STK Amp design

[toshiba tony]

Hello. Who's the audio gurus? Can't bring the make to mind but it uses a + and - supply. Fifty volts each side. There is a STK ic and at all levels there is instability. It sounded like the speaker cone rubbing but as it's an active PA speaker this took some stripping down. There is a voltage difference of 13v on one of the PSU rails. I always thought that in these types of amp the the outputs should be a lot closer. It's not Class D and each rail is clean. Could such a large (comparitively) difference in the rail voltage cause distortion? Decoupling caps are ok and even with the output chip removed its the same. One line 13v down. It's not SMPS just a torroidal. Cheers anyone.

[Michael Maurice]

The power supplies should be equal = and -. You must have 0V on the outputs.

Try removing the STk and see what the power supplies read.

In all probability the STK is faulty.

[kevinaston1]

What is the model number of the amplifier?

[toshiba tony]

Cheers both. Genelec 8040. I'm new to PA work but having worked for Sony and LG for many years I'm hardly a hobbyist. I would have thought that the split psu should be similar so I'm hoping once I balance these I may be ok. Both lines are clean and there's merely a centre tapped secondary and a bridge. The difference is odd. I even used to do H&H warranty work for a dealer. This is an odd one.

[jamesperrett]

Genelec 8040's are very high quality studio monitors. I must admit that I'm a bit surprised to find they use STK amps in them - do they look original?

[sortedradio]

I believe they use STK 442 amplifier modules in them, which are no more or less reliable than any other STK ic.

[toshiba tony]

When the owner told me how much they were it shook me, inside they are crap. A couple of tone correction ic's and the STK. I think once I sort the PSU out I may be ok. A 13 volt difference on the supply rails is baffling me more than the operation of the output stage. Straight forward centre tapped secondary onto a bridge and no regulators, just smoothers. The negative rail is about 40 and the positive is 28. That's with both rails unloaded. I once had an an H&H brought to me from an engineer. Strange PSU voltages, low distorted output. The centre tap on the transformer was floating. It was that nasty one with near mains on the illuminated green strip. Yes, it bit me. Can't remember the model but I remember the shape well.

[tony brady]

there was at least one Thorn amp where the centre tap of the transformer went to a 10 ohm resistor which then went to the 0V of the caps - this was to save on having fuses. have you got continuity between the transformer centre tap and the caps?

[McMurdo]

if the psu is ok, and the supply rails are unbalanced to 0V, either one side is more loaded than the other or you have a grounded bridge design, have you got the diagrams?
I'm assuming it doesn't use commutated rails.

[SiriusHardware]

Tony. do you have a dual output bench power supply on which you can 'Tandem' the outputs to produce a balanced + / 0 / - supply? See if the amplifier works as expected with an alternative supply?

With nothing at all connected to the transformer secondaries - not even the bridge - do you get the same AC voltage out of each transformer secondary (or out of each half of the secondary)?

You've looked at the smoothing electrolytics but are there also any low value capacitors (disc ceramic, etc) in the immediate area of the transformer secondaries or bridge(es)?

[toshiba tony]

Thanks all, I'm going to bring it in, I'm in the tv game, 45 years. The recording studio I've taken on like me to do repairs on site. Can't always be done. As I am new to them and on trust it is up to them, this balanced supply has thrown me. 0 centre tap is fine, off load it's still 13v down on one side. The circuit is so basic. I've been repairing Switch Modes since they came out, alongside PFC (active and passive) This one defies logic. Cheers.

[TonyDuell]

I assume you've checked the smoothing electrolytics. What happens if you swap them over, does the fault move to the other supply rail?

What about a single open-circuit diode in the rectifier? I assume it's a bridge rectifier feeding the 2 capacitors with the centre tap of the transformer going to the midpoint of the capacitors and system ground. One diode open would make one side a halfwave rectifier the other side full wave.

[toshiba tony]

Got it in the workshop, -35v cap defective. No hum on output though. It had to be something really mental, no swelling. ESR ok. I mean, what else could it be, cheers everyone.