HELP Repair advice!

[tedmich]

trying to fix an ancient QSC Series One 1400 amp, it blows the PCB mounted fuse for channel A almost immediately on startup. I measure ~1.2 ohms between the channel A FWB +/- : not good. 1.2ohms is there when fuses removed, power off, PT power tap disconnected etc. Channel B FWB reads open between +/- as it should.

schematic: http://www.qscaudio.com/support/libr...20One/1400.pdf

The B channel FWB looks newer, the Channel A FWB older so I pulled the A FWB,but it tested AOK and all the filter caps (2200uf/100v) read AOK on the board. I have the schematic and there are very few things that look like they can "fail closed" between the Channel A +/-75VDC rails so I am a bit confused.

Its one of the most recent S1 1400 models, revision P with the 220k upgrade hum balance pots and the 2 L brackets on the tranny, older units had 100k hum balance and 4 standoffs on tranny. It does have the funny dual octal sockets for accessories, shown in the related 1100 schematic:
http://www.qscaudio.com/support/libr...20One/1100.pdf

Hell its 25y old and very dirty, maybe something is shorting under the transistor heatsink as that's difficult to pull?

Any insight would be greatly appreciated!

[tedmich]

Its a USA 900 model?!

the amp internally looks much more like the USA 900 series, ie MJ15023/22 power transistor pairs not the Toshiba 2SB554/2SD424 pairs
also no TR1 clipping adjust trimmer and current limit trimmer TR5 is added, C12 2.2nF cap across hum trimmer (TR2).

So its a USA 900!

[Enzo]

Don't assume that unless you know it for sure. There were several versions of the 1400. Your link appears to have the collection of them, but though all their amps have that family resemblance, all it takes is a couple little detail differences to blow you out of the water. The first drawing in the link, the small one, has C12 across the 220k hum balance, and also has TR5. Other than the output transistor types, what other differences between that drawing and your amp?

Also, the switch from Asian transistors to the MJ15xxx series really wasn;t a change. DOn;t mix the two types, but you can pull the 2SD424s out and stick MJ15024s in their place without making any other changes except adjustments.

Your bridge rectifier measures shorted? But pulling it shows it is not? The output collectors on the heatsink are all at ground, if you didn;t realize that. And though that may be the case, that won't prevent shorted outputs on each side to short across the two rails. Got low resistance to ground from the +/- point of that bridge? The ballast resistors are accessible, so if they show shorted to ground, then their transistors must be.

And a couple shorted filter caps would do it, though I generally see one short, not two.

By the way, before chasing down any odd problems, check that hum balance for opens in the two 47ks between the +/- rails and for the pot itself open. I have replaced many of those trim pots. And if one of the 47k opens, it yanks the input of that op amp over to one rail.

Because of the cramped quarters in that amp and the layers of dust crud, I found myself using the layout drawing on these more than on many other repairs. Are the layouts available to you?

[tedmich]

thanks Enzo, I ran across one of your previous posts regarding the hum balance pots; I pulled both and they measure fine so I reinstalled them. The 47k all read AOK too.

The right/A channel FWB is at ground for its (+) leg, ~1 ohm above ground for its (-) leg.

The 0.22ohm ballast resistors are FP right? So they fail open eh?

oddly Q1 and Q2 on the A channel appear nice and silver down in their heat sinks while the Q1/Q2 for channel b (the one that works) are the uniform brown. Maybe this thing has been mono a long time (none but the both FWB appear to be resoldered)

I have the USA 900 layout which is very similar but NOT exact; I don't have the 1400 layout (or the service manual!)

Maybe I'll just pull all the channel A transistors and look for the shorted one?

[Enzo]

Those wirewound resistors can fail open, but if they do, they remove the shorted transistor from the circuit, right? If the resistor opens, you can no longer read through it to check what it is wired to. Ultimately you may have one shorted xstr on each side, or several, and only some may have burnt out the resistor. Always check all the resistors after all shorted xstrs are found and removed.

Here is the secret to parallel transistors shorted. Listen up, anyone fixing SS power amps. If you look at one side of the amp, the collectors are all commoned together, and the working end of all the ballasts are wired together. Any one shorted transistor makes the collectors and resistors shorted together. So instead of measuring from collector to the power rail, go down the row and measure resistance from the collector directly to the emitter pin of each transistor. If the transistor is shorted, you get a very low reading. But if you are on a good one, or at least a non-shorted one, your meter sees a short, because a parallel is shorted, BUT your reading is in reality reading out through the resistor on the good part, and back through a second resistor to the shorted one. That adds a half an ohm to the reading.

So, when you go down the row and see ohm readings like 0.7, 0.7, 0.8, 0.7, 0.2, 0.7, it is a pretty good bet that the 0.2 ohm one is the shorted one. Pull it and see if the rest no longer appear shorted.

That only works for EC shorts. In my experience, in the unfortunate circumstance of a collector to base short, it usually takes out the whole row of transistors.

And remember the filter caps common goes to the speaker, not to ground.


I am glad your hum pots are OK. Until you get hip to them, they can make it maddening figuring out why nothing works. Off the usual radar so to speak.


And if there is any damaged outputs, then always check the two drivers, the 22 ohm resistors and the emitter resistors of the drivers. (looks like 6.8 ohm on this one.)

[tedmich]

Enzo, thanks again, as per your advice I pulled the lowest CE resistances Q4/6/8, and the shorts appear to be gone. Q4 and Q6 are the Mj15023's and Q4 is shorted CE, while Q6 and Q8 (Mj15022) are shorted CE and CB. Will replace and see whats up.

Should I replace the driver transistors (Mje15032/Mje15033) as well to play it safe? They only cost $1.50...

[Enzo]

For a buck and a half, why try to be cheap, swap them. And don;t forget those resistors.

[tedmich]

the resistors all test (and look) good, but I will replace all the EL caps (except the 8 x 2200uf @ 100V PS caps)

There is even a 4.7uf 16v NP EL in the signal path from each op amp, but luckily I have some nice little Wima 4.7uf 50v film caps I can swap in.

[J M Fahey]

Enzo, excellent tip !!!!!!!!!!!!!!!!!!!

[tedmich]

according to Bob Lee, Applications Engineer QSC Audio Products, LLC
Secretary, Audio Engineering Society

the later model Series One 1400 are identical to the USA900 models, even though the PCB looks different.
see: QSC Audio Forum • View topic - QSC 1400 / USA 900

so the Series One 1400 model I am working on actually puts out 20% more power (240w RMS/8ohms vs. 200w RMS/8 ohms on older 1400's)

None of the QSC schematics for the 1400 reflect the actual circuit: the USA900 schem does!

[zero cool]

I have often found one of the bridge rectifiers to be blown in this series of amps. check those too. PITA to get off the circuit board. need a big iron!!!

[tedmich]

I have often found one of the bridge rectifiers to be blown in this series of amps. check those too. PITA to get off the circuit board. need a big iron!!!

thanks ZC, I initially thought it was shorted as the power transistor short looks the same, pulled the shorted channel FWB with a 50w iron, all is AOK. Those FWB have the rarer spade terminal orientation; most have the 3 parallel and one perpendicular
but luckily Vishay IR makes a series of 25/35A FWBs in this configuration, Newark has the 25A 400v units in stock, and they can work with a heatsink


the 35A versions are rarer, mouser charges $10 each

I put these TO3 heatsink on the bridges, with heat transfer compound after lapping them flat and going through the anodization and down to raw Al. I added a little silicone RTV to lube the sheet metal screw threads. The screws are stubby 3/8" automotive sized (12?) which fit the plastic lined hole on FWB perfectly. I ground the screw tip ~1/16 off to be sure it doesn't come close to the PCB and used a belleville spring washer to give even tension; they should cool well.

[Enzo]

So the circuit is the same as the 900. Fair enough. We know the board layout is different. The circuit may be the same, but are the part number callouts the same? (ignoring the layout)

[tedmich]

So the circuit is the same as the 900. Fair enough. We know the board layout is different. The circuit may be the same, but are the part number callouts the same? (ignoring the layout)

I believe they are, this is a table for the PS R7 values from the USA900 schematic; notice the various models are listed as "1400"


of course the jack panels are totally different; I'd much prefer the USA models switchable LP filters to the old Series One octal accessory sockets! The USA back panels are actually dual sided PCBs while QSC kept single sided main boards for a while longer...

[tedmich]

Success! The 1400 lives again! Thanks Enzo!

Will do some power testing tomorrow as the Bach Toccata and Fugue in D minor was a bit much for the neighbors at 10PM, hell even the pink noise was too much!

Glad I pulled the drivers as the heat sinks were hiding some MJE15030/1's which are rated 100v under the recommended parts: MJE15033/2. They were also very brown and dusty.

Just to make it easier for the next guy I stuck a corrected schematic inside the bottom cover...

Thanks again all, a quality ~1HP stereo amp off Ebay lives for ~$100; no wonder the same basic design was in production for 17 years (1983-2000)!

When (if?) its PS caps go, I finally found some decent ones that will fit (25mm dia / 28mm max!): ECO-S1KP472CA Panasonic 4700UF 80V ELECT TSUP (25mmx50mm) , they'll give 215% the old 1400 capacitance and 50% more than the USA900. Their rating (80v /surge 110v) should handle the ~75v rails AOK