Mater Volume Interaction Problem.

Silent Sound

Member
Messages
5,407
Okay, so I just built an amp and it works great minus a few quirks I'm trying to suss out. The one I can't figure out is it's a two channel amp. Let's say it looks like this amp, but with a Marshall Plexi side, instead of Top Boost, a 6au6 instead of EF86, 6U4GB instead of EZ81, and 6V6's instead of EL84's:


There are lots of other differences too, like some some component values, but the phase inverter is pretty much the same, and that's where I believe the issue is coming from. Basically, I installed a master volume at the end of the tone stack on the "top boost" side and whenever I turn that master volume down to zero, the "EF86" side starts distorting like crazy at any volume (kind of a cool and aggressive distortion sound too, very unlike the smoothness of the "top boost" channel's distortion, so I don't think it's channel bleed.). If I turn that master volume up to about 1 or higher, the "EF86" side cleans up again, assuming I have the first preamp volume for the "top boost" side set to zero so it doesn't output through both channels at once. Both channels are connected at the input. It doesn't appear to be effected by any other controls, just that "top boost" master volume.

So, 0-1 on "Top boost" MV = distortion on "EF86" channel at any volume, with the heaviest distortion occurring at zero and it quickly cleaning up around 1-1.5.

2-10 on "Top boost" MV = "EF86" channel clean at most volumes, like you'd expect.

So do you guys think this is a design flaw or could I have wired something up strangely? Any ideas of ways to fix this, or should I just leave it be? What say you? I know I should have uploaded a schematic of this particular amp, but right now I've got it written in pencil over several sheets of paper and my scanner won't pick up pencil, so I'm waiting to get the design figured out before copying it down to a single sheet with pen so I can scan it.
 
Messages
826
Very strange! There is a symmetry to the two inputs to the PI, so your master volume on the 'top boost' side should behave similarly to the volume on the EF86 side, with respect to the effect on the other channel.

Is there any chance you wired the master volume after the .01uF cap, instead of between that cap and the 220k resistor?
 

Silent Sound

Member
Messages
5,407
Very strange! There is a symmetry to the two inputs to the PI, so your master volume on the 'top boost' side should behave similarly to the volume on the EF86 side, with respect to the effect on the other channel.

Is there any chance you wired the master volume after the .01uF cap, instead of between that cap and the 220k resistor?
No, it comes straight off the treble pot and goes into that coupling cap. But now that you mention it, I don't have that 220K resistor there. Mine looks a tad different, like I said. Imagine this preamp:

But connected to that phase inverter using the .022uF cap instead of the .01.

Yeah, it is a strange problem. I've got a scope and have been tempted to scope around and see if I can find something, but haven't had the time to get around to that just yet. Also, I don't entirely know what I'd be looking for in this situation. I had some positive feedback in the amp to begin with, mostly because the chassis is so small to have so much going on inside it, but after using shielded wires for most anything going to and from the tag board, it seems to have cleaned up nicely. Still, there are a few lonely wires that I didn't shield, because I used up all my shielded wire, but they're not close to anything and the amp is really quiet now, so I'm thinking they're fine.
 

pdf64

Member
Messages
7,912
0-1 on "Top boost" MV = distortion on "EF86" channel at any volume
My guess is that something (eg power amp or a pre-amp channel) is going into ultrasonic oscillation.
This can usually be cured by good lead dress / grounding arrangement / grid stoppers
 

schmidlin

Member
Messages
5,257
Well, I think you should properly diagram-out exactly what we are looking at before asking such questions. Saying it's kinda like this and kinda like that is groping in the dark. Grab a paper and pencil and a scanner...
 
Messages
826
Do you have a small cap between the plates of the PI, like the .47P shown in the Marshall schematic. That is there for stabilisation purposes, but I think it is only really needed in conjunction with global negative feedback.
 

pdf64

Member
Messages
7,912
If the problem stops when the cut control is adjusted to its least trebley setting, then it may be the power amp that's oscillating.
 

Silent Sound

Member
Messages
5,407
Do you have a small cap between the plates of the PI, like the .47P shown in the Marshall schematic. That is there for stabilisation purposes, but I think it is only really needed in conjunction with global negative feedback.
Nope. I don't have that cap installed. I also don't have any negative feedback loops installed, because I have a PPIMV and a cut control after the PI. Both are dual ganged 1M pots that go to ground instead of the opposing leg of the phase inverter like on the Vox schematic. They're also isolated by coupling caps before and after their connection to the PI. Why before AND after? Originally, I wanted to do a fixed bias power supply, but my PT doesn't have a suitable tap for a fixed bias supply. So I tried a back bias method, but I couldn't figure that one out on my own, and there's so little information that I could find on that method, that I gave up and went with cathode bias.

And yeah, Schmidlin, I agree that would be a better route. But I think between these schematics and any questions, you should be able to get a clear enough picture to at least throw out some suggestions.
 

Silent Sound

Member
Messages
5,407
If the problem stops when the cut control is adjusted to its least trebley setting, then it may be the power amp that's oscillating.
From what I recall (I'm at work now) I don't believe any of the other controls, outside of that one master volume, had any significant impact on it. But I don't recall specifically turning that knob so I'll definitely give it a try. At the least, I guess I should scope it and see where the distortion starts. I hadn't thought about it being a power amp issue.

Hopefully I'll get some time to work on it tonight. Unfortunately, I've got a really busy week ahead of me, which is why I'm asking questions instead of just trying things out and learning from my own mistakes. Plus, I'm kinda getting sick of these tiny orders to Mouser. Their shipping is ridiculous! So I'm trying to get as many possibilities out there for when/if I have to order parts, so I won't have to make too many separate orders. I'd rather order ten parts I don't need than go back and have to order one more part that I do need.
 

donnyjaguar

Member
Messages
4,194
The design of the output section isn't sound. The position of the "master volume" control will upset the operating points of the phase invertor tube sections. Other issues too, like the .1uF capacitor on the Screen grid of the pentode. I think you'll find 22nF is a much better choice here - unless of course you plan on amplifying the sounds of blue wales. :)
 

pdf64

Member
Messages
7,912
The design of the output section isn't sound. The position of the "master volume" control will upset the operating points of the phase invertor tube sections
I agree that will happen and it might be viewed as a non-ideal performance characteristic, but it seems to work fine in 'some well thought of' amps, eg Matchless Lightning http://www.prowessamplifiers.com/schematics/post/matchless_lightning.pdf

The volume controls in a 5E3 may be seen to operate in a similar manner (by loading the preceding stage), also several of the level controls in Mesa 'Mk' series amps, eg http://schems.com/manu/mesaboogie/boogie_mkiv.pdf

Other issues too, like the .1uF capacitor on the Screen grid of the pentode. I think you'll find 22nF is a much better choice here - unless of course you plan on amplifying the sounds of blue wales
22nF seems to be a lower value than generally used, eg Voxes tended to use 220nF; how have you come to it?
 
Messages
826
By the way, the cap between the PI plates on the Marshall schematic should be 47pF. I don't know where Marshall got that decimal point!
 

Silent Sound

Member
Messages
5,407
Well guys, thanks for the help! I was able to get enough free time last night to scope it out and swap out the screen grid and cathode bias resistors to get the power tubes running in a more comfortable range. Upon scoping the phase inverter, I noticed that one side was running about half the voltage of the other. Digging in deeper, I found a tiny piece of solder had dripped and connected the control grid and anode of one side of the 12ax7 socket at the PI. Cleaned that up, and now that volume "interaction" problem is gone and I've got more gain and clearer tone (of course)! I didn't think a triode would work at all with the control grid and anode connected, but apparently it can. It just doesn't work well.

So, with my screen voltage now a comfortable 10 volts below my anodes on the 6V6's and a healthy 90% bias it sounds really good and runs pretty stable. PT gets warm, but never too hot to touch. All the controls work as expected... except for one.

Before yesterday, my treble seemed to work too well, almost like it had quadruple the sweep of the mid and bass pots on the plexi side. I didn't consider it an issue, as it worked. But now, after fixing everything else, my treble knob controls absolutely nothing! The mid and bass work fine, but treble doesn't even control the volume! Which is weird, because they're all wired up too each other, and seeing how the treble is a mix resistor, I would think it should do something so long as the pot was wired correctly and functioning. I measured the pot and it works fine enough and sweeps appropriately (240k Ohms one one side, 270k Ohms on the other). I checked all the wires and everything seems fine, no unexpected shorts or anything. It almost seems like the treble is stuck all the way up. Maybe while scoping around I shorted something and blew the treble cap. It looks fine, but it's a Silver Mica, which can be fragile. I wasn't working on that side of the board, so I'm really at a lost as to what's going on and what happened but I can replace it easy enough with a ceramic cap and see if that fixes it. Otherwise, I'm pulling the scope back out and crossing my fingers. Now, I just need the time!
 

Blue Strat

Member
Messages
30,197
Well, I think you should properly diagram-out exactly what we are looking at before asking such questions. Saying it's kinda like this and kinda like that is groping in the dark. Grab a paper and pencil and a scanner...

Agreed. Details are everything and the OP is probably describing how he sees it in his mind as opposed to how it really is.

No sense in trying to diagnose something we can't see.
 






Top