Pioneer QX-747

[shooter]

after a year the owner finally dug the amp outta his attic and onto my bench

says it cuts out, which it does, also won't play through the rear speaker or rear headphone jack
otherwise, appears the inputs are ok, tested AM,FM,AUX, will get to the rest in time.

I started on the PA board, found rail volts within reason, all 4 channels have Vac in and out, even when speakers intermittently "toggle" off.  Will be trying to "understand" the protection circuit next.  The manul is 82 pages, but I'm off to the printer in the AM to get the important pages printed.  here's what I have so far;

Very well engineered box, all wire wrapped   I no longer have my tools

[PRR]

When it "cuts out", do the relays clack?

What is the DC voltage on the output, no-signal?

How old are those electrolytics?

[shooter]

the relay does "kick-in", appears there is a fault, or my leaning a perceived fault.
I gave up before I made it in/out of the power transistors, but the drive to them looked "good enough" for now.  Audio is fine out the front speaker jacks.  there is switching and conditional requirements to get sound out the back set.

I'm confident I am the 1st tech inside since it left the factory in the 70's? 

my game plan is to blow out the minor dust, wash down all the mechanicals with cleaner, finish tracing the signal to the speakers.
All the DC, and AC are clean and stable, so once I get it operating normal, I'll probably just recommend run til fail. 
Probably only get an hr a day inside til Friday.

[PRR]

The protection circuits separate quick transients and audio signals from real problems with cap filtering. Failed caps will give erratic triggering.

[shooter]

Failed caps will give erratic triggering.

Thx, I'm gonna work the AC out the speaker jacks n see why the rear Never work, while I'm doing that I'll clip my meters to the rails and capture min/max.  Tonight will probably just be herbal beer and wire tracing, the backs pissed that I worked like a 30yr old.... from OUR generation

[shooter]

haven't made it to speakers, yet.  I updated the protection board with what trips it's trigger, and where it comes from, mostly.  I probed what is shown as 14.1V, nothing there, 0-20mV AC or DC.  did the same on the 13.9 line, there I had small audio, varies with volume, forgot to see if there was a DC offset.  I'll try and get a full DC check next.

The relay trips, with vol @0, no audio, max vol, any input selected.  while I was watching the 14.1 and the relay kicked, nothing changed with the 14.1 line.  the 13.9 had a transient spike, probably the relay emf, it doesn't appear to have a blocking diode, by design.

[shooter]

  was measuring the 14.1 at the wrong place
all the DC values are within reason on the protection board;
14.1 = 14.5
13.9 = 14.4
these are the "most" out, everything else is within a tenth V
hopefully I can monitor the 4 input transistors and see if any trip up the circuit when the relay kicks.
also gonna see if there is a cal to balance the 4 amps, 1 ch is out enough to raise suspicion, although, my thinking 3 are out since with 300mVac in, I get 290mV on 1 and 170mV-ish on the other 3

[shooter]

  I fixed the NO rear speaker problem, I read the manual
behind and under the AM antenna is a matrix plug, set for 2CH!  the schematic makes it look like a switch, which there is a 2ch/4ch switch.
hopefully i'll get through the cal's for balancing channels soon, guessing it's for baseline noise?, you adjust with no speaker and input DL with a 5.1k R.  there's an internal MV?, plus 8 pots for the 4amps

[shooter]

Cal's done, everyone was close.

spent an hr monitoring the protection board,  the relay driver, Q9, or associated parts seems to be failing.  Q8's base stays rock solid when the relay kicks out, the 14.1, about the same, twitches a couple tenths.  Q9's base drops out, anywhere from 0 to 3V then acts like it "wants" to charge back up.

I have some new 2n2222A, metal top as a close enough for test, scrounging for a new 100uf.  I'll probably replace C8 also since the -5.1 is reading -4.2, and by the time it gets to Q8's base it's -1.2v instead of -2.0

EDIT:  any cons about adding a blocking diode across the coil

[sluckey]

EDIT:  any cons about adding a blocking diode across the coil

Nope. Unless you connect the diode wrong. Then you will definitely be changing a relay driver! 

[shooter]

Thank you kind SIR. 
my changes didn't change anything, went back to the 4 ch in, only measuring AC. monitored the base of the 4 transistors, CH 4 seems suspect, but then the Wife intervened, will disconnect that input, and any others n see where that leads. realistically, Monday/Tuesday 

[PRR]

It is unlikely the transistors failed.

It is very likely the electrolytics are dried-up.

All through the amp!

While it is a beautiful machine, you have to wonder if it is worth a total re-cap, and labor.

[shooter]

While it is a beautiful machine

It is indeed!  well documented, well built, quality that you don't see much these days!
It probably is NOT worth re-caping it!  there are just to many, on the other hand it works and sounds great, except.....
my strong suit in my past life was to "work around" whatever I had to get the system up, short of real safety, fake safety was just fine to bypass

I did pull ch 2 & 4 feeding the protection board, problem still there, If I get 1 & 3 pulled with same results, I'll terminate those inputs just to make sure, then have a discussion with the customer.  This is more about borebum & fun than money and time for me

[shooter]

Trying to understand the circuit operation;
(Ref PA and Protection schematics above)

Here’s what I think I understand ? So far;

Balanced current flow “above n below” 0, yielding a constant Voltage drop, making Q1 happy.

Increasing current and holding R constant, makes Bigger voltage drop, causing Q1 to conduct?, dragging 14.1 down?

Thinking AC signal isn’t monitored with this circuit?

I did disconnect all 4 inputs, got 1 trip early, then 1hr+ of Christmas music, then Rear speaker started getting spitty/poppy, shut down, no heat anywhere.  Content the answer is as PRR called it, Caps.

(original config, fail rate was between 30 sec to 5min)

[PRR]

It does read AC signal.

Omit the 15K resistors for a moment-- assume the output is shorted, no voltage swing.

Current into the short causes voltage in the 0.5r resistors. Since it runs near class B, only one resistor at a time conducts.

If the sum of the voltages on the 0.5r resistors reaches 0.6V (after charging the 0.22u cap), it triggers, which ultimately cuts-out the speaker relay.

So the max current into a short is 1.2 Amps.

The 15K against the 3.3K allow higher current when there is output voltage swing. More voltage in the load is less voltage across the transistors, they tolerate more current. I've never seen quite this method and not going to analyze it today.

Resistor failure is unlikely (another reason I am not going to study resistor ratios).

If the 0.22u caps have gone dead, quite short-time transients would trigger the protection.

There is *another* protection system, Q5 Q6, which watches Average DC Level. All outputs are summed in 47K resistors, then filtered with 470u caps. This is a couple seconds. If this sum gets to + or - 3V (approx), shutdown is triggered.

If the 470u caps are dead, shutdown could be sudden, even on quite small signals.

(There is a flaw in this scheme. If Left goes to +10V and R goes to -10V, the sum is zero. OK, exact-opposite errors are unlikely.)

[shooter]

not going to analyze it today.

No reason to, you have helped, n called it well, 

I was guessing Q5 & 6 monitored the 7.5VAC (pin3), via the -.7, and Q7 dealt with the 30VDC, but I do see what you're saying.

[shooter]

so I'm putting together options for the customer, I'm looking at the PS board deeper than the output voltages when I notice the fuse ratings don't seem to have much correlation to "typical" circuit draw listed on right side

[PRR]

LM3886 will take that voltage, has much better protection built-in, and is available full-kit (with PCB) very inexpensively through eBay. You have the PT. See if LM3886 can be mounted on the heatsinks, and the PCB fit conveniently.

[shooter]

LM3886

I was noodling around with LM3900 for the protection board last night, I also have a good working 60W SS stereo PA brick, but I know the customer, Greg, his summer beater is a '59 ford hard-top convertible original stock..
He does restoration work, but only original equipment, even if that means having a new vein machined for the original water pump kinda thing.

My suggestion will be keep it as a keep sake, If you REALLY want a working, original (close), buy a donor, I'll rebuild the boards for a flat rate + parts.  We'll see, I have a sticker on my beater camper that says "a fool and his money are great customers:).  I also have one that says "apple pie makes you sterile" but that doesn't pertain to this discussion

[shooter]

The owner opted for shelving the amp, I suggested putting an elf on it
I also secured conjugal visits

Anyone that has more money than sense might consider a re-build of probably one of the better examples of great design from the '70's

[shooter]

I took the amp back, spent an hour looing at hard top convertibles in various stages of restoration.
Greg ask "what it would take", I gave him my "best guess", He said yup, sooo.. I will be doing a restoration
Probably pick it up in Feb sometime when I get back, and hope to return it next Christmas

[shooter]

should be getin the amp soon

got a ballpark ecap list for  the audio/ps/protection AND a wirewrap tool!
if you're curious, it's worth finding the pdf and seeing how old stuff was done well from the draftsmen to the full complement of engineers

[shooter]

not sure what I was doing when I placed my order

Did get 12 caps swapped out, I been soldering turrets to long!  They started looking better the more I did, long way to go!.
38 more caps on had, but I think I'll re-count what I should have, and order what I missed:BangHead:

[shooter]

Got the protection board re-capped.  time to dislodge the PA board, only 4 caps, AND I have them

[shooter]

Swapped out 4 more, tried my beta tester I thought I fixed last year.

all the transistors tested the same, max beta, and max leakage. owners manual does say in-circuit is probably only good for open/short conditions.  I'll probably do some no power meter checks since it's laid open

[shooter]

Had to take the faceplate off to get to TS / control boards, my 1mm hex was to small, my 1.5mm was to big!
went to Napa, asked for a miracle and got one   They had a 1.3mm.
So I have replaced all the caps in the audio path, >50  .  Gonna button up enough to power up and see. 
I left the PS alone for now since it was not dropping out, all volts were good.  Once the audio circuit is up, I'll re-cap the PS

[shooter]

I’ve been looking over the TS, below.  I do ok till I get to Q5/6 collector, then turn left.  Is C17/18 acting as NFB?

Any major problem using R33/34  as ½ a Vdivider feeding a tubes grid?  My next build is gonna be HiFi-ish and I'm bonding well with this SS project


Thanks

Dave

[shooter]

Fired it up, my heart sank!!!!  all I got was KA..KA...KA..
switched input sources and WALLA star spangled banner was just starting on my country station

30minutes and counting!  I'll get with owner and get a 2nd "draft" to re-cap the PS, possibly matrix/decoder boards also, then run full set of AC/DC checks and call it good!!!!!!

[PRR]

> Is C17/18 acting as NFB?

The whole tone-pot network, Rs and Cs, is a Baxandall.
http://www.learnabout-electronics.org/Downloads/NegativeFeedbackTone.pdf

[shooter]

Thanks PRR,
I buy like 2 things a year from ebay, the 1st this year, someone selling that board from a QX-800, which has L/R controls on 1 board instead of 2 like the 747.  If the pots alone are good, I got a good deal, they are 10 click detent.

at the 42min mark the amp glitched out, rear R channel, repeatable, so it's scope time.  before capping, any, or all, or random, so progress.  I still need to cap the matrix board, I'm hoping that's where I find the issue, I do not want to get tangled up in the sub-channel board.

[PRR]

After 40 years, I *just* realized why Peter used a center-tap pot. Without that, there is no DC path for the tube grid.

Nearly everybody else used a no-tap pot and threw a 1Meg on the grid to ground. Performance suffers very slightly but part-availability is so very much better.

Your transistor version has to add several parts for bias. Same idea, different details, and looks like 1/10th the impedance basically reflecting ~~1/10th the supply voltage with similar current.

[shooter]

looks like 1/10th the impedance basically reflecting

that's in my "next week" re-learning class
I still lack in interstage impedance matching, and TS impedance calc's are still incomprehensible.

I shot the main problem to channel 4, PA board, driver section.  got out my canned air, upside down, and when I goosed Q 7 & 8, boom, channel dropped out, slowly recovered, Q12 gets all scratchy, Q16 could care less, got no reaction from anything downstream including the big transistors.

I'll probably order all 4 transistors, coupling caps, and diodes, should have any R I might need.

I have the Caps to re-cap the PS board, so while I'm waiting, I'll re-cap
 

[shooter]

PS, re-caped.  found an error on the silkscreen with a cap PLUS! sign.  Per Sluckey, I have been photo'n  , that, the schematic, and parts foil-side, all say silkscreen is wrong.

2 hours till glich   least I could pick the music!
phono in is bad
phono into AUX works
FM works
noise floor is to high, all channels equally, about twice what I'd expect as '70s normal,  post Volume, where I've been playing  , scoping tomorrow

[shooter]

got ch4 working, no glitching, or dropouts.  Did trip all channels twice over about 5hrs, both times very close to power-on.  and of course I wasn't monitoring the B+/- rails

The 4 main "HV" caps are still original, monitoring the rails, end to end, for hours using min/max shows AVG ~59.6vdc rail to rail, min ~56.9, max 60.1  When the volts do slide down, AC ripple increases.   So I think i'll order 'em up just because, button it up and let the owner evaluate it over time.

[shooter]

I miss the days of self calibrating n diagnostic systems 

Went back to basics, confident the rail volts is ok, for now
Found the 30vdc failing, found an off board filter cap
Hopefully in the morn I can figure out whether it’s the supply or load

I opened the audio path at the control/TS boards, jacked audio in and it’s LOUD
So once I get the 30vdc sorted I’m back to the left half the amp.

aside; the owner's hobby is factory original restorations on 1958 convertible hard tops, has 5 in his shop
when I explain where I'm at, I just get this cheezy grin!   and I just say, ok, i'll keep at it

[PRR]

> restorations on 1958 convertible hard tops

He has good reason to understand why you can't take 10 cents an hour.

(My neighbor did a long-stored Hudson last year, and getting the money was worse than getting the clutch out.)

[shooter]

why you can't take 10 cents an hour.

  out in the sticks barter is still a method of payment, and he has a lift, heated garage......

[shooter]

finally got to use the wrap-unwrap tool   works just like my brain remembered 

Isolated the 30vdc supply, tripped out after an hour or so, love my fluke's min max feature, need new batteries

put another shopping list together, least most of the parts are pretty inexpensive.  still gonna ohm each of the 30v taps over time and see if there's a loading issue.
the 30v is notated 200mA, ohm says 150ohms for max I so I'm guessing many times 150 per tap is a good threshold

[shooter]

rebuild the 30vdc supply with xref parts, didn't work 
Soo, I went the LM317T route, no glitches in 30min play (4 hours bug free is my goal)
monitoring the 317, running 29.86vdc solid, 186mA, (the 6 goes to 8 occasionally, but settles back to 6)

while waiting for gremlins I still gotta chase down hisssss

inside the red has been replaced by the 317

[shooter]

found the hissss, ch 2 & 3 on the "driver 1/2" the PA. 
still gotta re-work PS so I don't have gator clips and shorting hazards.

I just picked up a pair of Bose 10.2 II speakers, looking fwd to warming them up through this amp, IF I ever get the bugs out