Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

March 28, 2024, 02:17:56 PM

Login with username, password and session length

Recent Posts

 

ampeg vt-22 PI

Started by ilyaa, July 08, 2014, 04:15:37 AM

Previous topic - Next topic

ilyaa

once again - fixing this one up for a friend.....

symptom was no/low signal and weird sound when turned on.

quick preliminary looked showed that the screen grid resistor on V5 (R42) was fried (low voltage there - resistor tested open and looked a bit burnt). ill change that.

BUT upon looking around the rest of the amp i also found that the PI voltages were bad. ~150V on pin 1 and ~330V on pin 6. now i remember having a similar issue on a deville i worked on. i also know more about long-tailed pair configurations now, enough to say that it looks like pin 6 is not conducting! both heaters are lit. i tried a different tube - same symptoms (a 12ax7, admittedly, not a 12au7 as it says on the schematic, but that shouldnt matter for this issue, correct?). ive resoldered all the pins and adjoining connections and checked the region and it all looks good. does NOT seem mechanical.......any other things i should check or am i just being sloppy with the mechanical stuff most likely?

J M Fahey

Is SSGuitar turning into TubeGuitar?  :(

Enzo

So what is happening at pin 8?

ilyaa

#3
J M - ive got a solid state ampeg SVT next in line!!

oh, and pin 8/3 is giving me 6.8V....a little low - but seeing as both plates share the same cathode current, how could the voltage there explain one conducting and the other not?

Enzo

The two triodes are supposed to share the common cathode, that doesn;t mean they do.  Until you tell me, I cannot assume the voltages at pins 3 and 8 really are the same.   A broken conection perhaps.  or a bad socket.  Things happen.

You cannot reliably measure the grid voltages on 2 and 7 with a meter to ground.  So you measure them from cathode to grid.  Then simple subtraction tells you the true grid voltage.  Are yours the same?

Have you verified resistance on the cathode, plate, and grid resistors?

ANy sign C11 or C13 is leaky?

ilyaa

voltages at pin 3 and 8 are around 6.8V

pin 2 to pin 3 is about 6.3 V
pin 7 to pin 8 is about 6.1 V

they are not the same but only 0.2V off or so....i know a small grid voltage/current controls a large anode voltage/current, but this difference seems to small to explain the imbalance, right?

all the resistors in the area (R27-R33) measure spot on.

no obvious signs that c11 or c13 are leaky - both grids on the power tube side of the capacitors are at about -60V....how could i tell if they were leaky otherwise?

i do have to say that compared to the other tube sockets on the board, this one does seem just a bit wonky. but the tube feels snug when its in there. just not AS snug as the other tubes on the board. only other strange thing is that all the tube connectors on the tube side of the socket are silver-ish looking except the pin 1 connector looks gold for some reason....or yellow.....???

ilyaa

#6
hmm i dono this ones baffling me -

all resistors in the area look good. i opened c11 and c13 on the PI plate side and that didnt make any difference. i took out the tube socket and gave it a good look and it looks fine! i might still change it out anyway and see if that makes a difference but i dont know......any other possibilities??

one other details: with no tube in the socket, both plates are even at about 400VDC...so the imbalance occurs only when the tube starts pulling current....

J M Fahey

DonĀ“t have an answer yet, but let me think 2 things aloud:

1) this is not a long tail PI  .
No long tail resistor anywhere, just a standard shared cathode biasing resistor R28 .

2) the fact that there is less than the expected voltage drop across it, and at the same time higher than expected plate voltage across R31 makes me think we are missing current across the lower triode, (pins 6/7/8), so I suspect a bad/open contact, either  socket contact 6 or 8.
Turn amp off, pull the tube, and tighten the socket contact with a needle.
Then replace the tube and remeasure.

ilyaa

its not called a long tail pair when they share the cathode resistor across two separate triodes? i thought thats what a long tailed pair was.....

its hard to mess with the socket contacts - im just going to wait until some new sockets come in the mail and just replace it and see what happens....

at this point ive also tried opening all the caps in the area to see if leakiness is the issue and that has made no difference, so im thinking it HAS to be a bad contact

J M Fahey

I think a new socket will solve that problem.

As of the inverter: in the long tailed one you have 2 resistors from joined cathodes to ground: the proper biasing one and another, *much* larger, which acts as a constant current source.
Being constant (total) current, when one triode raises consumption, the other one lowers by the same amount, so signal is coupled cathode to cathode without any other external path.
In this one, standard relatively low biasing resistor straight to ground only provides bias.
Signal is coupled to the 2nd triode straight to its grid, through R33 and C12 .
But now the second triode, which has its own gain, would make for a *huge*  imbalance.
So its gain is lowered to -1 (meaning its output is same value as input but inverted) adding local feedback resistor R32.
If it were an Op Amp which by definition has "infinite gain", R32 and R33 would have same value ... but a 12AU7 is far from that, only some 12X gain, so values must be adjusted so the stage with a real world 12AU7 has unity gain.
So instead of 1M/1M we have 820K/1M .

Roly

A Long Tailed Pair (LTP) is also called a Diff or Differential amplifier.  It is characterised by having a large value (e.g. more than 10K) un-bypassed resistor common to both cathodes (emitters, sources; the "long tail") to approximate a constant current sink.  In some circuits, particularly op-amps, it will actually be an active constant current sink (approximately an infinite resistor connected to an infinitely negative voltage).

Two stage sharing a common cathode bias arrangement doesn't have a special name, generally encountered as a tremolo modulator in guitar amps, and is characterised by having a small resistor value around half what you would expect with a single stage (e.g. 680 ohms rather than 1k2), and  normally bypassed so the stages are not AC coupled via the bias resistor.


{There is a rather strange and sad sidebar to the LTP.  Its inventor, a Brit by the name of by Alan Blumlein, was something of a victim of his own success.  Alan was a very bright fella who was just getting into his powers when WW2 started, and being a very bright fella he soon found himself working on top secret war projects.  As a result much of his best work was kept under wraps until long after his death.

In his short life he invented stereo, Ultra-Linear valve output stages, and made major contributions to television broadcasting and radar.  He was killed in 1942, aged 38, when the Halifax bomber being used in the development of the early "H2S" (so simple it stinks) radar system crashed.  He was granted over 120 Patents.}
If you say theory and practice don't agree you haven't applied enough theory.

ilyaa

crap -

new socket - same problem.....

dont know what other things to try - ill add one observation that i havent mentioned yet: when im measuring DC voltages to ground in that part of the amp, i feel a slight vibration through the meter. it feels hummy, like 60 cycle kind of stuff. as i move away from the section of the board with the PI it gets weaker and weaker and is totally absent when im measuring other voltages in different parts of the amp, even the ~500V on the power tube plates/screens.......????????

g1

Triple check that both side heaters are lit.
Remove PI tube, power off.  Measure resistance from pin 8 (at the socket not at the board) to ground.
Now measure resistance from socket pin 6 to point D of power supply resistors.

ilyaa

ooookayyy so things went in a slightly different direction -

i took another look at the tube and it reads "ECC 832" - i was assuming it was an 833 but no!! a quick internet search told me that the ecc 832 - or 12DW7 - sometimes has a non-standard pinout.....

the schematic asks for a 12AU7 - i dont have one on hand so i put it an actual 12ax7 and now the voltages on both PI plates are the same - but they are high - about 340V each - assuming this is because 12au7 is lower gain so the 12ax7 maybe running at higher quiescent current than schematic values - right??

but at least the imbalance seems to be solved!

in fact, ALL the tubes that should be ecc833 are ecc832.....sneaky sneaky! who was in here before me?!
but no other voltages seem funny........

i put in the 7027 power tubes that the amp came to me with and they started red-plating....uh oh....is that usually a tube problem or an amp problem? i only had one other 7027 on hand so i put it in alone and it seemed to be okay - lonely but not red-plating at least - i can use 6L6's in here, correct?



g1

Ok. the 12DW explains the PI imbalance, as you proved by trying the 12AX and finding it was balanced. 
Yes, you can plug 6L6's right in with this circuit.  Just have to be good enough to handle the high plate voltage.
Of the old tubes, not likely they are all bad, try them one at a time, probably the one or two that redplate are pulling down the bias for the others.