Tremolo Problems

[nastyoldtech]

I'm switching an amp over from fixed bias to cathode biased. It's got 6v6's in the output at about 410V on the plates. Right now the whole thing is set up sort of like a bf deluxe. I currently have it running a wiggler bias tremolo. I was looking for a similar type of tremolo that could be an easy fix for the move over to cathode bias and I came across the Vox Cambridge Reverb schematic. It's tremolo is almost identical to the one I'm using only it's used in a cathode biased amp.

Today I reconfigured the amp as cathode biased and reconfigured most of the tremolo to reflect the Cambridge's tremolo section minus a few resistor values ( just a little off)  it's pretty close. The amp sounds great but I get only a faint hint of tremolo out of it. Funny thing is when I turn the amp off with the depth dimed max I get an intense tremolo right after I flip the switch off and then the volume fades off.

So I'm thinking the Cambridge uses EL84's and I'm using 6v6's, not enough of a pulse going out from the oscillator to shut down the 6v6's? Why does it work right after I shut off the amp? Ugh, any suggestions would be a big help.

http://www.schematicheaven.com/voxamps/vox_cambridge_reverb.pdf

[jjasilli]

1.  The 270VDC B+ supply seems low with only 175VDC on the plate of the trem tube.  It's pretty common to see excessive voltage on the trem tube plate. So you might want to feed the plate B+ from an earlier higher voltage filter cap.  Often the screen cap is used. 

2.  You might want a larger depth (intensity) pot, so when it's turned up less trem voltage bleeds to ground.  Note that too much trem voltage swing can cause the power tubes to redplate, so be alert for that.  One way to experiment is to use jumper cables to A/B different value resistors between the pot's ground lug to ground.  This will save you from swapping pots to help find max intensity (but during this procedure you won't be able to turn the intensity fully down due to the temporary series resistor).  When you find the right max intensity it will = the 250K pot + the resistor value.  Then substitute the closest value pot.

Here's what I did in an el84 amp converted to cathode bias.

[tubeswell]

Try hooking it up like the GA17RVT. The ~330k resistor from the grid load resistors to ground, becomes your intensity pot.

[Baguette]

Hello Nastyoldtech,

Been there, done that.

As you say 6V6 need more amplitude to wiggle the bias compared to EL84. Copying the GA17RVT won't help since it has EL84 too... check the GA16T or 18T instead. However these amps have a much lower plate voltage (300VDC IRRC) so you'll have to finetune the circuit to your amp.

1/ Change the R40 680k resistor in serie between the oscillator plate and the trem intensity pot with a jumper.
This should resolve most of you problem.

2/ If you need even more amplitude, replace the .1 intensity cap to a .22. And you could also make the ocillator bias a bit hotter, replacing the 3.3k res with & 1k or 1.5k, and bypass it with a larger cap (like 220uF).

Depending on the voltages in your amp you'll have to fine tune the circuit your self. Changing the .1 cap is a good place to start. Too big and you can have a nasty distorted sound when the intensity is turned up. Too low and the trem doesn't go deep enuff.

I have a similar trem in my latest Tweed Deluxish amp and it's one of the finest trem I've heard, right up there with the Harmonic Vibrato.

Let us know your findings.

Cheers!

[nastyoldtech]

You guys, thank you for all your help. I really love how this forum pulls together to help someone out. I do appreciate it.

Baguette, I did some experimenting with larger tubes in with these wiggler tremolos a while ago and found that you needed to up the gain to get them to sound anything like they do with the little bottles. This was a little bizarre because I was getting basically nothing.

I jumpered that resistor and replaced the .1uf cap with a .22uf. It did the trick, thank you! It sounds just as good as it did in fixed bias.

But... If you turn the SPEED control all the way to it's slowest setting it turns the tremolo off completely, and when you turn it to go faster it takes a few seconds to kick in. Almost like it's ramping up like a Leslie.

Do you think it's the increased intensity cap or do I have to change the R37 resistor to some other value (right now it's a 100k, the schematic shows it as a 150k. I don't think 50k would make a difference?)

[tubeswell]

But... If you turn the SPEED control all the way to it's slowest setting it turns the tremolo off completely, and when you turn it to go faster it takes a few seconds to kick in. Almost like it's ramping up like a Leslie.

Do you think it's the increased intensity cap or do I have to change the R37 resistor to some other value (right now it's a 100k, the schematic shows it as a 150k. I don't think 50k would make a difference?)

It is not unusual for the LFO cycle to have to build up a bit at the start when you first dial it in, and changing the RC resistors will affect the cycle a wee bit. Using a footswitch helps slam it into action. I wouldn't be concerned about it.

[PRR]

> If you turn the SPEED control all the way to it's slowest setting it turns the tremolo off

The -slowest- setting? Then R37 (100K is too low) is totally swamped by R36 turned to 3Meg to get the lowest speed.

Punk-out at lowest speed is perhaps EDIT: C23 being weak: it has to bypass well to 2Hz, so it has to be a good 20-40uFd. If it is old it may be more like 10uFd now, which works at the faster speeds but not the lowest.

Stick 47uFd >10V across R39 and see what changes.

[nastyoldtech]

Your probably right Tubes, but the geek in me loves to get to the bottom of this stuff. I think I enjoy learning about how stuff works more than actually using the stuff I get up and working.  

I might try all the above just to see what's tops.

With the way it's setup right now I'm guessing something like this would work. jjasilli's pot idea say a 500k pot instead of the 250k, put that .1uf cap back in and put the resistor back in that was in series with the .1uf, but use that resistor to attenuate the output so's not the need for an oddball potentiometer hunt?

[nastyoldtech]

> If you turn the SPEED control all the way to it's slowest setting it turns the tremolo off

The -slowest- setting? Then R37 (100K is too low) is totally swamped by R36 turned to 3Meg to get the lowest speed.

Punk-out at lowest speed is perhaps C25 being weak: it has to bypass well to 2Hz, so it has to be a good 20-40uFd. If it is old it may be more like 10uFd now, which works at the faster speeds but not the lowest.

Stick 47uFd >10V across R39 and see what changes.

Hello PRR, you posted while I was typing. I'll try those in a minute here thank you! Oops you mean C23 = 25uf (being weak)?

[Baguette]

But... If you turn the SPEED control all the way to it's slowest setting it turns the tremolo off completely, and when you turn it to go faster it takes a few seconds to kick in. Almost like it's ramping up like a Leslie.

I hear you, I was having the same kind of problems 2 weeks ago when building my amp. Designing the oscillator might be a bit tricky, and it may happen copying an exact commercial circuit may not yield optimal results due to not so big voltage differences.

Most likely, the problem comes from te Speed control arrangement (a resistor in series with the Speed pot wired as a variable resistor). You'll have to tinker these two guys.
- The R sets the highest speed you can go. You have to make it high enuff so the wobble doesn't stop when the speed knob is all the way up (oscillator swamped), and low enuff so it doesn't rob some range.
- the pot sets the lowest speed you can go. Choose its value so when set to 0, the trem is still here.
The easiest way to choose those components values is to slap in a huge value pot (like 5M) and find the highest and lowest resistance values you have trem within.

One more thing to remember: the HIGHER the resistance, the LOWER the speed.

Happy tinkering,

V

[nastyoldtech]

That was it, spot on! I had a new 25uf cap on there and just subbed another 25uf in parallel for 50uf and a 1.5k cathode resistor. Full range of the dial and not really any ramp up like before. I tried a higher cap value of 100uf and the speed would go so fast as to shut off, so the 50uf cap was the perfect balance on the dial.

When I put that cap on there it seemed like I lost just a little bit of tremolo intensity. I kind of wanted a little more anyway? Now I'll mess with John's suggestion for a little more amplitude by way of that intensity pot. I might put that .1uf cap back and up the 250k pot to 500k or such testing for max intensity via the resistor method. Then tweaking it with that R40 resistor.

Thank you for all the help guys I really do appreciate it! 

[nastyoldtech]

You might want a larger depth (intensity) pot, so when it's turned up less trem voltage bleeds to ground.  Note that too much trem voltage swing can cause the power tubes to redplate, so be alert for that. 

Well... I tried a bunch of resistors off the pot to ground and in the end I threw a 1M and found that plus the 250k pot to be the best (I didn't have a 750K to make an even 1M pot). I left the .22uf cap in there and kept the R40 slot jumpered. It sounds amazing! I guess I'll have to go down a little to a standard 1M pot but upping it really is an improvement over the 250k pot.

I did not see any red plating so am I good to go in the long run going with that high a value (1M pot)? 

[jjasilli]

 

[Baguette]

I did not see any red plating so am I good to go in the long run going with that high a value (1M pot)? 

I'd bet you're safe. Anyways, I'm quite sure the 1M speed pot doesn't influence the trem voltage, just the rate of the LFO.
For fine tuning the maximum trem intensity, you can play with the .22 cap to the Intensity pot. Lower it a touch and you'll have less intensity.

Cheers

[nastyoldtech]

I tried the pot trick and it did not work. Last night I mistakenly put a log pot and the effect was really bad but even with a linear pot what happens is the effect gets bunched up at the end of the dial. So you get nothing from 0 to about 1/2 way and then it starts. Go lower and it's a little better until you go back to 250k and you get the full range across the dial. At least the way the circuit is setup at the moment.

I put the 250k pot back in an have begun experimenting with intensity caps again. I'm up to about .5uf and it's at the intensity that the 1M pot was at its fullest but I have the full dial available. I'm getting a more "thumping" from the speaker though? 

No red plating or any gremlins like that. Just thump, thump, thump, thump at how ever bpm it's set at. Is this normal? I don't remember it being like that in fixed bias but I don't remember my own cell phone number either so there you go!

[tubeswell]

AFAICT you have to get the LFO voltage swing so that it moves up and down the output tube bias voltage gradient as much as it can without hitting either end (i.e. without going into +ve grid voltage and without going so negative that the tube is cut-off). So the best thing is to try fiddling around with the LFO bleed-out/slope resistor (i.e. the one that comes between the LFO stage to the grid leak resistors). I'd start by putting the RC network of the LFO back to 'stock' values first tho.  Happy to hear other's views on this matter.

[jjasilli]

No red plating or any gremlins like that. Note that you need to check for redplating under full signal condition over time.  (It would help to have a dummy load and a signal generator.)  The AC signal will swing from + to (-) on the tube grid.  That will add to the DC trem swing voltage on the grid.  When that all goes full + it can cause the tube to lose bias for about 1/2 the signal swing, which could cause redplating.  

Also I agree that you seem to have a speed issue.

Tremolo design is not easy.  It might help to plagiarize a known design in a PP 6V6 cathode biased amp.  The Silvertone 1482 comes to mind.  It uses a weird trem tube, but it's a great sounding trem.  Schematics are posted on this Forum. 

[Baguette]

Tremolo design is not easy.  It might help to plagiarize a known design in a PP 6V6 cathode biased amp. 

Absolutely. Check the  Gibson Viscount. My design is similar to the Gibson one but I had to change some values to suit my circuit (which runs a smaller power tube cathode resistor and different voltages - 220 ohm versus 250 and 320V B+ versus 300V B+ for the Gibson).
I had to replace one .03 cap with a .02 due to part availability, and increased one .01 to .02 to lower the speed range a bit. The .3 cap feeding the Intensity pot went down to .2, but .1 gives much enuff intensity already. Also had to replace the 560K resistor with 470k (part availability). The LFO tube cathode on mine is 1.5k / 220uF instead of 1k / 20uF. The Speed control on mine is a 1M linear pot in series with a 150k resistor. I also made the footswitching arrangement ala Fender, with one of the resistor (the 560k on the schemo) connected to the LFO cathode instead of ground. Makes the trem turn on faster. Check the Valve Wizard article for a good explanation.

http://www.schematicheaven.com/gibsonamps/ga16t.pdf

[Baguette]

BTW the thumping problem may originate from the big cap feeding the Intensity pot.
Too big a cap (like .5) and you may end up with a thumping tone and other non desirable noise (like if the trem sound was being distorted when Intensity is turned up).
There's a sweet spot to find where the cap value is just right: no loss of intensity and no noise.

I was very happy to have a caps and a resistors sub-boxes around when I was tuning my trem. It's all about fine tuning.

[nastyoldtech]

I see they're using a .33uf as the intensity cap in that Gibson 16T? That's the value I thought would be just perfect but alas I don't have one on hand to test it out.

That Silvertone is using a pentode, high gain to shut'em down. Does impedance have a lot to do with this at the output? I had a CF setup on this thing and pulled it to configure it like the Cambridge. With fixed bias I had it setup a little like a 6G16 with that CF. When you remove the CF the intensity is weaker. I believe the CF drives the 6V6's to shut off better. I'm thinking with a pentode you'll get some more gain out of it but you'll loose out some of the intensity because of its higher impedance?

[nastyoldtech]

I messed with the speed pot resistor R37.

My observations,

1.) Too high a value and the speed will go so fast it turns off earlier on the dial. You then have nothing from that point to full CW position.

2.) Too low a value and almost the same thing happens but it also has the effect of having a hard time "kicking in" meaning a slow start up being a few seconds before the effect is audible.

I think I've got it about right with a 120K resistor. Is there another part of the circuit that sets the slow speed? C21 or C22?

Thank you again!