"New" (old) amp design and tremolo

[Tyrannocaster]

Okay, I am waiting for stuff from Mouser and I'd like to hear what I did wrong with this design.   I'm building a 5D3 (that's a Deluxe) Tweed preamp with a 5E5 Pro power amp but I would also like to add a trem to it. I'm doing this in such a way that I can do the preamp and amp first and then add the tremolo later; there's a separate board for it and of course it has its own tube which won't do anything until I'm ready.

I've looked at all the tremolo circuits I can find but I was really looking for something that would work on a preamp rather than the bias section of the power amp; most of those are for fixed bias amps, anyway. The Tremolux has a cathode biased setup very similar to my old 5D3 Deluxe but it isn't quite the same and I settled on the Vibro Champ circuit as possibly being the most portable. What do you guys think? The only thing I am really unsure of is that the VC has a 47 ohm resistor between the cathode bypass and ground - I left that out but wanted to ask about it. But the big question is "Will the tremolo even work in this part of the amp?"

The voltages in the schematic are stock values - I will be using a SS rectifier (not indicated on the schematic yet) and I'm not exactly sure what they are likely to be but I am guessing they will be in the ballpark. The PT is from an old Hammond tone cab that used 6V6s; although I have 6L6s in mine, past experience has shown that they don't draw that much more in this kind of configuration than the 6V6s and this PT has more than enough heater current available. I am guessing the final voltages will be a bit less than the 5E5 readings on the schematic.

The values in red are taken directly from my '55 5E3 and they are there because I like the way they sound; they are quite different from the stock values. Again, I am using the 5D3 preamp because mine sounds so good I have just copied it. It isn't the same as the 5E5, having only one tube, although it is similar.

So, let the corrections begin! I don't know squat about tremolos and am waiting to learn more.

[tubenit]

Here is an alternative idea.   I am also attaching the ExpressSCH editable version so you can VERY easily change it to whatever you want.

With respect, Tubenit

[tubeswell]

The 5E9A trem will work in a 5D3 circuit - you might want to adjust the size of the (10M) feedback resistor.

But seeing as how you've gone to a driver/cathodyne splitter combination for the PI, and seeing as how you want to wiggle the pre-aamp, then you might want to think about a roach trem.

If you want to go with a output stage wiggle, I have built the circuit in the attachment (albeit that its SE) and it works.

[HotBluePlates]

The Tremolux has a cathode biased setup very similar to my old 5D3 Deluxe but it isn't quite the same and I settled on the Vibro Champ circuit as possibly being the most portable. What do you guys think? The only thing I am really unsure of is that the VC has a 47 ohm resistor between the cathode bypass and ground - I left that out but wanted to ask about it. But the big question is "Will the tremolo even work in this part of the amp?"

What you have will work just fine.

The 47Ω resistor is the other half of the feedback loop; the loop requires a series resistor and the 47Ω shunt resistor to define a voltage divider determining how much feedback there is. You have no loop, so you don't need the 47Ω resistor.

The 12AY7 piques my interest. Were you keeping the 2nd half, with the grounded grid, to maintain the current through the 820Ω resistor, and to keep the same 0.01µF bypass cap? You could switch to a 7-pin single triode tube type, and probably buy a lifetime stash for the price of a couple of NOS 12AY7's. I've just done the same thing with the Standel build I just finished. If you do switch tube types to run a single triode, double the cathode resistor and halve the bypass cap to maintain the same bass roll-off point.

Which reminds me: a 0.01µF cap against 820Ω will have a -3dB point high enough that you might as well not have a bypass cap. I calculated a number almost as high as 9kHz, but that does include a simplifying error that will guess the roll-off a bit low.

[Tyrannocaster]

Cool; thanks for taking the time to post that. I need to get that image on my other computer (bigger monitor) so I can look at it with my own side by side. You would use one that varies the output stage then, which is something I was trying to avoid, although perhaps I shouldn't. I just didn't know how well the trem could be adapted to the cathode biased push pull output stage; I have had mixed results in the past doing that. Sometimes it worked, other times nothing I ever did solved the problem. Since I take that a more of a comment on my debugging skills you can see why I might try a different approach. My fixed bias tremolos have been less troublesome to get running. Still, having said that, it's not like I've built hundreds (or even dozens) of these.

Just my preference, but I wouldn't parallel those two preamp halves. The only reason that channel is even there is because you have to have it to get the weird and wonderful Tweed interaction, but AFAIC, there's plenty of gain in the first stage already.

But I didn't expect such a detailed response; thank you very much!

[Tyrannocaster]

What you have will work just fine.

The 47Ω resistor is the other half of the feedback loop; the loop requires a series resistor and the 47Ω shunt resistor to define a voltage divider determining how much feedback there is. You have no loop, so you don't need the 47Ω resistor.

The 12AY7 piques my interest. Were you keeping the 2nd half, with the grounded grid, to maintain the current through the 820Ω resistor, and to keep the same 0.01µF bypass cap? You could switch to a 7-pin single triode tube type, and probably buy a lifetime stash for the price of a couple of NOS 12AY7's. I've just done the same thing with the Standel build I just finished. If you do switch tube types to run a single triode, double the cathode resistor and halve the bypass cap to maintain the same bass roll-off point.

Which reminds me: a 0.01µF cap against 820Ω will have a -3dB point high enough that you might as well not have a bypass cap. I calculated a number almost as high as 9kHz, but that does include a simplifying error that will guess the roll-off a bit low.

Wow, I'm sure my replies will seem confusing since three of you posted at the same time and I didn't review everybody's posts before answering. Anyway, all I can say about that cap is that the amp sounds so much better with it that it does with the stock value that I left it in. I guess I should see how it sounds with nothing there, though. No point in a component that doesn't actually serve a function.

As for the second half, I kept it because of the strange interaction I get on my old 5D3, which I have been told involves some feedback between the sides. I DO know that if you build a 5D3 with just one channel it does not sound at all the same, and you lose the whole oddball effect of channel two. I don't even think of that channel's knob as a volume control; to me it functions as a "Q", or resonance, control. It's almost like having a wah-wah and moving the pedal to change the sound although you don't do it in real time. To me, this is well worth half a tube!

Thanks so much for the help!

[Tyrannocaster]

The 5E9A trem will work in a 5D3 circuit - you might want to adjust the size of the (10M) feedback resistor.

But seeing as how you've gone to a driver/cathodyne splitter combination for the PI, and seeing as how you want to wiggle the pre-aamp, then you might want to think about a roach trem.

If you want to go with a output stage wiggle, I have built the circuit in the attachment (albeit that its SE) and it works.

I have used half-tube tremolo designs in the past and found them to be much more finicky as far as getting them working. But that may be just coincidence. One time I got one to work beautifully in an 18 watt Marshall clone so I put the exact same circuit in another clone only to totally fail to get it to work even after a lot of component subbing and help from other, more knowledgeable amp people. That was so frustrating I don't want to go through it again. (That one was from a Matchless Hurricane; it sounded great in the amp it worked in.) I like the idea of your amp and I'm storing the schematic because I had played with the idea of a parallel single ended amp recently and I usually come back to things I stumble across.

[Tyrannocaster]

Here is an alternative idea.   I am also attaching the ExpressSCH editable version so you can VERY easily change it to whatever you want.

With respect, Tubenit

That is quite cool. What do you see as the advantage of varying the power stage over the preamp? Your circuit is very similar to what I had in mind, it just works on a different part of the amp than I was thinking of.

I just downloaded JSchem but don't know how it works yet, although I can view the schematic in it easily. Thanks for putting this in!

[HotBluePlates]

As for the second half, I kept it because of the strange interaction I get on my old 5D3, which I have been told involves some feedback between the sides. I DO know that if you build a 5D3 with just one channel it does not sound at all the same, and you lose the whole oddball effect of channel two. I don't even think of that channel's knob as a volume control; to me it functions as a "Q", or resonance, control. It's almost like having a wah-wah and moving the pedal to change the sound although you don't do it in real time. To me, this is well worth half a tube!

Thanks so much for the help!

Good point. I overlooked that when posting.

Consider then if you more often use the channel without the tone control, or the channel with the tone control. As drawn, the non-tone channel is being driven. I agree that adjusting the other channel's volume impacts the tone of the channel being used.

[tubenit]

I redrew the schematic I posted to more closely reflect what you are wanting to do.

One change is using a 33k input resistor instead of a 68k.

I would encourage you to consider using the spdt switch to parallel V1 as an option. Very cheap mod that will allow you to NOT have it
paralleled (similar to your amp) or having it paralleled for 30% more gain with NO increase in noise plus it will then allow the tone stack to
be more directly involved in your signal path.

So for a simple $4 spdt, you can have you current tone PLUS a really nice option of a more overdriven tone with the tone stack more engaged.

With respect, Tubenit

[Tyrannocaster]

Consider then if you more often use the channel without the tone control, or the channel with the tone control. As drawn, the non-tone channel is being driven. I agree that adjusting the other channel's volume impacts the tone of the channel being used.

Oh - I hadn't realized that. I don't think I've used the Mike inputmore than a few times in the years I've owned this amp; for some reason I was thinking the Tone control worked on both channels; that's why I thought it was a good idea to post the schematic. Stuff like this gets past me...

Thanks for that; I would have been going "Hey, where's my tone control?"

[Tyrannocaster]

I would encourage you to consider using the spdt switch to parallel V1 as an option. Very cheap mod that will allow you to NOT have it
paralleled (similar to your amp) or having it paralleled for 30% more gain with NO increase in noise plus it will then allow the tone stack to
be more directly involved in your signal path.

So for a simple $4 spdt, you can have you current tone PLUS a really nice option of a more overdriven tone with the tone stack more engaged.

With respect, Tubenit

That's a really good idea that hadn't occurred to me. And there's even an empty spot for one of those little switches right below the Volume control (I'm using the chassis of a gutted Acoustic 109 amp for this) that I had been thinking about: "Now, what can I do with THAT?" Actually, there are two of those (they held LEDs in the Acoustic), the second being under what will be the "Q" control, or the second volume control.

I can see that I have to familiarize myself with JSchem. Thanks for doing that work for me.

EDIT: I have been messing with JSchem (and redrew it to reflect the correct channel with the tone control) and it's really neat! (Although I wish UNDO worked.) I noticed that you added in the 47 ohm resistor on the trem circuit's tube bypass; I thought that was unnecessary.

[Tyrannocaster]

Another thing I forgot to mention is that in this amp I've decided to use a set of Mullard EL37s in the power amp, at least at first. (They're just sitting there gathering dust - might as well use them.) It's my understanding that in most respects these act similarly to 6L6GCs (I have used them like that in the past with no problems) but are there and considerations to take into account because of the tube difference? While they may be similar in many ways, they are certainly NOT 6L6s physically.

EDIT: Also, guys - thanks a ton for all this help.

[tubenit]

I noticed that you added in the 47 ohm resistor on the trem circuit's tube bypass; I thought that was unnecessary.

You're correct. It is not necessary. Just a drawing error being in a hurry.

With respect, Tubenit

[HotBluePlates]

Another thing I forgot to mention is that in this amp I've decided to use a set of Mullard EL37s in the power amp, at least at first. ... It's my understanding that in most respects these act similarly to 6L6GCs (I have used them like that in the past with no problems) but are there and considerations to take into account because of the tube difference?

Two big differences: heater current and transconductance.

The heater current for the EL37 is 1.4A per tube, compared to 0.9A per tube for 6L6's. So you're heater winding will need to support an extra 1A of heater current.

The transconductance is also higher for the EL37 (cited as 11mA/V or 11,000 micromhos in data sheets) than in the 6L6 (generally given as 6,000 micromhos in data sheets). That higher Gm is more typical of true pentodes like the EL34. The British way of specifying Gm more readily tells you that with a given change of grid voltage (say, 1v of input on the control grid) the EL37 will have a bigger change of plate current (11mA for each volt for the EL37, compared to 6mA for each volt in the 6L6).

What this means for you is that the EL37 will require less drive signal to push it to full power output than the 6L6. You should hit maximum clean power and move into distortion lower on the volume knob. It also means that the EL37 will reach its proper idle current with a smaller bias voltage, and consequently a smaller cathode resistor.

While you could use a different/smaller cathode resistor for the EL37's (or figure some tricky way to switch the resistors depending on output tube used), the larger value used for 6L6's will simply cause the EL37 to idle at somewhat less current than an "optimum" resistor. Assuming your PT can support the extra heater current draw, I'd say stick to the 6L6 cathode resistor value. It will cause the EL37 to idle a little cooler and maybe take a bit more input signal to drive to full power, but is otherwise very safe.

I too have some Mullard EL37's just sitting around that I need to use. I've just had too many other things to do that have kept me from building an amp around them.

[Tyrannocaster]

Thanks, HotBlue; the power tranny originally had heater current for not only a couple of 6V6s (if that was all I'd be in trouble!), but also 5 12AX7s 3 other smaller tubes whose numbers I can't recall; and of course, I won't be using the 5U4GB rectifier. (It's my understanding that even though that one doesn't run off the 6.3 supply, not using it will free up some small amount of extra current for the rest of the amp.) I think that in view of your info on the requirements of the EL37s I will start with the 6L6s first just to be safe and then try the others later. In general, I have found organ power supplies to be really stocked with heater current because of all the special tubes they typically have. I wouldn't use this one if I were running the output stage in fixed bias mode, but I think it should be adequate (if close to the edge) for this purpose although I suspect the voltages may be somewhat low compared to the Fender values for the Pro 5E5. I will be using a SS recto, which will help bring the voltage up, but until I get something actually connected I don't really know what to expect other than a guess.

That's interesting about the EL37 breakup and I am saving the comment to my scrapbook. I'll keep the cathode resistor values I have now, at least to start with. BTW, I will be putting separate resistors and caps on each output tube. It's something I found in the Mullard book of circuits they published in the 50s and they say it allows mismatched tubes to work better together (cathode bias) so since the parts are cheap I figured "Why not"? I did it on my last amp and it seemed to work well (I tried the EL37s in it but ended up switching to EL34s in the end because I wanted to sell it but keep the EL37s). I always have some tubes around that aren't matched as well as somebody might like.

[Tone Junkie]

Sorry for being off topic but Hotblueplates which 7 pin tube is like 1/2 of a 12ay7 tube.
Thanks Bill

[HotBluePlates]

Well, the 6AV6 is like half a 12AX7, and the 6AT6 is like half a 5751. Those are the two I have on hand for my amp.

The 12AY7 data sheet shows a mu of 44, plate resistance of 25k and Gm of 1750 micromhos at a plate current of 3mA. To exactly match the tube's performance, you should look for a tube that nearly meets all of these parameters.

After a pretty thorough survey of Frank's, I found that there isn't an exact single-triode version of a 12AY7. There are several that exactly match 5751 characteristics.

Tubes that equate to single-triode 5751: 7-pin - 6AQ6, 6AT6, EBC90; 9-pin - 6CN7, 6FM8
Tubes that equate to single-triode 12AX7: 7-pin - 6AV6, EBC91, 6BK6, 6DR4

A tube that's in between 12AY7 and 5751 specs, maybe enough to be a good sub for 12AY7: 6BN8. It's a 9-pin tube with a mu of 70 (like 5751), plate resistance of 28k (more like 12AY7) and a Gm of 2500. Antique Electronic Supply has them for $8.60 each, which surprises me. Some of the other tubes I listed above are as low as $1.85 each.

Another possibly interesting tube is the 6AB4/EC92. Mu is 60, plate resistance is 15k, and Gm is 4000. With typical guitar amp values, you'll come a lot closer to having gain equal to mu than with 12AX7's.