Twreck Phasor/Vibe/Fizz Effect

[John C]

My homebrew Trainwreck has some unwanted distortion that reminds me a bit of a phasor or Uni-Vibe style effect. The "effect" is present at any & all notes & strings when the amp volume set high for dirt/compression (with the guitar volume turned up also). I cannot hear the problem when the amp is clean, such as when the amp volume is at ten o' clock or less, or with the guitar volume rolled down enough to prevent the Twreck overdrive & compression. This phasor-like sound reminds of a Pro Junior's sound when cranked up, if you've ever heard one of those amps.

What I'm calling a "phasor" type sound might actually be the sound that "JimmyR" describes as overdrive "Fizz" in his "Anyone try a choke in an Express?" May 16th, '07 post.

I wold love some help with this, so any comments or suggestions, please let me know. Thanks!

John

[dartanion]

Try the Paul Ruby Buzz Fix.

Mr. Ruby hangs here so I'll let him chime in with the details. I have them somewhere but can't find it. I would point you to a few threads on the 18watt.com, however it is temporarily shut down.

[John C]

Hey, thanks. I'd love to hear about it.

[joe6v6]

John - Heres a link that should get you two the Paul Ruby Buzz fix page, http://www.18watt.com/storage/18-watter ... fo_311.pdf
to be honest what you describe sounds to me like a severly mismatched set of power tubes. Did you check the bias individualy? The reason I bring this up is an amp that I worked on actualy had a tremelo effect because of mismatched power tubes. Also lead dress is important, try messing with the presence wire & the wires to the tone controls. Any pics of the guts?

JOE

[John C]

I'm staring to wonder if there are actually two separate problems instead of one. In other words, I wonder if the tremolo effect is the result of one problem, and then I have a “fizz”/”buzz” as a result of another problem. I’m not sure about that though, and hopefully there’s just one culprit. I can tell you that I ordered the (EL34) tubes as a matched pair, but I certainly could try a different set tonight. I did, however, change the 82k resistor (using the “Kelly ‘90” scheme as my reference) to a 100k in the phase inverter, so now there is a 100k resistor going to each plate of the inverter. The voltage now measures about 6 volts across the two plates of the inverter (it was slightly higher before I changed the resistor; maybe 6.5V). The matched resistors seemed to produce a little less noise, some of which I might even say sounded a bit like “white” noise.

I have checked the bias for both and it seems fine; although I had to significantly decrease the resistor values for the bias circuit (based on the Kelly scheme) to get a bias supply going at all. That fact still makes me wonder a little, even though the bias supply and power supply voltages seem fine…

Thanks for the link! It looks great so far… I’ll have to finish & reread the article in a few minutes here… I appreciate everyone’s help.

[John C]

An update: I decreased the value of the caps that come off the plates of the PI and feed the power tube control grids. The "fizz" is gone, the vibe is still there (just a subtle/moderate sound), and I've lost a little tone. I changed the values from .02 to .001, and the results were interesting. The amp is now quite fun to play (still very dynamic), but again, some of the tone is gone. A little bit of top end seems to be missing and there is definitely less bass. I'll wait 'till tomorrow to do some more soldering...

Any comments or suggestions, again, please feel free.

[Shermock]

Hi John C,

I'm wondering a couple of things...

Could you post a gutshot so we could see what you've built? I'm thinking about lead dress, type of caps/resistors, etc...that may be contributing. You could have cathode bypass cap not really working well, the .002 or .1 coupling caps could be acting up, etc... Blocking distortion in the preamp...

How faithfully did you follow the layout, say, of Francesca?

Do you have 1.5K grid resistors on the power tube sockets?

By changing your PI caps to .01 you've removed bass frequencies from the mix, like you said, and it seems to help the "spitty." That may mean the power tube grid resistors need to be changed to a larger value. Some Marshalls had as big as 5.6K.

Y'know, one thing leads to the next. Say you have some oscillation generated in the 1st gain stage through the tone stack. That's going to disturb the already tenuous relationship between all the gain stages, and could lead to so many different interrelated problems...none of which may show up as an audible squeak or howl.

My earlier version of an Express would just plain "shut off" past a certain point on the volume knob. I happened to hook my meter up to the 1 ohm cathode/bias resistor on a power tube socket. As I turned up the volume I watched the tubes suddenly draw current when the amp would "shut off." I realized the power tubes were working very hard to amplify a signal that was ultrasonic. So I knew there was an oscillation somewhere in the extreme high freq.'s that I had to get rid of. I removed the bright cap on the volume knob and that did it.

Maybe you've done all this and the problem is still there. I would love to see a pic of your amp!

HTH

Shermock

[John C]

Well, I don't have a guts picture right now and that's probably a good thing! It's a little messy right now because I've been doing a lot of experimentation with the components. Certainly lead dress could be my issue, but before I made all the experimental changes, the layout was similar to Francesca and the buzz has been constant through a couple of major layout changes. I do in fact have the 1.5 grid stoppers in place and I did try the 5.6k stoppers with no difference. I can also tell you that I've tested the circuit by sending the output at all the different stages to another amp. The buzz doesn't "appear" until I get to the output of the power tubes (unless I just couldn't/didn't hear it prior to that).

Can anyone tell me if they have any experience with a specific part of the lead dress to concentrate on? For instance - keep the input leads away from the transformer leads, or something like that...

Shermock, could you hear the oscillation the bright caps were allowing in your amp? Seems like you were implying that you could not.

As I mentioned in an earlier post, I had to significantly reduce the resistors in the (fixed) bias circuit to get a bias supply to work. I can get the appropriate bias to appear now, but the bias pot (20k, as the scheme says) is rather touchy now. When I first start turning up the bias pot, I get nothing, nothing, nothing... and then all of a sudden I get strong amount of bias supply. I can dial it in with some careful turning, but I can't help but wonder if that's a clue that the real problem is somehow power/bias supply related.

[mdroberts1243]

Well, I don't have a guts picture right now and that's probably a good thing! It's a little messy right now because I've been doing a lot of experimentation with the components. Certainly lead dress could be my issue, but before I made all the experimental changes, the layout was similar to Francesca and the buzz has been constant through a couple of major layout changes. I do in fact have the 1.5 grid stoppers in place and I did try the 5.6k stoppers with no difference. I can also tell you that I've tested the circuit by sending the output at all the different stages to another amp. The buzz doesn't "appear" until I get to the output of the power tubes (unless I just couldn't/didn't hear it prior to that).

Can anyone tell me if they have any experience with a specific part of the lead dress to concentrate on? For instance - keep the input leads away from the transformer leads, or something like that...

Shermock, could you hear the oscillation the bright caps were allowing in your amp? Seems like you were implying that you could not.

As I mentioned in an earlier post, I had to significantly reduce the resistors in the (fixed) bias circuit to get a bias supply to work. I can get the appropriate bias to appear now, but the bias pot (20k, as the scheme says) is rather touchy now. When I first start turning up the bias pot, I get nothing, nothing, nothing... and then all of a sudden I get strong amount of bias supply. I can dial it in with some careful turning, but I can't help but wonder if that's a clue that the real problem is somehow power/bias supply related.

The bias supply certainly raises suspicions. I think I read somewhere that biasing too cold can cause a buzzing sound in the amp. The values in the Ceriatone kit and the modded Kelly schematic worked great for me.

The Francesca pics are great for guidance on the lead dress... my Ceriatone build worked out great by trying to faithfully follow the lead dress in those pics.

I played with a grid resistor on the 2nd gain stage of V1 to try and tame the amp from breaking into oscillation too easily, but all my problems with tubes and extreme sensitivity were solved by focusing on getting the voltages right according to those in the Francesca-modified Kelly schematic. See: http://ampgarage.com/forum/viewtopic.php?t=1408

[John C]

Mark, can you give me some numbers for your voltages? Where did they start and how close did you end up getting them to the numbers on the scheme? Mine are all a bit high, but I figured they were acceptable. The wall voltage at my house is about 124V+ AC, and the B+1 ends up being about 415V DC with the following B+ voltages similarly about 5% higher than the numbers on the schemes.

I, too, have have dropped that 150k resistor down quite I bit. I like 38k or 47k much, much more than the 150k. That really has been a major key to getting the clean-to-mean feature via the guitar volume. At 150k, it was "mean" all the time with the amp volume above 10 o' clock.

[mdroberts1243]

Mark, can you give me some numbers for your voltages? Where did they start and how close did you end up getting them to the numbers on the scheme? Mine are all a bit high, but I figured they were acceptable. The wall voltage at my house is about 124V+ AC, and the B+1 ends up being about 415V DC with the following B+ voltages similarly about 5% higher than the numbers on the schemes.

I, too, have have dropped that 150k resistor down quite I bit. I like 38k or 47k much, much more than the 150k. That really has been a major key to getting the clean-to-mean feature via the guitar volume. At 150k, it was "mean" all the time with the amp volume above 10 o' clock.

At 116V AC with EL34s installed my voltages ended up at: Raw DC 418V, B+1 393 VDC, B+2 383, B+3 302, B+4 285, B+5 271... these are within 3 to 4 V of the Version A1b schematic.

My problems stemmed from a colour-code mixup on the Ceriatone PT... I had wired the primary such that I was using the 100V tap with my nominal 120V AC wall voltage. B+1 was 480V and B+5 was 318VDC.

At the elevated voltages I went through my entire stock of tubes (both preamp and power tubes) suffering from various problems like extra sensitive to breaking into oscillation, microphonic, blocking distortion, 'spitting', etc. I was able to get stable performance and worried that this amp was way too sensitive to tubes. The best power tubes at the high voltage were chinese 6V6GTs... The amplifier sounded great though and that kept me from seriously checking out the voltages, as I assumed they were correct by design.

I think getting below 300V on the lower B+ voltages might be important. With the new (proper) voltages tube selection was a lot easier, although a lot of tubes seem to be microphonic.

[Shermock]

Hey man!

No, I couldn't hear that ultrasonic oscillation. I knew it was there because of watching the meter.

Are you any closer to a cure? This is interesting for the observer, more frustrating for the afflicted!

Shermock

[John C]

Hey, thanks, Mark!

Shermock: Although I don't necessarily know if I'm curing the true problem with the sound, I have made progress with the symptoms. And if you want interesting, try this: I've been able to nix the "fizz" by mismatching the control grid input caps (normally both .02 on the schemes). At the moment I have .001 on one tube and .1 on the other. It doesn't make any difference which tube gets the bigger or smaller cap. Secondly, I tried a choke in place of the 25W R and I prefer the choke. I don't know the specs on the choke (borrowed it from a Blues Deluxe), but it looks tiny and I have a big one on order already! I also tried the choke and the resistor in series, but didn't care for that.