Amp with high pitched static noise

[spacebiscuit]

The Amp is a T400 MOSFET Tower Amp by Class Ltd, London. I've found no information online, its a PA amp and I purchased it in the mid-90's and it was second hand then.

So I guess my next step should be to remove the MOSFETs and test these out of the circuit. I read up on the fakes and I should check that they are within range? If they are within range are they ok to use? Given that they are no longer made I guess I have no other choice.

With MOSFETs removed if I power up the amp the fuses still blow I know it's another component that I need to identify?

Is that a more logical approach?

[bigfathairyvika]

Being a PA amp, would I be correct in saying it is a mono amp?

If it is a mono amp, I would start by testing for drain to source shorts on all the mosfets.
If any are short circuit then remove all of them.
If there are any SCR's fitted check them for shorts.
If there are any other transistors , check them for shorts.
Then check all tracks are OK.
Then check for shorts across all capacitors.
Then check all resistors and diodes.
Recheck for a short across the power rails.
Do NOT skip any of these checks.
If all is OK. Disconnect the power rails from the amplifier board. Carefully apply power and measure the voltages on the disconnected power supply rails.

It sounds like, from previous posts, you have no way to test the mosfets.
The mosfets being unobtainable, does not mean that a modern replacement cannot be used. I have many times replaced mosfets and transistors with modern equivalents with no problems at all. They are available from RS, Farnell, CPC, Mouser and many other reputable dealers. And at only around a quid or so each.

Edit : Could you take a picture of the entire board /s so we are able to see the whole picture.?
Pictures of only parts of the amp only allow diagnosis of that part.

UPDATE: Just found the previous thread.
I noticed that there is also some fusible links in the gate circuit. These should be a short.

Further update: Looking at the PA board , there is no way that the fuses could blow without the fets installed and being short circuit. So , without the fets installed, the power terminals must have been connected incorrectly or the board has been shorted to the case metalwork.

[spacebiscuit]

It's a stereo amp - there are two of the boards and one is working so I can measure resistors and transistors and compare with the working board. I mostly have done this and the resistors are check out fine.

One of the small transistors was giving a reading on the good of 843 on the emitter and 808 on the collector. On the channel i am fixing readings are 774 and 748 respectively - not sure if that is significant?

I did read that some fakes had the wrong pins around the wrong way, according to the datasheet and marking not the board the pins are correct - G|S|D. I had read several guides that suggest in diode mode on a dmm one probe to source, one to drain, then move from source to gate then back to gate and it should open the circuit and provide a reading?

[bigfathairyvika]

Originally Posted by spacebiscuit

It's a stereo amp - there are two of the boards and one is working so I can measure resistors and transistors and compare with the working board. I mostly have done this and the resistors are check out fine.

One of the small transistors was giving a reading on the good of 843 on the emitter and 808 on the collector. On the channel i am fixing readings are 774 and 748 respectively - not sure if that is significant?

I did read that some fakes had the wrong pins around the wrong way, according to the datasheet and marking not the board the pins are correct - G|S|D. I had read several guides that suggest in diode mode on a dmm one probe to source, one to drain, then move from source to gate then back to gate and it should open the circuit and provide a reading?

Forget using your dmm to test mosfets are working.
The only thing you can check with a meter and a mosfet is for a definite short circuit. You cannot check if they work OK without putting in a circuit like a simple bulb switch. But his will not tell you if they are fakes. Fakes quite often are working fets but seriously under specified.

Also, you need to remove transistors and fets from the board to check them. No point testing whilst in circuit unless the unit is able to power up ok.

Note, that there are component checkers available very cheap at around 15 to 20 quid or less.

Have you checked for shorts from the supply rails to ground? And the supply rails to each other?
Disconnect the power supply connections first.

Have you checked for shorts between tracks?

As you have a working board , remove both boards and compare them track side and component side?
Are they exactly the same.

Also before disconnecting anything , take photos of connections so if you forget which wire went where you can check.

[Diabolical Artificer]

Your amp maybe something like the attached schematic. The amp is blowing fuses because you have a short circuit, IE a very low resistance between V+ & ground and maybe V- & gnd. Before you can power the amp up safely you have to find those shorts.

If you look at the simple schematic, Q 3 & Q4 could potentially cause a short (sc), there maybe other parts too that are sc. Which is why Mark suggested checking across the supply rails etc. Most checks are done with your DMM set to ohms, clip the black lead to ground, use the red lead to wander around checking, a good place to start are the fuses or the big caps.

Andy.

[spacebiscuit]

I think I may have found something, as suggested I have been poking around with DMM in diode mode. Where the negative plugs to the board it looks as if the track has is a break where it should make with the resistor R8.

I've checked on the good channel and there is continuity but not on the broken channel's board. It looks as if the track has been scratched.

See attached photos.

[G6ONEDave]

Is that blue capacitor connected like that on the working amp pcb. It could be either a modification or a mistake.

BTW you want to set your dmm to dc ohms not diode test. Diode test generally displays a voltage, such as volt drop across a semiconductor internal junction. You want to check for a short or very low resistance and you need the resistance range setting to do this. Depending on your dmm it will either have a 200 ohm range or it might be auto ranging. Would need a pic of the dmm to confirm.

Dave

[Diabolical Artificer]

Is that blue capacitor connected like that on the working amp pcb. It could be either a modification or a mistake

This was spotted a few posts back Dave & seems to be factory.

Any connector like those spade connectors are prone to joint or track damage due to them getting tugged about. Treat them as potentially suspect & re-solder them then test whilst jiggling them about. I dislike spade connectors, horrible things.

Andy.

[G6ONEDave]

The cap might be factory fit but why is it wired differently to the one on the right in the picture of post 22? The one on the right I can accept but the one on the left looks wrong to me and should not be going to just the isolated end of the 10k resistor like it does with the cut track. Of course I could be totally wrong with my interpretation.

Has the power supply been checked with both fuses removed, to see that it is producing +/- supplies. Problems will arise if any of the diodes in the bridge rectifier have gone short.

With the MosFets and fuses removed what are DC resistance readings between the supply inputs to the amp pcb (amp side of the fuses) and centre ground/earth/0V.

If there is a short circuit then look for an electrolytic capacitor that may be fitted from either supply rail to earth, that could be the problem.

If the amp is capacitor coupled to the loudspeaker load, check that that capacitor is also OK.

Dave

[bigfathairyvika]

I think a step back is needed.

According to the OP, with the track cuts and weirdly wired capacitor etc, the amp WAS WORKING.

The only thing that has changed is the mosfets and a case of connecting the power wires the wrong way round.

The OP was told to remove the mosfets before checking for shorts.

But as can be seen from the photos, they have NOT been removed. Unless of-course if the photos are old photos.

[spacebiscuit]

I've not yet removed the MOSFET's I plan to remove them later.

As stated in the previous post to this, all that has changed is the moronic move by myself switching polarity and the fitting of the MOSFETs.

What I can't be sure is if that track was cut previous to my repair attempts, but it sounds as if you guys believe it to be an intended modification?

But anyway, previous to the above the channel worked fine.

[spacebiscuit]

Both MOSFETs removed and the amp powers up and the fuses don't blow.

I then put both MOSFET back in - the two fuses blow.

I then removed MOSFET at TR3, with TR4 still in situ, the amp then blows the red FUSE 2 only - FUSE 1 remains good.

I do have replacements for the 2 x N-MOSFETS but I don't want to just fry them. I have watched several videos that show how to test a Mosfet with a DMM but consensus here seems to be that it is not possible.

What is recommended way to test a Mosfet, do I need something like this

https://www.ebay.co.uk/itm/374922420...mis&media=COPY

[bigfathairyvika]

That tester is one of several that I use for basic checking.
It will tell you if your mosfet is a mosfet.
But will not tell you if it is a poor quality or under specified version.
It will also only tell you what it thinks a device is.
You will need to check data sheets to be sure.
Sometimes it can read a faulty device as something else.

Useful for quick checking other components out of circuit.

From your tests it would have been far quicker and saved fuses if you just removed the fets and measured the resistance drain to source.
It sounds as if both fets are dead.

Have you measured the voltages yet?
Do NOT fit the new fets until you have:-
A) verified the supply voltages are correct.
B) Checked ALL the other components.

[spacebiscuit]

I've checked the the new N-MOSFETS which have not been installed.

From what I've read I shorted the G/D to discharge. Then black to S, Red to D - meter reads 585, then moved red to G then back to S and reading was '100''. Remains at 100 until i discharge.

I then checked the P-MOSFET that have been installed, I switched polarity - red to S and black to D and repeated steps as above - this time I just get 600, there in no lower reading when move G back to D. Does this indicate a blown P-MOSFET?

I think I have tested for short circuits and bad tracks, and readings are same for the other components as they are read on the good board.

How should I test the voltages - am I checking where the power is supplied to the board on the spade terminals?

[bigfathairyvika]

The measurements you have taken of the mosfets are meaningless.
As mentioned in several posts throughout this thread by several members you cannot test mosfets with a basic DMM.

I now give up and won't be posting on this thread further.

[spacebiscuit]

Fair enough I won't waste anyone's time any further here.

Rest assured this amp will be repaired, thank you for your input.

[Cobaltblue]

Thread reopened at OP's request

Cheers

Mike T

[spacebiscuit]

So I've taken a little break from this project, went away purchased myself a very basic component tester which will hopefully enable me to provide better feedback as I test the components of my board.

I tested the original 2SJ351 MOSFET, one of the replacements from order#1, and 2 of the new replacements from order#2 (different supplier).

The 3 replacements (including one which when installed is blowing fuses) provide similar results:

New #1
4.40v / 1.86nf / 1.4ohms

New #2
4.31v / 1.85 / 1.1

Replacement #1 (which blows fuse)
4/14 / 1.87 / 1.1

Original:
sends unit into self test calibration mode.

Given that the results of the original MOSFET I guess I can safely assume that it is toast?

Given that the results for the first replacement and the new replacements are similar I didn't fry it?

Can the above results be used to confirm anything from the expected spec given on the datasheet of the original Hitachi MOSFET, I'm trying to learn how to interpret the results in a meaningful way:

https://datasheet.octopart.com/2SJ35...eet-101743.pdf

I've not yet removed the complimentary paired N-MOSFETS - 2SK2220. These are still on the board and I do have a replacement pair.

Note the latest replacements have not been tried - still in their packets.

[Diabolical Artificer]

Typed a reply explaining the results in full, a wrong key stroke deleted 10 minutes of typing, so...

In short the original is as you say toast. the other parameters listed seem to indicate the first 3 results equal a functioning device, that doesn't mean these devices will work ok in your amp though. With so many fakes around it's hard to tell with your simple tester if you've bought good fets or duds. To test them properly needs a more complex test set up or a curve tracer capable of testing them at full working voltage.

Pop them in the amp & give em a try but try at least to limit the current going into the amp on first power up, hopefully that will reduce any damage, but probably won't, sorry.

Andy.

[spacebiscuit]

Is there any sense in removing the N-MOSFET to test if they are damaged. If damaged would they cause the current behaviour?