Bias vary Tremelo problem

[Toxophilite]

Hi all
I'm having a problem with the tremelo on my AB763 single channel deluxe build..
It's a bias vary tremelo from the 6G-11A I believe, It works well except it always has some nasty farty distortion in it, not a nice distortion. Even at moderate volumes (like 2 or 3 on the volume)
I've adjusted the bias different ways,(hot cold what have you) I've set it so the bias voltage to the 6V6s is right at -35 like the schematic


and still distortion


So I was trying changing the bias for some recording this evening, trying to lesson the distortion on the tremelo , I had the tremelo on by accident, and I noticed the variation was between about 30-45 mv..really HOT, when I turned the tremolo off the bias dropped down to 22mv on both tubes(they're matched well)...


THis doesn't seem right to me. and also seems a likely cause of excessive tremelo distortion


Could this be something amiss with the tremelo circuit?


Any idears??


Thanks

[Toxophilite]

i tried swapping in a new tube and just to make sure I tried swapping out the red LED and putting the bypass cap/resistor back on the cathode of the tremelo tube, no change in the sound.

The distortion starts about halfway up the intensity dial and gets progressively worse.

Here are some voltages from the tremelo tube:


1 -  201v
2 -  .3mv
3 - 1.8v


6- 420v
7 - 199v
8 - 202v


The amps bias is currently set at 25mv/ma at the 6V6 cathodes
and there is 424 vdc on the plates of the power tubes
the bias for the power tubes is about -37
I have new JJ 6V6gt's in the amp
The only number i can see that is significantly off is the B+ on pin 6 which is 56 over the rated amount
I'm taking from point A like on the 6G-11 schematic but obviously the deluxe has a higher voltage at that point

I really need to be able to have a clean tremelo (I generally play clean clean and more clean)
there are few songs I play that rely upon it. A waltz , a ballad and a pop number

Any takers??

[Willabe]

I've set it so the bias voltage to the 6V6s is right at -35 like the schematic

I had the tremelo on by accident, and I noticed the variation was between about 30-45 mv..really HOT, when I turned the tremolo off the bias dropped down to 22mv on both tubes(they're matched well)...

I don't know why turning on the trem would shoot the current draw of both 6V6's up that high.

More negative bias voltage will cool off the tubes, so when you turn on the trem it is shifting the overall bias down, less -bias (more tube current) along with varying it?

I don't see how that could happen. I remember that Tubeswell said that the reason output tube bias vary trem sounds the best was because it causes the tubes to run hotter when the trem is on. BUT I don't think he meant that much hotter. That's a lot. 

[Toxophilite]

That's what I was thinking
I was going over my circuit and I couldn't fins anything amiss..I will look again!

[Toxophilite]

With the tremelo on the actual bias supply seems to swing between about -25 and -55 v ,  every couple seconds it flashes at -102 (it doesn't climb there, just flashes quickly) but I suspect that's the meter
With no tremelo it's at about -39 right now with 25mv/ma on the power tube cathodes
I tried swapping in some 60s 6V6Gts , and I still get the distortion
I went over my tremelo circuit again and all seems to be in order


i wonder if it's how the tremelo circuit interacts with the actual bias circuit??

[Willabe]

Those -bias #'s seem more in line.

Does the distortion go away when you turn down the guitars volume?

[Toxophilite]

Do you mean the volume control on the actual guitar or the volume of the amp?

[Willabe]

On the guitar.

[Toxophilite]

hmmm
That's a hard one to tell. when i turn down the guitar, I of course lose most of the top end


I'm playing older Gretschs from the mid 60s with hilotron pickups (and one newer guitar with 60s hilotrons too..I like em)
They have really low output (2.8 - 3k ohms)


Maybe the distortion goes down. it's hard to tell.  It gets pretty quiet of course, When I turn the amp up and the treble up to compensate, ther's probably similar amounts of distortion


Ae you wondering if I'm playing a really high output guitar and driving the front end really hard?

[Willabe]

Ae you wondering if I'm playing a really high output guitar and driving the front end really hard?

No, I'm just pokin around.

I was thinking that with no trem at your mA setting you have a certain output volume.

Then with trem and it's then a hotter mA setting (?) could be a little louder, so turn down the guitars volume pot and maybe it's now clean(er) at the same trem volume/non trem volume.

I've not checked my 5G9 trem for bias and mA measurements with trem on.

But I don't hear any more distortion with the trem on. And I play/set the amp pretty clean, amp volume ~3 to 4 for most things, regular/early reissue 62' Strat PUP's, no pedals.     

[Toxophilite]

I'm pretty sure the amp doesn't any louder with the tremelo on


probably a little the opposite. the distortion is pretty unpleasant, and it's slightly more noticeable after the attack of a chord
a frappy mushy buzz


i've tried the amp plugged into a different cab too to eliminate the speaker being the problem(though that doesn't really make sense)

[sluckey]

With the tremelo on the actual bias supply seems to swing between about -25 and -55 v ,  every couple seconds it flashes at -102 (it doesn't climb there, just flashes quickly)

The dc bias voltage should not change like that. We need to see a schematic that shows exactly what you have. Maybe some pics too. Don't just show us a stock fender schematic.

[Willabe]

The dc bias voltage should not change like that.

~How much should they change, ~ -10dcv, -20dcv, and where would the change be centered?

If the amps bias was set at, say, -30dcv, would it vary from -30 to -50dcv? Or maybe a little loss on the from the -bias when trems not on, say -27 to - 47dcv?

I'm thinking that with different power tubes, to get enough depth, closer to shut off, then some would need a wider swing?     

[Toxophilite]

Okay here's the best I can do with what I have. I believe it to be complete. please point out any obvious errors
I can change/fix them
I might've messed up the rectifier PT area. the SS rectifier is a plug in job I made using 2 diodes and a 8 pin plug. it goes right into the old 5U4Gb plug on the hammond AO-43 chassis this amp is built in
I used a deluxe reverb schematic as a platform and then changed/added in all the necessary bits and pieces
Two things I didn't put in are the bias check 1 ohm resistors on the 6V6 cathodes and the red LED that is on the tremelo tube cathode instead of the  cap and resistor combo.
I hope to put some pictures up but I can't get one of the reverb circuit as it is hiding inside it's own parasitic chassis

[sluckey]

With the tremelo on the actual bias supply seems to swing between about -25 and -55 v

At what point in the circuit do you observe this?

Looking at your schematic one thing jumps out at me... You are powering the trem tube from node A (output tube plates) which does not have good filtering. Move that to point B (after the choke) and see if that makes a difference.

Looks like the bias pot is 750K??? That needs to be much smaller, about 50K. And the bias pot wiper connects directly to ground. This will allow the bias to be set to zero which is deadly for the output tubes. I suggest you put a resistor between the wiper and ground. Chose a value that will allow you to only lower the bias voltage to about 25V.

You built two of these amps. Are they both distorted with tremolo engaged?

[Toxophilite]

excellent you see the mistakes!


-Measuring the bias swing, just after the 220K resistors prior to the 6V6


- I am actually powering the trem from the same point as the power tubes (B)(i will fix that on the schematic)


- and there is actually a 22K resistor after the bias pot (which is 50K)  I was being lazy and you caught me!!
I will post a revised version


-Of the two amps only this one has reverb and tremelo. The other is a straight up 2 channel AB763 deluxe
They were meant to operate together for the specific purpose of using my space echo in stereo
This amp is the main amp.
The 2 channel is the wet sound and the secondary dry sound.
I was hoping to use the two of them at this big corporate event, my band (The Colorifics) is playing with my brother's circus(The Underground Circus) on the 14th

[Toxophilite]

Here's a revised version
Thanks Steve!


I put the bias resistors in and hopefully fixed all the other mistakes


To further clarify


I was measuring the bias swing from between the 220K and 1.5 resistors on the grids)?) of the 6V6s


One potential weak spot I know of in the tremelo circuit is the .05 cap going to ground from the intensity pot. I used an old cap here as it was all I had at the time. My digital multimeter doesn't always read old capacitors well. so I'm not sure how good it is or if it's leaky
I might swap in a new one, though I'm not sure how crucial it is (SWAPPED IN NEW ONE, NO DIFFERENCE)

THe other changes I have made within the last few months are:
-I swapped in a different OT to try it, It's from a baldwin amp that was running 2 EL84s, it's physically a little bigger than the original AO-43 one and it's 27-1 ratio. It sounds great with no tremelo..I can't remember if the tremelo was a big problem prior.
- I monkeyed with the dropping resistors after installing the reverb circuit. there is now 200vdc on the phase splitter which is about 20 v more than the original deluxe schematic.

[sluckey]

Two things I'd like you to try...


1. Put a 1N4007 diode across the intensity pot. The anode lead will connect to the end with the .1 cap from the trem circuit and the cathode lead will connect to the end that is fed by the bias voltage. This will kill the positive swing of the trem signal which may be causing the 6V6 grids to draw current (especially since you also have a LED on the oscillator). If no joy, remove the diode and try #2.

2. Remove the LED and put a 4.7K and 22µF cap on the cathode of the oscillator tube. My thinking is your trem signal may be too strong.

[tubeswell]

So I was trying changing the bias for some recording this evening, trying to lesson the distortion on the tremelo , I had the tremelo on by accident, and I noticed the variation was between about 30-45 mv..really HOT, when I turned the tremolo off the bias dropped down to 22mv on both tubes(they're matched well)...


THis doesn't seem right to me. and also seems a likely cause of excessive tremelo distortion

Its normal for the bias to run quite a bit hotter when bias-vary trem is switched on. Bias-vary trem varies the bias. With the LFO wiggling the output tube grids, the net effect is an increase in output tube current (even though the trem is driving both tubes into cooler territory for part of the LFO cycle, it is also going into grid-current territory for the other part of the LFO cycle. With the trem wiggle going, it takes longer for the output tubes to dissipate the extra heat than it does for the plate current to build up, compared to if the trem is inactive). You should bias the 6V6s quite a bit cooler to begin with - about 18-22mA tops. If you keep the tubes cooler, the trem should work better as well.

[Toxophilite]

I will try those two ideas this eve (off to work and rehearsal)


I had tried biasing the amp cooler but just to make sure I biased it at around 19.5 and it sill starts distorting around halfway up the intensity pot.


I have to pick up some diodes enroute! I'm out!

[alerich]

If it makes you feel any better Toxophilite my Princeton Reverb RI does the same thing. I changed R14 from 1Meg to 500K and I seem to recall that the distortion problem was less or nonexistent at 1M but the intensity was much less, too so it was a moot point. I changed the cathode resistor to a red LED and that did not seem to impact the distortion much. I like the intensity on 10 and the speed on 1 when I use trem. I really haven;\'t investigated it much since I don't use trem all that much but I'll be watching this thread with interest. Honestly, if I was using trem a lot I'd just use a BOSS Tremolo pedal. I have one from the predecessor to my PRRI that didn't not have trem. I think that pedal sounds every bit as good as Fender's bias vary trem (was never a big fan of the roach trem) and no distortion.

[Toxophilite]

I tried the too ideas
Put the diode across the pot as suggested anode to .1 cap side cathode to the bias supply, it actually slowed down the tremelo and if anything maybe added some distortion!


I also tried putting the cap and resistor back in the circuit instead of the Led, that didn't make a big difference, maybe slightly less tremelo, if anything, same amount of distortion

I've been doing some reading and it would seem that my tremelo swings too wide in it's bias varying?
Thus producing ugly crossover distortion??

[Toxophilite]

Could my high PI voltages be a problem 203 and 206  ?


Also should the tremlelo be affecting the voltages throughout the amp or just for the power tubes?
I get movement on the PI voltages (as much as 20 volts), the preamp voltages (a smaller amount, maybe 5-10 volts)etc

[sluckey]

Could my high PI voltages be a problem 203 and 206  ?

I don't consider that high.

Also should the tremlelo be affecting the voltages throughout the amp or just for the power tubes?
I get movement on the PI voltages (as much as 20 volts), the preamp voltages (a smaller amount, maybe 5-10 volts)etc

That's not right. Did you put new filter caps in this amp? Are you absolutely sure the amp is wired correctly?

[Toxophilite]

hmm a problem then
This amp has a brand new JJ can cap in it
All the components except the transformers and choke are brand new
I'm pretty sure it's wired correctly
There was much trouble shooting done originally
However, I'll start going over the whole shebang again and lift the boards etc to make sure nothings gotten under them
I'll report back on my findings if any

[shooter]

No help but a question, if I put my red meter lead between the .1uf and the intensity pot, blk lead on ground, meter set for VDC, what should I read?  btw, love your manufacturing process

[Toxophilite]

Is it a trick question?? And is that with tremolo on or off?

Thanks,,,I was getting rid of the patents and it was like they were handing me a line!
Couldn't resist messing with it.
though I forgot under- The Gun

[Toxophilite]

So far I can't find anything amiss, no solder drips under the board


The Pi voltage swing is actually more like 10 volts but it is definitely moving with the tremelo, as is the preamp though it's more just a couple volts moving in tenths of volts


I know Hoffman's deluxe uses a 6G11-A bias vary tremelo with a deluxe reverb circuit
I'm going to try and compare them and see if there are any significant circuit differences
maybe in the area of filtering.
On my amp the reverb circuit is an add on and like the original it takes it's power from the same points ('B" and 'A')
as the powertubes and tremelo   and the preamp respectively.
probably grasping at straws
i did notice the hoffman schematic has a 470k res. in front of .1 cap going to the intensity control whilst I have a 220K, maybe the higher value will tame things a bit


Does anyone have any insight into this?
And does my schematic look workable?

[Toxophilite]

I aslo noticed the bias vary on the Hoffman one channel does not have the .05 cap to ground on the wiper
but rather has an extra .1 cap to ground prior to the .1 cap that is connected to the intensity Pot

hmm this is the 6G16 bias vary tremelo!

[sluckey]

On my amp the reverb circuit is an add on and like the original it takes it's power from the same points ('B" and 'A') as the powertubes and tremelo   and the preamp respectively.

Please clarify this statement. The reverb driver should be powered from point B. Point B also feeds power tube screens and the tremolo. But the reverb recovery should be powered from point D, same as the preamp.

Look on the schematic at the cathode of the second preamp triode. It connects to point A. What does it really connect to? This should have nothing to do with your trem problem.

Really need to sort out the B+ nodes since other B+ nodes are being affected by the trem. What are the B+ voltage readings at points B, C, and D?

[shooter]

It's not a trick question, it might be a dumb question, I'm trying to learn, I get how the circuit operates, but not the details. My guess is that side of the intensity pot is purely AC,  but I'm wondering if my meter sees a "floating" DC like in an open circuit.   anyway, don't mean to distract from your real problems.

[sluckey]

There are two voltages present at that point. There will be the negative dc bias voltage and there will be a healthy low frequency (3Hz to 10Hz) AC sine wave (trem signal) riding on top of that negative dc voltage when the trem oscillator is enabled. The trem signal causes the negative bias voltage to vary up and down, centered around the static dc level. Not all meters can accurately measure the low freq AC voltage. With an analog meter you would see a voltage varying above and below the negative dc level at the slow trem rate. You might say the needle is wiggling rather than being steady. Most digital meters will quickly sample the voltage, then wait, then quickly sample again, etc, etc. This sampling makes the voltage just appear erratic. You might even think you don't have a good connection with the meter.

If you disable the trem oscillator, the voltage would be exactly the same negative dc that you would measure on the output tube control grid.

[shooter]

Thanks Sluckey, I knew I'd see the AC, just wasn't sure if I'd see the DC reference on that side of the pot.
sorry for the hi-jack

[tubeswell]

Thanks Sluckey, I knew I'd see the AC, just wasn't sure if I'd see the DC reference on that side of the pot.
sorry for the hi-jack

You have to have the speed turned all the way down to see the LFO voltage wiggle on most meters (unless you have an analog meter). At the higher oscillator speeds, the voltage ends up looking constant on the meter.

[Toxophilite]

Sorry Sluckey that was a late night typo/mental glitch , my reverb circuit is powered by B and D as it is supposed to be


my B+ nodes are ALL as shown in the the last version of the schematic I posted
And I have NOT moved them around as they were all correct


Sorry for the confusion

[sluckey]

my B+ nodes are ALL as shown in the the last version of the schematic I posted

I don't see any measured voltages for any nodes. I do see the original figures for the first node and node B, but nothing for node C and node D. I'd like to see some actual measured numbers.

[Toxophilite]

Here you go
biased to 24mv at the 6V6 cathodes(1 ohm resistors)

-38 6V6 grid bias supply


A- 422
B -417
C - 293
D- 274


I will amend the schematic this eve

[Toxophilite]

I did a little perhaps pointless experimenting and noticed  something
Don't know if it's relevant or not


the higher I have the amp biased the cleaner the tremelo gets
with New JJ 6V6s in it


If I bias it low 18-22mv (-39) - it really craps out and sounds like a small amp hitting the proverbial wall (compressed, distorted and farty)


If I bias it around 25mv(-37)  it distorts alot of which is more noticable after the initial attack


If I bias it REALLY hot 32mv (-35) (and very very briefly) - it's even a little cleaner(still distorting some  though)

[Willabe]

This may be a wild goose chase, but you said earlier that the power supply B+ dcv rail were moving with/at the trem's speed.

I wonder if they move as much or at all, when you bias the tubes hotter? 

[Toxophilite]

That's an interesting question
And I wish I'd though of checking that before I changed my tremelo circuit from a straight 6G11-A to be more like the 6G16 (avout 15 min ago)
I put in a 470K resistor in lieu of the of the 220K , added a .1 cap to ground between the resistor and the .1 cap to the intensity control and disconnected the .05 cap to ground on the intensity control


The voltage swing is definitely less with this setup than before though still present, maybe 4 swing (2 up 2 down) volts on the PI


It seem to run much better now with the distortion not nearly as offensive (no after note crapping out)
However I think there's less tremelo as a result


THis might've been the root of my problem(as Sluckey suggested). Too much tremelo, as the hotter biasing seems to lessen the intensity of the tremelo as well
there is also no speaker whup whup at higher intensities any longer




I'm not sure if there's another problem lurking and this just helps to alleviate the symptoms
Otherwise, so far everything has checked out with my schematic comparisons

[Willabe]

Where's it biased at now?

[Toxophilite]

It's biased at around 23 around -39 on the grids
I was hopping to squeeze a little more intensity out of the circuit
i'd rather have less intensity than nasty distortion when it's not wanted any day of the week
And when I use my super I rarely use the tremelo deeper than 6 or 7 really


I might try a a different sized resistor in that spot. Maybe 360K or 390K, to see if I can find an optimum comprimise

[Toxophilite]

I put a 360K resistor in place of the 470k that precedes the .1 resistor, give more tremelo, on the edge of getting crappy when maxed, tried a .047 cap in lieu of the .1 and it definitely had a deeper tremelo but it was starting to distort.


Maybe I expect too much from the bias vary tremelo?

[sluckey]

Are you using a JBL speaker? Didn't you have some problems with that speaker thumping (whump, whump, whump) with your bias vary tremolo? If so, try another speaker and report back.

That diode mod should reduce or prevent thumping, but you said the diode made the distortion worse and actually slowed the speed down. I don't see how it could possibly affect the speed. I wonder if you had the diode connected properly?

[Toxophilite]

I have  a JBL in my amp, but my bench speaker  that I've been doing most of my testing with is a 10" 50s English Rola


I'm pretty sure I had the diode hooked up correctly.
I even looked up diode polarity just to make sure I wasn't(!) getting it backwards
However, It's pretty easy for me to test again as i alligator clipped it in


I had the anode hooked up to the point on the pot where the .1 cap from the tremelo circuit goes
And the cathode hooked up to the point on the pot that the bias supply feeds in


or just talking of the intensity pot leads:


Basically anode to input
cathode to ground,
wiper ignored


yes?

[Toxophilite]

So I guess no more ideas on this
I was hoping to have a tremelo as deep as my super which has the "bug' type tremelo
As I have limited and mostly recent (past year or two) experience with this type of tremelo I'm really not sure what i should be expecting as far as depth. I do now I don't like it crapping out after the note


Could a change in OT produce that result (he said revealing the horrific extant of his ignorance)
I did for kicks swap in a different OT (from a 2 El84 Baldwin) it's 27-1 turns ratio and was originally driving an 8 ohm speaker
It's more vibrolux sized rather than the old one which was a little smaller(the exact dimensions of a deluxe OT actually) and 30-1 turns ratio
The 'new' one is hooked up right and working well in every other respect
I installed that a few months back just to try it

[sluckey]

I'm out of ideas and I don't have a similar amp to do any comparative testing. The only bias vary trem amps I have are cathode biased. Wish I could hear what you hear or see the amp.

[Toxophilite]

Thanks
I realize long distance trouble-shooting has severe limitations
There is nothing like being on the spot.
From what I've read it seems like it MIGHT be crossover distortion, which would seem to tally with the fact that it gets worse the lower the bias current is(or higher the negative bias voltage)
I'm going to go over my schematic/circuit yet again .

[Baguette]

IME bias vary trem can be a headache to fine tune. Often, copying a production amp schematic does not work ideally.
If you have distortion, it's because you have too much voltage swing at the trem output and I see two ways to cure that:
- higher the 220K up a notch or two (max should be 1M)
- lower the .1 cap a notch
Also IME the .05 cap from intensity pot wiper is useless and may be removed. A lot of Fender amps do not have it but some do for some reason.
Keep in mind those swaps will also lower the effectiveness of the intensity control (reduce depth) so it's a trade off thing and you'll have to choose a balance between huge swell and clean trem.

[trobbins]

The impact of grid conduction can be alleviated by lowering the PI to output stage coupling caps, and raising the 6V6 grid stoppers.   Low frequency roll off may change when lowering the coupling caps, so you would need to determine an acceptably low value.


LFO harmonic distortion could also exacerbate grid conduction by allowing the conduction time to be longer.  The LFO higher harmonics can be attenuated further by capacitor shunt filtering - such as a capacitor across the LFO output stage cathode follower 470k load, or adding an additional low-pass CR filter stage before the intensity pot. 


Hopefully you have a star ground arrangement where the trem and bias circuit grounds also star to the output stage cathode and screen star node.


Have you also tried checking if your feedback circuit is not applying too much feedback?  The OT specs will affect that feedback level, and a series high pass cap in the feedback leg can reduce the low frequency feedback that would be attempting to correct the 'distortion' contribution that the LFO signal is introducing to the output stage.