3stage + CF options

[Smokebreak]

Here are my first scribbly drafts. I've got dividers and pots everywhere, so I can see what works
My transformers got here today and I'm excited. I got a pair from RJ, Edcor 600CT and 8K 15W, and a Hammond 1760H(Deluxe 6.6K) to compare. The Hammond looks absolutely puny next to the Edcor. How can it possibly make a bigger sound?

[img2550]http://whiteghostshivers.com/images/smo ... heme2.jpeg[/img]

[img2550]http://whiteghostshivers.com/images/smo ... yout1.jpeg[/img]

[John_P_WI]

Ha, looks like the circuits I played with thru the 90's. Just a suggestion, you may consider changing the V1a plate load to 220k and leave the V1b plate load at 100k.

I think you will have problems with the 2nd stage as there is over 100 times gain with very little attenuation. It would probably be better to change the cathode to at least a 2k2. You'll easily throw 100 v peak to peak out of that stage with a 1 v input as shown and the 470k - 470k divider will cut this in half. IT is always better to build gain slowly across multiple stages. You could change the bottom leg 470k resistor to a trim pot IF needed or add a 220k r across the legs of the 1M pot.

You may need a big ass grid stop on the third stage, I'd try a 68k at least. You could also make the third stage cathode switchable to a higher value for "texture".

Just remember driving the grids positive sounds like ass, build gain slowly across the stages. Also, driving the PI too hard can cause a swirly phasey sound. You may need a "fizz" cap on the PI or smoothing caps.

Finally, IT looks like you are going for the California modded sound, are you going to incorporate feedback resonance?

Hehe, in all it should be a good start and sound fairly angry. The 6v6's will smooth a lot when pushed but should be fun.

[dave g]

Just a suggestion, you may consider changing the V1a plate load to 220k and leave the V1b plate load at 100k.

I would agree with this.

I think you will have problems with the 2nd stage as there is over 100 times gain with very little attenuation. It would probably be better to change the cathode to at least a 2k2. You'll easily throw 100 v peak to peak out of that stage with a 1 v input as shown and the 470k - 470k divider will cut this in half.

I agree about the swing being too much (I will elaborate more in the next paragraph) but I disagree with the cathode resistor. When choosing a cathode resistor value, you need to think "What do I want this stage to do?". There are really 3 choices. If you want maximum gain (as in voltage amplification, not distortion) and headroom (like on an input stage), or if you want a symmetrically clipping stage, center bias. If you want to clip the positive half of the waveform seen on the tube's grid, bias it hot. If you want to clip the negative half of the waveform seen of the tube's grid, bias it cold. Biasing two consecutive clipping stages cold is a bad idea because it leads to symmetrical clipping and cancellation of even order harmonics.

Just remember driving the grids positive sounds like ass, build gain slowly across the stages.

This is only half true. Ideally, you would actually draw a load line over the tube's curves and figure out how much swing you need on each stage's grid to get the amount of clipping you want. This is typically on the order of 4 to 6 volts if you've biased the stage asymmetrically. To clip the 2nd stage, you need all the gain you can possibly squeeze out of the first stage, but then you need to attenuate it massively in between the 2nd and 3rd stages. Driving the grids positive sounds like ass ONLY if you are getting blocking distortion. A small coupling cap (2n2) plus a big grid resistor (100k) like he's got on the second stage will alleviate this. Grid clamping distortion sounds GREAT when done properly.

The "trouble spot" in the schematic Smokebreak just posted is going to be the coupling network between the 2nd and 3rd stages. Just like you said, the swing on the plate of the 2nd stage will be way too much to drive the 3rd stage. Honestly, I would probably just put a 220k in series with the 22n coupling cap and then feed a 100k pot. You can feed the 3rd stage's grid off the wiper of the 100k pot.

Lastly, I'd ditch the CF master or whatever that is on sw2. Just feed the tonestack off of the 100k cathode resistor like a normal superlead. I highly recommend a PPIMV as well, as I think the combination of the preamp and PI distortion is key (see my last post in this thread).

[Smokebreak]

Ok thanks a bunch guys. I'll have to ask 1 question at a time here and do some processing.
So concerning stage2>3, it seems you both are saying the same thing, in that there needs to be more attenuation there, with even more care given when biasing stg2 hot. I see you both suggest lowering the stg2 gain pot. Is this to raise the high pass point to help prevent the blocking distortion? I think a 100k pot would do this at any setting, and a 220K across the 1M would do this variably as you go cw? Or are we talking more about loading effect (which will require an explanation on the basic concept, for me)?

I like the idea of using a100K pot, as that's my push pull dpdt pot, and I think I can find those with actual solder lugs, as opposed to the 1M I have which is pcb mounting. So if I go that route, we have off the plate : 220K>22n>100k pot>grid. Basically, I don't understand why we wouldn't need a divider at the wiper there.


The other arrangement would be 22n>1M w/220K>470k divider>3rd stage grid resistor. I guess I'm trying to see the differences here. It looks to me that the 220k series resistor is having the same effect as the divider network?
Ok that was 4 questions.

Edit: I wanted to try the "Jose" CF master with the clipping diodes, just because I've never tried it. I figure it can't hurt...it's on a switch and a dual gang pot, so minimally evasive on the front panel. I'll definitely put in PPIMV. I've got that with the normal master on another one and really dig the combo

Finally, IT looks like you are going for the California modded sound, are you going to incorporate feedback resonance?

Well, I've got the presonance control on the schematic. How do you guys like that vs presence and resonance?

[Reeltarded]

The CF master with diodes is the key to the glue that makes the metal molten.

Hawt!

[John_P_WI]

Smoke, regarding the heavy attenuation that I mentioned I use it to obtain more of a percussive sound, to keep things tighter. There is not as much compression as the tube is not railed out and slowly letting down giving the sustain.

As far as the Jose master, I'm confused by your mention of a dual gang pot, I have not read the whole thread sorry if I missed something.

I believe this link has a very similar schematic to yours, especially if you take out the gain and CF for the loop send.

https://tubeamparchive.com/viewtopic.ph ... highlight=


Here is another thread (same scheme) where Dailey talks about diode selection.

https://tubeamparchive.com/viewtopic.ph ... highlight=

This is fairly similar to what Friedman, Cameron etc were doing years ago - in fact Mark Cameron commented in the first thread linked.

OK got it on the presenance, have not incorporated it into any builds since Martin developed it.

[Smokebreak]

Thanks John. For the dual gang MV, my thought was to use it in an untraditional way, in an effort to save room on the front panel. With the CF and normal masters, and a switch, that's just too much real estate. So, I figured with the dual gang, I'll have one MV, and a switch to select between the gangs, which would each be wired to the different type masters. Then, I can just call the switch "crunch" or something.

Yea, that schematic is very similar, just a different arrangement of resistance. it's these things I'm most interested in with this build. I've added some extra eyelets to my board, which I plan on drilling today, so I can hear what these sort of placements do.

Thanks for the links! I think I've just been watching too much Metal Mania on Vh1Cl lately;)

[Reeltarded]

The masters work together in an unexpected way. A very good way. Some of the best sounds aren't driving the PI as usual, not nearly.

With the CF dialed back a bit from 100% and the simple master used to set volume range these amps have the most tremendous thump.

[Smokebreak]

The masters work together in an unexpected way. A very good way. Some of the best sounds aren't driving the PI as usual, not nearly.

With the CF dialed back a bit from 100% and the simple master used to set volume range these amps have the most tremendous thump.

Oh, you're suggesting using both w/o a switch. I hadn't thought about that. I guess Id need the switch for the diodes though

[Smokebreak]

and then how would you feed the stack? I'm not seeing how to wire them both up, less making all 4 connections on the switch, simultaneously

[Reeltarded]

You could go commando and switch all masters out.

[Smokebreak]

You could go commando and switch all masters out.

My neighbors would love me even more!

Ok I think I see how to implement both MV at the same time. I was overcomplicating it. Cathode>CF master>TS>treble>normal master>PI

[img:768:1024]http://i1053.photobucket.com/albums/s47 ... 8515ef.jpg[/img]

[Smokebreak]

My choke came in today, Hammond 155H - 5H, 50mA, 400VDC max. I'm using the 6V6 Plexi PS, and it has a 5K5W between rectifier/res cap and the choke, but I'm still concerned about the choke's VDC rating. I missed that when I ordered it. PT is Edcor 600VCT. I'm unfamiliar with how chokes are spec'd, so I'm wondering if this 400VDC rating is something I need to be concerned with. I also figure that the 50mA rating is fine, since it's not seeing plate current?

[Reeltarded]

That is a good question. I don't know the answer. Maybe ask in the tech section for visability.

[Smokebreak]

Drilled my chassis and made a board over the weekend.
What is the purpose of a 470n cap in parallel to the reservoir cap? I assume it's frequency dependent filtering, but don't recall seeing this before

[img2448]http://whiteghostshivers.com/images/smo ... 260002.jpg[/img][/url]