, it has the E & I typical of an AB design..

300V x 12mA is only 3.6W, plus a little more for screen, just call it ~4W. That's nothing, the tubes should live very long lives.

Really, the thing to do is a power test..  that should be instructive.

That is good idea, should be able to get to it this weekend.

[WimWalther]

Thanks, I will take your advice and leave it as is for now.  Are the two output transformers something seen often in older amps?

Honestly, it's not something I've ever seen before. Dual OTs are what we expect on a stereo, not a mono amp. As with the absurdly-large resistors, it seems to be another example of using what was available.

Normally, the tubes would be running parallel, all on a single OT.

As for the power, it was advertised as 12 watts back in the day but it sounds like that was just advertising........

"Music power".. heh. But really, testing is the way to proceed.

If you do a power test, check it at 8 & 16 ohms. And if it makes more at 16, then try 32 if possible. With the two OTs in series, I'm curious which load is preferred

[Pavedog]

Was finally able to make some time to test the output of the amp using an oscope and bench meter, surprisingly it is putting out a max of 14.03 Watts RMS @ 8 ohms. I tested it at 16 ohms and it was just under 10 watts.  Now that I have spent some time playing through the amp it is pretty good. You get almost a Marshall type tone on one setting and then a Fender twin type on the other. You can blend the two together and get some interesting tones.

Thanks for all the help! 

[johnnyreece]

Congrats!  One thing I wanted to chime in and mention that might help you feel better (I skimmed, so apologies if it was stated and I missed it):  On a cathode-biased amp, I believe "plate voltage" is actually plate minus cathode, so you're probably somewhere around 285, not 300.  Hope that'll help ya sleep at night if the plate voltage number had you worried.