Tremolux vs. Vibrolux tremolo?

[Wombaticus]

Anyone have any first hand experience with both of these tremolos?

The Vibrolux 5F11:


http://www.schematicheaven.com/fende...5f11_schem.pdf

and the Tremolux 5G9:

http://www.schematicheaven.com/fende..._5g9_schem.pdf

Note that the Vibrolux uses only a single 12AX7 half and the depth pot is driven directly from the plate, whereas the Tremolux uses both halves of a 12AX7, using the second half as cathode follower buffer before the depth pot. Both wiggle the fixed bias of a push-pull 6V6 power section.

Anyway, the question really is if there is a significant tonal benefit of going to the trouble of including the cathode follower?

Note also that my inclination is to go with a LND150 for the oscillator rather than use a tube in the case that I decide to do a Vibrolux style, which would frankly be the path of least resistance, and to be preferred unless someone tells me the Vibrolux tremolo sucks.

If I went tremolux style I'd probably break down and fit another tube, although I'm obviously not a purist when it comes to using sand for oscillators and source followers.


Thanks,

Paul

[R.G.]

Don't have an opinion about the merits of tremolux versus vibrolux, but MOSFETs make a great replacement for triode LFO oscillators **and** for cathode followers.

[Wombaticus]

Don't have an opinion about the merits of tremolux versus vibrolux, but MOSFETs make a great replacement for triode LFO oscillators **and** for cathode followers.

Well, sure. Cookbook stuff - well, ever since you helped to write that particular cookbook. Thanks RG.

[tubeswell]

Hi Paul

I built a 5G9 clone a couple of years ago and really love the hypnotic slam it gives - It is still my favourite tweed amp of all time. The fixed bias aspect makes it a bit cleaner than say a 5E3, but the LTP really drives those 6V6 grids well and the whole circuit hangs together particularly well I think. The CF stage really enhances the operation of the trem circuit

I haven't tried a 5F11, but I have built a AA1164, which has a similar single-triode LFO wiggling a pair of fixed-bias 6V6s, and also utilises a driver-cathodyne PI. The AA1164 trem is nowhere near as good as the 5G9 trem. It is noisier all round and hasn't got the same hypnotic quality. I imagine the 5F11 similar to a 'dry' AA1164. I am now in the process of modding my AA1164 to have a MOSFET source follower stage after the LFO stage - haven't finished that mod yet tho'.

Excuse the bad ad-hoc twiddling - this is the (wiggly) 5G9 backing the (lead) 5G9 on quietly with a 6G15 clone plugged in front of it. Gives you some idea of the trem intensity quality tho.

[Wombaticus]

Yes, that's the trem quality I crave! ... looks like I'm gonna have to redo my layout to incorporate a source follower, or just man up and shoehorn in another valve.

[Wombaticus]

Hi Paul

I built a 5G9 clone a couple of years ago and really love the hypnotic slam it gives - It is still my favourite tweed amp of all time. The fixed bias aspect makes it a bit cleaner than say a 5E3, but the LTP really drives those 6V6 grids well and the whole circuit hangs together particularly well I think. The CF stage really enhances the operation of the trem circuit

I haven't tried a 5F11, but I have built a AA1164, which has a similar single-triode LFO wiggling a pair of fixed-bias 6V6s, and also utilises a driver-cathodyne PI. The AA1164 trem is nowhere near as good as the 5G9 trem. It is noisier all round and hasn't got the same hypnotic quality. I imagine the 5F11 similar to a 'dry' AA1164. I am now in the process of modding my AA1164 to have a MOSFET source follower stage after the LFO stage - haven't finished that mod yet tho'.

Excuse the bad ad-hoc twiddling - this is the (wiggly) 5G9 backing the (lead) 5G9 on quietly with a 6G15 clone plugged in front of it. Gives you some idea of the trem intensity quality tho.

The 5G9 is certainly looking very appealing to me too. What kind of voltage are you putting on the plates of the 6V6s? 350-370 like with the 5E3? Also, what rating of choke is fitted in the original? Finally, how does the 5U4GB stack up against the 5Y3GT in the 5E3, voltage drop, sag, and otherwise?

Many thanks, Paul

[tubeswell]

The voltages are all pretty close to the Fender schematic/layout values. The choke is the standard 022699/125C1A Vintage Style Fender Choke. Not nearly as saggy as the 5E3, but the trem slam is better with the 5G9 topology, and if you put a hotcake etc in front, it is one helluva rockin' machine. I found I preferred a .047uF coupling cap on the 'inst' channel (instead of 0.1uF) and an option for a 1.1uF or 2.2uF bypass cap on the first stage is better with a 12" speaker at full blast. I did use an OT with a 5k primary. The stock amp uses 8k. Anything between will work

[Wombaticus]

The voltages are all pretty close to the Fender schematic/layout values. The choke is the standard 022699/125C1A Vintage Style Fender Choke. Not nearly as saggy as the 5E3, but the trem slam is better with the 5G9 topology, and if you put a hotcake etc in front, it is one helluva rockin' machine. I found I preferred a .047uF coupling cap on the 'inst' channel (instead of 0.1uF) and an option for a 1.1uF or 2.2uF bypass cap on the first stage is better with a 12" speaker at full blast. I did use an OT with a 5k primary. The stock amp uses 8k. Anything between will work

Hi Tubeswell,

If you have a scope handy, could I impose on you to take a couple of measurements for me? Or perhaps you know the answers offhand...

I've decided to do a wombatified 5G9, with a pure FET version of the tremolo (depletion mode device for the oscillator amp and an enhancement mode for the cathode/source follower). I've got an LTSpice model up and running, but it would be very very helpful to me to know what the peak-to-peak voltage swing of the valve trem at the power tube grids is in your unit when it is on the maximum depth setting, with the speed setting on both min and max. Also, what is the range of frequency for your oscillator?

I know it's asking rather a lot, so if it's too much trouble, don't bother. But if it' is possible, it would be very much appreciated. I want to get as close as possible with the model first, because I don't particularly relish tweaking in the confines of tight tweed combo style build.

If I can get something very close to a nice-sounding valve trem behaviour, I'll gladly post a layout for a little drop-in FET based tremolo board that would be a fun and easy mod for many projects.

[tubeswell]

Unfortunately I don't have a scope. Its on my to-do list - so one day I would gladly oblige.

[Wombaticus]

Don't have an opinion about the merits of tremolux versus vibrolux, but MOSFETs make a great replacement for triode LFO oscillators **and** for cathode followers.

At the end of the day I went for the all FET route for the trem. The build is a 5G9 tremolux http://www.schematicheaven.com/fende..._5g9_schem.pdf with the tube tremolo replaced with an all FET arrangement - a LND150 for the low frequency oscillator and a STF3NK80Z as a source follower: http://www.wombatamps.com/Tremolo_FET.pdf

Since this is the first time I've implemented a trem of any sort, I'm pleased to report that it works... more or less. As expected, it *is* going to need just a little bit of tweaking. Right now it oscillates fine at higher speeds, but if I make it too slow, the oscillations die out before reaching the end of travel of the speed knob - I'd like 'em to go slower. It could also use some more authority in the depth department..it just doesn't have that hypnotic quality.

The strangest thing so far though is that if I quickly reduce the speed (or it is just right I suppose), and I hit the right note or something, the trem suddenly takes on a much deeper and wildly spacey character...like a reverb gone wild, or delay - kind of fun actually but not right. Once this has started, I can then turn the volume knob on the guitar all the way down (or even unplug it), and I can still hear the trem whupping like a helicopter (it's not quite like a motorboating sound)... apparently something gets coupled to the LFO, and the LFO becomes audible without any further input from the guitar. I can control the speed with the speed knob, although I can kill it by reducing the depth or speed enough. I haven't had much chance to troubleshoot.. had no sooner finished the amp than had to take the wife out to dinner. But I'm assuming this is probably this just a layout/lead dress issue, although a quick bit of chopsticking didn't help. I suppose I may need to shield one or more of the runs to the speed or depth pot -- I'll continue to experiment as time permits, but any insight or advice would be most welcome.

Cheers,

Paul

[Wombaticus]

Interestingly, the helicoptering seems to have something to do with the plate voltages/bias on the 6V6s. I was running a NOS 5Y3GT and getting about 360V on the plates when this was happening as described above.

I tried swapping in a 5U4GB and the plate voltage went up to 385V -- but the depth of the tremolo improved, presumably because of the change in bias, and with a minute or two of fooling around wasn't able to induce the helicopter effect. Unfortunately though, with the higher voltage the 6V6 plates began to redplate, and my bias control circuit was already as cold as it would go. I'm going to to have to swap a resistor in that circuit if I want to run the 5U4GBs with this amp. Or maybe first I'll put in my JJ 6V6s, which are damn hard to redplate.

[tubeswell]

Interestingly, the helicoptering seems to have something to do with the plate voltages/bias on the 6V6s.

That's right. With a bias-vary trem - if you change the bias voltage - you change the trem (and vice versa). Cooler biasing gets more slam

Another way of increasing the swing could be to up the source follower resistor (to say 220k?) - but you may also need to adjust the bias voltage accordingly

I'm going to to have to swap a resistor in that circuit if I want to run the 5U4GBs with this amp. Or maybe first I'll put in my JJ 6V6s, which are damn hard to redplate.

It might be the same with the JJs. I'd keep the 5U4G, but adjust the bias voltage and see how you go. With bias-vary trem, you need to run the output tubes colder at idle (say 18-22mA), because they heat up to about 30mA when the trem kicks in, especially at lower trem speeds.

[Wombaticus]

That's right. With a bias-vary trem - if you change the bias voltage - you change the trem (and vice versa). Cooler biasing gets more slam

Well, I understand how the bias point might influence the evident depth of the trem, but it's less clear to me why I would get the weird audible oscillations and presumed coupling and/or feedback effect at one voltage vs. another...
I'll just have to continue to experiment. Next step will be to alter my bias circuit to allow a colder bias appropriate for the higher plate voltages, and see if the helicoptering stays gone, or if it comes back... I haven't had very much time to play with it yet. I believe it doesn't hurt to have the 5G9 voltages a bit higher than in a 5E3, and in any case I'd like a tad more headroom from this amp than from my deluxe-ish build, so I probably will stick with the 5U4GB recto. Besides, although I have a selection of standard 5U4s floating around, I have one Russian 5U4GB recto close equivalent (5U3C) in particular with a big coke bottle form factor that is just too cool looking not to finally put to work: Winged "C" (SED) 5U4-G

[tubeswell]

Well, I understand how the bias point might influence the evident depth of the trem, but it's less clear to me why I would get the weird audible oscillations and presumed coupling and/or feedback effect at one voltage vs. another...

Probably something to do with the slightly uneven plate swing from the LTP hitting the 6V6 grids, and if the bias is at the right point the trem picks up on the difference and plays with it - but I know the effect you are talking about, and biasing will change that

I have one Russian 5U4GB recto close equivalent (5U3C) in particular with a big coke bottle form factor that is just too cool looking not to finally put to work: [/url]

bingo - I got exactly the same sort in my 5G9, and they are good rectifiers

[Wombaticus]

Played with this just a bit more tonight....I've now tried this with a fairly wide range of B+ voltages, power tubes, and bias settings from quite hot to quite cold. Some combinations are definitely better sounding and more generally stable than others, but I can always trigger the effect eventually.

I've discovered that there is a huge bias shift that takes when the phenomenon takes hold. In one particular instance, I can watched it go from a 19mA/385V plate current/voltage with the trem operating normally, all the way to 50mA/330V (yikes!) when the helicoptering kicked in... that's right, the current draw increase was dramatic enough to sag the B+ a full 55V (and a good thing, too)! However, the bias supply voltage itself does NOT seem to be affected... it must be high amplitude oscillations within the LFO itself, or perhaps some kind of bias runaway within the power tubes themselves?

Whatever is causing the bias shift, it is not a loose connection.... For one thing, I can begin and end the helicoptering at will by adjusting the speed or depth knobs.