Trem-O-Nator 2 (another go at inserting the TON on my AC30 "light" build)

[Jennings]

Some may recall my efforts a few months ago to use the spare triode of my single channel AC30 to add the Term-O-Nator circuit to this amp.  It worked, but I simply couldn't get it to sit right no matter where I tried inserting it (after the first triode or after the tone stack), and no matter what pot value I used.  Result was the same: massive volume loss as you turn up the intensity, and the trem only really became clear after the pot was over half way.


Anyway, I thought I'd give it another crack...and also happened across this thread where the problem sounded identical to the issues I was having:


https://el34world.com/Forum/index.php?topic=28666.0


So just like the above thread I tried changing the pot to a 5K and moving my insertion point to the cathode of V2a (see attached schematic).  Well, I'm definitely onto something!  I have MASSIVE thumping!  So big in fact even with the amp's volume at zero that I have to use an attenuator or risk damaging my test speaker.  So, with the attenuator in play I investigated further...Using the footswitch kills the thump, but without the footswitch, or with the footswitch and the trem engaged the thump is enormous and I can see the glow of the power tubes even pulsing in time!  Ignoring the thump, the volume loss to signal is much better, and the pot sweep of the 5k in this position seem great.  So I'm getting somewhere. 


Any suggestions as to what to try next?  Perhaps splitting that 56k into two resistors and inserting between them?  I realise it's all a bit trial and error here to some degree.

[stratomaster]

If I'm understanding your description, you're functionally shorting the cathode follower's cathode to ground regularly via the LDR. This is a very bad idea. There is relatively high voltage DC present here that you are grounding. The potentially speaker damaging pops and power tube pulsing visibly on non bias tremolo are huge warnings that you're not on the right track.

However, I think if you add adequate decoupling that you might be able to make this location work for a tremolo.  Try a large cap (0.1-0.22μF) in place of the wire that currently links the 56k and 100k resistors in the cathode.  Then add the 250k Intensity pot connection to the 100k slope resistor input side.  Break the link between the wiper and 3rd lug on the Intensity pot and instead ground it.  Now you'll have essentially a pre-tonestack master volume that is controlled by the tremonator oscillator.

[dickjonesify]

Intrigued by this. I just built the Trem-o-nator into my amp. There are two channels and I tried it on both. On the 5F6A Bassman side, it works beautifully right on the treble wiper before it goes to a coupling cap and then phase inverter.

The other channel is the Dumble ODS clean channel. I have had similar issues as you trying to figure out where to introduce the signal and what to adjust 

[Jennings]

I think if you add adequate decoupling that you might be able to make this location work for a tremolo.  Try a large cap (0.1-0.22μF) in place of the wire that currently links the 56k and 100k resistors in the cathode.  Then add the 250k Intensity pot connection to the 100k slope resistor input side.  Break the link between the wiper and 3rd lug on the Intensity pot and instead ground it.  Now you'll have essentially a pre-tonestack master volume that is controlled by the tremonator oscillator.

After the quick insert position test I tried a quick test of the above using a .15uF cap…the position seems to work well trem/signal wise, but I get the same big pulse/thump (which you can visually see in the power tube glow and hear with speaker excursions).


So, at the weekend I’ll re-examine the trem circuit wiring (it’s on a mini board), ground points, lead dress and B+ node. I’m considering putting it on its own node.

[Jennings]

More head-scratching to come.  No wiring errors in the trem circuit.  I've updated the circuit diagram, including showing my grounding points and insertion points more clearly.

[ac427v]

Your original circuit in the first post was 1 connection away from trem nirvana.
I suggest back-tracking to that circuit but make the trem connection to the Treble pot wiper. That's it!

The 50k value of intensity pot only works on the ab763 circuit which originally had a 50k load resistor. Vox, Bassman, Marshall circuits needs a much higher value pot to keep from draining  signal to ground. I found a 250k B (linear taper) or Reverse Audio Taper gives a good range of control. The pot must be wired as in your original post (as a variable resistor). With NO CONNECTION TO GROUND. The goal is to maintain all of the signal except for what flows to ground through the LDR during trem operation.

[Jennings]

I thought the same myself originally...my first attempt at inserting the Trem-O-Nator was exactly as you describe...250k pot and at the treble pot wiper.  Just resulted in large signal loss as I increased the tremolo intensity.  That's what kick-started the experimentation.  I tried a 500k pot.  No improvement.


Anyway, I think I've cracked it!  Last night I removed both wires from the intensity pot to the vactrol and tested the amp.  No problems at all, so I know the amp and the oscillator are both working fine in isolation.  So today I examined my audio path insertion connection, and realised the turret I'd used was actually connected under the board to the V2a cathode - doh!  So I corrected that and berated myself for missing it the first time.  I also moved the HT supply to node B, rather than node D, and it's ground to point 2 in order to separate it from the rest of the preamp HT nodes and grounds. 


Tried firing it all up, and bingo!  Everything worked well, and sounded great.  Signal volume loss / imbalance issues look to be sorted...trem sounds lush...so I tried a 250K B pot, but that didn't work well here.  Basically the trem only appeared towards the very end of the turn, so very little range of control in the full sweep of the pot.  Not mmuch better with a 250k A for comparison.   Both were wired as a variable resistor.  Switched back to my 50k log pot and that just seems to work nicely in this position/configuration.  Much better balance between trem on and trem off signal, good range of control in the pot sweep (mostly the second half of the turn...might try a 50K linear at some point...or a 100k...don't have any to hand), and none of the almost volume pot type behaviour I'd experienced the first time I'd experimented with adding this trem circuit.


So, updated schematic attached showing what's worked for me.

[dickjonesify]

Yeah, “reverse audio” taper seems to be the go-to for trem depth. It’s an awkward control.

I found myself wanting more depth out of it so I just ditched the depth pot altogether and I love it haha.

[pdf64]

Attempting to modulate the signal level by loading it down at the cathode of a cathode follower seems counterproductive, because the source impedance there is pretty low, eg around 1k-2k.
So a 1k5 load only causes a -6dB reduction.
Rather than V2a cathode, it might be better at V1b anode (=V2a grid), or V1a anode, as the source impedance there is about 40k, and a load of 1k5 would cause -29dB level reduction.
So anode then blocking cap  (0.15uF) then depth pot then vactrol optoresistor.

[ac427v]

Great that you got it sorted. But sorry that my advice sent you down a rabbit hole. Great that you were smart and persistent enough to move forward anyway
I think the Trem-O-Nator is much smoother than the traditional Fender roach-based trem. I especially like using it in a two channel amp where a power tube bias trem (my favorite sound) affects both channels.

[Jennings]

Attempting to modulate the signal level by loading it down at the cathode of a cathode follower seems counterproductive, because the source impedance there is pretty low, eg around 1k-2k.
So a 1k5 load only causes a -6dB reduction.
Rather than V2a cathode, it might be better at V1b anode (=V2a grid), or V1a anode, as the source impedance there is about 40k, and a load of 1k5 would cause -29dB level reduction.
So anode then blocking cap  (0.15uF) then depth pot then vactrol optoresistor.

This is exactly the cause of my head-scratching the first time I tried inserting this trem.  I originally tried inserting it at the treble pot wiper, and then at the input to the volume control.  With various sizes of pot value.  In both these locations the first quarter turn of the intensity control just lowers the volume of the amp audio signal drastically, way before the trem effect kicks in as the pot is further turned.  I've not been able to find a way of inserting the trem in those locations without having large volume loss for the first part of the intensity pot turn.  Maybe that's fairly normal and I just don't like roach driven trem?  I only have one other amp with a typical Fender blackface style preamp and vactrol approach...in fairness that has a fair volume difference between trem on and trem off.

[Jennings]

Great that you got it sorted. But sorry that my advice sent you down a rabbit hole. Great that you were smart and persistent enough to move forward anyway
I think the Trem-O-Nator is much smoother than the traditional Fender roach-based trem. I especially like using it in a two channel amp where a power tube bias trem (my favorite sound) affects both channels.

No apologies required...this is all something of a tinkering and experiment on my part.  The amp I'm really happy with, and it's just all about finding where the trem might work best in it.  So all suggestions and options are helpful and fun to try.

[astronomicum]

In both these locations the first quarter turn of the intensity control just lowers the volume of the amp audio signal drastically, way before the trem effect kicks in as the pot is further turned.

I am wondering if the LED might be in a constant ON state, lowering the dark resistance of the LDR. Does your meter measure MIN, MAX and AVE? If so, lift one end of the LDR and with the LFO operating, measure the resistance across the LDR: MIN, MAX, and AVE. If you are using a Vactrol, cover it to protect it from light and repeat your tests.

[pdf64]

... I originally tried inserting it at the treble pot wiper, and then at the input to the volume control.  ...

I suggest not to use either of those nodes.
The source impedance at the treble wiper varies with tone control settings and can be very high, eg maybe 250k.

And 'at the input to the volume control' implies 'after a 500pF cap', so for much of the audio band, that's high impedance too.

[ac427v]

I am trying to understand why this circuit works.
The VTL5C1 rating is 50 M ohms when LED is dark.
Does this extremely high value make a source impedance of 250k irrelevant?

[astronomicum]

The VTL5C1 rating is 50 M ohms when LED is dark.

Regardless of what the ratings indicate, in my experience, you will get nowhere near 50M dark. And when it is in circuit and operating, the dark resistance is more like 100K. If the LED is in any state of ON, dark resistance will be even lower.

[Jennings]

The VTL5C1 rating is 50 M ohms when LED is dark.

Regardless of what the ratings indicate, in my experience, you will get nowhere near 50M dark. And when it is in circuit and operating, the dark resistance is more like 100K. If the LED is in any state of ON, dark resistance will be even lower.

This has been my experience too…the LED dark/light isn’t an instant reaction, so there’s a period of time between on beats where the light is fading out rather than instantly dark. Couple that with the recovery time of the optoresistor and you have a sort of working “off” resistance between trem beats which is in the hundreds of k if you’re lucky rather than the full M values of the photoresistor.

[Jennings]

... I originally tried inserting it at the treble pot wiper, and then at the input to the volume control.  ...

I suggest not to use either of those nodes.
The source impedance at the treble wiper varies with tone control settings and can be very high, eg maybe 250k.

And 'at the input to the volume control' implies 'after a 500pF cap', so for much of the audio band, that's high impedance too.

That makes a lot of sense…should I perhaps use a much bigger blocking cap from the V1a anode, before the 500pf coupling cap and connect at the join?

[Jennings]

In both these locations the first quarter turn of the intensity control just lowers the volume of the amp audio signal drastically, way before the trem effect kicks in as the pot is further turned.

I am wondering if the LED might be in a constant ON state, lowering the dark resistance of the LDR. Does your meter measure MIN, MAX and AVE? If so, lift one end of the LDR and with the LFO operating, measure the resistance across the LDR: MIN, MAX, and AVE. If you are using a Vactrol, cover it to protect it from light and repeat your tests.

Likely this is the case…I’ve yet to measure as such, but the volume drop effect isn’t there if you use the footswitch to disable the oscillator, even if you’ve pre-set the intensity pot to the area the volume drop kicks in (first part of the turn). It only comes into play when you then turn the oscillator on, bringing the LED and Photoresistor reaction into play

[pdf64]

That makes a lot of sense…should I perhaps use a much bigger blocking cap from the V1a anode, before the 500pf coupling cap and connect at the join?

As per reply 9, I suggest it might be better to apply the modulation at V1b anode (=V2a grid), or V1a anode, as the source impedance there is about 40k, and a load of 1k5 would cause -29dB level reduction.

So, superimposed on to the existing schematic, anode then blocking cap  (0.15uF) then depth pot then vactrol optoresistor.

[Jennings]

As per reply 9, I suggest it might be better to apply the modulation at V1b anode (=V2a grid), or V1a anode, as the source impedance there is about 40k, and a load of 1k5 would cause -29dB level reduction.

So, superimposed on to the existing schematic, anode then blocking cap  (0.15uF) then depth pot then vactrol optoresistor.

Ah yes, sorry…understood. Me not paying due attention. I’ll try that next.

[Jennings]

We have great results!  I inserted the tremolo at V1b anode, as suggested, and it sounds great with an acceptable volume drop level when engaged or disengaged via the footswitch.  Happy with that result, so thanks to pdf64 for steering me     I suspect it might also work well then at the anode of V1a accordingly, but haven't done an a/b test. 


I could play further with the optoresistor types and LEDs to perhaps improve on the now acceptable volume loss, but I suspect I might be chasing my tail and I'm happy with what I've now got.    Amended schematic attached for those interested.

[sluckey]

Are you sure it's connected as shown in the revised schematic? V2A doesn't seem to be biased correctly. What dc voltage do you measure on pins 7 and 8?

[passaloutre]

I think your cathode follower needs to be biased

[Jennings]

Great spot gents...many thanks.  Luckily that's just a drawing mistake on my part with the blocking cap position.

[ac427v]

Congrats on fitting Trem-o-Nator into another amp!-   I bet it sounds great!

[sluckey]

That'll work mo' better.