Arcam Delta 290 Keeps Blowing Output Mosfets

Recently got an Arcam Delta 290 - believe it to be a later model with IRFP240 mosfets mounted vertically on a heatsink and built in phono preamp. Originally it had no power and a blown fuse. Checked the output mosfets found Q1 to be shorted, removed Q1 and Q2, replaced the fuse and the amp powered on ok.

I've ordered 10x new IRFP240 from Farnell and replaced both Q1 and Q2 the amp then came on ok for a few seconds light went green and then a few seconds later I smelled something burning and the light went orange. Checked and Q1 and Q2 are both shorted dead again and also R1 and R1A. Went through the whole circuit checked all resistors and diodes and replaced a few resistors that were damaged. None of the transistors or diodes were showing shorted so I replaced Q1 and Q2 again - this time the amp came on again for a few seconds and then immediately blew the fuse just before the light should turn from orange to green. Again Q1 and Q2 are dead.

I do not really know where to go from here - I do not really want to keep replacing Q1 and Q2 if it isn't going to fix the problem but I can see no faults or shorted components that would explain why these keep burning.

Maybe someone here knows more about this model and can point me in the right direction or maybe Farnell are selling garbage IRFP240 chips? Am I correct in my assumption that all the output mosfets run off the same Power supply so that if there was a PSU issue surely it should be burning all of the output mosfets?
 
Welcome to the forum!

I have no familiarity with your amp but I can try to help. I'm attaching a schematic found on the web for any members who wish to contribute.

Do you have an any equipment beyond a multimeter? A 'scope?

Initially, I suggest keeping Q1 ad Q2 out of circuit. Check the Zobel network (R3 and C4), as its failure could cause oscillation and transistor damage. If you probe C4's lead, you should find about 5.6 Ohms back to L1. Similarly, try to confirm continuity of C4's lead to ground. Confirm C4 value ad ESR if possible.

Set bias pot VR1's wiper to the R16 end; this will set bias current to minimum. Apply power, and measure Q5 collector to emitter voltage. It should be roughly 0.9V. Inspect Q4 and Q5 voltages to see if they reasonable, i.e. a few tenths of a volt base to emitter on each. Should be roughly 0V at R1/R1A re ground. Output may have evidence of oscillation since Q1/Q2 are absent. Try advancing bias pot to confirm voltages across R11 and R12 increase smoothly. If all looks good, reinstall Q1 and Q2, but make sure bias pot VR1 is at minimum before applying power. You might want to investigate Dim Bulb Testers for added caution. Many here are advocates.

Good luck!
 

Attachments

  • delta_290.pdf
    1 MB · Views: 93
First off thanks for the link to the other thread - I did check there but i'm not sure its the same fault! I did pull Q14 and it tests fine as do Q4 and Q5. Also the other thread didn't come to any conclusion/solution so one can only assume the OP either gave up or figured it out and never posted back.

I have done the testing with Q1 and Q2 out of the circuit - the amp powers on but never switches to the green light - presumably as it can detect that Q1 and Q2 are missing. I've checked R3 and C4 and both test fine out of circuit and I have got 5.6ohms back to L1 and C4 one leg is connected to ground.

I've adjusted RV1 as suggested and measured Q5 between collector and emitter (one probe on collector, other on emitter) and i'm seeing around 40v not 0.9v. - there is 0v on R1/R1a referenced to ground. I've tested Q5 and Q4 out of circuit with a transistor tester and they both check out ok. Also adjusted RV1 whilst checking voltages across R11 and R12 and they do increase from a minimum of around 2.5v up to around 4v.
 
Hopefully you didn't pay Much for the 290 . These are frankly, rather Poor sounding contraptions. Mine certainly is
Use/biasing of the 240s' is completely inept imo.
A simple Stoopid LM3886 chip amp Amplifier is another world better performing/sounding.
Also a fraction of the foolish Arcam pricings.

Mine has been sitting unused for 15 + years now.
I'd put it into the trash bin if I hadn't paid for it new.
 
Hi @CRC2018

Bare seems to have had unhappy experience with his amp and I'm not discounting his critique. I will also observe that the output stage design is very unusual, as is the bootstrap drive Zener (BOOTL). Also noteworthy are the service notes alluding to oscillation tendencies. Would you advise the serial number of your amp? You should probably install all stability related modifications noted in the service notes. The notes do refer to failure of transistors due to oscillation.

A oscilloscope would be very valuable if available, and a distortion analyzer, etc.

Legibility of the schematic is a challenge for my eyes and I misread Ref Des D5 as "Q5" and thought it referred to the the nearby transistor--- which closer scrutiny shows is in fact Q14. It's encouraging that VR1 produces change at R11 and R12. Would you confirm actual installed values at R17, RV1, R16, and R55? I interpret those schematic values as 4.7k, 1k, 82K, and 100k, respectively.

That you observe 0V at R1/R1A is also encouraging. As additional confirmation, would you report servo voltage at IC1 pin 6? More comments later.

Thanks.
 
Hopefully you didn't pay Much for the 290 . These are frankly, rather Poor sounding contraptions. Mine certainly is
Use/biasing of the 240s' is completely inept imo.
A simple Stoopid LM3886 chip amp Amplifier is another world better performing/sounding.
Also a fraction of the foolish Arcam pricings.

Mine has been sitting unused for 15 + years now.
I'd put it into the trash bin if I hadn't paid for it new.
try with another preamp
 
Serial is D29008775 so should be a later model.

R17 - was 3.3k but I changed it to 4.7k yesterday which dropped the voltage across R11/R12 from a minimum of around 2.3v to a minimum of 1.9v
RV1 - 1k
R16 - appears to be 41k
R55 - 67K

I do have a cheap digital scope yes.
 
Right so Pin 6 on IC1 is reading 0v referenced to ground. I've reinstalled Q1 and Q2 now and we appear to be getting somewhere - that is they aren't self destructing immediately anymore! The amp is however not switching over to green light ready status and stays on an orange light - operating the speaker on/off switches does nothing the relays aren't clicking so I think it is still in protection mode. Voltage across R1 (problem channel) is 0mv and across R101 (good channel) is 3mv.
 
Does advancing RV1 bias pot affect voltage across R1? Careful! ;)

Have you determined which of the power amp schematics is applicable? I haven't as yet. VIPROT signal in page 12 schematic looks worth scrutiny. As does MUTEL signal in the page 15 schematic.
 
I believe it is the first power amp schematics on the PDF you linked earlier as it originally came with IRFP240 mosfets and that is the one referring to those. I believe earlier versions came with different mosfets and a different heatsink layout.

I did try advancing the RV1 pot slightly to see what would happen and whilst I did get it up to 2mv referenced to ground on R1/R1A I wasn't seeing anything across R1/R1A.
 
PDF page 7 seems to specify R16 = 56K, R17 = 4.7K, R55 = 100K. Is this what you have? Is this what's present in the working channel? Does working channel also have same MOSFETS you're applying in problem channel?

If you need to revise any caps, polystyrene will probably be hard to find. You can substitute NPO/COG ceramic caps. They have superb distortion; avoid other ceramic dielectric types in the signal path.

IC501 on page 13 should be at ground (pin 6) when protection is in the clear-to-go state. Probing its input at pin 2 may lend insight.
 
The channel which didn't blow up its mosfets also has IRFP240 but these are made by International Rectifier and the new one's i've obtained from Farnell are made by Vishay - specs appear identical when comparing the datasheets.

The channel which still has its original mosfets has:
R116 - 56K
R117 - 3.3K
R55 - 120K

The bad channel has:
R16 - 56k
R17 - 4.7k (I replaced from 3.3k)
R55 - 120K

I'll check out IC501.

This is why this is such a puzzle - I am replacing the mosfet's like for like but they did keep burning!
 
And amp output observed at R1/R1A remains at ~0V? Check voltage across R25 (near VIPROT). It should be 0V, indicating Q13 is not conducting. This examines network near Q13.

If there's no voltage across R25, look for voltage across R523 near the protection IC, and at R522 if there's voltage present across R523. The near term objective is to find why -0.8V appears at pin 2.
 
Last edited:
Yes it was still at 0v across R1/R1A. So I was probing around R25 as you suggested I must've slipped with the probes or something near there and heard quite a loud bang, immediately unplugged everything and visually inspected the area could see nothing so plugged it all back in. Turned on a few seconds later light went green, speaker relay buttons working and amp appears to now be working normally. I don't know what has happened but clearly the loud bang has fixed something and the amp is now working normally - i've had it running with some speakers for the past hour and it's working normally good clear sound no distortion!

Thank you very much for all the help you provided!