6v6 plexi-ish build questions

[Colossal]

I don't see the mixing resistors and cap where V1A and V1B output join. Usually each has a 400K resistor with the bight channel's resistor bypassed with a 470pF cap. Don't know if that would introduce hum though or reduce the blanket over the speaker sound.

I saw that but not sure how close Don was aiming for a true Plexi preamp.

[donzoid]

Wow, that's a lot of great stuff to digest.

Background is - I have played JMP's for 30+ years and owned others (like the JCM800) and the two JMP 100 watters I still have are pretty stock plexi design...1959 super leads that had no master when they came out of the factory. Of course I'm old now and went deaf long ago so I've had MV's put in them, used brakes but the tone stacks are stock. My favorite one is my 50watt (also with a MV installed) so all your suggestions are good...As Wyan said, I'll try lots of them...I just wanted to see how close this could come, but using 6v6 instead of say KT66. Hell even my Express clone w/ 6v6 is too loud in clubs so I have to throttle IT down in most rooms.

Great stuff though and I am going to make some of these changes to try to make it like stock plexi meets JTM45 and a bit of bluesbreaker thrown in too. (with a rectifier and cathode biased that's kinda what I was imagining.) It doesn't have to sound like ANY of those, but it needs to be good like all of/any of those are! hindsight being 20/20, I should have asked TAG before landing on a layout/schemo! Or many of these would be in there already.

[donzoid]

1st and easiest I'm going to do after re-reading these, is the screen resistor changes, because as I brought that value up, it became "strident" (brittle?) and stiff and when the screens were much closer to the plates, it broke up easier and was more full/thick.

I also thought the cathode bypass cap should be lowered so that's confirmation as I believe that will change my bass response curve. I'll do the screens first though.

[donzoid]

I've re-done the screen circuit to: 820R's on each screen, 200R in the line (where the 6k used to be) and decided I'd better start listening to Dave more, so I put a 2.5K in series with the choke. 350v at plate, 330 at screen. Its already starting to sound "swampier" in terms of fullness/body and less stiff and strident (as I put it before). Totally subjective terms, I know. On the TMB channel its still got more hum than I want so next up is elevating the heaters.

[Colossal]

I've re-done the screen circuit to: 820R's on each screen, 200R in the line (where the 6k used to be) and decided I'd better start listening to Dave more, so I put a 2.5K in series with the choke. 250v at plate, 230 at screen. Its already starting to sound "swampier" in terms of fullness/body and less stiff and strident (as I put it before). Totally subjective terms, I know. On the TMB channel its still got more hum than I want so next up is elevating the heaters.

Don,

Maybe a typo above? Did you mean 350V plates and 330V screens? Glad to hear you are making progress; well done.

[Jana]

I have been skimming--maybe this was addressed.

Some possible insights into screen voltage being higher than plate voltage.

This happens with some amps that use smaller bottle tubes (6V6, etc.).

I have seen it in some fender amps.

Why?

Ohms law will give a simple explanation for it. The small bottle amps typically have an OT transformer that is uses smaller gauge wire for the primary windings. As a result, there will be a higher DCR in the primary. If you measure the DCR of a 50 watt Marshall OT, for example, it will be ~ 40 ohms or so. If you measure the DCR of the typical 6V6 OT with an 8K primary, the DCR can be somewhere in the range of 100 to 300 ohms (making some guesses here based on old memories--don't quote me).

Then, take into account that the screens on a small bottle usually draw much less current that a big bottle, there is very little voltage drop across the choke.

Because of the DCR of the OT primary, the voltage drop across the primary of the OT is sometimes more than the voltage drop across the choke! So, what you wind up with is a screen voltage that is higher than the plates. It is the nature of the beast. What you can do, which is sounds like you already have, is to increase the resistance between the plate node and the screen node (choke plus resistor, etc.).

I would not try to adjust the screen voltage with the individual screen resistors. If you do that, then you are messing with the compression/feel of the amp as well as changing voltages. Chose the screen resistor value based on the "feel" you want (larger value, more compression--I would stick with 1K). Then adjust the choke/resistor value to get the voltage you want.

This has been the monthly drive-by post by Jana.

[xtian]

That's awesome, Jana. Not 10 minutes ago I was working on a 1950s Oahu style suitcase amp with 6V6s in push-pull. Screens are at the SAME node as the OT CT, and are about 10 volts higher than the plates. Was seeing arcing in one power tube when applying signal.

[donzoid]

Thanks Jana! I don't "usually" like to get drive-by's but this one is an exception.

In this case maybe the compression effect was unwanted (though that seems unlikely) I think I just went too far with it and stuff was getting wacked out.

Putting in the choke resistor as Dave suggested early on is much better and the "kerrang" effect is better (which I kinda like on Marshall). So with the 200R feeding the two 820's I am getting about the same as a 1k on each...seems better and yes Dave, sorry that WAS a typo its 350 plate 330 screen not 2xx. I'll go edit and correct it now.

[donzoid]

Scrounging around here for the parts to build the heater elevation...in the meantime, I have (yet another) question. Would dropping the value of the 8.2k grid resistors get me more power output? Seems like it would, by allowing the grid to see higher voltage swing. Or, am I better off adjusting the cathode resistor/cap values (or maybe both). Cathode biased amps are new for me, sorry for the stupidity. I am seeing 1.5K on other variations, quite a bit lower.

Mind you I "must" be deaf (hey I told you already I still have 2 100watt JMP's even though I "never" use 'em) 'cause the bosslady says this amp is louder than I think. But clearly there is more headroom to be had - at least I can get this to the level of a deluxe? Power and OT are pretty close to one of those in size and spec (I think).

[Colossal]

Scrounging around here for the parts to build the heater elevation...in the meantime, I have (yet another) question. Would dropping the value of the 8.2k grid resistors get me more power output? Seems like it would, by allowing the grid to see higher voltage swing. Or, am I better off adjusting the cathode resistor/cap values (or maybe both). Cathode biased amps are new for me, sorry for the stupidity. I am seeing 1.5K on other variations, quite a bit lower.

Don, the 8k2 are the grid resistors and that value is typical of the pure 18W EL84 designs. 1k5 is typical of the 6V6 based Fender designs, although the old 100W Plexis used 1k5 and only on two of the four power tubes! Mark Huss used 5k6 which is often seen on the 50W Plexis. The grid resistor is used to prevent oscillation and should definitely be included. The output coupling cap, grid leak resistor, and input grid resistor form a filter and their respective values decide how fast the power amp recovers from a change in the idle condition, especially during overdrive.

If you are looking for more output and a fuller/fatter sound, as I mentioned above in my things to try post, consider changing the output coupling caps from 0.01uF to 0.022uF, lowering the grid leak resistors from 470k to 220k and change the PI plate, cathode, and tail values to what I listed. The values in your amp were chosen to drive EL84 tubes which are sensitive and easily driven. 6V6s take a bit more of a push and the values I listed will drive the output tubes harder. But keep in mind, it's a package deal: change it all. Everything affects everything else.

Mind you I "must" be deaf (hey I told you already I still have 2 100watt JMP's even though I "never" use 'em) 'cause the bosslady says this amp is louder than I think. But clearly there is more headroom to be had - at least I can get this to the level of a deluxe? Power and OT are pretty close to one of those in size and spec (I think).

This amp should be at least as loud and probably louder than a Deluxe with the right setup. My 6V6 Plexi was built very close to Mark Huss' amp, high plate voltage and all and it was terrifically loud amp, albeit fixed bias.

If you want more power...

PI Inverting Plate (left side): change from 100k to 82k
PI Non-inverting Plate (right side): leave at 100k
PI cathode: change from 820R to 470R
PI tail: change from 47k to 15k
PI input 2: increase bypass cap from 0.02 to 0.1uF
Output coupling caps: change from. 0.01 to 0.022uF
Grid leak resistors: change from 470k down to 220k
Power tube input resistors: change from 8k2 down to 1k5-5k6, tune by ear

[M Fowler]

I've had good results with this layout, you could adjust for 6v6 tubes.

[donzoid]

Thanks Dave for 'splainin it. I have all you stuff on a text file for 1 at a time implementation.

EDIT: Based on what you said in your last post, I'll just "do it all"...

Mark, I looked at that as one of my final candidates.

What do you think of this one? I was a hairs breadth away from it vs. the one I did use. Only using cathode instead of fixed bias. I pulled it off PPwatt.

[donzoid]

Finally getting back to this after pulling a 30 hr night/day to do that gig, go home, unload gear, re-load a van full for the collectors market, go set that up, etc. etc. etc. Man I'm too old for this shit.

OK I just want to confirm I have the "right" components identified for the changes now that I have rounded up most/all of the new ones to put in Dave's changes. Maybe after that, I'll put in Statorvane's too.

If it's a change not mentioned on this graphic, then I "actually did" figure out where it goes

[Colossal]

Working from left to right on your diagram...

Q: Output couplers to 0.022uF
A: Yes

Q:Does it matter which one gets changed to 82k?
A: Yes, absolutely! It is the one on the right. Note: if you don't have a low value such as 10-15k for the PI Tail resistor (you say you have 18k), then consider NOT changing that 100k plate load resistor. The 18W PI is set up to have a balanced output, the values are not arbitrary. If you look at a real Plexi, you will see the 82k/100k plate loads but a 10k tail. Without getting technical, where the tail resistor value is low, say less than 15k, the balance begins to worsen, so the inverting side plate load is lowered to compensate. When the tail resistor is higher, >15-22k, the plate loads are often the same size. So, long story longer, like I said previously...it's a package deal, don't just change one thing incrementally looking for changes, go all or none.

Q: Output grid leaks, correct?
A: Yes. Lower those to 220-270k. I like 220k. Do this change along with increasing the 0.01s to 0.022uF.

Q: PI null input cap, correct?
A: Yes. Increase that to 0.047-0.1uF. I really meant to say 'slave input'.

Q: I have an 18k. Close enough or find a 15k.
A: Really best to find a 15k.

Q: Change to 470k.
A: I know you meant 470ohm, but yes. Lower that to 470R to go more with the traditional Plexi values. After changes, you should have:
470R bias
15k tail
470k-1M grid leaks
0.1uF slave input cap
82k inverting side plate resistor
100k non-inverting side plate resistor

Q: 0.68uF straddles this 820R?
A: Yes! You might even want to put that on a switch. Place a 100k resistor between the cap and ground. Use a SPST switch to short the resistor. This stops switch popping. Great mid boost.

[donzoid]

Cool! And again I get hobbled by one friggen' resistor. Went to Norvac today. 15K resistors, out of 1/2 watt, out of 1 watt, out of 2 watt, out of even THREE watt...LOL. I "will find one" in the reclaim bin. And the heaters are elevated, but I am going to just do all this at once before I see how many DC volts I am floating but it should be fine, 180K from screen supply, heater CT/47K/22uF-100v cap, those last two to ground just as diagrammed. I'll measure it once I power up again. Gotta lotta mods to do here, stay tuned should have them done by tomorrow. Thanks again Dave and yes, I "meant" 470R not 470K...dang it.

EDIT: Just found a 15K in the reclaim bin, measures 16K but hey, it's the only one there, so that's what she be.