Schematic for Ridicule

[Tim Hall]

Ok, I'm throwing this out there for your ridicule. Actually I'm hoping someone experienced might give it a quick glance to see if I've made any grievous errors (or done anything that looks plain dumb). It's nothing original, of course, and most of you will immediately know where it comes from.

The last thing I have to figure out is getting the right B+'s. Can anyone explain how you go about doing this? I figure in principle it's just an Ohm's law type problem, but I'm not so sure how it works with so many things going on. I estimated the DC off this PT would be around 345V with 120VAC at the primary...hope I did my math right.

Thanks for any help.

Tim

[mykey]

Ok, I'm throwing this out there for your ridicule. Actually I'm hoping someone experienced might give it a quick glance to see if I've made any grievous errors (or done anything that looks plain dumb). It's nothing original, of course, and most of you will immediately know where it comes from.

There's nothing really wrong with it, but you might get some POPS when you close those switches. Don't be afraid to improve on the (mesa boogie) design.

[Enzo]

OK.
The 25uf cathode cap on the second triode is drawn over to the switch under the first triode, but it is grounded there. This would be pointless, so I have to imagine you meant that cap to connect ABOVE the switch, ie the switch cuts them both in and out for gain. On the other hand, wired together that way, then teh switch was open, then the two caps would form an unwanted - I assume - negative feedback loop. Better would be a switch for each - make the thing a DPST or use a relay.

I don't know why the coupling caps from the PI to the output grids are so mismatched. Is there a rationale for this?

I don't know what resistors you need in the B+, but it is indeed Ohm's Law. Figure on average about 1 or 2ma per triode. SO lets say about 3ma for the 300v node. That means 3ma has to drop 20v across the last B+ resistor. That yields a 6667 ohm resistance. 6.8k the closest standard value.

The second tube runs on the 320v node, dropped from the 345v B+. Assuming 3ma for it as well, and remember the current for the 300v node runs through this resistor also, we get 6ma dropping 25v. 4167 ohms, so 3.9k or 4.7k or a closer tolerance part of suitable value.

Those are of course just examples. Slap them in and see how close it comes, then tweak for best results. Is there some reason for those particular voltages? I'd be thinking that the voltage itself was not so important - 20-30 volts either way might not make a lot of difference. Decoupling the stages is the main deal here. and for that, the actual voltage doesn't matter. A couple of 5k resistors ought to be fine, whatever they provide volt-wise.

I get 339vDC from 240VAC, but close enough.

Mykey is right, you will definitely get pops. For example the cathjode cap - lets just look at the first stage one for now. WHen the switch is closed, that cap has to instantly charge up to the volt or two sitting on the cathode. POW! If you ground the lower end of the cap with a high value resistor (47k, 100k, whatever), then it will always be charged up through that resistor. The switch essentially shorts across the resistor. That doesn't change the voltage across the cap, but the resistor prevents the cap from having much of an effect. SHort across the resistor and the cap- all of a sudden "appears" in the circuit. Without a pop.

Any unterminated cap can make pops the same way. High calue resistors across the switches is one way to fight it, but there are other ways to configure rge circuits with pops in mind.

[Tim Hall]

Enzo, I had indeed intended to switch BOTH cathode caps out (and not create any negative feedback). Thanks for explaining the B+...I get the idea now. I wasn't sure if you calculated the the voltage drop for the last node over both power resistors or not, and I wouldn't have thought to add the current draw for the last two nodes when figuring the V drop on the second node.

Popping I definitely don't want. I'm still trying to decide if it's worth the effort/components to have all the switching...I'm really not sure how big a difference it'll make to be able to cut all these caps. I hadn't seen this done before with the cathode caps, and wasn't sure if it was a good idea. Any opinions on the value of this? The idea is to go from Fenderish to Marshalish.

The idea behind the mismatched coupling caps after the PI was to have asymmetrical bandwidth hitting the power triodes...thought this might add a little something to the distortion quality(?) Probably a bad idea. Any thoughts?

I really appreciate you guys looking over the "design." I'm learning a lot from this forum. If anyone has any other ideas on improving it, I would really like to hear them.

[Tim Hall]

...Revised

Revised

[pdf64]

I suspect that the tone/volume control arrangement you've got will be a bit muddy and limited. Have a look at the 6G2 brown princeton / 6G3 brown deluxe schematics for the best way forward.
The 56k grid resistor on the 3rd stage will restrict the gain / output of the 2nd stage, which may/not be your intention.
Hope that helps - Peter

[Tim Hall]

Peter,

Thanks. The 56k grid resistor is indeed intended to keep the gain down. That's also why I've chosen 5751's (plus I personally like the way they sound). That 12AU7 doesn't have much headroom. I want it to get dirty easily, but not with the volume on "2".

I will definitely take a look at the 6G2 and 6G3 schematics. The tonestack I have there was lifted from a 5F2. The 5F2 schematic showed a 250k tone pot, which I suspect should be 500k-1M(?)

********
Hmmmm...just reviewed the brown schematics. I really don't know which I should go with. Any opinions on whether I should go with the tweed or brown Princeton tonestack are welcome. I honestly have never played either of these amps.

[Don Symes]

You may want to use a BIG resistor across those switch terminals - at least 470k, up to 1M.

[Arthur B.]

The idea behind the mismatched coupling caps after the PI was to have asymmetrical bandwidth hitting the power triodes...thought this might add a little something to the distortion quality(?) Probably a bad idea. Any thoughts?

I suspect this wouldn't be a good idea. Any even hamonics genrated by this assymetry will cause the bass to be muddy sounding.

[pdf64]

Hi Tim, try
http://www.ampwares.com/ffg/
for the schematics, there you'll see that the princeton 5B2, 5C2, 5D2 and 5F2-A all use the same or similar tone/vol arrangement to the 6G2 brown princeton, whereas the 5E2 and 5F2 use the arrangement you've got on your diagram, which won't give much/any treble boost with the tone on full, so the tone control will probably just adjust between muddy and dark.
Re 56k v3 grid resistor, it will form a 130Hz high pass filter with the 0.022uF v2 coupling cap, which may/not be what you wanted.
Is the cathode resistor on the ecc82 large enough? It might not give enough bias voltage to keep the dissipation in safe range. Also, think about experimenting with increasing it to move the stage more towards classB, to simulate a full size amp? Whatever, common cathode resistors are most always bypassed so think about putting a cap across it.
Hope that helps - Peter

[Tim Hall]

Thanks for all the input. Peter, thanks for pointing out the 130Hz cut off...I think 72Hz is probably better. I keep wavering on the bias resistor...will probably just have to listen to different values there. The 10k load I have now falls just under the max dissipation curve at any bias.

I have a little amp right now that uses a similar output, and I tried bypassing the cathode resistor. I recall not being impressed the way it sounded with a bypass cap. I could have been imagining it, but it seemed to flatten out the response to pick attack/modulation a little.

Well, I suppose I need to start sourcing parts, and draft up a layout. Unless anyone else has any comments to add

[Bruce / Mission Amps]

I still think in the present configuration, the total gain from the preamp sections will have so much gain, with respect to the grids of the power tubes, that this amp will be very difficult to control, overdrive way too quickly and sound very compressed when turn up only a little bit above normal bedroom volumes.
I'm only guessing here but I wouldn't be suprised if the total theoretical gain of these preamp stages (before the driver) doesn't add up to something like 3000 to +4000.
So think about this, a set of power tube pentodes are looking for a max peak to peak swing (out of the PI/driver) of say, 15vac-20vac PP (before one tube gets driven into class B cut off)... and you drive the first preamp with 100mvac from a guitar signal...
.100vac times 3000 to 4000 = 300vac PP to 400vac PP... of course none of the stages could possibly swing anything like that so the very second stage will be distorted and start clipping with the volume up to mid- mid high.... and the third will be hard clipped into a square wave with the volume control still set lowish and the PI driver will try and send nearly pulsing DC into the power tube's coupling caps.
From what I've read, I think these power tubes bias at well under 10vdc so it doesn't take much signal to run one or both right into bias cutoff.
My experience has been that this can sound bad when taken to any high level. I think your preamp design will be able to make it bad.
Bad to me might be just what you are looking for..... I've heard god awful stomp boxes do similar things to decent sounding amps (and tone) yet some youngsters get a teenage woody over it!!

However, I am frequently suprised at how good or bad something sounds in the real world compared to my expectations and a fairly simple, as this one is, so it should be easy to mod when it is up and running.

[Tim Hall]

Hmmm...thanks for this perspective Bruce. I'm definitely not looking to build a chainsaw, but I do want some grind. I had originally toyed with bringing the Rp values down to 91k and eliminating the Ck on the second stage. I may revisit those ideas.

[Tim Hall]

Attenuation

Well, after doing the math I realized that going from 100k to 91k plate loads makes hardly any difference in gain. Anyway here's the scheme I've settled on, except that I may want to attenuate the signal some just before it hits the power triodes.

I was thinking I could simply up the value on the 1k5 stoppers, but I'm not sure what value I should go with. That is to say, I'm not sure how the math works out. Would I just figure it as simple voltage drop? Or is there a better way to pull the signal down a little?

Parts are arriving, and I've polished up a chassis nice and shiny! Hope it sounds decent