schematic critique: power supply

[yunger]

here's the latest evolution of my amp that i've been working on for the past few months. it's been a great learning platform. i'm trying to learn more about power supplies so i'm digging into it on my amp. what do you think?

[ThermionicScott]

I don't see any glaring problems. I always like to under-fuse my amps, just to err on the side of caution, so you might play around with smaller ones. (My 5E7 clone has a 1.5A in it, my 5F2A has a 0.5A, and neither has popped yet.)

You might also play around with attaching the heater CT to the output tubes' cathode, on the off chance that it reduces hum.

- Scott

[Bruce / Mission Amps]

The only thing I see as a potential issue is the total gain of all your preamp stages will have so much excess gain as to turn your cathode biased EL84s inside out and spit them up.
You'll only need 15-17vac of drive in this to blast the whiskers off the EL84s.
Hopefully your 22k "pre-PI load" resistor will keep it under control.

[Merlinb]

I would also add to Scott's comments that you really ought to have a fuse in line with the B+. It's all too easy to accidentally burn out a PT or OT without enough fusing.

[J M Fahey]

What does "rated@68mA" mean?
It's written on your PT.
If that's so, your PT is way underpowered for the job.
Your amp needs around 120mA at idle, a lot more when driven.
Your 2A primary fuse will burn *after* the rest of the amplifier.

[yunger]

thanks for the input guys!
i will indeed make the safety changes suggested by scott, merlin and jm.

scott, the heater ct is on the rk of the power tubes on the amp, just not shown on the drawing. that was a very helpful move.

bruce, you are definately right about the el84's sensitivity. the 22k may be a bit too high to cut down the preamp gain into the power amp. i have a 25k pot there for now so i can dial in the right amount of preamp/power amp breakup in a band situation. i originally had the tone control between the two triodes and realized that all i was doing was amplifying clean sound as well as noise just to attenuate the crap out of it before it drove the el84's into obnoxious territory (i'm thinking this may be why almost all fender pro jr owners complain about noisy amps.) i figured i might as well take advantage of the cascade for a subtle preamp crunch. i could always up that 22k load resistor if i wanted a clean, less noisy preamp. otherwise, i think a matchless spitfire type would have been more efficient and benificial for clean->overdrive.

jm, the power transformer is from a fender pro jr. it's my understanding that it is rated at 68ma. i may be wrong. would the amp drawing more current than the transformer can handle be a safety issue or tone issue or both? i mean, the amp works as is but i have nothing to compare it to. maybe this is why i havent been satisfied with the bass response of the amp. once i change the transformers, all that will be left of my old pro jr is the cabinet and frame..... but like i said, it was a great learning platform and tons of fun. i definately wouldn't have learned as much as i have if it wasn't for the direction and help from a lot of selfless people here.

[ThermionicScott]

JM, where are you getting the 120mA figure? This is just two EL84s and a couple of 12AX7s.

[J M Fahey]

Hi Thermionic Scott.
Shortest answer could be "I remember measuring that", but since my memory can (and does) fail often, I checked just in case:
http://www.drtube.com/datasheets/el84-philips1969.pdf
There, on Page 4 , under "Class AB2 , 2 tubes in push-pull, with 300mA plate voltage, it states:
No signal: 2x36 mA (Plate) + 2x 4 mA (Screen)
At clipping (still sine wave): 2x46mA+2x11mA= 114 mA +, say, 4 or 5 mA for the preamp, we are practically reaching those 120 mA.
When overdriven, quasi-square-wave output, the plates *but very specially the screens* will "eat" even more, easily 50% more, so a 120mA rating is on the ballpark, by no means excessive.
A 68 mA rating would be definitely poor, even dangerous.
Anyway, in this particular case, if the PT was already feeding a pair EL84's, it will continue doing so.
If you happen to have a 2xEL84 amp these days, please post what you measure, it will be interesting.

[yunger]

so would a hammond 270daz 260-0-260@104mA r.m.s. be more appropriate or should i save my money until if/when my pt shorts? btw, i replaced the 2A fuse with a .5A.

[ThermionicScott]

Hi Thermionic Scott.
Shortest answer could be "I remember measuring that", but since my memory can (and does) fail often, I checked just in case:
http://www.drtube.com/datasheets/el84-philips1969.pdf
There, on Page 4 , under "Class AB2 , 2 tubes in push-pull, with 300mA plate voltage, it states:
No signal: 2x36 mA (Plate) + 2x 4 mA (Screen)
At clipping (still sine wave): 2x46mA+2x11mA= 114 mA +, say, 4 or 5 mA for the preamp, we are practically reaching those 120 mA.
When overdriven, quasi-square-wave output, the plates *but very specially the screens* will "eat" even more, easily 50% more, so a 120mA rating is on the ballpark, by no means excessive.
A 68 mA rating would be definitely poor, even dangerous.
Anyway, in this particular case, if the PT was already feeding a pair EL84's, it will continue doing so.
If you happen to have a 2xEL84 amp these days, please post what you measure, it will be interesting.

It's not class AB2, but it will draw that ~120mA at full output. It should be a lot cooler at idle.

I wonder if the Blues Jr's PT is underspec'd like this so that it will "sag" more when pushed. I agree that it's bad design to use wimpy iron just for that purpose.

- Scott

[J M Fahey]

Sorry, it was a typo, I typed

"Class AB2 , 2 tubes in push-pull,

and should have written ""Class AB , 2 tubes in push-pull,", which is literally what is written in the datasheet.
I was switching back and forth between the answer page and Adobe Reader screens.
I could not cut and paste, which is what I always do to minimize typing mistakes, because it's a scanned old datasheet, no text but a full graphic image instead.
The current consumption calculations and estimates still hold.
Besides, it's roughly what I had measured earlier on the bench.

[ThermionicScott]

Yeah, sorry about pouncing there.

The Princeton Reverb PT is similarly underspec'd, as Stokes noted a long time ago. Makes you wonder why more of them don't blow up...

- Scott

[J M Fahey]

Well, maybe it's part of the magic, and I'm not kidding.
When I started making amplfiers, in the late 60's, I knew this old guy who knew *all* about audio transformers.
He knew no Fender or Marshall, his world was Marantz, HH Scott, Harman Kardon, British Hi Fi amps, etc.
I started cloning Bassmans and Bandmasters, which when compared to the originals, were "too loud - too clean - too hi fi" and I got to be known as the whiz kid (in those days, he he) to see for *bass* amps.
One day a guy called, in distress, he had a Fender with shorted tubes, "repaired" with a piece of aluminum cigarette paper stuck into the fuse holder: both transformers burnt.
Obviously I sent them for rewinding to this old guy, stating I wanted them rebuilt with the exact same winding, wire, iron, etc. and making an extra couple for me, same exact specs.
He whined about the poor iron, thin wire, low turns count (low inductance=poor bass), poor interleaving (no real high frequency).
Well, end of the story: they were "the real deal" ........
The only major amplifier I know which used Hi Fi transformers (Partridge) was Hiwatt, and you know their loud, clean, bright sound, even distorting.

[yunger]

i wanted to see what the unloaded B+ voltage was off the bridge rectifier and here's what i read:
232vdc (unloaded)
325vdc (loaded by the circuit)
this can't be right! i read it with my dmm, black lead to ground. what's up?

[J M Fahey]

Hi yunger,
I'm sure you are disconnecting your bridge rectifier positive terminal from the rest of the circuit, and measuring there.
*If* you do that, "unloaded" means no filter cap attached, you have the transformer secondary full wave there, your meter reads the average value of that=232V "DC" (which it is not).
When you "load" that, the filter cap charges to the peak value, minus some volts because of the load, =325V DC, which if you are very nitpicky you should call "DC+ripple".
In fact, when I test a just opened failing amp, measuring a "DC" voltage very close to the AC rating of the transformer , or measuring in an SS amp, say +40/-28V , is a surefire indication that the power supply big filtering (in the SS example the negative side one) is open, very often because vibration cracks the pin solder or the copper around the pad, or in a worst case, the pin is torn from the capacitor case.