I've been working on finishing a slight variation of a Deluxe Reverb. The biggest differences are the tone-stack and independent bias adjustments. The first pic on the following webpage is a nearly up-to-date schematic of it: www.tduryee.com Note: I've since removed the master volume
The general tone is nice, but when I turn it up, even before getting fully into overdrive, there's a fizzy static when I play. I've tried all kinds of things: Snubbing caps around the Anode resistors, putting 68k on all the grids, experimenting with getting more and less gain from different stages, and lots more. Currently I have the reverb side-chain and the tone-stack entirely disconnected/bypassed. I still get the fizz.
A few other hints:
(When everything is connected) If I turn up both the trebble and the presence, the amp produces an extreemly high pitched screech whether playing or not. Also, reguardless of tone controls or volume, turning up the reverb causes uncontrollable feedback/noise.
Anyone have any ideas?
The general tone is nice, but when I turn it up, even before getting fully into overdrive, there's a fizzy static when I play. I've tried all kinds of things: Snubbing caps around the Anode resistors, putting 68k on all the grids, experimenting with getting more and less gain from different stages, and lots more. Currently I have the reverb side-chain and the tone-stack entirely disconnected/bypassed. I still get the fizz.
A few other hints:
(When everything is connected) If I turn up both the trebble and the presence, the amp produces an extreemly high pitched screech whether playing or not. Also, reguardless of tone controls or volume, turning up the reverb causes uncontrollable feedback/noise.
Anyone have any ideas?
Hi, I usually stick to hi-fi audio as opposed to guitar equipment. Even though both do many of the same things, there are different goals. My initial thought are, check the power supply filter caps. From your description some type of feedback is taking place. With high gain amps too little or improperly functioning power supply filters can create a path between stages. Without seeing the actual build, it is possible that a component in one stage is close to and is coupling to one in another and causing the problem. Coupling caps are known to do this. In high gain amps, parts placement can become critical. It appears from your description that there is more than one problem. I'm not certain, but is it possible that some of the reverb issues could be phase related. Is it possible to get the tank hooked up wrong? Not in my area of expertise, but I've seen similar problems in transformer coupled stages. The power amp portion seems fine. If the feedback was present at all times I would suspect that the negative feed back was actually positive and the output transformer secondaries needed to be swapped. A quick check on an O-scope would tell. I hope this gives you some new things to check and may lead to a fix.
My guess is that your problems could be electro-mechanical, due to the reverb side-chain. The fizzy thing sounds as if something is sticking. I once had a similar problem that I couldn't cure in the amp and it turned out to be a misaligned speaker magnet, The speaker coil was rubbing on the pole-piece. And that high-pitched screech sounds to me awfully like acoustic feedback. Is the speaker near to the side-chain?
Thanks for the ideas.
I've done some more testing and here's where I'm at:
I ruled out speakers as a culprit by trying two different cabinets with the same result. I forgot to mention in my original post that this is a head, not a combo, and I have the speakers intentionally about 4 feet away to avoid interaction.
I designed the layout to give as much space as possible between caps so I don't think that's it either. Also, for testing purposes I've removed as many unnecessary caps and series resisters as possible (snubbers, cathode bypass, etc.), bypassed the tonestack, bumped the 6v6 grid resisters up to 5.5k, and generally simplified this thing as much as possible.
I updated the schematic here:
www.tduryee.com
I originally built this amp about 2 years ago and it turns out I had covered most of the noise, probably thinking it was just a byproduct of the Baxandahl tonestack in a high-gain amp. There was a snubber on V3a&b and cap paralleling the M.V. to ground. With those removed, the amp is unstable even at "1". If I move any object near the V3b grid wire, there is a definite antennae effect / interference, so my next test will be to replace that with a shielded cable. I suspect there's more to it though.
I can't see it being a schematic design flaw at this point because if it is, the Fender Deluxe would never have worked. It's probably a layout issue, so I'm trying to isolate the main culprit to deal with it.
As far as the reverb tank: Yes, it's possible to hook it up backward but I've been careful not to. I'm guessing the problem with the reverb stems from this other issue though, so I'm bypassing the side-chain to work on one thing at a time.
That said, I'm a little curious about how the mixing circuit works. What is the effect of the 3.3M in parallel with the 10p?
Also, are there impedance matching concerns when coming off V3b anode to the splitter instead of using a cathode follower?
I've done some more testing and here's where I'm at:
I ruled out speakers as a culprit by trying two different cabinets with the same result. I forgot to mention in my original post that this is a head, not a combo, and I have the speakers intentionally about 4 feet away to avoid interaction.
I designed the layout to give as much space as possible between caps so I don't think that's it either. Also, for testing purposes I've removed as many unnecessary caps and series resisters as possible (snubbers, cathode bypass, etc.), bypassed the tonestack, bumped the 6v6 grid resisters up to 5.5k, and generally simplified this thing as much as possible.
I updated the schematic here:
www.tduryee.com
I originally built this amp about 2 years ago and it turns out I had covered most of the noise, probably thinking it was just a byproduct of the Baxandahl tonestack in a high-gain amp. There was a snubber on V3a&b and cap paralleling the M.V. to ground. With those removed, the amp is unstable even at "1". If I move any object near the V3b grid wire, there is a definite antennae effect / interference, so my next test will be to replace that with a shielded cable. I suspect there's more to it though.
I can't see it being a schematic design flaw at this point because if it is, the Fender Deluxe would never have worked. It's probably a layout issue, so I'm trying to isolate the main culprit to deal with it.
As far as the reverb tank: Yes, it's possible to hook it up backward but I've been careful not to. I'm guessing the problem with the reverb stems from this other issue though, so I'm bypassing the side-chain to work on one thing at a time.
That said, I'm a little curious about how the mixing circuit works. What is the effect of the 3.3M in parallel with the 10p?
Also, are there impedance matching concerns when coming off V3b anode to the splitter instead of using a cathode follower?
I swapped out tubes in all positions without the problem getting much better or worse. The grid solder joint is good, but the other end of that line has me thinking. When I turn the volume knob it sounds like a dirty pot and touching it's case with my needle-nose pliers causes crackle noises. Does it seem reasonable that the pot would be causing problems?
Hi, You bet. Either it got, cruddy inside, is worn out, damaged or possibly shorted inside to the case. I've had ones right out of the factory be no good. See if perhaps you can hang one along side it and swap the wires. It doesn't have to be a perfect match, just reasonably close. All you want to do is see if the problem goes away. Another possibility is that the case of the pot isn't grounded. I had one amp oscillate badly (at about 100KHZ) because of that.
Bruce
Bruce
I swapped out the pot and it didn't change the main problem. It is however slightly less sensitive to touch and doesn't make quite as much noise when turned. My circuit ground is tied to a ground bar rather than running through the chassis. Do you suggest tying the pot-case to the circuit ground or would that add more risk of ground loops?
If you have an oscilloscope you might want to check for oscillation. I noted that none of the 12AX7A have grid stopper resistors on them (1K resistor right in series at the grid pin of the tube socket) and this can be a problem particularly if the volume pots are any distance away from the tube they are driving. (Master volume, clean channel gain, dirty channel gain, etc.) Other locations like the phase splitter stage also benefit from this approach.
Note: I've worked on more than a few diy fender clones that had one or more oscillating stages. The above changes fixed the problem - careful lead dress and a compact design in some cases will too, but the grid stopper is pretty bullet proof.
Pot cases if not effectively grounded to the chassis by their bushings should always be grounded.
Note: I've worked on more than a few diy fender clones that had one or more oscillating stages. The above changes fixed the problem - careful lead dress and a compact design in some cases will too, but the grid stopper is pretty bullet proof.
Pot cases if not effectively grounded to the chassis by their bushings should always be grounded.
90% of the time when I see guys having that problem it is because the output transformer primary is reversed.
If you ever use the old fender schematics they generally show 3 wire colors on the primary, Blue, Brown, and Red CT.
Current transformer manufactures do not necessarily follow yesteryears transformer color codes.
I beat my head against the wall several times with the exact symptom, Once I flipped the brown and blue leads, problem solved.
Not saying this is the problem in your amp, but it is worth checking, I just restored a Marshall SL 100W a few weeks back and had the same problem.
Trout
I think that's a vey good idea.
I wasn't thinking of positive FB, because I'm more used to 'HiFi' PP amps using UL or pentodes, where the amount of NFB applied would usually be enough to cause permanent oscillation if it is accidentally connected as postive FB.
With a guitar amp, though, the lower level of FB that (I think) is used may be just enough to make it highly unstable, verging on oscillation, if it's connected the wrong way around. Then, the slightest stimulus could make it scream. It would also make the amp very sensitive to noise, hum, etc. that would not normally be noticeable.
I had built a couple 5E7 bandmaster clones about a year ago that drove me absolutely nuts with squeal. I had upgraded the output transformers to the much larger Bassman OT's.
The layout was very clear to connect Blue to V4, Brown to V5. But that was the 1950's transformer, not the 2007 model.
I finally out of desperation lifted the NFB wire off at the output jack. No more squeal, But this also made the presence pot almost non-functional.
When I finally got around to working on them again, I was amazed that was all that had been wrong.
Another givaway is the OP's other symptom.
If you lift the NFB you get very fizzy distortion, and if you hit the strings hard it will indeed have a certain crackle to the initial hit on the string.
I would think if he is applying positive feedback these symptom would worsen?
On higher gain amps Like the Marshall Super Lead 100W it sounds dreadful when lifted or reversed, I can speak from 1st hand experience on that one!
I did notice that his schematic does not give any color code orientation on the output primary. This might very well be the problem.
Trout
Well I added grid resistors on all but the reverb driver tube. I think it helped slightly, but not much.
Trout: You had me excited there about the OT but I checked the connections and they were correct. The OT is a Hammond 1645 with Brown going to V5 and Blue to V6. Just to be sure, I swapped the primaries... having trouble hearing now.
I've been trying to run this at full MV and roll the gain up to test. One thing I notice however is that cutting back on the MV and turning the gain up is much more stable but I don't even get into overdrive before I get that fizzy distortion.
Also, I just updated the schematic again to show the OT wiring more acurately.
www.tduryee.com
Trout: You had me excited there about the OT but I checked the connections and they were correct. The OT is a Hammond 1645 with Brown going to V5 and Blue to V6. Just to be sure, I swapped the primaries... having trouble hearing now.
I've been trying to run this at full MV and roll the gain up to test. One thing I notice however is that cutting back on the MV and turning the gain up is much more stable but I don't even get into overdrive before I get that fizzy distortion.
Also, I just updated the schematic again to show the OT wiring more acurately.
www.tduryee.com
Do you have a scope, and if so have you looked to see what stages might be clipping when this "fizzy" noise starts to become evident.
One of the things I am wondering about is how you sum the reverb and main out together - this circuit has over 20dB of loss, so you need to make sure that previous stages are not clipped when you attempt to make up that loss.
One of the things I am wondering about is how you sum the reverb and main out together - this circuit has over 20dB of loss, so you need to make sure that previous stages are not clipped when you attempt to make up that loss.
The combination of an O-scope that doesn't impress me much and guitar notes instead of a signal generator made this a little more time-consuming than expected, but it appears there's really very little clipping(not surprising: it doesn't sound like clipping).
There is a bit of a noise that shows up on the waveform at the summer junction. The more clear result was when I turned it up to the point of getting that high pitched noise. Beginning at the point where the side-chain splits off there's 14.67khz eardrum-shattering squeal. At the return of the side-chain it's 12.8khz. It acts like a diode: it's not there until you reach a certain gain level, then it jumps to about .3V. Turning the gain up further sends it up in the whole numbers.
I tried it without the reverb tube and the only difference was the frequency jumped to about 16.5khz.
I'm going to start wearing ear-plugs for testing...
There is a bit of a noise that shows up on the waveform at the summer junction. The more clear result was when I turned it up to the point of getting that high pitched noise. Beginning at the point where the side-chain splits off there's 14.67khz eardrum-shattering squeal. At the return of the side-chain it's 12.8khz. It acts like a diode: it's not there until you reach a certain gain level, then it jumps to about .3V. Turning the gain up further sends it up in the whole numbers.
I tried it without the reverb tube and the only difference was the frequency jumped to about 16.5khz.
I'm going to start wearing ear-plugs for testing...
Ok, more info:
I don't know how I missed this but on V3b, the source is about 280V and the plate about 225V...
So the fizzy part makes sense: It was clipping most of one half of the wave.
Excuse my inexperience with o-scopes. The wave was not a clear sign wave before, but for this kind of clipping I would have expected to see a mostly flattened top half of the wave.?.
Anyhow, I dropped the cathode R to about 550 just to test, and now with the MV down, I can max the gain and it definitely distorts rather than buzzing. It's a bit harsh but I'll worry about salt and pepper later.
Changing that hasn't changed fixed the original problems though: With MV up and gain above approx. 1, the amp goes unstable.
It's probably relevant to mention that turning the reverb up even without the reverb driver tube causes uncontrollable feedback. Also, the MV doesn't change that, it just changes it's volume.
I realize this reverb split/sum loses a lot of gain, but doesn't it sound like V3b is getting driven too hard?
I don't know how I missed this but on V3b, the source is about 280V and the plate about 225V...
So the fizzy part makes sense: It was clipping most of one half of the wave.
Excuse my inexperience with o-scopes. The wave was not a clear sign wave before, but for this kind of clipping I would have expected to see a mostly flattened top half of the wave.?.
Anyhow, I dropped the cathode R to about 550 just to test, and now with the MV down, I can max the gain and it definitely distorts rather than buzzing. It's a bit harsh but I'll worry about salt and pepper later.
Changing that hasn't changed fixed the original problems though: With MV up and gain above approx. 1, the amp goes unstable.
It's probably relevant to mention that turning the reverb up even without the reverb driver tube causes uncontrollable feedback. Also, the MV doesn't change that, it just changes it's volume.
I realize this reverb split/sum loses a lot of gain, but doesn't it sound like V3b is getting driven too hard?
I notice that V3a and V3b share a cathode resistor. This results in a potential feedback loop: signal current through V3b results in signal voltage across the cathode resistor which V3a amplifies and feeds back into the grid of V3b. The reverb knob controls the gain of the loop.
You might try replacing the 0.68u bypass cap with something bigger just to see if that helps. Separate cathode resistors might be necessary in the end.
Otherwise, I think you still have another problem which might be related to layout. IOW, another loop that includes the MV.
You might try replacing the 0.68u bypass cap with something bigger just to see if that helps. Separate cathode resistors might be necessary in the end.
Otherwise, I think you still have another problem which might be related to layout. IOW, another loop that includes the MV.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.