trinityamps.com

Not surprised at all. As I noted in my earlier post I've heard this in a number of amps, and even in some live recordings, with EL84's and even EL34's, generally in "Marshall-ish" circuits. I didn't notice it in the TC-15 I build for Fortis, nor is it noticable in my Hot Cat 30 clone.

After reading this last update I give my JCA20 a hard run and there is a touch of this effect in that amp too, but only a touch. It has to be really wound out to be noticable at all, and I think that amp needs to be reigned in a bit to be at it's best so its not really an issue (for me).

_________________
SG + Brit tube head = Nirvana

I tried Paul Ruby's suggestion of a resistor across the speaker jack but did not have the right cap to add it

the straight resistor did not seem do much except get really hot - ha ha. Did not really address the swirly buzz I have going when the amp is wound up

Not sure if adding the cap will do anything different but I have one coming and perhaps it will help. Also tried the Guitar Player mag filter like I posted earlier in this thread. Did not help either

I have also since rewired the board to the Trinity Layout vs the original Ceria-tone layout thinking that perhaps the old PI placement at v2 might have had some effect on the buzz - amp is very slightly quieter now, probably due to my more careful soldering etc, though the buzz remains.

Have the parts for a Ruby mod and will try that too at some point.

I'm also almost done my Trinity Plexi kit which is built to the "brown" layout and will be interesting to see how that amp performs

If there was a good sound clip of the way your amp sounds, it would help figure out what we're actually trying to deal with here, as the descriptuve terms are somewhat ambiguous. Otherwise it feels, in the words of Bruce Springclean, like "we're just dancing on the dark".

_________________
Great tones come in small glass jars!

There was earlier in the thread. Here it is again. If you listen carefully you can hear a fizzy buzz especially as the chords decay

http://snowboardmaterials.com/tmb_fizzz.mp3

OK, this is a good clip. A clip had been posted previously that wouldn't play properly on my non-Mac laptop. To me this sounds like a pretty clear case of the "Paul Ruby buzz", that gets fixed with the dual diode-zener combination. Sometimes when people complain about buzz, they mean a buzzy quality within the distortion tone, which is the kind of thing you would use the Zobel filter for. This insect like buzz trailing at the end of the note is what Paul Ruby describes. You just need to experiment a bit with the zener voltage. I would generally try one or two volts above the EL84s' cathode voltage at idle.

_________________
Great tones come in small glass jars!

I have tidied up my lead dress. Added shielded cable on all volume controls. Lowered grid leak/reference resistors on the power tubes to 100K. Increased grid stop resistors on the power tubes to 100k. None of these changes seem to help with the buzzy overtones when cranked. The only thing that seems to help is dialing back the PPIMV to around 7-8. I am really at a loss here. I am going to order a new set of EL84s to try next.

Any recommendations? I have JJs now.

_________________
Les

http://www.soundclick.com/unimind

Maybe some that have a little more clean headroom that do not break up as early.

_________________
Les

http://www.soundclick.com/unimind

Try a pair of 7189s, which are pin-compatible and have a very close spec to EL84s. You could also use 7189As, if you make a small wiring change on the sockets. Apart from being rated for higher voltages, they are actually tetrodes inside the glass, and have a later breakup and less likelihood of buzzy overtones. Here I'm assuming that your "buzzy overtones" sound different from the kind of buzz in Lindz's clip, which appears to be the kind investigated by Paul Ruby.

_________________
Great tones come in small glass jars!

Hey Les

Thought I would update this thread now that I had some time to work on my amp

A couple weeks later and I installed the "Ruby mod" on my amp and it worked like a charm ... I am pretty stoked!

Installed 14 watt zeners (measured 12 on my amp) w the 1N4007's and voila.. very little swirly buzz on the decay of notes and chords as in my original sound clips - there is a TINY amount still but I am quite happy with it and as other posters suggested, a little gives some pretty cool overtones

I'd suggest you try it as it was a pretty simple install and I think we were chasing a similar issue when I listen to your clips.

Sweet. I have some new tubes coming and then I was going to try the Ruby mod. I was waiting to see how it worked for you. Looks like I will need to do it as I am pretty sure we have the same issue.

By the way, did you see my post on your thread about the 470k resistor on the TMB gain pot? It should help you.

_________________
Les

http://www.soundclick.com/unimind

From my Bad OT??? thread (viewtopic.php?f=7&t=3038):

_________________
Les

http://www.soundclick.com/unimind

I don't agree that it's necessarily to do with V1a directly. It's just the total amount of end-to-end gain in the amp is so high that it will easily go into self-oscillation at high frequency. Especially given the higher-frequency emphasis of the bright channel. Your task is to work through the amp and reduce the amount of signal being radiated from the PI and power amp into the sensitive preamp front end.

One other thing - are you applying NFB to the tail of the PI at all? That's something that can complicate this kind of problem.

_________________
Great tones come in small glass jars!

That's the thing. I am not really seeing anything obvious. I have 'chopsticked' things a lot and even tried to move (reposition and re-solder) things around to no avail. I am not sure what to try next.
Yes I am. When I switch it off though it actually makes it worse.

_________________
Les

http://www.soundclick.com/unimind

What change are you refering to Phil?

Just work through it methodically. Consider using shielded wire for every longish signal connection, and gradually elimenate all possible ways signal could get radiated into the front end of the preamp. After that you may also try placing 5pF caps between the anode and grid of some of the gain stages, to help kill ultrasonic oscillations.


The reason I asked is that with bigger Marshalls the PI output becomes severely imbalanced when they go into clipping, due to part of the NFB being applied to the PI tail. Under linear conditions that would make the PI appear more balanced, which was Leo Fender's original intention. When an amp goes into clipping, the normal rules of NFB fall apart and the result is sonic mayhem. So what happens if you apply NFB only to the "B" input of the PI, but *not* to the PI tail? While this may not cure your parasitic oscillation, it might stop one power tube getting driven far harder than the other when the parasitics kick in and apparently overload the PI.


Just compare the pin-out diagram of a 7189A to a 7189 or regular EL84, and you'll see what I mean. 7189As have slightly different pin connections.

_________________
Great tones come in small glass jars!

I put in new EL-84s (MESA EL-84/6BQ5) and a new rectifier (Electro Harmonix 6CA4). My voltages went up quite a bit. Here are the current readings:

• Mains: 121ac
• B+: 362
• V1 pin1: 147 pin6: 162
• V2 pin1: 170 pin6: 274 pin8: 172
• V3 pin1: 212 pin2: 65 pin3: 101 pin6: 208 pin7: 72 pin8: 101
• V4 pin3: 12 pin7: 352 pin9: 338
• V5 pin3: 12 pin7: 356 pin9: 339
• V6 pin1: 311ac pin3: 366 pin7: 311ac
  

I also wrapped some foil around a piece of cardboard to use as a sudo-case shield. There is still some fizzie swirl if I set the PPIMV above 8 but with it at 8 the amp sounds amazing. The shielded cables and sudo-case shield have helped reduce the swirl a bit but I guess I may just need to try the Ruby mod next.

New clip:
http://www.unimind.us/Plexi_Noodle1_wet.mp3
http://www.unimind.us/Plexi_Noodle1_dry.mp3

_________________
Les

http://www.soundclick.com/unimind

I have now rewired some more. Upped the mix resistors to V2 to 1M to reduce the signal entering V2. Lowered the PI grid reference resistors to 220K to reduce the signal entering the PI. Tried the 'Ruby Mod' (had no affect). Tried the 5pf caps across the plates of V1 A and B (had no affect). Tried removing the V2A cathode bypass cap and the bright channel mix cap to reduce some high end (which I now know did not reduce any high end it just allowed the low end back in). The one thing that has helped very slightly is re-routing the OT secondary wires. That seemed to have a slight affect on the fizz. I have tried other things like using 2.7K resistors on the V1 cathodes, adding 470K grid reference to V2A, changing the cap across the PI plates to 47pf but none of these really had any affect on the fizz. They just lowered the gain a little and/or changed the sound and feel but the fizz is still there.

Here are the latest pics. If anyone has any thoughts or suggestions please let me know.

_________________
Les

http://www.soundclick.com/unimind

Two more pics:

_________________
Les

http://www.soundclick.com/unimind

Les - as part of my learning and chasing issues in my one amp I came across this thread that may be of help. Guy went to town trying to chase what he called "harshness" in his amp when it was dimed. He's working with a Wreck style amp so some comments are not exactly specific to an 18 watt but much would still apply

Good reading for sure and I am sure you can get something out of it...

Sounds like a host of things made small incremental improvements but this part may be of interest if you have not tried it yet (or perhaps in conjunction with other things)

"Just before I buttoned the chassis back up, I thought I might as well try the 'big cathode bypass cap' trick on the EL84s. I squeezed in a 2200uF cap soldered across the existing 220uF one - what a change! This is almost as significant as the split load PI. Much, much smoother distortion sound...."

http://ax84.com/bbs/dm.php?thread=434311

Since I fried the OT on my Ceria-tone clone of Unimind's brown layout (faulty speaker connection) I decided I would take the opportunity to completely rebuild it to see if I would still get the buzz without the Ruby clamp on the amp, assuming that my layout, lead dress or some oddball component was causing the problem.

Interestingly it still buzzes after replacing every single part on the amp except the PT and the pots - I completely rewired and replaced every single component, modded the board to match the Trinity layout, and paid better attention to lead dress etc - I am curious if the PT may be the issue?? I mentioned in another post that my MM PT in this amp is rated as 315 watts (vs 290 for Trinity PT's) .. Also it is a Ceria-tone steel combo chassis and it has no sides - perhaps signals radiate around the chassis a little differently .. Perhaps I repeated some layout and lead dress mistakes?? It is a bit microphonic on V1 so I may mess around with the leads to that socket a bit

It buzzes a small amount less than before but is still quite noticeable especially if I use my attenuator to dumb it down to sane volumes (We-ber Mass)

Oddly the Trinity kit I made does NOT buzz, and the amps are very similar now -- perplexing to say the least. I'm going to put a Ruby mod back on later this week and see if it kills the buzz like it did on the initial build

Thought it might interest those that are following this thread