Trem in nondescript Japanese amp

[AZJimC]

Howdy Y'all!
Ive been in the periphery for a long while now, but I finally got into a couple old amps, and now seek advise.

I'm working on an old Japanese amp that has/had trem. The oscillator is a 6av6 (Diodes unused), and it's plate is fed B+ via the "Depth" pot.. It does oscillate and thump, etc.  I want to modify this of course, Cathode is hardwired to first input triode on 12ax7 cathode, which has 4K resistor, and 10Mfd cap to ground.

The question up is, what direction should I take it?
Considerations are,
A. I'd love to swap sockets and use a 12ax7 for trem, hoping I can get another gain stage that way.
B. possibly best to seek a solid state oacillator, because I really like very slow trem, at very low intemsity, and tubes seem to drop out at lower than 3 beats a sec.
C. I have what I need to use 12ax7 for this spot, so if I go with that, whats the most effective circuit for single triode trem?

Thanks up front,

Jim

[sluckey]

The Marshall 18 watt (model 1958 or 1974) has a very nice single triode tremolo that modulates the first preamp tube's cathode. You would probably have to do a bunch of tinkering to get it below 3Hz though. I've messed with a bunch of trem circuits and 3Hz seems to be about the bottom end. Maybe a 4 stage cap circuit could do lower freqs???

[AZJimC]

Hi, Steve (if I'm recalling correctly),
I'm thinking about just getting this mess sorted out for now, and a working trem, then worry about the slower version another day. I'm certain I'd have to lay out some cash for a solid state solution, and I need a working amp for now... The 18w may be an answer, I'll give it a look, and thanks for your response.

[PRR]

> very slow trem, at very low intemsity, and tubes seem to drop out at lower than 3 beats a sec.

For "slow", just increase the caps (all of them, including the cathode cap). With modern caps (low leakage) you can go hundreds of times slower than usual trem rates.

The tube has no lower limit.

HOWEVER- tube or transistor, the usual R-C oscillators have *start-up* times related to frequency. Also excess gain, but too-much excess gain causes a non-Sine glitchy waveform. A kick-start can reduce but not eliminate start-up delay. (And we often prefer to kill the oscillator when not in use because it is hard to avoid stray thump.)

Function-generators (also digital schemes) can launch a sine from a cold start "perfectly".

[eleventeen]

PRR is correct, a phase-shift oscillator (with the three caps in series) is a very unreliable starter, and possibly the slower the freq you have dialed up, the slower it will be. It will also be slow with a weakish 12AX7. The caps knock out a load of gain, a phase shift oscillator designed for a 12AX7 may never start with a 12AU7 and might be iffy with a 12AT7. (I have done this experiment)

[AZJimC]

Well, I've been underway on this amp, gotten the circuit as is sorted out, and I think I may keep the 6AV6 oscillator that's in it, and I have an extra hole and plenty of heater power to add on if I like. I have learned that the output tubes Ive got, one of them shorted now, have been getting 500v to plates, and 300+ on screens, as well as about 7 volts on heaters, so I guess the tranny is plenty big.

I had a 38w 150R resistor that Ive installed as a sag resistor, and that has it tamed a bit, now about 480 on one tube, I'm awaiting some High voltage handling el84's (maybe EL84M or ???) to get final voltage measurements.
Took out three jacks and built a Hi/LO pair, and placed a pot in the extra hole for gain, replacing the current vol pot in the circuit, and then wiring the existing vol as a MV.???? maybe a cathode resistor switch?

Also took out existing front footswitch jacks, and put a switch in trem marked hole to shut down trem totally.
In the old footswitch hole marked ECHO I placed a pot to soon be reverb dwell. The current circuit uses V1b to send reverb, and a 6AV6 for recovery. I feel I need to get the reverb send boosted a bit before the dwell will be of much use.

so far, it's very gainy with a lot of headroom, and the V1 I have seems pretty hot. I used an older one, and it lost a lot of edge. It bangs the heck out of the 15" in this combo. I get happier with every capacitor change, as I move the whole box to 18w marshall specs, or as close as I can get with the add ons, and valve differences.

Thanks again, and I'm watching for good hints,
Jim

[AZJimC]

Okay, I've been looking over the various PPIMV schemes, and like some others Ive tried the cross line type, with the same crappy results. The 250K version seems highly praised, and transparent.

My lil smoker has the 130ohm cathode resistor shared with both cathodes, and i've added a 47mfd 100v cap since it had no bypass cap originally. Here's my confusion, The Lar-Mar is normally used on fixed, or grid biased amps, and most say it can also be used on cathode biased as well. I understand it replaces the 220k to bias current, so at full tilt it's as if not there. Now I'm looking at adapting that to my amp. I have no 220k resisters Post PI, however I do have the 470K pair to ground, in the path after the PI coupling caps. I have been using an 18w Marshall combo schematic, and the results so far are very cool.

Question is, should I replace these 470k's with the twin pots, or add it as if I had Fixed bias and allow the bias end to float, or tie it to ground? Seems to me that the end result either way is a 250K-250k cross line MV, since the two ends of the pot are interconnected? oh well, I'm easy to teach, in that I know how goofy I can be.

Thanks,
Jim

[sluckey]

should I replace these 470k's with the twin pots

Yes, but I would use dual 500K pots.

[AZJimC]

should I replace these 470k's with the twin pots

Yes, but I would use dual 500K pots.

I knew you'd say that, :-) and I just ordered the 250k.... hmmmmm darnit!

[sluckey]

250K will also work, but the 500K will put the same load on the PI as the 470K. I doubt I could hear any difference but others might. If I had a dual 250K on hand but would have to buy a dual 500K, I'd just use the 250K.

For your cathode biased amp you will simply connect the mv pots just like a typical volume control, ie, coupling caps to the top lugs, bottom lugs to ground, wiper to grid stoppers.