Traynor amp hybrid.

[Nixon101]

I think it looks pretty good. That is exactly where I was imagining it might go when an LND150 SF was being discussed. There is very little extra wire length to the power tube grids, and the signal level is high, so less chance of noise from the power supply. If there is a problem, maybe some fast rectifiers or bypass capacitors would help. As far as heat dissipation from the tube, it's only about 2W, so it's not likely to be a problem.

You have the trem to play with yet too, no?

The amp has a real bias varying tremolo. Which as I understand, is quite uncommon. I think fender used it on a couple amps back in the day, maybe brownface, but everything blackface and newer does not have it. (I know you most likely know that already and could probably school me on amp history) Anyways... i knew it would be an obstacle, but after thinking about it logically, I figured it should work. (Dare I say this might be the first amp with the mojo of the SSS style driver tube, with bias varying trem?) I did test it out, taking the grid bias voltage off the trem pot, and the driver tubes grid (and cathode, thus power tube grid) voltage wiggled up and down like it should. Although this was before jumping in the whole circuit, when I was getting the weird bias readings, but now that is figured out, I'm expecting the trem to operate as per usual. Not sure if I'll get to testing it out today yet or not.

[martin manning]

I believe the bias vary trem system should work; I think I said that somewhere before (on pg. 2). I've never heard of anyone doing a power tube bias vary trem with CF driver... you may indeed be the first!

[Nixon101]

I believe the bias vary trem system should work; I think I said that somewhere before (on pg. 2). I've never heard of anyone doing a power tube bias vary trem with CF driver... you may indeed be the first!

I believe you did say that. I don't use, or even like tremolo really. But it just sounds so good on this amp. I didn't want to sacrifice it.

[Nixon101]

Well I don't know what's going on.... when I had the circuit jumpered in, my voltages were correct. Now that I have permanently swapped the wires over, they are back to being wonky. I threw the mains switch, standby still off just to read voltages and I was getting really weird readings. Bias voltage was slowly dropping the whole time, one half of the driver tube was lower than the other, and a positive grid bias of about 2 volts. That was with the same tube I used before. I swapped tubes and the negative voltage went back up (more negative), stayed steady, but still positive grid bias on the driver tube. I thought no way I fried that tube. Put the first one back in and now it's acting like the second one, steady voltage, but still not biased correctly. I honestly don't know what I did wrong. The only thing I can think of that I did differently was when I mocked it up with jumper wires, I took bias voltage before the trem pot, now it's through the pot and from the wire that originally supplied bias voltage to the power tube grids. I'm still getting a proper reading of -42ish volts before the 1 Meg resistors, but at the grids and cathodes are not correct. When I said the voltage was slowly dropping when I tried the first tube, it dropped to about -30 before I shut ot off and swapped the tube. I'll post before and after pics if anyone can see something obviously wrong that I did.

Edit.... after thinking about it, since when I tested it I just jumped the bias voltage in and jumped the coupling caps over to the driver tubes grids, so I actually had the bias voltage coming in with my new 1 Meg grid resistors AND the original 100k's in parallel. And that's when it was operating correctly. Now the 100k' are out and bias is only through the 1 meg's. Are these bias resistors too large? Should I swap them out for 100k? I realize that my bias arrangement is different than you guys are normally doing, I used your bipolar supply for the plate and cathode voltages, but kept my original bias supply to inject onto the driver grids. Where my driver tube grids, and cathodes, are "decoupled" for lack of a better term. How do I know what the proper values of the bias/grid leak resistors are supposed to be?

Edit...again.... I think I answered my own question. I just jumped 100k's in parallel with the 1 meg's and all the voltages seem to he correct again, driver tube grids with about a negative 1.5v bias with respect to cathodes. I'd like to know what's the reason, the science, and the logic behind choosing values for these resistors? I somewhat understand why we choose bias resistors when they are on the power tubes grids, but as a driver tube operating as a dc coupled cathode follower, I have no idea.

[Nixon101]

Just for experimentation, I made a voltage chart and paralleled different value resistors across the 1 Meg's I have on the board. I still for some reason don't think I am supposed to have too low of a value as wouldn't I lose a lot of signal to ground? Anyway with the weird voltages I did take it off standby and can hear a bit of hiss through the speaker, so I know it's working and passing signal. I didn't plug a guitar in as it's getting late and I wish I could just figure it out and know it's working properly before doing so. If it's gotta be 100k then so be it, but is that normal?

I also tried a cv4024 as the driver, and with the 1 Meg, the grid and cathode voltages biased to be about equal. With another 1 Meg in parallel for a total of 500k, I got proper bias voltage. I'd assume the cv4024 with 500k grid leak resistance would be better than the 12ax7 with 100k?

[martin manning]

I don't think using 470k instead of 1M is a problem, in fact other versions of this circuit have used that value, or use the 100k the amp has now. Funny I was going to suggest that you parallel the 1M grid leaks with something like 100k back when you were having this issue with the CF powered up on its own, but then you reported success connecting it up so I didn't. My thought there was to lower the impedance to AC ground at the grid, and I'm thinking that is still the issue. The grid leak is there for the same reason as always (to provide a reference DC voltage for the grid), but you don't want it to be too small because it will cause excessive loading of the previous stage, and that changes its AC load line and distortion characteristics. I don't see why you can't use the existing bias supply to bias the drivers, leave the bias trem setup exactly as it is, taking the bias for the drivers off the wiper of the trem intensity pot.

[Nixon101]

I don't think using 470k instead of 1M is a problem, in fact other versions of this circuit have used that value, or use the 100k the amp has now. Funny I was going to suggest that you parallel the 1M grid leaks with something like 100k back when you were having this issue with the CF powered up on its own, but then you reported success connecting it up so I didn't. My thought there was to lower the impedance to AC ground at the grid, and I'm thinking that is still the issue. The grid leak is there for the same reason as always (to provide a reference DC voltage for the grid), but you don't want it to be too small because it will cause excessive loading of the previous stage, and that changes its AC load line and distortion characteristics. I don't see why you can't use the existing bias supply to bias the drivers, leave the bias trem setup exactly as it is, taking the bias for the drivers off the wiper of the trem intensity pot.

Your right I have seen others that use 470k, but I'm sure most schematics out there I remember seeing 1 Meg, I think I even seen one that used 2 Meg. So that was my reasoning thinking something was still "wrong". I think I needed someone to just tell me to stop being a baby and play it and see what happens, which is what I'll do now. I'll experiment with different values and see what difference it makes lowering those resistors, and how the driver performs at different bias levels. I did a calculation of a high pass filter to see how much of a difference 100k would make and it shouldn't be too noticeable. I've learned a lot over the past 4 or 5 years tinkering with amps, but it's still just a drop in the bucket compared to what you and others around here know. This was a little out of my element, but knew it was something I wanted to try and tackle, and understand. And I'm getting there slowly.

[martin manning]

This is a fun experiment, and I'm looking forward to hearing about your results, sound-wise. You are right, the RC corner point will have some effect on the frequency response, but the AC load line will also be rotated clockwise as you lower the load from 1M to 0.5M to 0.1M. That means the positive swings on the PI outputs will clip sooner, so the positive and negative swings in the summed output will be clipped sooner.

[Nixon101]

This is a fun experiment, and I'm looking forward to hearing about your results, sound-wise. You are right, the RC corner point will have some effect on the frequency response, but the AC load line will also be rotated clockwise as you lower the load from 1M to 0.5M to 0.1M. That means the positive swings on the PI outputs will clip sooner, so the positive and negative swings in the summed output will be clipped sooner.

So before I experiment (ya I know, hurry up already) I was reading about cathode followers and input impedance. What I keep reading is that to make the grid leak resistor "appear" larger than it is, is to bootstrap it to the cathode. Is this something I should be considering if I don't want to load the stage before too much? In pretty much everyone else's driver design, the grid bias resistors are connected to the cathode supply resistors via a 820k resistor (or whatever value was chosen), does this make it appear bootstrapped? Where my design the grid bias voltage comes from a different source than the negative cathode voltage, thus not being bootstrapped? For some reason, I keep thinking this is one of the reasons my circuit is not operating quite like everyone else's, but I also lack the knowledge and understanding of this circuit, to know whether thats right or wrong. I tend to overthink things, but that comes from a desire to fully understand what I'm trying to learn. This thread has veered way off course from the original topic (my fault), but has turned into a diary of my thought process trying to understand what is foreign to me. And for all I know, this bootstrapping concept may not be applicable at all to what I'm doing (overthinking), but I appreciate the help and patience.

[martin manning]

What I keep reading is that to make the grid leak resistor "appear" larger than it is, is to bootstrap it to the cathode. Is this something I should be considering if I don't want to load the stage before too much? In pretty much everyone else's driver design, the grid bias resistors are connected to the cathode supply resistors via a 820k resistor (or whatever value was chosen), does this make it appear bootstrapped?

No, the 820k is just part of the divider for the HV-derived bias. There is no bootstrapping effect there. Your PI was working into 100k grid leak resistors before, so that shouldn't be a problem.

[Nixon101]

This might sound stupid, but I was so hung up on these higher values I see in the sss schematics, that I didn't even think that 100k was originally in there. And that it should be fine. Like I said, overthinking.....

Anyway, I did try out a bunch of different combinations. It was definitely louder with 1 Meg or 500k, rather than 100k. Not much but noticeable. What I did find though that no matter what bias I had on the grids, whether it be -2v, 0v or +2v, i didn't hear a difference at all. Although I'm not really sure what I'm supposed to be listening for. I settled for the cv4024 with 560k.

My first impressions of adding the driver stage....
(I had the chassis out the whole time so I didn't have the reverb hooked up, and I hate a clean signal without even a hint of reverb, so I hated every second of this.....half joking...)

I noticed maybe a bit more top end sparkle, a bit more clarity. Sustain and compression, all the things everyone says it adds, just barely noticeable. The amp was pretty clean and punchy to begin with, but to be honest, I was a little underwelmed. But I think alot of it has to do with missing that sweet, sweet reverb. I was also under the impression i might get a little more out of the volume dial before it starts to break up, which didn't happen. And the bass still got flubby higher up on the volume dial. After I was done experimenting, I soldered the resistors on the board and just buttoned it up and put it away, so tomorrow will test with reverb. I wouldn't say it's a night and day difference, but it's not useless either. Was it worth it? Well for me, it wasn't really about the end result. I learned a lot along the way, thanks to you mostly, and that's what I really wanted to accomplish.

[martin manning]

What I did find though that no matter what bias I had on the grids, whether it be -2v, 0v or +2v, i didn't hear a difference at all.

I've become convinced this bias anomoly is just a meter loading problem. The best measurement of the grid voltage will be from the bottom of the grid leak to ground, which is just the bias voltage, of course. A good direct measurement of the bias would be from the bottom of the grid leak to the cathode, but measuring from the top of the grid leak to the cathode as you were doing before should be pretty close.

My first impressions of adding the driver stage....
I noticed maybe a bit more top end sparkle, a bit more clarity. Sustain and compression, all the things everyone says it adds, just barely noticeable. The amp was pretty clean and punchy to begin with, but to be honest, I was a little underwelmed. But I think alot of it has to do with missing that sweet, sweet reverb.

Well, now you know. Another listen with the reverb should be interesting. How about the trem? Does it work pretty much as before?

I was also under the impression i might get a little more out of the volume dial before it starts to break up, which didn't happen. And the bass still got flubby higher up on the volume dial.

Those things could be happening in the earlier stages then. Maybe you can make some adjustments to coupling caps for the bass, and, if you can get access to a scope, find out where the clipping starts as the volume is increased.

[Nixon101]

What I did find though that no matter what bias I had on the grids, whether it be -2v, 0v or +2v, i didn't hear a difference at all.

I've become convinced this bias anomoly is just a meter loading problem. The best measurement of the grid voltage will be from the bottom of the grid leak to ground, which is just the bias voltage, of course. A good direct measurement of the bias would be from the bottom of the grid leak to the cathode, but measuring from the top of the grid leak to the cathode as you were doing before should be pretty close.

My first impressions of adding the driver stage....
I noticed maybe a bit more top end sparkle, a bit more clarity. Sustain and compression, all the things everyone says it adds, just barely noticeable. The amp was pretty clean and punchy to begin with, but to be honest, I was a little underwelmed. But I think alot of it has to do with missing that sweet, sweet reverb.

Well, now you know. Another listen with the reverb should be interesting. How about the trem? Does it work pretty much as before?

I was also under the impression i might get a little more out of the volume dial before it starts to break up, which didn't happen. And the bass still got flubby higher up on the volume dial.

Those things could be happening in the earlier stages then. Maybe you can make some adjustments to coupling caps for the bass, and, if you can get access to a scope, find out where the clipping starts as the volume is increased.

I've read other threads where people were saying the same thing about reading the voltages on the grids of the phase inverter. I'll have to agree, that is what's going on. The trem works flawlessly. Absolutely no issues with that. Also yesterday's test I used a cab that I dont normally use with that amp, so that could be throwing off my first impressions too.

The amp has been pretty much gone through already tweaking the voice to my taste, which wasn't really far off the original schematic. The biggest thing I did to achieve my tone (think srv, punchy, clean with a bit of hair) was raise the preamp voltages. It will still break up when dimed, but it's a pretty clean, punchy, smooth sounding amp already. It's my favorite amp really. Some people aren't really fans of the weird baxendal/James tonestack. But I like it. The one thing I still really want to try is finding a nice old set of real tetrode sylvania 6ca7 (in my price range lol) as apose to the Philips branded 6ca7, which are actually mullard xf2's. When I bought the amp about 2 years ago.it still had all the original tubes it shipped with back in 1970.

When you mentioned oscilloscopes before, I looked around and the one I was thinking about buying last year locally to me, is still for sale. A tektronics 465b. $200 cad with no probes. Maybe the guy is flexible on price lol.

[martin manning]

The trem works flawlessly. Absolutely no issues with that.

Great! So this arrangement could be applied to any SSS amp. Bias varying trem sounds sweet.

The one thing I still really want to try is finding a nice old set of real tetrode sylvania 6ca7 (in my price range lol).

EH 6CA7's would be something you could try for less money. They get good reviews, and are currently running around $60 US for a matched pair.

When you mentioned oscilloscopes before, I looked around and the one I was thinking about buying last year locally to me, is still for sale. A tektronics 465b. $200 cad with no probes. Maybe the guy is flexible on price lol.

That's a great analog scope, and the price seems pretty reasonable if it's in good shape and everything works. Tek's of that vintage are reparable too. You can get some cheap probes that will work for $20-$30, but Tektronix probes would be better, as the scope will know if you connect an x10. There are some nice digital scopes out now in the ~$300-$400 range, and some under $200 that may or may not be worth your time. Any scope is useful, and a great learning tool.