Quad 303 smoking! - UK Vintage Radio Repair and Restoration Discussion Forum

alanworland · 2nd Jan 2018, 10:36 pm

Had a session with my early Quad power amp using a DaDa upgrade kit.
This basically involved changing output capacitors, power supply cap, a few electrolytics, resistors/presets and a couple of diodes on the driver boards.
All went ok and powered it up to set 67v then set the halfway volts to 33.5v on each driver board. The plan was to let it warm up and check volts again.
Before I got that far I was aware of the 33.5v was dropping, although 67v was still there, within a short space of time smoke appeared! Pulled plug. It came from the 0.33 ohm resistor on the driver board.

Disconnected supply from each driver PCB and 67v is still ok.
Right channel lost one of the output 2N3055's, TR1, and a driver transistor TR105 from the PCB
I am making enquiries with regard to TR105 (RCA 38496) but going through the circuit and checking orientation of diodes and electrolytics I can't find what may have caused it.
My concern is that once the faulty components are replaced what would be the safest method of powering it up? Not wishing a repeat of the fault!

Alan

Diabolical Artificer · 3rd Jan 2018, 7:49 am

Smoking isn't good for you or your amp, ouch! I always, well mostly always power anything up on a variac and lamp limiter after making changes, but you could use a current limited power supply if you have one that goes that high.

After making changes go through your work again and again to see if you've done something daft, it may not be obvious the first time. things like solder bridges, excess flux can cause issues and check components actual properties, is the 0.33 ohm R actually 0.33 ohm.

I guess this is an emitter resistor, I've had these go short and very high.

Andy.

trsomian · 3rd Jan 2018, 9:07 am

I agree with the earlier response. Power it initially from a current limited power supply, with the current limit set to 100mA or so. Even if it is only a 30V supply, major problems will probably be evident at that voltage. When it is happy at that, move on to the internal supply. Not sure about the Quad, but sometimes one can remove the DC fuse(s) to isolate each power amplifier. As to the fault, carefully inspect the modifications made looking for errors, and check for short circuits with a meter

Radio Wrangler · 3rd Jan 2018, 9:16 am

The 303 has a regulated power supply, but you may want to use an external lab type supply so you can wind down the current limit to something less damaging.

The 303 was designed to push the 2N3055 type device as far as it could go. The reason for the regulated supply was so the power amp rail could be set as high as they dared, but without having to build in any margin to allow for mains voltage variation and surges. And even then they had to go to a somewhat selected higher voltage variant.

We now have plenty of far higher voltage transistors, but the problem is that they've become too fast. The 'Quad Triplet' circuits around the output devices rely on the power device being much slower than its drivers. Narrow the gap in speed and the stability goes out of the window. So you really are stuck using early-type 2N3055. Oh, and if you try to restore the speed difference with faster drivers, then the stray coupling in the wires to the power transistors will get you.

When teaching, I've used the 303 amplifier as an example of the consequences of reliance on unspecified parameters.

David

alanworland · 3rd Jan 2018, 12:02 pm

Thanks for replies, it must be something I have done as it was working before I 'fixed' it! I shall have another inquisition!
I can run the driver boards off a limited power supply so will do that once I have changed components.
Would it be ok to just replace TR105 or should it be replaced with TR106? are they a matched pair?

Alan

Nuvistor · 3rd Jan 2018, 12:49 pm

A word of caution about newer 2N3055 transistors, seems they are different from the old ones.
https://www.vintage-radio.net/forum/...d.php?t=138407

fetteler · 3rd Jan 2018, 12:50 pm

Did you make sure that the quiescent current adjustment pot was set to its minimum position whilst you were letting the amp warm up?

Cheers,
Steve.

alanworland · 3rd Jan 2018, 1:29 pm

I must say that I think I should be ok with the 3055's as I have a few from 30 plus years ago marked in exactly the same manner (SM 2N3055H 9107)
With regards to the quiescent current I had the pots set to midway but didn't get as far as measuring anything before my problems occured.

Alan

Mooly · 3rd Jan 2018, 3:57 pm

Turn the pot initially to give MAX resistance which corresponds to minimum bias.

Use a DBT (dim bulb tester) for powering up, it saves so much grief. A 60 or 100 watt GLS lamp is ideal.

2nd Jan 2018, 10:36 pm	#1
alanworland Heptode Join Date: Feb 2011 Location: Southend, Essex, UK. Posts: 803	Quad 303 smoking! Had a session with my early Quad power amp using a DaDa upgrade kit. This basically involved changing output capacitors, power supply cap, a few electrolytics, resistors/presets and a couple of diodes on the driver boards. All went ok and powered it up to set 67v then set the halfway volts to 33.5v on each driver board. The plan was to let it warm up and check volts again. Before I got that far I was aware of the 33.5v was dropping, although 67v was still there, within a short space of time smoke appeared! Pulled plug. It came from the 0.33 ohm resistor on the driver board. Disconnected supply from each driver PCB and 67v is still ok. Right channel lost one of the output 2N3055's, TR1, and a driver transistor TR105 from the PCB I am making enquiries with regard to TR105 (RCA 38496) but going through the circuit and checking orientation of diodes and electrolytics I can't find what may have caused it. My concern is that once the faulty components are replaced what would be the safest method of powering it up? Not wishing a repeat of the fault! Alan Attached Thumbnails

3rd Jan 2018, 7:49 am	#2
Diabolical Artificer Dekatron Join Date: Jul 2009 Location: Sleaford, Lincs. UK. Posts: 7,670	Re: Quad 303 smoking! Smoking isn't good for you or your amp, ouch! I always, well mostly always power anything up on a variac and lamp limiter after making changes, but you could use a current limited power supply if you have one that goes that high. After making changes go through your work again and again to see if you've done something daft, it may not be obvious the first time. things like solder bridges, excess flux can cause issues and check components actual properties, is the 0.33 ohm R actually 0.33 ohm. I guess this is an emitter resistor, I've had these go short and very high. Andy. __________________ Curiosity hasn't killed this cat...so far.

3rd Jan 2018, 9:07 am	#3
trsomian Hexode Join Date: Oct 2010 Location: Box End, Beds. UK. Posts: 271	Re: Quad 303 smoking! I agree with the earlier response. Power it initially from a current limited power supply, with the current limit set to 100mA or so. Even if it is only a 30V supply, major problems will probably be evident at that voltage. When it is happy at that, move on to the internal supply. Not sure about the Quad, but sometimes one can remove the DC fuse(s) to isolate each power amplifier. As to the fault, carefully inspect the modifications made looking for errors, and check for short circuits with a meter

3rd Jan 2018, 9:16 am	#4
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,905	Re: Quad 303 smoking! The 303 has a regulated power supply, but you may want to use an external lab type supply so you can wind down the current limit to something less damaging. The 303 was designed to push the 2N3055 type device as far as it could go. The reason for the regulated supply was so the power amp rail could be set as high as they dared, but without having to build in any margin to allow for mains voltage variation and surges. And even then they had to go to a somewhat selected higher voltage variant. We now have plenty of far higher voltage transistors, but the problem is that they've become too fast. The 'Quad Triplet' circuits around the output devices rely on the power device being much slower than its drivers. Narrow the gap in speed and the stability goes out of the window. So you really are stuck using early-type 2N3055. Oh, and if you try to restore the speed difference with faster drivers, then the stray coupling in the wires to the power transistors will get you. When teaching, I've used the 303 amplifier as an example of the consequences of reliance on unspecified parameters. David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

3rd Jan 2018, 12:02 pm	#5
alanworland Heptode Join Date: Feb 2011 Location: Southend, Essex, UK. Posts: 803	Re: Quad 303 smoking! Thanks for replies, it must be something I have done as it was working before I 'fixed' it! I shall have another inquisition! I can run the driver boards off a limited power supply so will do that once I have changed components. Would it be ok to just replace TR105 or should it be replaced with TR106? are they a matched pair? Alan

3rd Jan 2018, 12:49 pm	#6
Nuvistor Dekatron Join Date: Aug 2013 Location: Wigan, Greater Manchester, UK. Posts: 9,433	Re: Quad 303 smoking! A word of caution about newer 2N3055 transistors, seems they are different from the old ones. https://www.vintage-radio.net/forum/...d.php?t=138407 __________________ Frank

3rd Jan 2018, 12:50 pm	#7
fetteler Octode Join Date: Jul 2011 Location: Staffordshire Moorlands, UK. Posts: 1,479	Re: Quad 303 smoking! Did you make sure that the quiescent current adjustment pot was set to its minimum position whilst you were letting the amp warm up? Cheers, Steve. __________________ Those who lack imagination cannot imagine what is lacking...

3rd Jan 2018, 1:29 pm	#8
alanworland Heptode Join Date: Feb 2011 Location: Southend, Essex, UK. Posts: 803	Re: Quad 303 smoking! I must say that I think I should be ok with the 3055's as I have a few from 30 plus years ago marked in exactly the same manner (SM 2N3055H 9107) With regards to the quiescent current I had the pots set to midway but didn't get as far as measuring anything before my problems occured. Alan

3rd Jan 2018, 3:57 pm	#9
Mooly Octode Join Date: May 2011 Location: Lancashire, UK. Posts: 1,351	Re: Quad 303 smoking! Turn the pot initially to give MAX resistance which corresponds to minimum bias. Use a DBT (dim bulb tester) for powering up, it saves so much grief. A 60 or 100 watt GLS lamp is ideal.