Tannoy Precision 8D active studio speaker dead...

[LektroiD]

Tannoy have ceased production of this model and will no longer service them. Neither will they release a service manual for this model. So I'm stuck with a pair of expensive studio monitors that serve little more purpose than a door wedge.

So far I have replaced the regulators, the rectifier, the LM3886 power amp chips, but can't seem to bring it alive. The driver is fully working, as I tested it with the other amplifier from the other speaker. I just don't know what to replace next.

Any ideas where else I could be looking or where to test next? I'm at a dead end here.

[wraper]

Measuring voltages? What is on mute pin?

[wraper]

testing != blindly replacing the parts.

[nitro2k01]

First rule of troubleshooting, thou shalt check supply voltages!

[LektroiD]

testing != blindly replacing the parts.

It's the only option, since there's no schematics or service manuals.

First rule of troubleshooting, thou shalt check supply voltages!

I have no idea what the supply voltages should be. The correct voltage is coming from the regulators, but I have no idea how it is distributed after that, as I have no info...

So really the only way I can go about this is by blindly changing parts that may be damaged. Just wondered if there was a common problem with active speakers, or if anyone knows about this particular speaker?

I have no idea where to go next, which is why I'm asking here...

[krivx]

I have no idea what the supply voltages should be. The correct voltage is coming from the regulators, but I have no idea how it is distributed after that, as I have no info...

You can figure out likely supply voltages from the IC datasheets and ratings on the filter capacitors. You don't need to know exact values, it's unlikely a voltage rail is a little off. It's fairly likely a supply has been pulled low or disconnected somewhere. What's are the voltages at the power amp IC supply pins?

[wraper]

testing != blindly replacing the parts.

It's the only option, since there's no schematics or service manuals.

wtf LM3886 datasheet is for? I have a schematic for less than a 5% of devices I repair, still able to measure somehow.

[wraper]

And yeah, replacing diode bridge??? That can be checked for sure. Just by checking AC voltage before and DC after, checking internal diodes with multimeter. Is there some voltage after diode bridge on filter capacitors at all? With such an approach you may end up reworking all pcb with just a blown fuse at fault.

[Yago]

Compare voltages/waveforms etc to the working one?

[Lightages]

A friend of mine had a bad set of Tannoys. After looking at his pictures of the circuit board I suggest resoldering every joint. They were very poorly built. This fixed his problem.

[simingx]

And yeah, replacing diode bridge??? That can be checked for sure. Just by checking AC voltage before and DC after, checking internal diodes with multimeter. Is there some voltage after diode bridge on filter capacitors at all? With such an approach you may end up reworking all pcb with just a blown fuse at fault.

... or a faulty power switch, blown line fuse, defective power cord etc...   

[Riotpack]

a chip amp like this should be simple to repair.
Start at the power supply, is there both positive and negative voltages on the amp IC?
Is there any protection replays after the IC going to the speaker? (download the datasheet and find the pinouts) Can you input a signal straight into the chip amplifier input pin? (lift factory coupling cap and feed signal in to test)

Usually after confirming power supply you simply start at one end and work through the amplifier stages until the signal ceases.

[Riotpack]

Here are the schematics for the 6D which uses the same chip amp - I am unsure of the differences but it may help.

[LektroiD]

Here are the schematics for the 6D which uses the same chip amp - I am unsure of the differences but it may help.

Thanks for the file, not sure how you managed to find it, but I wish it was as simple as the 6D, which only seems to have 2 power chips, there's 3 boards which all interconnect on the 8D, one is the power, then what appears to be a filter board, and a digital board. Looks like a pretty complex circuit on the 8D

[LektroiD]

I've done some tests, the 7915 is taking -36.0VDC in and throwing out -26.66VDC. This is a part I have already replaced, so I'm wondering if there's something else in line with it that might be causing the spurious reading?

I can't do live tests for long as the 3x LM3886 chips are roasting without heat syncs. The rear panel provides heat syncing, and I can't access anything when it's bolted to it, as you can see in the pic above...

[wraper]

I've done some tests, the 7915 is taking -36.0VDC in and throwing out -26.66VDC. This is a part I have already replaced, so I'm wondering if there's something else in line with it that might be causing the spurious reading?

I can't do live tests for long as the 3x LM3886 chips are roasting without heat syncs. The rear panel provides heat syncing, and I can't access anything when it's bolted to it, as you can see in the pic above...

Either it is dead or you measure it wrong. Common (GND) terminal of the 79xx is not the same as on 78xx regulators and is not connected to the heatsink tab.
First of all check if LM3886 is using single supply or dual supply, if don't know how, measure if pin 4 is connected to GND. Single supply will have output electrolytic capacitors of 1000s uF too. Then measure (GND referenced) if there is a power on the pins 1,5 V+, and negative supply on pin 4 if it isn't connected to GND (using dual supply) . Also measure voltage on the pin 8 (mute). I'm asking this because if the controller mutes the power amp, you wont get any sound even if amplifier is completely ok.

[LektroiD]

Seems both the 5W resistors are short, and sit before the regulators. Now I need to source a couple of these in the UK... Anyone know where I may buy a couple?

Would using a higher wattage mess things up?

[wraper]

They do not become short, only open. They might be short only if there is something dead in parallel. unsolder and check if they are 100 and 220 ohm and if there still is a short without them.

[LektroiD]

After a few tests and comparing to the other speaker, there should be 36V/-36 going in and 20V/-20V coming out of the 5W ceramic resistors. On the faulty board, it has the same voltage into the ceramic resistors as what's going out (36VDC) - would this not suggest a short? Whatever it is, something is not right there... I'll desolder them now and see what results I get...

[wraper]

After a few tests and comparing to the other speaker, there should be 36V/-36 going in and 20V/-20V coming out of the 5W ceramic resistors. On the faulty board, it has the same voltage into the ceramic resistors as what's going out (36VDC) - would this not suggest a short? Whatever it is, something is not right there... I'll desolder them now and see what results I get...

That suggests there is no substantial load after those resistors, not that they are faulty. You shall check resistors by measuring resistance  . This only indicates that something is not ok after those resistors.

[wraper]

BTW, the problem might be artificially introduced as well when you blindly changed parts there. Compare with working one if you didn't solder wrong parts or in wrong direction by accident.

[LektroiD]

Aha! Resistor checked out ok, but the 100µF electro behind it had an ESR of 0!!!

Unfortunately I only have 100µF in 35V, which wouldn't last long with 36V going through them. I'll grab some on my next shopping spree...

[LektroiD]

*update*

I put the 100µF/35V in as a temperary measure just to test, now I have 24V going in, and 9.6V coming out, so looks like I'll have to replace the negative reg again (I have plenty of those in stock).

Stay tuned though, chances are, with that reg throwing out more than double the required voltage, I would imagine there may be further damage...

I'll order some new 50V or even 63V electro's (just to make sure) and go from there...

[wraper]

Aha! Resistor checked out ok, but the 100µF electro behind it had an ESR of 0!!!

Unfortunately I only have 100µF in 35V, which wouldn't last long with 36V going through them. I'll grab some on my next shopping spree...

What, 0? the less the better, don't know how precise your esr meter is thou, usually it should read about 0.1-0.5 OHm. However capacitor shouldn't affect anything at all in this situation unless it is shorted. However it definitely is not.

[wraper]

now I have 24V going in, and 9.6V coming out, so looks like I'll have to replace the negative reg again (I have plenty of those in stock).

It might be overloaded as well. The best would be to desolder it and check using some power supply to power it. Thou voltage before it should drop lower than in working amplifier. So likely it is blown.

[LektroiD]

Strange, I just fired the amp on again, and it is back to -36V going into the reg (and -24V coming out the other end... Hmmmm! 

[wraper]

Does positive regulator output normal voltage? Considering -24V instead of -15V, likely there is some opamp IC already blown too on that power rail. Where did you get that 7915, not ebay by chance? Do you still have original IC?

[LektroiD]

The positive reg outputs 15V as expected, although it has 36.0V going in. Unlike the other speaker which at that point has 20V going into each of the regulators. I would assume it is one of the components before the regulators which is not functioning properly. On the working speaker, the voltage going in to the ceramic resistors is 36V but lowers the output to 20V, on the borken one, there is 36V coming out of the resistors...

I got the regulators from Farnell in the UK (I have a good stock of components from reputable sources, as I'm building a lot of audio/music projects).

[wraper]

The positive reg outputs 15V as expected, although it has 36.0V going in. Unlike the other speaker which at that point has 20V going into each of the regulators. I would assume it is one of the components before the regulators which is not functioning properly.

I'm 99% sure that everything is ok before regulator IC, however it have a very light or no load instead of normal. Maybe opamps already blown because of the increased voltage on negative rail, so no load left.

[LektroiD]

hmmm looking at the PCB picture. the screw at the top left, corrosion? something leaked onto the PCB?
may i suggest if the leak somehow screwed some of the components? or components screwed = leaked something? capacitor?
any other dried fluid trails ?

more pictures pls :p

It is covered in various coloured glue, some soft & gooey, some hard and brittle. I have cleaned most of it off. This is Tannoy's ridiculous idea of PCB design - whatever possessed them? It has made it really hard to identify & solder/desolder components.

That said, there are a few traces where the solder mask has worn down to bare copper, but I have probed them and the continuity is ok.

I'm 99% sure that everything is ok before regulator IC, however it have a very light or no load instead of normal. Maybe opamps already blown because of the increased voltage on negative rail, so no load left.

I'm a little confused: The voltage drop happens at the ceramic resistor on the working unit, this is before the regulator. What components after the regulator could change the input to the regulator on the faulty unit?

The thing that concerns me is neither the negative or positive rails are functioning as usual. I would understand if it were just one of the rails.

PS. The schematic for the 6D on the previous page has a similar input section to the 8D, but missing a few components. For example, on the 8D there are two 10n ceramic capacitors at the input between ground and each of the rails, this is not shown on the schematic.

[wraper]

Those resistors are there to reduce power dissipated on the voltage regulators. when there are significant load, voltage will drop on them. As input voltage of the voltage regulator reduces, power dissipated on it reduces too. Blown load will impact both regulators as most of the load likely are powered from both of them. Dual +-15V power usually is used if there are opamp ICs.

[LektroiD]

I understand why it uses a bipolar supply, just surprised both rails are having the same issue, I'd understand if it was one or the other, but both regulators are fed over-voltage.

Anyway, here are a couple of photos to depict what I'm working with here... As you will see the board is covered in nasty glue/gunk.

[LektroiD]

Top view...

[wraper]

As I see from the pictures there is nothing that regulators could power on this board. Therefore, either another board was disconnected (power goes to it over the cable) while you made measurements or another board is already damaged.

[wraper]

Likely this particular board have no issues other than blown regulator.

[wraper]

I understand why it uses a bipolar supply, just surprised both rails are having the same issue, I'd understand if it was one or the other, but both regulators are fed over-voltage.

They are not fed with overvoltage. There is no voltage dropping on the resistors because of the absence of load after the voltage regulators. 36V is completely ok for 7815/7915. As I said, both rails have the same load which is powered between them.

[LektroiD]

As I see from the pictures there is nothing that regulators could power on this board. Therefore, either another board was disconnected (power goes to it over the cable) while you made measurements or another board is already damaged.

Yes, there are two other boards powered by this, a filter board stuffed with LM837's and a digital board for S/PDIF input.

The only thing on the power supply board that might draw power is the chip amps LM3886. These all warm up when power is on, in fact they get very hot quickly, but that's probably the blown reg causing that. The fact that they heat would suggest to me that they are working.

I'll swap out the reg and see what results I get...

[LektroiD]

Regulator swapped, same results! This time I made sure all the boards were interconnected to ensure there is a load, but still I have ±36V coming into both regulators instead of ±20V

[wraper]

Regulator swapped, same results! This time I made sure all the boards were interconnected to ensure there is a load, but still I have ±36V coming into both regulators instead of ±20V

So did the output become -15V as it should be? As I said, another board(s) which is powered from this regulator is dead or some wire is broken. It was obvious from the beginning that replacing regulator IC wouldn't change this in any way. Check if there is +-15V on power pins of the opamps located on two another boards..

[LektroiD]

Still the same output from the -reg as before (these are rated for a max 35V, so maybe it's unhappy about receiving 36V).

I tested the pins 2, 6, 9 & 13 on a few of the LM837's, no negative voltage going in at all. I also checked the interconnect header, there is no negative voltage there either. Looks like something is open somewhere...

[wraper]

but ... from the picture i could not make out how the -15v is going to the cables ...

Obviously on top layer  . My guess if voltage regulator is ok, there might be something in parallel to it what is leaking a small current. As load is absent, voltage goes up.

[wraper]

Capacitor won't cause this. It could leak only to GND, therefore no voltage increase possible. IMO that is likely not fault, the fault is absence of load and voltage will return back to normal when load appears again.

[Refrigerator]

You should take both of those big caps and clean all the snot around them because that glue is known to absorb moisture and become conductive/corrosive ( i see there is some of that on component leads and it seems pretty bad ).

[LektroiD]

took C22 out to check, ESR was 0.5r, I swapped it with a newer cap anyway. Going to clean off the remainder of the snot around the components to see if that reveals anything, although the rubbery stuff in places is near impossible to remove...

[wraper]

Are you guys crazy or what? Leave that gunk alone. It is used for holding parts, if you remove it with brutal methods, PCB won't look any better and it is not a cause of increased voltage. Even if it is low quality and became slightly conductive, it's resistance would not be nowhere close to be able increase output voltage by a tiny bit. IMO it's pretty useless to do anything on power amp board at all for now. To check if voltage is regulated correctly, solder something like  0.5-1W 1k resistor between regulator output and GND. If it returns back to -15 volts, just leave this board alone  and search why there is no power on another board.

[Refrigerator]

Here.
https://www.youtube.com/channel/UCc--CFzOjEViwUwrMCOGVhg
Go to this channel and watch all the audio amplifier repairs, this should help.

[LektroiD]

OK! I'm back after the weekend, I wasn't AWOL just a busy weekend, and with nothing to report, I didn't see the need to post... I have every bit of enthusiasm to get this speaker working once again, and appreciate everyone's keenness to help out too, very much appreciated. I hope all you guys had a good weekend too...

The latest:
As suggested, I put a load resistor between the ground and output of the regulator... 10K was the nearest 500mW resistor I have in stock, if 1K is preferable, you'll have to wait until I can source one, as I'd imagine a 250mW would smoke, although I can try if you think it would make a difference. Anyway, now I'm down to -20V out of the regulator, but the input side remains at -36V (not -20V I'm getting on the other unit).

Here's a few pics of the filter & digital board. - Ignore the trimmer in one of the pics, it hasn't fallen out of anywhere, I just used it to prop the board level so I don't get any weird depth of field issues...

[wraper]

Anyway, now I'm down to -20V out of the regulator, but the input side remains at -36V (not -20V I'm getting on the other unit).

Forget about input side voltage, It will drop with proper load. 2k 0.25W would do as required power decreases with resistance increase. First of all you should check is there is -15V (-20 or whatever  ) on the connector of the other board which is connected to the power amp board. Or just check continuity. It might be just broken wire or trace, This 47R resistor looks suspicious too. It has either broken terminal or there is solder blob on it, check it.

[LektroiD]

1.5K load resistor in, and we're down to the required -15V on the output.

there's no -V going out of the header from the power board, only +V.

There were a few broken wires which is why you'll notice 3 wires soldered onto the back of the header, but these are all fine now.

The 47r was just covered in Tannoy snot, tested in circuit: 46.7r

[wraper]

Then there is broken trace on the top side of the power amp pcb. I see some suspicious things between tho voltage regulator heatsinks (might be broken trace).

[wraper]

Check marked places.

[LektroiD]

Well spotted!

I have rerouted the trace and we now have power, proper input and output voltages at the regulator, expected voltages at the header, and the power LED on the front of the speaker is now operational once again.

However, no sound... I'm going to check out all of those opamps on the filter board next (unless there are other suggestions)...

[wraper]

Well, one of my first suggestions about LM3886:

Also measure voltage on the pin 8 (mute). I'm asking this because if the controller mutes the power amp, you wont get any sound even if amplifier is completely ok.

[LektroiD]

Mute pin ( is just bouncing around a few .mV. Same response on all of them

Pin 4 and 7 both go to ground.

[wraper]

Are you sure that pin 4 goes to GND? It is negative supply and should not be connected to GND. Even in single supply connection it cannot be connected to pin 7 in principle. Also I see thick negative supply trace on the top side of the PCB going behind LM3886's along the edge of the PCB.

As of considering normal dual supply connection (but check first pin 4, there must be -35V, must measure continuity to ceramic resistor on 7915 input):
Voltage on mute pin must go below zero to approximately -2.6V according to how I understand datasheet (it is current, not voltage driven). If current is not sinked, voltage will stay at zero and mute is on. I suppose all those pins are connected to one point, likely through resistors and further go to connector. Check where traces go and parts connected there.

[LektroiD]

Just tested with the front soft power switch connected and activated, there is -1V going into mute, and pin 4 was only earthed while there was no power, it is in fact, -36.6VDC

[wraper]

Just tested with the front soft power switch connected and activated, there is -1V going into mute, and pin 4 was only earthed while there was no power, it is in fact, -36.6VDC

That means it was not earthed, you just made wrong measurements somehow. I guess LM3886s should be working then. I would try to put some audio signal directly to power amp pcb. Maybe just touch audio input with tweezers or small screwdriver while touching metal part of it and listen if the hum appears in the speakers. Just be very careful to not short something, or some part may explode. I'm going to business trip today, so my further help might be delayed if any.

[LektroiD]

It's not easy to do any tests while the speakers are connected, the extra short cables are fixed into the cabinet (I don't want to tug at them too much as I know how flimsy these things are). The back needs to be virtually closed before I can attach the cables. This is one of the worst designs I've seen when it comes to maintenance, not to mention all the snot the boards are covered in. Considering these were £1000 each when new, I certainly won't be considering Tannoy when I next upgrade the studio.

Hopefully this pic will give an indication of the constraints I'm working within...

[wraper]

I see you have oscilloscope. Connect it instead of speakers.

[wraper]

Or you can check with oscilloscope if signal is present on power amp input instead. If not, check preamp.

[LektroiD]

Good call... I'll feed in a 1Khz square from my FG, see how far into the circuit it goes.

[LektroiD]

Fed a 1KHz square in at +/-1V, I got a sine out of one of the chip amps (pin 3 - output), the others displayed nothing. Then, due to the tight space I'm working in, I managed to short out two of the pins on one of the other chip amps, nearly welding the probe to it.

Now the sine I'm getting is modulated with another lower frequency sine, this added carrier sine is present on nearly all the connections of the chip amps.

[Yago]

Is the lower freq 50Hz?
Could have knackered a diode or cap in the PSU section and AC is getting through.

[LektroiD]

Back after waiting for some new test probes to be delivered. I seem to have misplaced my Tektronix probe adaptors which are safer for probing chip pins - those standard hook things are too clumsy for chip pins. Anyway...

I seem to be getting a waveform back into the power board, but I'm getting strange results from the chip amps...

There is a waveform going into two of the chip amps, but the output is flatline (zero/zilch/nada).

Mute pin has -1.06VDC going to it

I hope the sketchy illustration clarifies what's going on.

[Yago]

Saw this and thought of you Lektoid! :
http://www.ebay.co.uk/itm/Tannoy-Precision-8D-Amplifiying-Unit-tested-and-working-/191411958647?pt=DE_MusikInstr_Lautsprecher_Monitore&hash=item2c910ab377

Might be handy for the future to have a spare!

[LektroiD]

Saw this and thought of you Lektoid! :
http://www.ebay.co.uk/itm/Tannoy-Precision-8D-Amplifiying-Unit-tested-and-working-/191411958647?pt=DE_MusikInstr_Lautsprecher_Monitore&hash=item2c910ab377

Might be handy for the future to have a spare!

It would be perfect, except shipping only in Germany. 

[krivx]

I would contact and ask about shipping. Most sellers will, especially if you have good feedback already.

[LektroiD]

I asked about shipping, and he got back to me saying it would be somewhere around €20 depending which courrier, I asked if he could give me the exact shipping cost so I could buy it and not had a reply.

I would rather get the unit I have up and running to be honest, this evilBay thing is like chasing my own tail.

Anyone got any suggestions regarding the tests I did today?

[LektroiD]

may i suggest, since you did find a PCB track problem. would it be good to first completely eliminate the rest of the "spots" so that then you can maybe repair all the possible broken PCB as round 1? in a 25 yr old stereo drawmer compressor, i had to isolate channel 1 from channel 2 in order to find and repair all the PCB track related problems, maybe similar problem in this PCB too. so did that gunk actually corrode the track around the regulator? or it was something else?

Looks like I ripped the trace up when I swapped out the regulators. That's fixed now and I haven't noticed any other problems on the PCB

[LektroiD]

R23? I just checked for continuity, and yes it's tied to ground

[LektroiD]

did you managed to fix the amp? just curious ...

I battled long and hard, but eventually caved in and went for the evilBay option. Now the other speaker seems to have a slight issue making intermittent noises (without audio input), probably one of the crap caps drying out. I'll have a look when I can next muster the strength to get it off the stand (they seem to weigh about 15 tonnes, and I can't lift too much after a car accident last year).

Anyway, here they are in all their glory...

[pinyoro]

Beautiful studio!!

[LektroiD]

dang man .... if i were you ... i would try to get my hands on a meyer HD1 (or something similar analog sounding, silk tweeter etc) ... your gear seems to have so much analog sound in it ... you definitely need a better pair than those imo

btw ... that akai sampler with the floppy ... does it still work?

I use the analogue inputs on the speakers, not the digital.

After nearly 25 years working as a studio engineer, I found the dual concentric design much easier to find the sweet spot than separate tweeter/driver combination, which is why I have favoured Tannoy for many years. Up until the 8D's were working I was using my System 800 passives (which are now on back on my hifi separates next door). Once you have tuned your ears to a particular brand, it's absurd to move to another; just to spend the next 6 months tuning your ears to them. In many cases, speaker placement and room mode tuning is much more important than the speaker itself. Therefore, money is much better spent treating the room acoustics once you have found your weapon of choice. If the Meyers work for your ears then that's great, but I have no plans to jump ship any time soon.

Yes the sampler is fully working, also has USB card, which apparently is quite rare... However, I was thinking of selling the sampler as I don't use it much.

[LektroiD]

I have a problem with the replacement unit. This is an intermittent fault so may be why it wasn't picked up when it was sold.

Occasionally (usually after it has had No audio fed into it for a while), it drops by a few decibels, the only way to kickstart it is to send a few short bursts of high volume audio to it. Then it works normally, but this is not ideal as I have neighbours who I'm sure don't appreciate bursts of high volume audio emanating from my property.

I have tried the amp in both units to ensure it is not a faulty driver or cable.

I'm suspecting a dead opamp, but I don't want to look into repairing just yet as I'm in discussion with the seller to try to resolve this, and the unit may need to be returned in the same condition as it arrived. I just want some ideas as to whether this may easily be repaired, and whether to ask for a part refund on that basis. Provided it can be repaired.

[KhronX]

On these models (actually, both Reveal and Precision 6D/8D's), the most common issue is one or both of the 5W ceramic dropper resistors ahead of the voltage regulators failing open-circuit.

The brown glue drying out and turning corrosive and conductive is a ticking time-bomb, and the switches can go wonky too.

Most recently, the 80Hz high-pass switch on one of my Precision 6D's had kicked the bucket, after i took the back off to take care of an intermittent contact on the tweeter output connector.
Nothing a squirt of contact-cleaner and 20-30 operations couldn't fix

You can find the schematics here:
http://91.121.194.115/tannoyschematics.zip

The filter board is almost definitely the same. The only (electronic) difference between the 6D and the 8D is that the 8D has two bridged LM3886's driving the woofer, where the 6D only has a single one.

I hope this doesn't violate any forum rules, but there's a considerable-ish thread on this subject over on the badcaps.net forums:
http://www.badcaps.net/forum/showthread.php?t=16644

[Messtechniker]

After reading all this, I am glad I decided at the time (mid 2010)
to go with the passive Precision 8s (without the very expensive "D")
for my voice recording studio and instead completely refurbish an old Quad 303
I happened to have. I fully agree - once you are used to concentrics
as I have been all my life starting with Tannoy HPDs, you would not
want to change.

Seeing such bad manufacturing, I would be inclined to completely throw out
this Chinese crap and roll my own electronics for these speakers.
Time permitting of course.

Anyway, I hope you get this sorted out.

Yours Messtechniker.

[KhronX]

In case you've read through the thread about this over on badcaps.net, rolling my own electronics is exactly what i plan to do in the very near future 

On the bright side, i caught my Precision 6D's for a steal (a bit under 200eu shipped), with one of them dead - i most definitely could never have afforded them (or been willing to pay that much, ie. retail or close) otherwise.

[Messtechniker]

Quote from page 2 of this thread:

The positive reg outputs 15V as expected, although it has 36.0V going in.

If this should be normal for this piece of electronics then beware!
The 78xx data sheet specifies 30 V max. for the 7815 as recommended,
and 35 V as absolute maximum rating. These figure could even
be much lower for LM78xx regulators of unknown origin.

Pretty bad by design....

[KhronX]

If memory serves, that reading was taken with no load (ie. neither the filter board nor the S/PDIF i/o board connected).

The analog filter board draws around 60mA from the +/-15v rails, and the digital board draws another 100mA or so - those white ceramic resistors will drop the voltage the regulators see at their inputs (even further).

Quote from page 2 of this thread:

The positive reg outputs 15V as expected, although it has 36.0V going in.

If this should be normal for this piece of electronics then beware!
The 78xx data sheet specifies 30 V max. for the 7815 as recommended,
and 35 V as absolute maximum rating. These figure could even
be much lower for LM78xx regulators of unknown origin.

Pretty bad by design....