Yamaha DSP A2070
 

In article ,
Don Pearce wrote:
Here's the diagram of the offending amp.
http://s139.photobucket.com/albums/q...kies/A2070.jpg

Ah, they have followed the application note in the data sheet

http://www.datasheetcatalog.org/datasheet/nec/UPC1225H.pdf'

pretty well. They have dropped some of the good stuff, though, that
could well do with going back in. The output stability choke and
resistor plus Zobel network, for example, which you will find on page
10. Anyway, there are not too many bits to check, the bases and
collectors of all those transistors will do, and compare them with a
good channel. Most likely it is the IC that has keeled over.

I've got the amps out - bit of a fight to do so.

Two stereo amps - identical circuits but sort of mirror image layout.

Only near suitable power supply I have to hand produces +/- 15 volts at up
to one amp. One pair of amps is clean and appears to work ok at a rough
check - the other pair oscillates at both high and low frequencies, and
one output pair gets very hot. Both with no input but 8 ohm loads. All the
transistors check out ok on my tester and removing either of the ICs
doesn't stop the other oscillating. Given there are really no common
components between the amps I'm a bit lost.

--
*In "Casablanca", Humphrey Bogart never said "Play it again, Sam" *

Dave Plowman London SW
To e-mail, change noise into sound.

Yamaha DSP A2070
 

On Wed, 18 Nov 2009 23:34:46 +0000 (GMT), "Dave Plowman (News)"
wrote:

In article ,
Don Pearce wrote:
Here's the diagram of the offending amp.
http://s139.photobucket.com/albums/q...kies/A2070.jpg

Ah, they have followed the application note in the data sheet

http://www.datasheetcatalog.org/datasheet/nec/UPC1225H.pdf'

pretty well. They have dropped some of the good stuff, though, that
could well do with going back in. The output stability choke and
resistor plus Zobel network, for example, which you will find on page
10. Anyway, there are not too many bits to check, the bases and
collectors of all those transistors will do, and compare them with a
good channel. Most likely it is the IC that has keeled over.

I've got the amps out - bit of a fight to do so.

Two stereo amps - identical circuits but sort of mirror image layout.

Only near suitable power supply I have to hand produces +/- 15 volts at up
to one amp. One pair of amps is clean and appears to work ok at a rough
check - the other pair oscillates at both high and low frequencies, and
one output pair gets very hot. Both with no input but 8 ohm loads. All the
transistors check out ok on my tester and removing either of the ICs
doesn't stop the other oscillating. Given there are really no common
components between the amps I'm a bit lost.

We are talking about the circuits in the jpg? They really oscillate
with the ICs taken out? That makes no sense. One thing that does
bother me on your schematic is the 18pF capacitors (C131 & C133) in
parallel with the 33k feedback resistors. I can't imagine what they
are there for, and they certainly won't be helping the stability.

d

Yamaha DSP A2070
 

In article , Don Pearce
wrote:
On Wed, 18 Nov 2009 23:34:46 +0000 (GMT), "Dave Plowman (News)"
wrote:

In article , Don Pearce
wrote:
Here's the diagram of the offending amp.
http://s139.photobucket.com/albums/q...kies/A2070.jpg

Ah, they have followed the application note in the data sheet

http://www.datasheetcatalog.org/datasheet/nec/UPC1225H.pdf'

pretty well. They have dropped some of the good stuff, though, that
could well do with going back in. The output stability choke and
resistor plus Zobel network, for example, which you will find on page
10. Anyway, there are not too many bits to check, the bases and
collectors of all those transistors will do, and compare them with a
good channel. Most likely it is the IC that has keeled over.

I've got the amps out - bit of a fight to do so.

Two stereo amps - identical circuits but sort of mirror image layout.

Only near suitable power supply I have to hand produces +/- 15 volts at
up to one amp. One pair of amps is clean and appears to work ok at a
rough check - the other pair oscillates at both high and low
frequencies, and one output pair gets very hot.

It oscillates symultaneously at *two* frequencies - low and high? What are
the frequencies? Sounds like one may be driving the other via something
like triggering 'protection' or dragging down the rails.

If using with a bench supply, I'd recommend using some caps as well to
ensure the rails are low impedance in the audio and rf ranges. If you don't
do that, what you see may be the amp interacting with the rails your bench
supply and wiring are providing. That can be quite complex if your bench
supply also has something like foldback limiting.

Both with no input but
8 ohm loads. All the transistors check out ok on my tester

Did you take the transistors out of circuit to check?

and removing either of the ICs doesn't stop the other oscillating.
Given there are really no common components between the amps I'm a bit
lost.

We are talking about the circuits in the jpg? They really oscillate with
the ICs taken out? That makes no sense.

That does make is sound like the problem is in that area. So far as I could
see when I looked at the diagram there would be no supply of current for
the bases of the o/p devices if you lifted out the IC. So what would define
what voltage and current they see? You may simply be relying on leakage
effects, etc.

Is there anything actually connected to the bases of the two o/p devices
when the IC is removed?

What is the voltage between the two emitters of the o/p devices? I am
wondering if it is going above the level that should be triggering the
sensor device, even if only briefly. So producing a relaxation oscillator
at the rate set by the response delays in that part of the circuit.

One thing that does bother me on
your schematic is the 18pF capacitors (C131 & C133) in parallel with the
33k feedback resistors. I can't imagine what they are there for, and
they certainly won't be helping the stability.

I'm less sure of that. I have often found that a small shunt cap there
improves the performance of a design as it corrects loop changes elsewhere.
No idea in this case, though, as I don't really know the details of the IC
behaviour.

FWIW I have in the past seen power amp oscillations due to effects like.

A) The cap between the output bases acting like a resonant load+antenna

B) Not having RF caps between the power lines and/or ground to stop the
rails at the amp waving up and down at RF.

I have no idea if these would be relevant here. But they show that there
are a number of ways for HF oscillations to arise.

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

Yamaha DSP A2070
 

In article ,
Don Pearce wrote:
Only near suitable power supply I have to hand produces +/- 15 volts at up
to one amp. One pair of amps is clean and appears to work ok at a rough
check - the other pair oscillates at both high and low frequencies, and
one output pair gets very hot. Both with no input but 8 ohm loads. All the
transistors check out ok on my tester and removing either of the ICs
doesn't stop the other oscillating. Given there are really no common
components between the amps I'm a bit lost.

We are talking about the circuits in the jpg? They really oscillate
with the ICs taken out? That makes no sense. One thing that does
bother me on your schematic is the 18pF capacitors (C131 & C133) in
parallel with the 33k feedback resistors. I can't imagine what they
are there for, and they certainly won't be helping the stability.

Each amp still oscillates with the other IC removed.

--
*When blondes have more fun, do they know it?

Dave Plowman London SW
To e-mail, change noise into sound.

Yamaha DSP A2070
 

On Thu, 19 Nov 2009 10:39:06 +0000 (GMT), "Dave Plowman (News)"
wrote:

In article ,
Don Pearce wrote:
Only near suitable power supply I have to hand produces +/- 15 volts at up
to one amp. One pair of amps is clean and appears to work ok at a rough
check - the other pair oscillates at both high and low frequencies, and
one output pair gets very hot. Both with no input but 8 ohm loads. All the
transistors check out ok on my tester and removing either of the ICs
doesn't stop the other oscillating. Given there are really no common
components between the amps I'm a bit lost.

We are talking about the circuits in the jpg? They really oscillate
with the ICs taken out? That makes no sense. One thing that does
bother me on your schematic is the 18pF capacitors (C131 & C133) in
parallel with the 33k feedback resistors. I can't imagine what they
are there for, and they certainly won't be helping the stability.

Each amp still oscillates with the other IC removed.

Hang on. Do you mean that the oscillating amp still has its IC fitted?
You can only stop oscillation completely by removing both ICs?

d

Yamaha DSP A2070
 

In article ,
Don Pearce wrote:
On Thu, 19 Nov 2009 10:39:06 +0000 (GMT), "Dave Plowman (News)"
wrote:

In article , Don Pearce
wrote:
Only near suitable power supply I have to hand produces +/- 15 volts
at up to one amp. One pair of amps is clean and appears to work ok
at a rough check - the other pair oscillates at both high and low
frequencies, and one output pair gets very hot. Both with no input
but 8 ohm loads. All the transistors check out ok on my tester and
removing either of the ICs doesn't stop the other oscillating. Given
there are really no common components between the amps I'm a bit
lost.

We are talking about the circuits in the jpg? They really oscillate
with the ICs taken out? That makes no sense. One thing that does
bother me on your schematic is the 18pF capacitors (C131 & C133) in
parallel with the 33k feedback resistors. I can't imagine what they
are there for, and they certainly won't be helping the stability.

Each amp still oscillates with the other IC removed.

Hang on. Do you mean that the oscillating amp still has its IC fitted?
You can only stop oscillation completely by removing both ICs?

Yes. Both amps oscillate. Removing either IC obviously stops that amp
oscillating but not the other one.

d

--
*Some people are only alive because it is illegal to kill.

Dave Plowman London SW
To e-mail, change noise into sound.

Yamaha DSP A2070
 

On Thu, 19 Nov 2009 11:10:26 +0000 (GMT), "Dave Plowman (News)"
wrote:

In article ,
Don Pearce wrote:
On Thu, 19 Nov 2009 10:39:06 +0000 (GMT), "Dave Plowman (News)"
wrote:

In article , Don Pearce
wrote:
Only near suitable power supply I have to hand produces +/- 15 volts
at up to one amp. One pair of amps is clean and appears to work ok
at a rough check - the other pair oscillates at both high and low
frequencies, and one output pair gets very hot. Both with no input
but 8 ohm loads. All the transistors check out ok on my tester and
removing either of the ICs doesn't stop the other oscillating. Given
there are really no common components between the amps I'm a bit
lost.

We are talking about the circuits in the jpg? They really oscillate
with the ICs taken out? That makes no sense. One thing that does
bother me on your schematic is the 18pF capacitors (C131 & C133) in
parallel with the 33k feedback resistors. I can't imagine what they
are there for, and they certainly won't be helping the stability.

Each amp still oscillates with the other IC removed.

Hang on. Do you mean that the oscillating amp still has its IC fitted?
You can only stop oscillation completely by removing both ICs?

Yes. Both amps oscillate. Removing either IC obviously stops that amp
oscillating but not the other one.

Ok, so you have two possibilities. Either both amps have the same
fault, or there is a single fault in some common circuitry. It is
worth having a look at the power supply caps. If they aren't
decoupling as they should it is quite possible that they could cause
the amps to oscillate.

Just to be sure, could you try one of the known good amps in this
board to see if you get oscillation.

I'm reminded of a car fault years ago - no sidelights. Having examined
the wiring loom, switches, fuses etc, I found the real problem which
was that every bulb was blown.

d

Yamaha DSP A2070
 

In article ,
Jim Lesurf wrote:
Only near suitable power supply I have to hand produces +/- 15 volts
at up to one amp. One pair of amps is clean and appears to work ok at
a rough check - the other pair oscillates at both high and low
frequencies, and one output pair gets very hot.

It oscillates symultaneously at *two* frequencies - low and high? What
are the frequencies?
One is approx 100 Hz - the other supersonic and only visible on the
'scope.

Sounds like one may be driving the other via
something like triggering 'protection' or dragging down the rails.

Both rails read the supply voltages.

If using with a bench supply, I'd recommend using some caps as well to
ensure the rails are low impedance in the audio and rf ranges. If you
don't do that, what you see may be the amp interacting with the rails
your bench supply and wiring are providing. That can be quite complex if
your bench supply also has something like foldback limiting.

It's a pretty basic home made bench supply - just using a pair of old RS
cards. The connector on the PCB is a standard type and I had a suitable
plug so made up a loom to use that. And as I said one pair of amps is
fine, but not the other. I'd say my loom is in fact shorter than the one
in the amp from the power supply. But I can easily add some smoothing
caps.

Both with no input but 8 ohm loads. All the transistors check out ok
on my tester

Did you take the transistors out of circuit to check?

No - that's the next thing.

and removing either of the ICs doesn't stop the other oscillating.
Given there are really no common components between the amps I'm a
bit lost.

We are talking about the circuits in the jpg? They really oscillate
with the ICs taken out? That makes no sense.

That does make is sound like the problem is in that area. So far as I
could see when I looked at the diagram there would be no supply of
current for the bases of the o/p devices if you lifted out the IC. So
what would define what voltage and current they see? You may simply be
relying on leakage effects, etc.

Is there anything actually connected to the bases of the two o/p devices
when the IC is removed?

No.

What is the voltage between the two emitters of the o/p devices? I am
wondering if it is going above the level that should be triggering the
sensor device, even if only briefly. So producing a relaxation oscillator
at the rate set by the response delays in that part of the circuit.

I'll check later and report back.

One thing that does bother me on
your schematic is the 18pF capacitors (C131 & C133) in parallel with the
33k feedback resistors. I can't imagine what they are there for, and
they certainly won't be helping the stability.

I'm less sure of that. I have often found that a small shunt cap there
improves the performance of a design as it corrects loop changes
elsewhere. No idea in this case, though, as I don't really know the
details of the IC behaviour.

FWIW I have in the past seen power amp oscillations due to effects like.

A) The cap between the output bases acting like a resonant load+antenna

B) Not having RF caps between the power lines and/or ground to stop the
rails at the amp waving up and down at RF.

I have no idea if these would be relevant here. But they show that there
are a number of ways for HF oscillations to arise.

I'll be playing with it later today. Luckily it's not urgent.

--
*Change is inevitable, except from a vending machine.

Dave Plowman London SW
To e-mail, change noise into sound.

Yamaha DSP A2070
 

"Dave Plowman (News)" wrote

Yes. Both amps oscillate. Removing either IC obviously stops that amp
oscillating but not the other one.


I suspect that Jim is correct, and that the oscillation is due to your bench
supply and it's wiring. Try decoupling the supply rails to ground as close
to the board as possible with both largish electrolytics and ceramic caps of
100nF or thereabouts.

David.

Yamaha DSP A2070
 

In article , Dave Plowman (News)
wrote:
In article , Jim Lesurf
wrote:
Only near suitable power supply I have to hand produces +/- 15
volts at up to one amp. One pair of amps is clean and appears to
work ok at a rough check - the other pair oscillates at both high
and low frequencies, and one output pair gets very hot.

It oscillates symultaneously at *two* frequencies - low and high? What
are the frequencies?
One is approx 100 Hz - the other supersonic and only visible on the
'scope.

Is the 'approx 100Hz' actually mains ripple frequency? Or just close to
100Hz? To me it sounds like one oscillation is causing the other via an
effect like PSU loading or relaxation oscillations.

Sounds like one may be driving the other via something like triggering
'protection' or dragging down the rails.

Both rails read the supply voltages.

Read how? Easy for a dc meter to show an everage that doesn't indicate
variations.

If using with a bench supply, I'd recommend using some caps as well to
ensure the rails are low impedance in the audio and rf ranges. If you
don't do that, what you see may be the amp interacting with the rails
your bench supply and wiring are providing. That can be quite complex
if your bench supply also has something like foldback limiting.

It's a pretty basic home made bench supply - just using a pair of old RS
cards. The connector on the PCB is a standard type and I had a suitable
plug so made up a loom to use that. And as I said one pair of amps is
fine, but not the other. I'd say my loom is in fact shorter than the one
in the amp from the power supply. But I can easily add some smoothing
caps.

FWIW I would always use caps neat the amp board to stop the rails being
waggled at HF. A bench PSU and wires from it can have too large an
impedance at RF to keep the rails presented to the amp steady.

Similarly, if using a meter to watch the rail current I'd always shunt the
meter with something like a 1microF non-electrolytic. Otherwise spurious
oscillations are likely in my experience.

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html