Another DC coupled amp (Ampeg bass combo) - Page 2 - UK Vintage Radio Repair and Restoration Discussion Forum

Boater Sam · 18th Jan 2018, 6:19 am

Been caught out like that before and I always say it will be the last time, never is though.

Always check previous repairs, not everyone is as proficient as yourself.
I've had valve bases wired in mirror image, droppers fitted upside down, wrong windings on mains transformers used for rectifiers, all sorts.
There are muppets out there.

Ted Kendall · 18th Jan 2018, 8:54 am

Quote:

Originally Posted by Boater Sam

There are muppets out there.

Yes, indeed.

Radio Wrangler · 18th Jan 2018, 9:23 am

They would have been all bodge attempts to find the real original fault, the mis-loaded resistor from the factory. Muppets for sure.

I wonder how many amps were turned out with a similar fault?

The Blomley architecture would make a monkey out of anyone assuming an ordinary circuit.

David

ITAM805 · 18th Jan 2018, 5:56 pm

Thanks guys, but despite this apparent fix, the DC offset of 475mV is still there

With the Mosfets still out of circuit the +/-16V rails were now back and stable with the variac at 240V,

But when I reinstalled the Mosfets the +16V rail now goes up and past 16V if I wind the variac much beyond 160V, i.e. 20V at 200V input? How can this happen, keeping in mind nothing on this amp seems to work as expected!

The drivers Q1/Q4 both read good on my tester, I haven't removed Q3 but a diode test suggests it's good? The gate/source voltage are approximately 4V on Q5/8 and 3.75V on Q2/7 (at 160V input)

So, no closer than I was in post#1, any idea guys?

Radio Wrangler · 18th Jan 2018, 6:13 pm

The opamps will have internal diodes from their output pins to their supply rails (ESD protection)

If something tries to drive current into the opamp output pin and drive it beyond the 16v rail, the diode will turn on and route current into pulling the supply up too high (to the detriment of the zener)

Possibly one of the opamps has a damaged output stage. They've been somewhat overrun.

On that circuit, the current-mirror output stage can be isolated from the earlier stage, and with the centre of the resistor string grounded, you can look to see how healthy the output stage is.

Remember that this is a CURRENT output stage. If there is a little offset CURRENT, it can look like a significant offset voltage if there is only a high DC resistance load. The loop with the opamp should fix the offset, but if something is still dodgy you may still see the offset.

David

ms660 · 18th Jan 2018, 6:30 pm

I wonder if it's a bad ground somewhere, might be worth measuring any voltage between transformer CT and the junction of the two 16 volt zeners when the 16 volt rail rises.

Lawrence.

ITAM805 · 18th Jan 2018, 6:33 pm

David thanks, lots to consider there? I should have said the DC offset is into an 8R dummy load.

Thanks Laurence, I shall check that

ITAM805 · 18th Jan 2018, 8:30 pm

Regarding the rising 16V rail - I have to admit to a schoolboy error, I used a 15V zener in series with a 1N4007 to get the 16V, but I put it in back to front

The DC offset is now the only issue, hopefully!

s22ns22ns22n · 21st Feb 2020, 7:14 pm

I came across the discussion concerning the Ampeg bass combo https://www.vintage-radio.net/forum/...d.php?t=142940

The amplifier is said to be of the Blomley type. That is NOT correct. The schematic reveals that the bias circuit feeds the rectifying transistors (Ampeg Q1 and Q4) so that both output halves supply signal within a small interval around zero (+/- quiescent current). This leads to the usual class-B cross-over distortion/gm-doubling problem.

The Blomley principle is to ensure that at any given moment ONLY one output half supplies signal, while the other supplies a constant quiescent current. This requires that a small voltage interval exists (output of Ampeg U5:B) where both rectifiers are switched off. Of course, this interval should be as small as possible.

18th Jan 2018, 6:19 am	#21
Boater Sam Banned Join Date: Dec 2013 Location: Middlewich, Cheshire, UK. & Winter in the Philippines. Posts: 3,897	Re: Another DC coupled amp Been caught out like that before and I always say it will be the last time, never is though. Always check previous repairs, not everyone is as proficient as yourself. I've had valve bases wired in mirror image, droppers fitted upside down, wrong windings on mains transformers used for rectifiers, all sorts. There are muppets out there.

18th Jan 2018, 9:23 am	#23
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,876	Re: Another DC coupled amp They would have been all bodge attempts to find the real original fault, the mis-loaded resistor from the factory. Muppets for sure. I wonder how many amps were turned out with a similar fault? The Blomley architecture would make a monkey out of anyone assuming an ordinary circuit. David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

18th Jan 2018, 5:56 pm	#24
ITAM805 Nonode Join Date: Jul 2010 Location: Folkestone, Kent, UK. Posts: 2,172	Re: Another DC coupled amp Thanks guys, but despite this apparent fix, the DC offset of 475mV is still there With the Mosfets still out of circuit the +/-16V rails were now back and stable with the variac at 240V, But when I reinstalled the Mosfets the +16V rail now goes up and past 16V if I wind the variac much beyond 160V, i.e. 20V at 200V input? How can this happen, keeping in mind nothing on this amp seems to work as expected! The drivers Q1/Q4 both read good on my tester, I haven't removed Q3 but a diode test suggests it's good? The gate/source voltage are approximately 4V on Q5/8 and 3.75V on Q2/7 (at 160V input) So, no closer than I was in post#1, any idea guys?

18th Jan 2018, 6:13 pm	#25
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,876	Re: Another DC coupled amp The opamps will have internal diodes from their output pins to their supply rails (ESD protection) If something tries to drive current into the opamp output pin and drive it beyond the 16v rail, the diode will turn on and route current into pulling the supply up too high (to the detriment of the zener) Possibly one of the opamps has a damaged output stage. They've been somewhat overrun. On that circuit, the current-mirror output stage can be isolated from the earlier stage, and with the centre of the resistor string grounded, you can look to see how healthy the output stage is. Remember that this is a CURRENT output stage. If there is a little offset CURRENT, it can look like a significant offset voltage if there is only a high DC resistance load. The loop with the opamp should fix the offset, but if something is still dodgy you may still see the offset. David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

18th Jan 2018, 6:30 pm	#26
ms660 Dekatron Join Date: Apr 2011 Location: Cornwall, UK. Posts: 13,454	Re: Another DC coupled amp I wonder if it's a bad ground somewhere, might be worth measuring any voltage between transformer CT and the junction of the two 16 volt zeners when the 16 volt rail rises. Lawrence.

18th Jan 2018, 6:33 pm	#27
ITAM805 Nonode Join Date: Jul 2010 Location: Folkestone, Kent, UK. Posts: 2,172	Re: Another DC coupled amp David thanks, lots to consider there? I should have said the DC offset is into an 8R dummy load. Thanks Laurence, I shall check that

18th Jan 2018, 8:30 pm	#28
ITAM805 Nonode Join Date: Jul 2010 Location: Folkestone, Kent, UK. Posts: 2,172	Re: Another DC coupled amp Regarding the rising 16V rail - I have to admit to a schoolboy error, I used a 15V zener in series with a 1N4007 to get the 16V, but I put it in back to front The DC offset is now the only issue, hopefully!

21st Feb 2020, 7:14 pm	#29
s22ns22ns22n Diode Join Date: Feb 2020 Location: Copenhagen, Denmark Posts: 1	To Blomley or not to Blomley I came across the discussion concerning the Ampeg bass combo https://www.vintage-radio.net/forum/...d.php?t=142940 The amplifier is said to be of the Blomley type. That is NOT correct. The schematic reveals that the bias circuit feeds the rectifying transistors (Ampeg Q1 and Q4) so that both output halves supply signal within a small interval around zero (+/- quiescent current). This leads to the usual class-B cross-over distortion/gm-doubling problem. The Blomley principle is to ensure that at any given moment ONLY one output half supplies signal, while the other supplies a constant quiescent current. This requires that a small voltage interval exists (output of Ampeg U5:B) where both rectifiers are switched off. Of course, this interval should be as small as possible.