Tremolo Help

[JazzGuitarGimp]

Sorry if this has already been discussed - I've arrived late to the party I'm afraid. Have you tried a few different tubes in the oscillator circuit? A weak tube will cause the oscillator to behave erratically. There is a minimum gain requirement for it to work first time, every time.

[tubeswell]

Is the board made of that black papery stuff by any chance?

[C Moore]

This gets more discouraging as I go.
I have had a few friends play the Amp/Head.
They say it sounds "great"....with the Trem not working...so the potential is there.

Back to the Trem Problems.
I pulled the pots and jacks...lifted the board...and everything really looks OK.

I trimmed all the leads a bit shorter...but there was really nothing long enough...or bent enough to have caused any kind of short...not that I can SEE.
When I first started chop-sticking the baord...it was REAL sensitive. Just a light touch and the Trem would cut in and out. I replaced some stuf...and then I had to push harder on the board to effect the Trem.
Then it cut out all together...and nothing would make the Trem work.

I guess I will just button this back up and see what happens.
How likely is it that...somehow...the glass turret board is conductive...and is somehow causing a short...from one turret to another...or something like that.?
This is all beyond my knowledge of electronics. Just not sure what to do if this does not work after I get it back together.
Thank You

[Firestorm]

Hate to say it, but if it's not an assembly problem, it's a design problem. These things can be awfully finicky.

[martin manning]

How about the 6SL7 tube? Is it strong?

[C Moore]

Yeah...tried several SL7.
They are either Modern Day Tung-Sol...or JAN NOS. They work as V1 or as the PI...but the same tubes will not work in the Trem Circuit.
I wonder if Fire is correct.? And that I have some weird...proximity...lead dress...kind of a thing.
Well...will put it back together...if it does not Trem...I will build a different Trem Circuit.
I realize what Martin is saying...that there is nothing "wrong" with the circuit on paper...but I think I just need to change my luck at this point.
So...I guess I can just steal something from a Brown-Blonde-Black Face Fender
One of their bias vary circuits...that uses a dual-triode.?
Thanks

[martin manning]

One last thing you could try if it still doesn't work is increasing the plate resistors to 470k and the cathode resistor to 2k2. That will increase the gain a bit, which might help it to oscillate. The amplification factor on that tube is only 70, so it needs all the help it can get.

[Phil_S]

Hate to say it, but if it's not an assembly problem, it's a design problem. These things can be awfully finicky.

If you are using a 6SL7 as the oscillator tube, it may not have enough gain to trigger. The octal 6SQ7 has gain factor rating of 100, but it is only single triode. You may not care for this suggestion, but I'd try to shrink that hole to install a 12AX7. There are adapters made for octal cut outs or you can improvise with a small piece of sheet metal and re-drill it. Good luck.

edit: here is Aiken's article on the topic...pay attention to the gain discussion. The tube's mu rating is just a rating and even a 12AX7 could possibly be biased below the 29 minimum he states (I guess). I'm working on the assumption that the higher the mu rating, the easier it will be to get the tube to trigger. After a quick look at Aiken, maybe you simply need to bias the tube differently? I'm not particularly quick when it comes to electronics theory and the calculations. I also see now that you are using the Ampeg design and the 6SL7 should work. Article here... http://www.aikenamps.com/PhaseShiftOscillators.html

[Firestorm]

It doesn't have to be a dual triode; the 5F11 circuit works great. But I think a factor 70 tube should have enough gain if you want to do the two triode circuit. The trick is getting the LFO to turn on. There's a certain genius in the blackface circuits (these drive a neon bulb which you don't want) but the LFO is turned on and off by the application of raw negative bias to the LFO grid. These things never fail to start (problems with them are elsewhere) . You could try adapting that.

[C Moore]

You guys are awesome...all of you.
This chassis had octals...which I kind of like...so I went with those...but I am not married to 8 pins.
But...sense I was using an octal...i looked for a Trem Circuit that used 8 pins. So I found this circuit in an ols Ampeg. I built one other amp with this tube and trem circuit...and it works great.
BUT...the power tubes were cat biased...and the Trem went straight to the power tube grids if that matters.
As long as I can stay with one dual triode (due to space) I do not mind switching to AX7...and getting a Surf Bug if I can make it fit.
I will take all this in...and then ask you guys the best way to do what I decide on.
Thanks again for all the help and suggestions.

[cbass]

Something else to consider if you have resoldered a lot you can get flux residue on the board that can be conductive.Happened to me once.
SO you may want to check for stray voltages on the board and clean it real good

[Phil_S]

...BUT...the power tubes were cat biased...and the Trem went straight to the power tube grids if that matters.

I did a wise-ass build one time, a Vibro-Champ circuit with fixed bias. I had a wild hair because I used an Antek torriodal PT that had a winding that made for a perfect bias supply. I never did get the tremolo to work. I finally gave up and made it cathode bias and then the tremolo worked fine.

That Ampeg, IIRC is a floating paraphase design. If you used fixed bias, then I think you were just tempting fate. I don't have any clue what is so different that it worked once, but not a second time. Make that problem amp cathode biased and expect it to solve the problem. If it does, then you know it was a design problem for sure.

[C Moore]

Hey Phil -
If you go to the third post in this thread...I have a link to the original schem (for that other amp I built).
Is that the type of PI you are talking about.?
With all the work I did to get dual bias to work (courtesy of the help from TAG) I kind of hate to change to cat bias now.
But can I find I different Trem Circuit that will work with a SL7.?
So far I have been looking at Fender:
6GL6
6G11
5F11
6G2
I do not have enough electronics knowledge to know if any of those will work in my amp...with a 6SL7 as the oscillator.?
Thanks

[martin manning]

As I previously pointed out, the max heater-cathode voltage for a 6SL7 is spec'ed at 90V. The Fender circuits use the oscillator to drive a cathode follower stage which has no plate load and a large cathode resistor. This means a high cathode voltage, as much as 250V at the 400+ volts used for the supply. By the way, what is your supply voltage?

[Phil_S]

Hired Hand:

Did you post a schematic of exactly this problem amp? The thread is too long to easily back track. Other comments follow.

1) Martin knows what he's doing. His knowledge of tube amps eclipses mine by leaps and bounds. He continues to focus on the h-k voltage. I suggest paying attention. Your amp is outside the proper operating parameter for h-k voltage.

2) I don't believe this is a matter of whether you are using a 6SL7 or 12AX7. Though the 6SL7 tube has a lower mu (70 vs 100) it is still in the relevant range and can be biased properly to provide enough gain. As noted above (probably by Martin), it is harder to get the 6SL7 to work, but harder doesn't mean it won't.

3) To answer about the PI, the Ampeg schematic shows what's called a floating paraphase inverter. Take a good look at it. The second triode grid (bottom in the drawing) is decoupled from the 470K/510K by a .02uf cap; also the grid is grounded via a 470K/300pf R/C. Without getting all technical about this topology, the second triode is slaved to the first triode. To the best of my knowledge, Fender tremolo circuits all (except the single ended VibroChamp which can't have a PI) use a LTP (long tailed pair) PI, which is a "gainer" PI. My grasp of the technical aspects of this point run out of air at this point, but I am thinking there is something to it. I suspect you have mixed and matched the play doh and your end product came out brown ;-( because you didn't think about how the color wheel applies.

4) Look here for some practical info http://www.lynx.bc.ca/~jc/tremoloCircuits.html I suggest paying special attention to what he says at the top of his page about voltage swing, quoted below:
"Note: each circuit presented here is recommended for use with a particular type of power tube load and sufficient/necessary voltage swing - the circuit needs to match the input swing requirements of the tube to the output of the Tremolo circuit ... these particular Tremolo circuits are derived from amplifier examples that seem design-stable ... (*) it should be said that one way of exploring the pairing of the Tremolo circuit with your amp is by varying the Bias of the power tubes and noting how much the Tremolo effect has on the amp at full intensity - when you've settled on a level you like take a measurement of the bias current and compare it to what you would normally have set it at, they should not be too far off in value for a good Trem circuit match to your output stage ..."

FWIW, my SWAG on this is that there is more than one problem to be solved to get your tremolo to oscillate properly and I further guess that it is a combination of a poor match of your trem circuit to the power section AND the h-k problem that Martin has noted.

With respect to a "good match" if you have a 6V6 amp, the attached Fender diagram is probably the most relevant. Because you are using a 6SL7, it might help to study the attached from RC30 to get an idea of the plate and cathode resistors you might select to insure enough gain.