Modded Tremolux

[johnnyreece]

Ah but your single triode Jet trem might’ve been the best answer.

Maybe. But the Tremolux paraphase inverter might not get along with the J12 trem. I don't know. Looks like the trem would also couple through that 6800Ω to the lower PI grid. I have no idea what might happen with that. Could be a new discovery. Could also be a heartbreaker!

Yeah, I think I'm smacking enough changes on the poor Tremolux that maybe I'll try this particular idea another day. I really do appreciate the discussion, though!

[johnnyreece]

So, I've played around with the tremolo section to try to get more range. I wish I could find an oscillator calculator or formula that *doesn't* require all three resistors and capacitors to be the same value so that I could experiment with this...Anyhow, does this seem like it might work? Might be another "slap it together and see how it sounds" type-thing.
Any input is appreciated!

Tremle Dee Tremolo Mod.jpg

[johnnyreece]

So, I had a little time this weekend and found some errors on my schematic and layout. Got those sorted out, and these are my "As-intended-to-be-built" schematic and layout. Feel free to share any thoughts you have on why it might or might not work, or if you think something could be better!

[LOUDthud]

A word of warning on this amp. With that switch changing the bias on the output stage, B+ is going to be moving around. Because of the way that the Tremolo "Depth" control is tied into the Phase Inverter, if B+ is too high, no signal will get through the Phase Inverter to the Power Tubes. Solution, lower that tail resistor in the Phase inverter.

[johnnyreece]

Thanks for the heads-up! Do you think something like 4k7 would be better?

[Colossal]

What is your goal for the bias switch?

[johnnyreece]

Just being able to switch between hot, mild, and cold bias. Still deciding what values. May go with 2x500 and a 1k. That would give me 250, 333, or 500. Just something to play with. It would also give options if using different output tubes (thinking about maybe using 6K6 at times, or if I beef up transformers, 6L6). I wouldn't be switching on the fly, or anything like that.

[LOUDthud]

Thanks for the heads-up! Do you think something like 4k7 would be better?

You are going to have to experiment to get it working with all combinations of output bias. A higher resistance Depth control might be needed. The Depth control pulls the cathodes positive and that starts to bias the PI colder reducing the gain. This is a unique circuit among Fender amps.

[johnnyreece]

So, I discovered a couple things. One, I had my layout backwards (d'oh!). Got that mostly updated and will repost. There were a couple minor changes made, but I think overall, the design is about the same. I finally got off my butt and figured out how to program my CNC to drill my turret board, so that's cool. I plan to eventually type up a tutorial (because I had a lot of trial and error to get it figured out). I went from DIYLC (for design) to Inkscape (to convert to gcode) to Candle (to send to the CNC). All free software. None of it was particularly hard to do, but getting the sequence right took some doing for my thick skull. I think it turned out okay, though.

[johnnyreece]

I know it's been a long time since I've posted on this, and there's not much to update, but here goes: My original PT turned out to be toast after preliminary testing. So, I bought a new one with multiple secondary taps from Weber (it was cheap!). With that in mind, I'm incorporating a high/low B+ standby switch. Unfortunately, the PT was not a stand-up type, but had wires coming out of each endbell, and they didn't' make a conversion kit for those, so I made my own feet and drilled the endbells. Looks okay in my book. I've also got the board made up. I've been working on my CNC software, and I think I'm ready to experiment with faceplates soon. The plan is to paint the back of a piece of plexi gold, carve the graphics, then fill with black; kinda like a Marshall. Oh, and most importantly (hah), I decided on a name: Tremolead.

One thing I've been wondering about...with the addition of this PT, I'll be able to use higher power tubes, including EL34, KT66, etc. I'm wondering if I should maybe add screen resistors? If so, is there a good compromise value I should use? I truly appreciate any input!

[Stevem]

Swapping around to different b+ voltages calls for more then just adding screen resistors which should be done anyway.
You need.
A higher wattage rating Ot.

You need switchable cathode resistors.

The use of a EL34 requires pin 1 to be tied to pin 8, and many times since pin 1 is not used in a 6v6 or 6L6 type tube that pin is used as a 1 pin terminal strip/ tie point.

[johnnyreece]

Thanks for chiming in. I'm using a JTM45 OT, so I'm covered there. I'm also using switchable cathode resistors (set up to change between 250, 333, and 500). I'll also make sure to tie 1 & 8 together if I end up going this route. The question still stands about the screen resistors, though. I'm thinking a minimum 1k? I don't really use EL34 much (and really never in Fender-type designs), so I'm hoping to get some input from those of you with more experience on the Brit side of things.

[johnnyreece]

I've got a schematic draft drawn up (please note, it's only had a cursory review - will check later for accuracy). I'd appreciate if anyone wanted to put a pair of eyes on it. The schematic doesn't make note of the voltages, but the PT has dual secondaries (330-0-330 and 360-0-360). As of now, the schematic doesn't include screen resistors, but that will quite possibly change at some point. Anyhow, thanks everyone who's still following along with this! I've got the board assembled, and am hoping to get moving a little faster soon, but my wife is opening a candy store, so that's taking a lot of my *fart around with electronics* time.

[johnnyreece]

Well, after a long time waiting, I finally got this together over the weekend. While everything *works*, it needs a few tweaks. Here goes:

-Bias resistors will need modification. The 250/333/500 setup isn't quite going to cut it. With a 5Y3 and 6K6GT tubes, they're running at 108% dissipation with a 500 ohm cathode resistor. With a 5U4GB and 6L6GC, the highest I could get to was 71.3%. So, I think I'm going to go with a setup to get me 176/375/600. I temporarily jumpered all of the resistors together to get 195 ohms, and that was good enough to get the EL34s I had in at the time to 90%, so 176 should be good for EL34/6L6. Hopefully, 600 will be high enough to tame the 6K6s. Guess we'll see!

-Tremolo section needs work. The speed switch works, but at its highest speed, it kills the oscillation. Might try tacking in another cap to see if that helps. Also, there's no action on the intensity until the pot's turned up over half, then it all hits at once. It may be that I have an audio taper pot in there instead of linear. Will check it out.

-On startup, voltages approach 500VDC in my power string, settling down after the tubes warm up. My caps are rated to 450V. Is this an issue?

I think that's pretty much it. Reasonably quiet, except when cascading the channels. There are three very distinct different sounds with this, which is nice. It's got the standard tweed-sounding input, a Marshall-like input, and then the higher-gain option with the flip of the switch. I'll try for a little further assessment after I get the bias sorted out. All things considered, I'm happy with the result so far. I'll try to snap some pics after it's finalized. Will also update the layout/schematic to the "as-built" version when I'm satisfied with the "as-built" amp.

Edit: I measured the intensity pot last night and, as suspected, it's an audio taper. I'll swap in a linear one soon. Speaking of pots, I'm also betting my speed pot is only 500k vs. 1M that the schematic specifies. I think that may well be why my speed gets so fast that it kills the oscillation.

[johnnyreece]

Parts arrived this weekend, so I made my (hopefully) final changes. Brief rundown:

-Bias resistors were modified. Long story short, I was able to get the EL34s to 93.2% dissipation, and the 6K6GTs down to 101.2%. I think that's close enough for me. Anecdotally, people run 6K6 in a 5E3 Deluxe with no modifications just fine, which is way harder on them than what I'm doing.

-Tremolo section - Still had some issues even after subbing in proper pots. I upped C15 to .02, which helped to an extent, but I still was losing oscillation on the high and low end. I replaced the tube with a stronger 6SL7, and got the low part to tighten up. I ended up goosing the tube a little by upping the plate resistor from 220K to 330K. Problem solved!

Here are my last questions, if anyone has any thoughts:

-My startup voltages still exceed my filter cap rating 450VDC. Is this an issue?

-My EL34 tubes are weird...when you look inside the holes of the plate structure, it appears that the grid wires glow blue. This isn't the normal blue glow like one might see in the glass. Fun fact, if the tremolo is on, the glow oscillates on and off with the beat of the trem, if that helps.

-My tremolo is satisfactory, however, there's not a huge difference between the speeds. I put my frequency counter on it, and on slow my range is 3.1-6.21Hz. Fast is 3.96-9.26Hz. Would it be worth it to increase R25 to slow down the slow range a bit more? Practically, how low can this frequency really go?

I'll try to get some pics taken of the finished product posted soon. With 6K6GTs in, I can crank the volume as much as I want, even through my 4x12 cabinet. With the big bottles, I've not extensively given them a run-through, but it is significantly louder. Through the same 4x12, it got uncomfortably loud before I could get fully turned up. Of course, this is in my little band room in my house. A more open venue would likely handle the higher power better.

Thanks for taking a look!