high gain amp design

[tedmich]

That's the thing, it's so flexible it's whatever I want it to be. Generically, it's a high gain channel switching amp. In the end, it'll have three or four channels (4 stage high gain, 3 stage high gain, clean and maybe a crunch channel), series/parallel FX loop, programmable options for each channel, foot switch, quad bias, and the list goes one. I should really just finish it up and show everyone.

It's taken alot of time because I've had to design alot of stuff for it, and when I started the project, I knew little more than how to read a tube schematic and *nothing* about MCUs. Here's the tally so far: MCU channel switcher code/boards, relay boards, output impedance selector board, quad bias supply/board, DC coupled MOSFET followers to drive the 4 x KT-88s (still have to test if the MOSFET drivers are worth keeping), effects loop (that's the current task, it's on it's third design, and almost done being prototyped). Then I have to to see if adjusting the cathodyne PI is needed/desirable, design a clean channel (hopefully that will be easy), then button up the chassis design (90% done, but I have to talk to the sheet metal shop around the corner and see if they can make the bends I've specified, otherwise it's off for re-design - again) and have it burnt. I'm almost to the point of just building the damn thing and adjusting stuff later. The design supports that option, but I'm stubborn.

And yes, even with all this fancy shit, it will still sound good. Hence the prototypes.

getting a little obsessive aren't we Mike? They never finish painting the Golden Gate eh? Hope you post clips when done!

[beinz]

interested to hear how it sounds... if im still alive.
unfortunately i have the same problem, but cant use the excuse of a complicated design, mine's pretty simple, but its my first, so i want to be as sure as i can first.
quad biasing sounds cool!

[defaced]

Thanks for the encouragement guys.

I wish it were just obsession, but finding quality time to do quality engineering is hard these days. I'm sure it would help if forums didn't exist, but then again, I wouldn't know how to do all this stuff if it weren't for these forums, so it's like a self eating watermelon (term courtesy of RG). Since I've got next week off, I plan to get some of those items finished up and the sheet metal burning the first of the year. This has been a goal at this time of the year for two years running now though, so we'll see. I'll make a thread and such to show off my awesome skillz (ha!), clips, etc, when the time comes.

Quad biasing is cool, but IMO, not necessary unless you want to mix tubes, and then dual bias works fine. I have debated removing it, but I put so much work into it and it is a unique feature that I'm hesitant to remove it. I have posted the schematic here before: http://music-electronics-forum.com/t16704/#post135894

If you want to hear what it sounds like today, look up the Fryette Deliverance. Since I now own a D120, I may end up putting a different preamp in this amp, or modding it to get a slightly more "raw" feel to it, not sure yet.

[beinz]

teee's got my wheels turning about capacitor voltage capacity. exploring the cheaper route of shopping out my components from mouser, i have no trouble finding 630v caps in the smaller values, but as you get into higher values (10uf+) electrolytics, its harder to find these in minimum 500v. there are plenty available at 450v.
as i look at a couple of schematics, my question is, how close to the max voltage rating is a good idea to run b+ at?
for example, the slo100 schem has a pt with 360-0-50-360, this should put b+ in the 500v-520v range, give or take. the schem has the electro's rated at 450v or less in these larger values.
marshall 2204 has a slightly lower b+, but caps rated at 500v.
i understand economics plays into things some as far as what certain manufacturers used, my concern is making sure i am not selecting components that are prone to failure because of a minor difference in voltage values. i could certainly find caps from f&t or jj, ect, in these voltages, but this requires more dollars, shipping, ect.

if i aim to be running b+ in the 425v-450v range, will 450v caps be sufficient, or am i playing with fire?

any thoughts?

[diagrammatiks]

you put 2 capacitors in series to get over 500v when you need to.

I wouldn't run 450 b+ with a 450vdc cap.

There are 500v rated caps and above they are just a bit more expensive.

[beinz]

thats kind of what i was thinking, may just have to eat it and get the pricier stuff. i dont want a board squished with a million caps just to get to desired voltage.
thx for your help.

[beinz]

do you know anything about the caps at justradios?
i stumbled across this site a while ago. they have a good range of caps at higher voltages for a reasonable price.
this site is geared towards old radios, but would these sound/work good in a guitar amp?
i would be curious to know more about their specs, life expectancy, ect..
im assuming that given temp, esr, and ripple specs are in line, a cap is a cap, and these should work fine?

[diagrammatiks]

i've never heard any complaints about them from people that have used them.

[big_teee]

Did you check out Weber Amp Parts?
https://taweber.powweb.com/store/capord.htm
B_T

[beinz]

good to know, thx.

[tedmich]

do you know anything about the caps at justradios?
i stumbled across this site a while ago. they have a good range of caps at higher voltages for a reasonable price.
this site is geared towards old radios, but would these sound/work good in a guitar amp?
i would be curious to know more about their specs, life expectancy, ect..
im assuming that given temp, esr, and ripple specs are in line, a cap is a cap, and these should work fine?

I very much like how polystyrene caps sound, and Justradios has some good ones. For power supply caps, ultra-long lifespans (CDE 381EL) handle higher voltages better than the std 2000h ones, snap ins are the best made/cheapest as axials are used nowhere other than a few high dollar tube amps... When doubling up caps for higher voltage there are calculations for so called "equalizing resistors" across the caps terminals, with 220k being a default value (the math often puts this lower) these also serve to drain the cap when unplugged, which makes servicing safer.
Capacitors - Multiple Capacitors
Engineers don't usually say 2 x 250v =500v, much closer to 2 x 350v = 600v

good cap refs
http://www.cde.com/catalogs/AEappGUIDE.pdf

also remember that electrolytic (EL) caps do not age well, in an amp OR on the shelf. I buy fresh, recently made EL caps from reputable (non SE Asian) dealers, usually CDE/Panasonic/Nichicon. (older film caps work fine!)

FYI motor RUN film caps are sometimes available at high AC ratings (275vac) for cheap on ebay,
MOTOR START CAPACITORS CBB60 SH 20uF 240VAC | eBay
these are high quality metalized PP film motor run caps which should handle 600-1000VDC for $2

[beinz]

thanks for the words, some different things to think about, and btw, MERRY CHRISTMAS! (since were almost there!)

[beinz]

hey guys, ive got another question if you dont mind.
in all this parallel vs series deliberation, i am unsure about how a resistor will work in one of a couple of senarios.
in my 2204 schem, depending on the particular layout, (i have attached the ceriatone as an example)
we have either dual section cans, or single 50uf caps paralleled to equal 100uf.
this example (ceriatone) uses 2 50+50 cans with one of the cans running a 10k/2w resistor between positive terms. the other can uses a 220k/3w res from 1 pos to neg, i believe this is for draining the caps?
being that a 50+50 is equal to a 100uf cap, if i were to use just 1 100uf/500v cap for example with 1 positive term, and 1 negative term, how would i place the 10k resistor? would both ends of the resistor go to the 1 positive term of the 100uf?
or the other option of running 2 220uf/350v caps in series for example to equal out to 110uf/600ish-v,
same question, how would i place the 10k resistor? same as the 50's, from one pos to the other pos?

the easiest would be to use 2 100uf rather than 4 50uf, fewer caps, smaller, less space ect. just unsure of how this will connect? i could use cans i suppose, but would prefer to stick with newer, smaller caps if possible.

[big_teee]

hey guys, ive got another question if you dont mind.
in all this parallel vs series deliberation, i am unsure about how a resistor will work in one of a couple of senarios.
in my 2204 schem, depending on the particular layout, (i have attached the ceriatone as an example)
we have either dual section cans, or single 50uf caps paralleled to equal 100uf.
this example (ceriatone) uses 2 50+50 cans with one of the cans running a 10k/2w resistor between positive terms. the other can uses a 220k/3w res from 1 pos to neg, i believe this is for draining the caps?
being that a 50+50 is equal to a 100uf cap, if i were to use just 1 100uf/500v cap for example with 1 positive term, and 1 negative term, how would i place the 10k resistor? would both ends of the resistor go to the 1 positive term of the 100uf?
or the other option of running 2 220uf/350v caps in series for example to equal out to 110uf/600ish-v,
same question, how would i place the 10k resistor? same as the 50's, from one pos to the other pos?

the easiest would be to use 2 100uf rather than 4 50uf, fewer caps, smaller, less space ect. just unsure of how this will connect? i could use cans i suppose, but would prefer to stick with newer, smaller caps if possible.

The 220k is a bleed resistor.
The 10k is the Dropping resistor for the B+ Rail.
If you look at the reissue 2204 Schematic, you will see that the second 50/50uf is also strapped together like the first.
http://mirror.stones-amp-inn.de/mars...l_50w_2204.pdf
If you look on sheet 2 you will see it on the top of the page.
That allows an additional 50uf of filtering on the EL34 Screens.
I Tried wiring mine both ways and ended up with the reissue strapped version.
So in Reality You could buy 1- 100/100uf Cap Can and accomplish the same thing as the 2 50/50s.
That would work out well if you are trying to save space, in a small Chassis.
T
**Edit Note:
If you strap the second Can 50/50uf or use one 100uf Cap.
You would take off the wire Labeled C on the Ceriatone Layout.

[beinz]

is there any efficiency or lack of by going with 1 x 100+100 can as opposed to 2 x 50+50? if not, then this would be the way to go for me.
i was thinking of this sort of backwards, but it now is simple and makes sense.