Amp with high pitched static noise

[G6ONEDave]

Check the N channel fets as well out of circuit. Any faulty output device can and often does cause excess current flow.

Do you have access to an oscilloscope by any chance? It does not have to be an all singing, all dancing type. Anything that has a 5Mhz to 10Mhz frequency rating will do for audio testing. If you don't own one, perhaps you could borrow one from work, depending on your employment. Possibly with a work colleague to show you how to use it, if you don't know.

Dave

[spacebiscuit]

So, I took the N-MOSFET x2 out - both give me a respective reading as follows:

Vg 406 / 416
Cg 536 / 539pf
RDS 0.9 / 1.1 ohms

The tester I am using is a "T7 multi-function"

https://www.amazon.co.uk/ARCELI-LCR-.../dp/B07V688CJP

There are 3 other transistors I guess I should also test so I'll work on removing those next and testing.

i think previously someone suggested to check the power supply to the board, I get 240v on AC, I also checked DC and got around 100v. I checked the working/good channel and got the same readings (!). At the time of my c*ck-up I did switch the power supply over to confirm that I had not fried the psu.

Regarding an oscilloscope, unfortunately I do not have access to one, nor do I know anyone who has one.

[Diabolical Artificer]

They sound ok. You might as well test all the transistors (Q.s), however make sure you know how the fit on the board it VERY easy to put checked Q's back in the PCB the wrong way. Take some high quality closeup pics. Same goes for wiring.

Andy.

[G6ONEDave]

Is that +/-100vdc or in my opinion +50/0/-50. When measuring any secondary voltages in the amp, the black (negative) of the voltmeter needs to be connected to the 0v (ground/common) as opposed to the -50v or any other voltage point.

Dave

[spacebiscuit]

Originally Posted by G6ONEDave

Is that +/-100vdc or in my opinion +50/0/-50.
Dave

Ok so I guess I measured this wrong, I left the probes on the + and - wires - same as when I tested AC. I'll re-test - sorry if this is such an amateur error!

I removed two more transistors:

ZTX853 and BC556 - both seemed ok in the tester

nfg 213 / 397
1.5/ 2.3ma
Ube 584 / 683 mv

However when putting them back in and referring to the photos I took are they wrong:

The ZTX853 looks like the collector and emitter are switch and the BC556 looks like base leg is where the collector should go on the board?

I need to check this against the other working channel - but doesn't look right to me or am I missing something?

[Diabolical Artificer]

It's hard for me to tell, are pic 1 & 2 different boards? Compare it to the other un-ouched board or draw a schematic, that will tell you if they are right or not.

Andy.

[spacebiscuit]

Photo 1 and 2 in the above post are the same board, just taken from different anglss.

I checked the good working board and these transistors are installed the same way around!

[Diabolical Artificer]

It's probably due to that's the way the board was printed. Again as I said in post #66 if you draw a rough schematic of what those two Q's connect to, that should give you reassurance. Looking at those pics you can see the traces easily, wouldn't take long with a vintage graphical interface AKA pencil & paper.

Andy.

[spacebiscuit]

Retested the 2 PSU connections to the boards today, 220VAC and 50VDC on both channels - with probes correct way around this time.

I have tentatively soldered in the replacement pair of P-MOSFETs today - the result is Fuse 1 (green led) is blowing, fuse 2 (red led) remains intact which is progress of sorts as the pair I tried previously were tripping both fuses.

I think the fuses are one for the P-MOSFET pair and another for the N-MOSFET pair.

[Station X]

You need to reverse engineer this and produce a circuit diagram. Then you'd know which fuse feeds what.

[Diabolical Artificer]

With the amp off & no fuse in the fuse holder, meter set to resistance, one lead to ground, the other lead on one side of the fuse holder, what you get? If the answer is under 10r you have a short. Take the Pfets out, repeat.

You can't power this up until the short circuits have gone, don't even think about it, Pfets are in short supply as are vintage fets, expensive too, blowing em up willy nilly is daft, might as well set fire to £20 notes.

As I've mentioned several times a schematic will help, it takes an hour with a bit of scrap paper to rough out what you see, then when you think it's right, draw it up neat.

Andy.

[spacebiscuit]

I'm going to build a bulb limiter as was suggested and I believe I have sourced a genuine pair of MOSFET's at least the seller reassured me that they are and I have means to raise a claim if they are not fakes.

So whilst they are on order I'll try and draw the schematic as advised.

[G6ONEDave]

This might be a silly comment but are those replacement fets fitted the right way round? The Nfet will be connected to the fused +50v output of the power supply. The Pfet will be connected to the fused -50v power supply output. To confirm this compare with the working amplifier, which should act as a guide for you.

Follow Andy's advice and don't refit any fets until all shorts have been removed from both sides of the +/- supply lines. Only when you are happy that the shorts have gone, try powering up without the fets fitted. Do the fuses still blow or are they alright?

If all appears to be well then and ONLY then refit the fets, making sure that they are in the right way round. Remember that all transistors are polarity sensitive, they are not known as the fastest fuses on 3 legs, for no good reason!

Once fitted recheck for shorts on the supply lines to earth. If the shorts are back again, undo the fet fixing screws and check again for shorts. This might seem an odd thing to do but there could be a short via the insulating washers and fixing screws to the heatsink. If this test does show a short, then you need to replace the insulator washers with known good ones. After that refit the screws and washers and check again for shorts.

Please report back with your findings.

Dave

[JonSnell]

In my experience of amplifier of this age, if the power MosFet is not short circuit, it is OK.
J35s will read as short circuit sometimes and by connecting the Gate to the Source, that will discharge the capacitor and the gate will re open.
Most use a 15/18k 1W resistor as a current source to bias the output stage up to 50% of supply rail.
These fail, regularly.

[Diabolical Artificer]

I'm going to build a bulb limiter as was suggested

A lamp or bulb limiter might save you, but as Dave says transistors will often die in micro seconds, too quick for a fuse or any other protection to save them, therefore EXTREME caution is needed. Most transistor amps have layers of protective circuits, yours doesn't. Before switching the amp on, check, double check then check again.

J35s will read as short circuit sometimes

I've had that before on a similar amp Jon, good point.

Finally, I admire your persistence in trying to fix this yourself , but sometimes it's best to hand it to someone else with more experience, though you won't get that experience without having a go, Andy.

[spacebiscuit]

Just to add - when I fried the latest round of fuses yesterday the MOSFETS were soldered on the tips of their pins and were not screwed down. Is this ok to do - obviously I have no source or speakers plugged into the amp - just testing if the amp blows its fuses.

As for the MOSFETS being the right way around, I've observed the letters by the solder points on the board which match with the datasheet for the original Hitachi parts. If the pins are the wrong way around at manufacture it is hardly a positive sign I guess.

As an aside, the amp took some beatings in the past, I used to use it for dj gigs in the early 2000's - only ever used two channels at a time but would fill a decent sized hall with a couple of hundred people with sufficient sound. It never let me down, it would get hot and I might put a fan in front of it but otherwise it was a solid work horse.

In terms of perseverance, I tend not to give up, often to the detriment to quality time with family, but I need something to keep me sane and if nothing else I am leaning here as I go.

I'd not heard of a MOSFET before this experience

[Diabolical Artificer]

Most mosfets AFAIK have the same pin out - GDS - left to right, with not plastic packages the drain is the metal tab. There are two basic types too, depletion & enhancement, but that doesn't apply here. Not having them screwed down shouldn't make a difference if the body's are insulated from the heatsink and each other.

You need to find WHAT is causing the fuse to blow, the fuse blows because it's exceeding it's current rating simply put, which is because the power supply is trying to shove excess current somewhere, usually caused by a short circuit or zero ohms. Use your meter to find it.

You've started on the fixing amps journey by swapping parts, which is easy, now you to have to use your brain ans diagnose faults, not so easy but as you say keeps you sane, it can also drive you completely nuts. Having a rough idea of how your amp works will help too, it's all about moving current about, electrons and all that jazz.

Andy.

[spacebiscuit]

Originally Posted by Diabolical Artificer

it can also drive you completely nuts

That's the point when I am usually found at the bottom of the garden taking care of the problem with a hammer. This would have been an outcome already if I didn't already have a working channel and therefore can technically be used still

[G6ONEDave]

Apart from right way round re the actual device pins, I was trying to point out that it matters where the Nfet and Pfet are fitted in the output stage. If the output pair are connected in the wrong half of the amp then they will usually either do nothing or can choose to go short circuit. This is why it is important to make sure that they are connected to the correct place within the circuit.

To try to explain the Nfet should have one of it's pins connected to the +50 (plus) volts. The Pfet should have one of it's pins connected to the -50 (minus) volts. Now if you were to swap the Nfet and Pfet around so that they are now in each others circuit position, they either do nothing or can go short circuit at power up, the result will depend on the drive circuit and available current from the power supply.

With all the removal, fitting, removal and fitting, it is possible that the 2 fets have become mixed up in terms of where they actually fit within the circuit. This is why I suggested that the working channel should be used to check which fet is in which position. Not just by seeing that the Nfet is on the left of the working channel, so it must be on the left of the faulty channel. That may not be the case as on the faulty channel the Nfet could be fitted to the right.

This is why it is so important to check what is actully connected to the fet pins of both channels. The faulty channel will in circuit connection terms be a carbon copy of the working channel. However it could be laid out as a mirror image or as a straight copy. A photo showing the whole pcb would show how it is laid out.

Hope this helps a bit, if not please ask.

Dave

[John_BS]

The circuit posted in 45 is so dreadful it's making me feel ill!

JOhn :>(