Your Favorite Tremolo Circuit

[patlaw]

If you were designing an amp that includes tremolo, what circuit would you use? The Sunn Sentura II tremolo requires an unobtanium roach and a three-gang 25K pot. Old Fenders use roaches with neon bulbs. Do you have a circuit without a roach that you like?

[Enzo]

Why assume the Sunn circuit MUST have a three decker pot? They made all three feedback stages variable, but certainly no Fender ever needed more than one stage variable.

Fender also made their share of "bias wiggler" trem circuits. Can't offhand think of a particular, but there were multiple ones.

[patlaw]

Why assume the Sunn circuit MUST have a three decker pot?

I'm just reading the schematic. While I've never had a problem fixing a tremolo circuit, I would have a real problem standing in front of a classroom explaining how to design the oscillator portion of the circuit. Maybe one day I'll dig into oscillator design using discrete components, but maybe not.

[Helmholtz]

I don't have a favorite tremolo circuit, because I never cared for the effect (don't even know if the tremolo in my Fender amps works).

Tremolo is just simple signal amplitude modulation, so could easily be realized with a pedal before the amp.
The waveshape of the LFO signal might matter, though.

I'd try to avoid a roach and maybe look for a Fet solution instead.

[mhuss]

The Fender 6G5 Pro-Amp modulated the signal at the PI, as did the 6G14 Showman and 6G8 Twin. My favorite, though, are the "bias wiggler" ones (as Enzo referred to them) like the 6G9-A Tremolux.

In general, the 6xx-A Fenders seem to have the sweetest trems, for my 0.02.

[Chuck H]

Like Helmholtz I'm not a tremolo user. I mean, sure, in the right application I might (?) but it's never been a go to effect for me. That said...

I've spent enough time time perusing the particulars of "tone" here and abroad to know that bias wiggle trems seem to be the preference among aficionados. The last amp I built did use a bias wiggle trem and though it was never quite right at higher effect settings (depth) it does have the benefit of taking the entire preamp signal to the trem. It does sound very "natural" if I can use a subjective term. This kind of circuit also has the benefit of mitigating shut off/turn on effect that other trem circuits can have because the tubes natural tendency to amplify relative to voltage drops and current kind of force a soft off/soft start action. This is probably why this sort of circuit is preferred. Of course this could be simulated with an outboard circuit or a preamp based circuit with enough effort. But like you, I might not know exactly how. But since bias wiggle trems are relatively easy to build into a design there's probably no reason to redesign the wheel. But I will say this...

If the rest of the amp is intended as a relatively clean amp using high plate voltages on the power tubes getting a nice trem envelope will be difficult or impossible. That was my issue in the build I mentioned. Relatively high plate volts means the useful wiggle is somewhat limited. So bias wiggle trems are probably best reserved for more moderate or even low-ish plate voltage power amplifiers. These design considerations should be made in advance relative to other goals for the design.

JM2C

[loudthud]

The advantage of the roach type tremolo is that it changes the gain without injecting the LFO signal into the single ended guitar signal. Some amps do this like versions of the AC30 and they have to filter out the LFO with a string of 4 to 6 caps and resistors. JFETs have limited breakdown Voltage so not great for tube circuits. The bias wiggle is great because the LFO signal cancels in the output transformer. The bad thing about the Sunn circuit is that when the light bulb fails, the photo cell goes to high resistance and you lose the guitar signal. In the Sunn circuit, the photo cell is in series with the signal. In the Fender circuit, the photo cell shunts the signal.

Funny Story: Last time I spoke with Gerald Weber he told me he had applied for a patent on a tremolo that injected the LFO into the phase inverter. I opened up one of his books and showed him the Fender 5E9-A. We got into an argument and I won when I pointed out how the Intensity Control is wired straight to the Phase Inverter cathodes. One of the guys from the Bumble Bee group (they work with Gerald) came up and told me that was the only time he saw Gerald lose an argument. I spent a couple of hours looking for Gerald's patent application and didn't find anything.

Link: https://el34world.com/charts/Schemat...5e9a_schem.pdf

The 5E9-A also cancels the LFO signal in the output transformer. It's a little tricky to set up, sensitive to B+. You might be able to make a transistor current source for the PI and wiggle that.

[Helmholtz]

To replace an obsolete roach this might be worth a try: https://export.farnell.com/advanced-...3168773?st=ldr

https://www.farnell.com/datasheets/16352.pdf

[Mick Bailey]

I've used NSL32 series optocouplers quite a bit and they work fine so long as the LDR is operated way below its maximum voltage, as they don't last too long otherwise. I've used a lot of DIY roaches with a shrink wrapped LED/LDR combination and they work really well - the LED will replace a neon in most amps. I bought a load of Chinese Vactec copy optocouplers and they also worked fine, but they're not lightproof! I had to mask off the leads and spray them with black paint.

[nevetslab]

Chris Squire, bassist for YES had a pedal made for him over in the UK that he used on All Good People, which was my favorite sounding Tremolo/Vibrato effect. Early on at CenterStaging, that pedal was brought to me when Chris and some of his mates were in at CenterStaging, and I didn't yet know this was that wonderful-sounding effect until I had found the fault and repaired it. When I plugged my bass into it to have a listen, I instantly recognized that sound, and here was that pedal that he used for that. Still kicking myself for not stopping right there and cobbling up the schematic of the circuit, as there weren't a huge pile of parts involved. 9V powered circuit, all discrete. No name or details on the pedal, just an obscure little pedal box that worked really good.