Ampeg BA-115 help needed

[smackoj]

hello gents; I picked up an Ampeg BA 115 chassis advertised as 'untested'. I did a thorough visual inspec. and see or smell nothing burnt or disconnected. I put it on my current limiter and get the led power light on but no juice (light) appears thru the limiter bulb. I pulled the 2 B+ wires off the brd and got 27 vac on each. I pulled the J22 ground wire off the board and got 7 vac. The schemo says it should have 40 vdc plus and minus on the rails and it gets divided by resistors to make 16 vdc plus and minus for the preamp. I can't really find a good place to check for 40 vdc with the brd still installed because there is so much hot glue goop on the power resistors.

Should I be thinking that the PT is bad because of the 7 vac on the ground wire?  Thanks

http://www.guitarkitbuilder.com/img/BA115_schematic.pdf

[shooter]

the led power light on but no juice

That's a basic indicator at least the + half is good

7 vac on the ground wire

That might send you down the wrong path.

I can't really find a good place to check for 40 vdc

check on d10/D13 directly and look for the +/-16, if that's good, the 40 is probably good

have your meter all clipped in and set up, THEN powerup long enough to get reading, 30sec?

[PRR]

> 27 vac on each

27V rms AC gives 38V DC. This is OK.

This DC should appear at the rectifier +/- pins and at the big caps. Also at the S pins of the power output devices(be careful not to short!). These are ALL glopped?

> I pulled the J22 ground wire off the board and got 7 vac.

Probably a bad idea (can kill stuff). All it tells you is that the two loads are not exactly balanced. We know that: the LED hangs on just one rail(+). The voltage you will see in this condition depends on too much stuff to work out.

You do not say what is WRONG. It doesn't blaze the lamp-limit. The LED comes on. You have good-enough AC voltages off the PT. I would put back that J22, then check for DC at the speaker outputs. Over 1V, do not connect speaker. Under 0.2V suggests the power stage has no gross fault at no-load condition.

[smackoj]

Thanks guys;  'what is wrong'  I just didn't see any light thru the limiter bulb so I didn't want to try anything else accept testing voltages until I asked for help.

I will test the source on the chips and the DC at the spkr jack asap and report back.

thanks again

[PRR]

A happy solid-state amp may not pull enough current to even dim-glow a lamp.

The lamp is just-in-case the amplifier is a near/dead-short on the wall plug. You passed that test. Move on.

[smackoj]

I finally got my limiter lamp back from my wife and the results are:

.21 vdc at output to spkr

Q2  +24 vdc
Q5  -36 vdc

[shooter]

Q2  +24 vdc   Q5  -36 vdc

That's at the (S)ource on each?

can you get on the +/- 16 by Q3/Q4.  as long as you're probin, what's the bias?

[smackoj]

I did as you said Shooter and found the following:
bias voltage at the trim pot is neg. 1.35
bias voltage at R41 is +1.35 vdc
R34 should have +40 vdc but I only found 24 v
At R50 there should be a plus 16 v but I found only 13 v
and I did find -16.5 at R42 but it's open because there is 0.00 vdc on the other side of it. there is some slight discoloration on the brd where it looks like the resistor smoked.

thanks again

[PRR]

Does it hum?

Get hot?

[smackoj]

Yes it does hum quite a bit thru the spkr

No it is not getting hot

thx

[PRR]

I think the BIG caps C6 C14 4700u 50V should be replaced.

The voltages are consistent with a failed C6. (You get a "DC" near AC RMS rather than 1.414 times RMS.)

The amp is over a decade old and not built to last forever.

C6 and C14 each do the same thing for their side of the power. If one has failed, the other one is likely to die soon.

[smackoj]

replace the big caps. got it. thanks

did you see where I found R42 open? 3 comments up from your last post.

thanks

[shooter]

WAG for R42, Q1, Q3 C20.  a base to emitter short might draw enough current to take out a 1K, c20 would most likely short then open, but usually you find the results all over the amp.

[smackoj]

I am not sure what 'WAG' means shooter.

thanks, jack

[shooter]

in this context, Wild Ass Guess

[PRR]

> I found R42 open?

Is that dead-nuts-dead-ZERO? Or just very small?

The way that too-clever scheme works, a just-right wrong voltage between rails will put that point real near zero. Get the big stuff fixed, then look at the small stuff.

[smackoj]

dead nuts zero on the meter

the power caps are ordered

[shooter]

Not to hijack or get into the weeds, but the "circled" circuit is really hurting my face trying to understand it, any ideas or wisdom

[sluckey]

Looks like a variable NFB circuit used as an automatic gain control. Probably used as a compressor. The LDR part of the opto-coupler is parallel to the feedback resistor for U3-A. The output of U3-A is sampled from the right side of C10. As the signal increases, the opto-coupler shines brighter and the LDR resistance goes down. This reduces the feedback resistance which decreases the gain of U3-A. Just the opposite happens when the signal at C10 decreases. So, weak signals get amplified more, and strong signals get amplified less.

[smackoj]

Thanks for the high tech info SL

[shooter]

Thank you!, I kept getting hung up on the steering diodes n missing the fbk line from c10

[sluckey]

The actual feedback path is directly from pin 1 thru the LDR and back to pin 2.

[PRR]

> the "circled" circuit is really hurting

Gain-control, not instant NFB.

When not limiting, gain is set by NFB at 47K:10K.

The resistors set 3/4 power rail references. If the Output passes 3/4 of the rail, the opto's LED lights. This shines on the LDR across the 47K and reduces gain. The LDR is selected for faster attack than decay. It is an Audio Limiter, though the time constants are pretty arbitrary. It is not a "hard" limiter which is why the threshold is 3/4 of the rail (perhaps 80% of clipping). If you push hard, it will clip, but not easy.

Below that, the JFET appears to be a crude power-on mute. The JFET is normally "on" (when no voltage anywhere). The Gate resistor pulls it to "off" when -16V appears. But not right away. This gives time for the rest of the system to thump and stabilize before signal is passed to the output stage.

[shooter]

The actual feedback path

Ya I got that, normal op amp closed loop, the rest above, that messed me up.

the JFET appears to be a crude power-on mute

like that part

[smackoj]

A "crude power-on mute" works fine for me because I'm a crude bass player.  IOW any note will do as long as it's somewhere close to one on the page. 

[smackoj]

well gents she's got new filter caps and a new 1k in the R42 slot. hit the switch and no hum and a very faint hiss so I'm real happy with the
end result. Couldn't have done it without you guys so muchas gracias amigos. here are a few pics of the amp just for general interest.

[shooter]

I'm real happy with the end result

so the rail volts are good?   Did you get the bias set per schematic?

[smackoj]

Shooter;  I haven't tested voltage or bias yet. I will report back asap.

thx, jack

[PRR]

I would want to see near 40V both sides ASAP.

For bias, I would be happy if at idle, the heatsink was somewhat warm. The bias scheme here is dubious, MOSFETs are not as fussy as BJTs, there's no exact optimum. If there is "some" current flowing, enough to warm up, not enough to be hot, that's good enough.

[smackoj]

Is the bias set by just the one trim pot?

The second pg of the schemo has a note stating to set the bias at 5mv across R61. Does that sound right?

thanks

[shooter]

Looks right, it's a .1ohm R, can't see much else for it but convenient measuring of current

 

[PRR]

Yes, 0.005V/0.1r is 50mA and a fine bias. And will throw 4 Watts of heat, about 1/10th the heat of FULL ROAR, so "warm" not "HOT".

[smackoj]

hi; I tested voltage at the source of the output chips and got  -37.5 and +37.5
BUT I screwed up trying to change the bias with the trim pot and blew a fuse. I replaced the fuse and put it on the current limiter and found a short. I looked closely and R42 was burnt open again.  I will replace it and try again. I didn't realize the trim pot was so sensitive. My bad

[shooter]

I looked closely and R42 was burnt open again

Just center the pot for now without power on should be 250ohms either side of wiper.
with a new R42, measure your +/- 16VDC and see what you have.

[PRR]

I would just short the dang bias pot. That will set a minimum bias. Bias is not critical in this amp. As you found, it may be hypersensitive. Also pots go bad, which in this case will send the bias HOT, and as you see, blow-up things.

[smackoj]

Well....I should have left the bias pot alone. I replaced R42 with another 1K slapped it back together, put it on the cur limiter and CRAP still a short in the amp. I tested the source pins on the pwr transistors and got something like half a volt dc. So, sorry to say I need more help figuring where to look for the short. Possible one of the transistors before the output or maybe even fried the output trans? None of that stuff smoked but I'm no expert on what it looks like when output trans go pooof.

thank, jack

[shooter]

what it looks like when output trans go pooof.

Now you do!

I was paid to make equipment work, fast, so with corporate money I would replace a lot of parts in a short time. $30 in parts n a couple hours labor vs $1,000s per hr down.

this would be my "shotgun" list.
Q1, 2, 3, 4, 5  C20, 31, and the 4 zeners.

You can, with power off, measure ohms to ground at both +/- 40, then start pulling parts, when the ohms jump, you're probably close.  as long as you're there check the pot, swap or

just short the dang bias pot

Take lots of pics, close up so you don't put a good part in backwards.
I would expect the + and - rail to have about the same ohms to ground so if you have many K on one and very few on the other, that's your start.

[smackoj]

Hi; I pulled the 2 output mosfets and tested. Both look gone on the meter so I have ordered replacements via Mouser. I will take a look at the resistance to grd on the rails tomorrow and report back. I'm not really sure what is meant by "short the trim pot"? A little more info would be helpful.

thanks again

[shooter]

Both look gone on the meter

IIRC, Igfets don't ohm like transistors, but short is short if that's what you're seeing.

[smackoj]

Hi Shooter; I tested the 2 power fets using this "video" method. Let me know if this is not an accurate method? Thanks

https://www.youtube.com/watch?v=RBJGOOTEwfU

[shooter]

That looks good to me, See, before you know it you'll be the goto guy for SS repair
In SS you can expect the un-expected, you also usually have a cascading failure, like when 1 transistor fails, it takes others with it, so hang in there.  with the FETs pulled and no power, what do you get ohms to ground at both 40 and 16 taps?

[smackoj]

I checked all the points on the schemo where it showed either 16 v plus or minus and 40 v plus or minus. I got apprx 1k at a lot of the spots but back by the power Rs it had close to 2k ohms.

jack

[shooter]

apprx 1k at a lot of the spots

I just dug out my 60W SS amp, similar to pg2 of yours and I get >100K at my 40v taps, more in line with what I'd expect.

[smackoj]

Hi gents; I have been waiting for the power mosfets (Q2, Q5) to arrive so I can replace them. Should have them tomorrow. I have tested all the other semi conductors and they all seem sound. I ran a voltage check again and I am getting 17 vdc plus and minus at the appropriate spots on the board. I am also getting 36 vdc plus and minus where the schemo shows 40 volts. Possibly the PT doesn't put out quite enough voltage for the higher lines?

I would like to know a little bit more about PRR's suggestion to "short the trim pot". Does that mean run a wire to ground from the trim pot?

Thanks for hanging in there with me.  Jack

[sluckey]

I would like to know a little bit more about PRR's suggestion to "short the trim pot". Does that mean run a wire to ground from the trim pot?

NO! That means put a jumper across the outer lugs of the pot. DO NOT CONNECT ANYTHING TO GROUND.

[smackoj]

Thanks SL I rec'd the proper power mosfets (found exact match on the 'bay'). I will report back soon.

[smackoj]

update on the Ampeg BA 115: I have replaced the 2 output mosfets twice.. It is operating without issue except for one BIG issue, no output at the speaker. I have the proper voltages on the pcb and to the source pin of the output mosfets; my best guess is that the solder pads under the output mosfets look to be starting to break down? What should I d o to test if the signal is not getting thru the output fets? should I check for AC and DC at the speaker wires?

thanks

[sluckey]

Did you solder the two mosfets in the correct positions?

[shooter]

check your headphone jac also, it provides spker ground with nothing plugged in, can temp a ground also

[smackoj]

According to the schematic I have them correct. The IRFP9140 is the P channel in Q2, and the IRFP140 is the N channel in Q5.  The source pin on the P9140 has +37 vdc and the P140 has -37 vdc on the source.