Maggie 9304 amp redplating--coupling cap value?

EastPoint

EastPoint

Recycler
Subscriber
I have a Magnavox 9304 console amp that was partially rebuilt--the previous owner replaced the 4 coupling caps but did nothing else. He put in .47uf and .152uf on each channel. I fired it up, and it sounds divine, but one of the EL84s is redplating. I tested the tubes--they're good. I am under the impression that the coupling caps should be something like .47 and .047, or .1 and .01. Is it likely that the redplating is due to his choice of cap values? What values should I go with? Thanks in advance!
 
Register to hide this ad
What is the B+ and what is the cathode resistor value or values if it uses one per channel . i use .1 uf and .047 uf for coupling caps .
 
See this thread: http://www.audiokarma.org/forums/showthread.php?t=516047 and check out post #4 and #29 and #30. Gordon (guru!) does a good job of explaining it. In my 9304 amp, I went up to .1/.01, then later upped that to .2/.02 and I thought I noticed an improvement in bass performance, slightly more bass, slightly lower, and better control. I think the 10x differential is important to keep the phase splitter balanced. The values in your amp are not 10x.

As far as red-plating specifically, I'm sorry I cannot help in that regard.
 
Last edited:
Cap value may affect bass response, but it cant cause redplating, unless the caps are leaky and throw off the bias voltage.
Check the voltages on the output tubes.
 
stick that tube in another socket and if it still redplates, it's the tube. If it don't, it's the socket or capacitor or both.
 
Sounds more like leaky coupling caps than wrong values, but I probably wouldn't run caps that large in one. The stock transformers aren't going to give you freqency response that low anyway.
 
battradio said:
What is the B+ and what is the cathode resistor value or values if it uses one per channel . i use .1 uf and .047 uf for coupling caps .
Click to expand...

The B+ is 341 volts. Which resistor(s) are the cathode resistor(s)? I've attached a picture of the amp's underside.

cademan said:
stick that tube in another socket and if it still redplates, it's the tube. If it don't, it's the socket or capacitor or both.
Click to expand...

I swapped the tubes around, and now nothing is redplating. I don't have anything playing through it, but it would still redplate regardless, right?

gadget73 said:
Sounds more like leaky coupling caps than wrong values, but I probably wouldn't run caps that large in one. The stock transformers aren't going to give you freqency response that low anyway.
Click to expand...

Should I just go ahead and swap in .1uf and .01uf caps and go from there? Or should I go with .2/.02?

Olorin67 said:
Check the voltages on the output tubes.
Click to expand...

Which pin(s) do I check? I'm still kind of new to this.
 

Attachments

  • P6030001.jpg
    P6030001.jpg
    109.8 KB · Views: 46
Cathode resistor is the 100 ohm ceramic jobber.


152 on the cap means .0015uf. I can't read the other one, but if its 473, that is .047 uf. Should be no problems with those values.

The way the numbering system works is a bit non-obvious. The first two digits are the value, and the third number is how many zeros follow it. That gives you value in picofarads. 152 would be 1500pf, or 1.5nf, or .0015uf, depending on how you like your units.
 
The caps say 473k and 152k, so I assumed they were .47 and .15. Is the "K" irrelevant? That is definitely confusing.

How do I measure the voltage to the tubes to determine if one is indeed getting too much?
 
off the top of my head I don't know what the K means. I'm sure its something obvious and relevant, but I can't come up with it.

473 should be .047uf, 152 is .0015. I think thats stock values, or close to it.

You could measure ground to each pin on the tube and post your findings. You can also compare it to the numbers on the schematic to see if they are reasonably close. I expect they'd be somewhat higher than schematic, but one will probably show considerably different than the others if there is something leaky.

Its also just possible that you had a bad connection that was solved by moving the tube around physically.
 
There is a generally mistaken impression that the engineer who designed any particular piece of electronic gear was an idiot and had no idea what he was doing. If the original cap was .1uF then .47 has GOT to be better because it will "make more bass." Well it turns out that this is entirely untrue. The coupling capacitor in a tube amp is part of a high pass filter in conjuction with the grid resistor of the next stage. The two passive components TOGETHER determine the slope of the filter!

For instance a .1uF cap and a 50k ohm resistor have a -3dB point of 31.8 Hz; a much smaller .047uF cap and a 470k resistor have a -3dB point of 7.2Hz. So you can clearly see from this example that it is not just the size of the cap that determines the slope of the filter but the RELATIONSHIP of the two components that is relevant.

Another mistaken impression is that one can just make the cap as large as possible and it will have no adverse affect on the performance of the device. It turns out that an amplifier with feedback has to be very carefully engineered with the time constants of the coupling caps and grid resistors selected so that at no time does the phase relationships of the filters exceed 180 deg. or the amplifier will turn into a oscillator. This frequently seen when someone replaces a coupling cap with too large a capacitor and the amp "motorboats" as a result of the amp becoming an oscillator.

There is also the mistaken impression that just adding capacitance will continue to "add bass." This is like saying that the amplifier is a quart bucket but if I can dump a gallon of water into it and it will hold it all. That's just naive.

So I hope it is understood that the engineer who designed the gear was not actually an idiot and may have had very good technical reasons for designing the gear the way it is.
 
gadget73 said:
You could measure ground to each pin on the tube and post your findings. You can also compare it to the numbers on the schematic to see if they are reasonably close. I expect they'd be somewhat higher than schematic, but one will probably show considerably different than the others if there is something leaky.
Click to expand...

Should I do that with the tubes in there or not?
 
Palustris said:
There is a generally mistaken impression that the engineer who designed any particular piece of electronic gear was an idiot and had no idea what he was doing. If the original cap was .1uF then .47 has GOT to be better because it will "make more bass." Well it turns out that this is entirely untrue. The coupling capacitor in a tube amp is part of a high pass filter in conjuction with the grid resistor of the next stage. The two passive components TOGETHER determine the slope of the filter!

For instance a .1uF cap and a 50k ohm resistor have a -3dB point of 31.8 Hz; a much smaller .047uF cap and a 470k resistor have a -3dB point of 7.2Hz. So you can clearly see from this example that it is not just the size of the cap that determines the slope of the filter but the RELATIONSHIP of the two components that is relevant.

Another mistaken impression is that one can just make the cap as large as possible and it will have no adverse affect on the performance of the device. It turns out that an amplifier with feedback has to be very carefully engineered with the time constants of the coupling caps and grid resistors selected so that at no time does the phase relationships of the filters exceed 180 deg. or the amplifier will turn into a oscillator. This frequently seen when someone replaces a coupling cap with too large a capacitor and the amp "motorboats" as a result of the amp becoming an oscillator.

There is also the mistaken impression that just adding capacitance will continue to "add bass." This is like saying that the amplifier is a quart bucket but if I can dump a gallon of water into it and it will hold it all. That's just naive.

So I hope it is understood that the engineer who designed the gear was not actually an idiot and may have had very good technical reasons for designing the gear the way it is.
Click to expand...

Yours is a pretty condescending response - I don't think anybody here is asserting that the design engineer is or was an idiot. Where are you getting this negativity?

The thread I referenced, with specific posts in the discussion, provide a context for my comments about coupling capacitance. (That's why I put the reference there.) Yes there is dialogue about slope and open feedback, but also consideration of the saturation point of the transformers and the hypothesis that bass response may have been purposely limited in the original design out of concern for turntable rumble and bass feedback inside a console cabinet, which is a tip of the hat to the original engineer, actually. The new values Gordon suggests (he's an experienced professional; I am not) are based on his modification and testing for improved stand-alone use.

And this is exactly why I referenced that thread and those posts. While I'm not sure it will help with the apparent red-plating problem the OP is experiencing, I thought it perfectly addressed his question about coupling capacitor values, particular when the ones in his amp seem to be so far off the mark.

Some of us are simply hobbyists and we appreciate an informative and educational response when we can get one.
 
The role of coupling capacitor values was well-explained by Palustris. They have nothing to do with current drawn by tubes - which is the reason overload (red plating) occurred. (And admittedly somewhat OT, nevertheless what he said regarding tweakers and urban legends are unfortunately too true. One finds that all the time. :sigh:)

Measurements are better done with tubes in and operational (necessary safety measures observed!). Grid voltage accuracy will depend on meter loading (preferably use a 10M or higher input dvm) and the rest on circuit topology. (I am not familiar with your amplifier's schematic.) It is also puzzling that overloading disappeared when you swiched tubes. Were the sockets etc. clean? Could you have dislodged some dirt/conductive matter when exchanging tubes?
 
So I fired it up and tested all the EL84 pin voltages with the tubes in.

Tube 1:
Pin 2: 0v
Pin 3: 12.5v
Pin 4: 12.5v
Pin 5: 12.5v
Pin 7: 325v
Pin 9: 324v

Tube 2:
Pin 2: 0v
Pin 3: 12.5v
Pin 4: 12.5v
Pin 5: 12.5v
Pin 7: 325v
Pin 9: 324v

Tube 3:
Pin 2: 0v
Pin 3: 12.5v
Pin 4: 12.5v
Pin 5: 12.5v
Pin 7: 320v
Pin 9: 324v

Tube 4:
Pin 2: 0v
Pin 3: 12.5v
Pin 4: 12.5v
Pin 5: 12.5v
Pin 7: 320v
Pin 9: 324v

While I was testing it, one tube briefly redplated, but when I swapped it for a different EL84, it didn't redplate anymore. I wonder if it was just a dirty tube socket? Should I douse the sockets with Deoxit? I'm listening to it right now, and no issues.
 
None of those numbers make me see anything obviously wrong. Plate, screen, and cathode voltages are pretty much what I'd expect to see. The grid (pin 2) is also at 0v, which is where it belongs. If you haven't, try cleaning the tube sockets and pins to see if that helps with the problem. It may be as simple as dirty connections.
 
Those coupling cap values are indeed stock, at least for the 93-04-10 like I have. The volts referenced are also right on or just a touch high perhaps, but nothing really out there.
 
The problem is your tubes are drawing different amounts of current (not matched) not the coupling caps. If your meter reads current, use the transformer shunt method of measuring current (will tell you what each tube is drawing, not just the pair). Switch tubes until you have the closest match in your pairs. If you cannot come up with suitable pairings, then you must get more tubes until you can.
 
guitbanger said:
The problem is your tubes are drawing different amounts of current (not matched) not the coupling caps. If your meter reads current, use the transformer shunt method of measuring current (will tell you what each tube is drawing, not just the pair).
Click to expand...

Can you describe the method in greater detail, i.e. where to make the pos and neg meter connections to read the current draw for each tube.
 
jheu, clamp on to the ot center tap with one lead and the plate with the other. A fused meter that reads current will tell you exactly what your current draw is. No guessing, easily and quickly repeatable. Don't blow up your meter (make sure your clip leads are up to the task of gripping securely). This way was taught to me by the late Ken Fischer of Trainwreck fame back in the early eighties. Works at the bench, backstage, or on the bus! lol
 
You must log in or register to reply here.
Back
Top Bottom