Understanding Early Tremolo circuits

[phydauex]

Hey All,
Been working through an amp based on a Brown Deluxe (6G3?), and was wiring up the tremolo circuit. I used the design from the Tremolux (5G9) that uses the cathode follower. So far it works, all though it seems to be very fast at about 1/2 way up on the speed knob, haven't had the time to go in and mess with it. So, I have several questions, that reference bias modulation tremolos:

1) Is there a reason that most higher powered amp designs (ie 6L6) use the cathode follower style tremolo and 6V6 designs use the single triode stage? I mean any specific reasons why that would be other than inertia of designs.
2) What does the cathode follower do to the single stage triode style tremolos, I guess to put it more simply, why is it there.

Just trying to get a better grasp of exactly what's going on and why.
Thanks for any insights and experience. I'm trying to build this 6v6 based amp with the throbbiest, swampy tremolo and still have good gain/crunch in the amp like the Brown Deluxe does.

[txstrat]

Well, I can't explain what's going on inside the circuits.
BUT, I've built a 6G3 (Deluxe) and a 6G11 (Vibrolux) and I must say the 1 tube tremolo of the Vibrolux is way better (more lush) than the half tube tremolo of the Deluxe IMHO.
Although the half tube trem of the 6G3 is still better than the optocoupler trem of the later models IMHO.
If I had to describe the sound of the power tube bias trem I'd say it sounds like you can chew on it... just great.

Hope this helps

Matt

[tubeswell]

You can slow down the speed range by increasing the capacitance in the oscillator RC network. To get a significant change in speed range, you only need to change 1 of the caps, e.g.; change the first cap from the plate from .02uF to .03uf (say - by solder-tacking a .01uF in parallel with the .02uF).

[phydauex]

Yeah, I noticed that done of the Tweed/Brown trems had a .03uf cap in there and some had the .02,. I'll do that first.
I agree with you Txtstrat, just about every bias trem no mater how cheap the amp, sounds better to my ear than any of the opto circuits. Even an old Multivox SE amp that had a great single tube trem (6AU6? can't remember now). I was just curious as to why some people say higher gain amps work better with the CF design.
Thanks Tubeswell for the reminder.

[jimboyogi]

Hey All,
Been working through an amp based on a Brown Deluxe (6G3?), and was wiring up the tremolo circuit. I used the design from the Tremolux (5G9) that uses the cathode follower. So far it works, all though it seems to be very fast at about 1/2 way up on the speed knob, haven't had the time to go in and mess with it. So, I have several questions, that reference bias modulation tremolos:

1) Is there a reason that most higher powered amp designs (ie 6L6) use the cathode follower style tremolo and 6V6 designs use the single triode stage? I mean any specific reasons why that would be other than inertia of designs.
2) What does the cathode follower do to the single stage triode style tremolos, I guess to put it more simply, why is it there.

Just trying to get a better grasp of exactly what's going on and why.
Thanks for any insights and experience. I'm trying to build this 6v6 based amp with the throbbiest, swampy tremolo and still have good gain/crunch in the amp like the Brown Deluxe does.

Here's my amateur take on this.

6V6's run with a lower bias voltage, and also usually a higher value of grid leak resistor. 6L6's have a higher bias voltage, and lower value of grid leak resistor.

This means that to "wiggle the bias" to 100% (or 50%, or whatever proportion of max. input grid voltage), the tremolo oscillation voltage required to wiggle a 6V6 is lower than that required for a 6L6. So automatically a 6L6 amp needs more trem voltage swing.

In a 1 tube LFO, the output from the plate is cap coupled into the grid leak resistor path to ground. So the output grid leak/trem intensity pot form the AC load for the LFO tube.
Because 6V6's can have a higher max. grid leak resistance, then the AC load for a 6V6 amp is usually set at a higher impedance to ground, and therefore the LFO tube can produce a reasonable output voltage swing, enough to vary the smaller 6V6 bias as much as is required for the trem effect.

But if this 1 tube is used with the lower value of grid leak resistance required for a 6L6, the larger AC load on the LFO tube lowers it's max. output voltage swing. Combined with the higher voltage swing required to swing the larger 6L6 bias, 1 tube is just not enough to get the required trem effect on 6L6's.

So for the 6L6 amps, a cathode follower is used between the LFO stage and the grid leak/trem intensity pot resistance network. The cathode follower forms an impedance buffer that allows the LFO to produce it's full output voltage swing, and therefore enough voltage to wiggle the higher 6L6 bias voltage adequately.

[phydauex]

Ah Soo! your Amateur take makes most sense. After posting this, I found a tech page on Aikens site that went pretty heavy into the design calculations. While I kind of glazed over starting to read it, the conclusion was similar to your very good explanation. I have it set up currently with the CF, and could get away with just the one triode. I may leave it as it is as long as having to extra voltage available doesn't overdo it. I also could do Tubeswell's (is that right) mosfet version of CF trem and put that 1/2 a triode back into gain chain. Thanks for the layman's description, I had a sense that was the case, but as usual, can't always articulate it.

[EFK]

Most of the early (pre 1960) Valco-based trems used only a single triode and set it up so that the cathode of the oscillator was linked to a preamp or paraphase (shared) cathode. This also can give you a good swampy pulsing and doesn't act on the bias of the power section. The intensity control was a pot in the B+ line feeding a oscillator plate R - this also made for a very effective intensity, and I use this a lot (although now I use those 2W PEC pots, while the originals were just standard 250K pots). The Gibson GA20T also made use of this B+ intensity setup although that used a 500K pot. The whole shebang only "costs" a half a 12AX7, which is great.

[soundmasterg]

Most of the early (pre 1960) Valco-based trems used only a single triode and set it up so that the cathode of the oscillator was linked to a preamp or paraphase (shared) cathode. This also can give you a good swampy pulsing and doesn't act on the bias of the power section. The intensity control was a pot in the B+ line feeding a oscillator plate R - this also made for a very effective intensity, and I use this a lot (although now I use those 2W PEC pots, while the originals were just standard 250K pots). The Gibson GA20T also made use of this B+ intensity setup although that used a 500K pot. The whole shebang only "costs" a half a 12AX7, which is great.

Got any Valco schematics to point me to on that one? Sounds like an interesting circuit to try out!

greg

[tubeswell]

Gretsch G6159, Supro S6698, 1624T, S6699 etc

[jimboyogi]

This trem system sounds very interesting!
How does it sound, and why did most amp companies not use this system?

[EFK]

All of those schematics are of the later (post 1960) Valcos which utilized both halves of a 12AX7 and a more conventional intensity control. I have never seen a published schematic of the earlier "organ tremolo" as they called it, most dating to ca. 1957 through mid 1959 amps. It isn't an over the top spectacular trem, but if you have things set up right it can get pretty swampy and makes for a nice simple trem that doesn't take up much tube real estate. It took me a while to figure out how they were implementing it - I had to keep buying old amps!

I suspect it is not used much anymore because of the modern stigma of putting the @ 350-370V supply voltage across a pot, but I've never had a problem using the PEC pots (although they're probably overkill, as I think the pot only has about 1/2 watt across it at most) and as I see it, assuming that the chassis is properly grounded as well as the pot casing, I can't see any more of a danger in putting the B+ across an appropriate pot as compared to say a switch or any other chassis mounted item.

If you're a heavy trem user, you might want something more complex, but if you just want it as an occasional effect and have a half 12ax7 to use, it's pretty cool.

[trem]

Most of the early (pre 1960) Valco-based trems used only a single triode.... The whole shebang only "costs" a half a 12AX7, which is great.

Always wondered how to spell 'shebang'.