Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: stratomaster on October 16, 2023, 11:06:35 pm
-
I've incorporated the Trem O Nator circuit on a 63 Showman where the traditional optical tremolo did not respond to any of the usual tick/thump remedies.
I believe I've got the circuit implemented correctly, however I get a substantial volume drop when I increase the Intensity beyond 4. The volume continues to drop with increased Intensity.
Also, the circuit seems to lose it's stability/tracking and stops trem-ing if the Speed knob is moved too quickly.
Has anyone encountered these problems?
My current iteration is built per schematic except for a home-brew optocoupler to try and rule out the VTL as the culprit. The symptoms are the same with the VTL or my water clear red LED shrink-wrapped to a GL5537-1 hack.
I've tried lowering the plate resistor to 220k with no luck but a potentially decreased tremolo effect depth--hard to tell the difference for sure.
Would a resistor in series with the LDR (at wiper of the Intensity pot) be an appropriate countermeasure or should I look elsewhere?
For reference I used 2deaf's layout for avoiding under board wiring (attached).
-
I found the following information on the amp garage forum.
It appears that the type of VTL5C1 I was using is a known problem in this circuit and that NOS or Xvive units operate in the correct range for use in this circuit. This sheds some light on the problem--I am assuming my homegrown coupler is similarly deficient.
-
Removing the 10k cathode resistor improved the tracking with varying the Speed knob.
Xvive vactrol on order.
Despite this thread not attracting much interaction I'll continue to update as this documents a case of the a Trem-O-Nator implemented exactly as described by the schematic and proven layouts but problems still persisting. Others may find themselves in the same boat, so a thread like this could be useful to them.
-
Cool info. I've not built or immersed myself in a tremolo circuit, but I've read lots about them here.
I have built my own ldr vactrols, for channel switching. Ebay and Amazon have lots of them, but some cross referencing through Google is needed to get good info. I ended up finding ldrs with sub K ohm to 3M on/off range, frickin perfect for channel switching.
-
Xvive vactrol on order.
Despite this thread not attracting much interaction I'll continue to update as this documents a case of the a Trem-O-Nator implemented exactly as described by the schematic and proven layouts but problems still persisting. Others may find themselves in the same boat, so a thread like this could be useful to them.
Please do update your results as there have been instances of wonky results posted here in a few Trem-O-Nator installations & sounds like this could be the explanation. I am hoping to try that circuit myself & have wondered about the various versions/manufacturers.
-
Cool info. I've not built or immersed myself in a tremolo circuit, but I've read lots about them here.
I have built my own ldr vactrols, for channel switching. Ebay and Amazon have lots of them, but some cross referencing through Google is needed to get good info. I ended up finding ldrs with sub K ohm to 3M on/off range, frickin perfect for channel switching.
Making sense of the datasheets for these LDRs in the context of a tremolo is no easy task--especially because of unit to unit variation and no consistency even within manufacturers for what data they actually publish about each model.
On paper the 5537-1 looked good for this application. Published response time is similar to that of the VTL5C1. And the dark resistance seemed high enough. It's in parallel with at most a 50k resistance, so there are diminishing returns over ~500k.
In the Fender circuit what the Intensity pot is doing is essentially controlling how much of the 50k path to ground is in parallel with/bypassed by the LDR. So at the low end of the Intensity, the wiper to ground lug resistance dominates the paralleled resistance calculation since the 5537-1 only gets down to about 25k according to the datasheet. I was thinking my problem would be relatively low Intensity, not overall volume loss.
What must be happening is the LED isn't going full dark, or the LDR response time isn't as advertised, or some combination of the 2. The LDR only needs to get up to ~25% of it's published dark resistance for the volume drop to not be noticable. Combined with a slower LFO, care of an additional 0.01uF cap, I thought I was in the clear. Response time should be ~30ms and the LFO is less than 10Hz (100ms/cycle), usually much less.
But something isn't cooperating.
-
Despite this thread not attracting much interaction I'll continue to update
I agree that that should not deter you. My guess is that others, like me, hold off if we do not have something useful to add. We wait for someone with knowledge on the subject. Once that happens its usually off to the races :laugh:
-
Some experimenting:
Used a 5mm red diffuse LED to monitor the oscillator.
Visually, it doesn't look like the LED goes to full dark. I took video on slow motion and that was my impression both in person and on playback.
I also took resistance measurements across the LDR (of my homegrown coupler). My meter wasn't fast enough to determine a max, but the average resistance at the slowest speed (2.8Hz) was around 6.5k with a minimum around 3-4k. This also seems to confirm the LED isn't going to full dark.
Finally I wanted to monitor the current in the LED with two different plate resistances and with/without the 47k cathode resistor. My findings were as follows:
Used a 470k plate resistor with a 1 Meg on alligator clips to get 330k. Floated the ground lead of the 47k and connected/disconnected with alligator clips.
330k plate
Current w 47k cathode: 425 micro amps ac
Current no cathode: 430 micro amps ac
470k plate
Current w 47k cathode: 319 micro amps ac
Current no cathode: 320 micro amps ac
Visually the LED is brighter with the 330k plate Resistance.
The "sweep" of the pulse appears smoother with the 47k in place, but this could be a placebo effect.
If the LED isn't going dark enough, then I don't know how a "proper" VTL5C1 is going to help. :dontknow:
-
Put 2.2k across the LED to bleed that last mA of current without making light. Experiment.
-
Put 2.2k across the LED to bleed that last mA of current without making light. Experiment.
Anything in parallel with the LED is also cathode to ground. I had issues with tracking while varying the Speed knob with 10k and 20k in parallel with the LED. No tracking issues with 47k or with the resistor removed. More importantly, the large volume drop was present with the smaller parallel resistors.
Normally I'd just put a trim pot in series with the LED and control the current that way, but now I'm also upsetting the bias of the triode.
I'm thinking the oscillator bias and the LED/vactrol need to be decoupled somehow. Perhaps the spare triode needs to be made a cathode follower or other buffer circuit to allow the vactrol to be tweaked with current limiting LEDs without disturbing the oscillator.
-
Gif of LED in oscillator attached. Hopefully it plays.
-
Installed an Xvive VTL5C1 and the results are much improved. With a 470k plate resistor the LDR varied from around 13k to 52k. A 330k resistor resulted in 10k to 32k. This was a much deeper tremolo, but the volume noticably dropped off with increased Intensity.
I settled on a 390k plate with a 47k cathode resistor. I'm sure there is more optimization to be had by playing with the cathode resistor and perhaps cycling through a few more VTL5C1s, but I'm satisfied for now. The tremolo sounds good. Sits between a good Blackface and a good bias trem. The best feature is the lack of extraneous noise, clicks, ticks, and thumps. This allows me to hard wire the tremolo ON, and use a switched Intensity pot for the gain boost--but the Intensity rolled down is quiet enough to pass for the LDR being out of the circuit.
Really wish I could use a 220k to 330k resistor for increased intensity, but the volume drop is just too obtrusive for my purposes.
I'm going to call this one tentatively resolved.