Harmonic Tremolo Oscillator Challenges

[FastNBulbous]

Hi there--I just finished building a 6G12 Concert scaled down into a Princeton Reverb-sized single channel 20-watt amp using a Deluxe Reverb output transformer.  The main reason I embarked on the project was to build a combo with the famed brown harmonic tremolo--the Princeton chassis' 4 12AX7 slots made it pretty easy to fit the 4-triode non-A tremolo version. 

I've got the amp up and running and sounding as it should, happy with filter and other voltages, it's just that the tremolo has been surprisingly finicky.  The main issue is that above about 3/4 on the 3MRA Speed knob, the oscillation stops and the effect disappears.  Additionally, it's a bit slow to start up and once killed by cranking the speed, it must be restarted at the switch.  I'm a bit mystified as the build was very close to stock values (and I've also studied Sluckey's very helpful annotated Revibe schem, which this build matches http://sluckeyamps.com/revibe/revibe.pdf)--the only noteworthy changes in the oscillator circuit are biasing with a red LED (very close to the 4k7 resistor specified) and .02 caps in all 3 oscillator positions. I've been scouring existing topics and haven't found anything that seems to describe this issue.

Here's what I've tried so far:
-Higher and lower cathode values, including the original 4.7k. The effect seems to eventually stop kicking in altogether with both higher (white LED) and lower (1k5Ω) values.
-Swapping back in the .01uf caps - almost seems worse and harder to start up.
-Increasing plate voltage on the oscillator triode, which starts out pretty low with a 470k plate resistor - high enough voltage and the effect disappears.
-Connecting one of the 1M speed resistors to the bias supply like in the AB763 (instead of the oscillator cathode) for a different jump start; seems about the same.
-Known-good very strong 12AX7 swapped into the oscillator tube position; seemingly no effect.

Is my only option to increase the 100k Speed knob tail resistor until the full knob range is usable?  The max fast speeds I'm getting aren't that fast compared with other Fender circuits employing similar oscillators that could definitely go faster without killing the oscillation.  I'm also curious about increasing the overall tremolo intensity--I've reduced one 1M resistor in the 6G12 to 470k as in the Revibe circuit but not sure if there are other opportunities there.

[Jalmeida]

Subscribed…

[FastNBulbous]

So I've moved on from initial tweaks and got the amp into a room with the cab for some actual playing.  After some experimenting I soldered in a 220k tail for the speed knob, which is the fastest I can get at top speed without it dying.  Also getting some pumping above 3 o'clock on the intensity knob--looking at increasing intensity caps or maybe the 4.7m resistor.  Trem sounds good, though, just wish it had a wider ranger than it's got, but maybe I'm just being wishful with the circuit.

[PRR]

> .02 caps in all 3 oscillator positions.

Stepped values work better; lower minimum gain. Lets first confirm Fender's 0.02 0.01 0.01 series. Then try 0.03 0.015 0.015

No, you do not get wide range when you vary just one part in a 3-section network. (If I sound snotty it is because I started on early synths which would LFO from 1000Hz to Days.)

[HotBluePlates]

.. I just finished building a 6G12 Concert ... harmonic tremolo ...

... it's a bit slow to start up and once killed by cranking the speed, it must be restarted at the switch.  ... (and I've also studied Sluckey's very helpful annotated Revibe schem, which this build matches http://sluckeyamps.com/revibe/revibe.pdf) ...

Your build places the "Vibrato Footswitch" in the location used by Sluckey's Revibe drawing?  Because it's different than where Fender placed their footswitch.

   -  Basically, these oscillators are slow to start by nature.

   - The 6G12 Concert has the footswitch in a location that starts/stops the oscillator.

   -  The Revibe circuit moves the footswitch so that the oscillator is always running; no more slow-starting.  Instead, it mutes/un-mutes the output of the oscillator.

   -  You might be able to modify your oscillator to shock-start with the bias supply voltage, as found in the Deluxe Reverb.  That is, if you choose not to follow the Revibe plan of moving the footswitch to a different location in the circuit.

[FastNBulbous]

Your build places the "Vibrato Footswitch" in the location used by Sluckey's Revibe drawing?  Because it's different than where Fender placed their footswitch.

No, the switch on my build is in the 6G12 location.  Since my issue is the oscillator petering out at both ends of the speed control I'm inclined to keep it in this location for the time being, since closing the switch in this arrangement provides that jump start--not sure how this would or could be achieved if the switch was in the "mute" location and the oscillator stopped.

[sluckey]

Put the original cap values in the oscillator. .02 plate, .01 middle, .01 grid. See if the oscillator drops out on the high end. Then if you want slower speed replace the middle cap with .02. Should have .02, .02, and .01.

I'm unclear on which LED you are using. White is not good. Use a red or yellow. They give a good bias voltage for the cathode.

The oscillator should be strong with no drop out at the high or low end of speed.

Now, with the footswitch set to turn off the tremolo (contacts closed), measure the oscillator plate voltage. Also measure the B+ voltage on the other side of the 470K plate resistor. What have you for both measurements?

[FastNBulbous]

The B+ node supplying the oscillator tube measures 307v, and the plate is 159v, both of which seem reasonable compared to the 6G12 and your (Sluckey's) Revibe circuit. And yes, I'm using a red LED with a voltage drop very close to the 4.7k resistor.

In response to Sluckey and PRR's suggestion to return the caps to .01, I've done so and the LFO is maintaining at all speeds--doesn't crap out at highest and lowest speed settings, although the thumping is worse and affects probably half of the speed range.  Not crazy about the limited range, but I already have a switch installed in the vain hope that I would be able to get even slower than .02/.02/.02, and now have that configured to offer me .02/.02/.02 as a switchable option slower than .02/.01/.01. 

At this point I can learn to love what I have, speed-wise, but the LFO thumping/pounding at higher (>noon) speed settings with max Intensity is something I'd like to be able to improve if possible.  I already increased the cap on the Intensity input leg from .047 to .1, which helped some, but did start reducing the harmonic tremolo bass frequencies.  Is there anything else I can do to reduce LFO thump besides reducing the Intensity knob below 3:00?

To PRR's point

(If I sound snotty it is because I started on early synths which would LFO from 1000Hz to Days.)

, I think I was coming from a similar (pedal) perspective and expecting it to be just as simple and easy to get a much slower LFO, when in actuality the wiggle room is pretty limited in this circuit.  The timbre of the tremolo is great, though, so I can force myself to get happy with a limited speed range. 

Thanks for the replies and help!

[sluckey]

The red LED causes the LFO signal to be much greater than the 4.7K/25µF. This is the root of the thumping. So reducing the LFO signal after the plate should be the solution.

Experiment with using a larger value to replace the 4.7M between the plate and INT pot. Or use a smaller value INT pot, maybe a 3.3M.

[FastNBulbous]

Thanks, these suggestions were really helpful in evaluating and making decisions.  It was the easiest to clip in >4.7M resistors and caps in parallel with the existing intensity cap (I wanted to avoid ripping out the intensity pot if I could avoid it since it's got mounted components). I also subbed in the 4.7k/25uf cathode combo and was surprised that the thumping was still there--should have but didn't try just the resistor by itself.  After some ABABAB-ing with an 8.2M and 10M resistors in place of the 4.7M, I decided to keep the 4.7M and keep the circuit as it is, thumping and all.  Why?

Swapping in the higher resistors attenuated/removed the thumping, unsurprisingly at the cost of tremolo depth, but to my ears didn't sound any different from rolling back the intensity knob a bit with the original 4.7M resistor installed.  Yeah, I'd love not to have a control that has thumping at one end, but the tremolo depth sounds REALLY GOOD with intensity cranked--with an 8.2M resistor, this level of depth is no longer an option, AND the first half of the intensity range is very subtle to nonexistent.  The other thing is that the thumping doesn't increase with the volume control--with the gain cranked and the tremolo maxed, there are some crazy sounds, especially at high speeds--very Leslie-like. 

So, barring some new revelation, my current understanding is that you can't get that deep of tremolo without the thumping, and that it can be worth it to have the option available. Appreciate the guidance on things to try, definitely got me there.  And the prodding to return the oscillator caps to stock definitely solved the startup and unwanted stopping issues.  Thank you thank you, all!  This amp was a fun project and sounds really nice.

[Dave]

Look at the Vox AC30 version of the harmonic tremolo. At the output of the circuit there is a string of capacitors with resistors to ground at the junction of the capacitors. That is a high pass filter. It is there to remove the thumping/pumping/noise. Add that filter or some variation thereof that you are happy with and it can solve your problem.


Dave