Hi,

I'm a bit puzzled by this blown power amp, as the design of these things is
not really my forte.

(I have posted the schematic over on alt.binaries.schematics.electronic)

The problem is around the two Lateral Mosfets Q16 and Q21 (EXC10N20 /
10P20).
R36 from drain to -70v is complete toast, R35 from the N channel to +70v is
OK.
R33 on the P channel gate is also fried, as is R30 on the other gate, but
not quite as drastically.

Now, R36 has been replaced before by someone, but I'm fairly sure they
didn't replace either of the Mosfets or any other components.
Also, none of the SCA / 2SC output devices or their emitter resistors appear
to be faulty, though I have so far just done a cursory check on these and
the usual suspects on a blown output stage.

I have removed both Mosfets, and on a (low voltage) Peak Component Tester,
and they show as both working, which may or may not prove a great deal
really.


Now, I had one of these in a few months ago, and came across a similar
fault, but only R36 AND R33 were burnt to a crisp, indicating I thought, a
blown P Mosfet.
Again this checked out as working on my Peak tester, but I ordered another
from Allen and Heath anyway, which actually took about 3 months to arrive.
Replacing the two resistors and the Mosfet effected a cure, and the unit has
not been brought back in again.


The question is, what failure mode might be causing these symptoms? I don't
fully understand what these Mosfets are actually doing in this circuit, or
whether something else in the circuit is causing these failures. Why is the
gate resistor on the N channel burnt?

Another major problem is actually getting these Mosfets - you can't, as far
as I can tell, and Allen and Heath no longer supply parts, instead having
subbed out parts supply to various "service centres", who can sell you a
complete amplifier PCB, but not individual components.


Sorry if this is a bit long winded.



Cheers,

Gareth

"Gareth Magennis"

The problem is around the two Lateral Mosfets Q16 and Q21 (EXC10N20 /
10P20).

Another major problem is actually getting these Mosfets - you can't, as
far as I can tell, and Allen and Heath no longer supply parts,

** Semelab BUZ901P and BUX906P are the same devices made in the same
factory.

http://uk.farnell.com/semelab/buz901...83?Ntt=buz901p

http://uk.farnell.com/semelab/buz906...85?Ntt=buz906p

Can you post that schem somewhere other than ABSE ??

Like most here, I have no access.


...... Phil

"Phil Allison"
"Gareth Magennis"

The problem is around the two Lateral Mosfets Q16 and Q21 (EXC10N20 /
10P20).

Another major problem is actually getting these Mosfets - you can't, as
far as I can tell, and Allen and Heath no longer supply parts,


** You did not look very hard ..

http://www.profusionplc.com/pro/gex/...teral%20mosfet

The " R " suffix is of no consequence for your amp, just an improved
version.


.... Phil

Gareth Magennis wrote in message
...
Hi,

I'm a bit puzzled by this blown power amp, as the design of these things
is
not really my forte.

(I have posted the schematic over on alt.binaries.schematics.electronic)

The problem is around the two Lateral Mosfets Q16 and Q21 (EXC10N20 /
10P20).
R36 from drain to -70v is complete toast, R35 from the N channel to +70v
is
OK.
R33 on the P channel gate is also fried, as is R30 on the other gate, but
not quite as drastically.

Now, R36 has been replaced before by someone, but I'm fairly sure they
didn't replace either of the Mosfets or any other components.
Also, none of the SCA / 2SC output devices or their emitter resistors
appear
to be faulty, though I have so far just done a cursory check on these and
the usual suspects on a blown output stage.

I have removed both Mosfets, and on a (low voltage) Peak Component Tester,
and they show as both working, which may or may not prove a great deal
really.


Now, I had one of these in a few months ago, and came across a similar
fault, but only R36 AND R33 were burnt to a crisp, indicating I thought, a
blown P Mosfet.
Again this checked out as working on my Peak tester, but I ordered another
from Allen and Heath anyway, which actually took about 3 months to arrive.
Replacing the two resistors and the Mosfet effected a cure, and the unit
has
not been brought back in again.


The question is, what failure mode might be causing these symptoms? I
don't
fully understand what these Mosfets are actually doing in this circuit, or
whether something else in the circuit is causing these failures. Why is
the
gate resistor on the N channel burnt?

Another major problem is actually getting these Mosfets - you can't, as
far
as I can tell, and Allen and Heath no longer supply parts, instead having
subbed out parts supply to various "service centres", who can sell you a
complete amplifier PCB, but not individual components.


Sorry if this is a bit long winded.



Cheers,

Gareth



Can you repost schematic as gif/pdf or whatever on a regular hosting site

Schematic is he

http://postimage.org/image/4zramf0lh/





Gareth.

"Phil Allison" wrote in message
...

"Gareth Magennis"

The problem is around the two Lateral Mosfets Q16 and Q21 (EXC10N20 /
10P20).

Another major problem is actually getting these Mosfets - you can't, as
far as I can tell, and Allen and Heath no longer supply parts,

** Semelab BUZ901P and BUX906P are the same devices made in the same
factory.

http://uk.farnell.com/semelab/buz901...83?Ntt=buz901p

http://uk.farnell.com/semelab/buz906...85?Ntt=buz906p

Can you post that schem somewhere other than ABSE ??

Like most here, I have no access.


..... Phil





Thanks, that's some of the info I was hoping for.

http://postimage.org/image/4zramf0lh/


Gareth.

"Gareth Magennis"
Schematic is he

http://postimage.org/image/4zramf0lh/


** Comes up blank.



..... Phil

"Phil Allison" wrote in message
...

"Gareth Magennis"
Schematic is he

http://postimage.org/image/4zramf0lh/


** Comes up blank.



.... Phil





Hmm, not much good at this, I think that was an HTML link, works on my PC
though.

Try this, might actually be a picture.

http://s17.postimage.org/3o3gjr0z3/pa_cp_amplifier.jpg


Gareth.

"Gareth Magennis

http://postimage.org/image/4zramf0lh/


** Comes up blank.


Hmm, not much good at this, I think that was an HTML link, works on my PC
though.

Try this, might actually be a picture.

http://s17.postimage.org/3o3gjr0z3/pa_cp_amplifier.jpg


** Blank again.


.... Phil

On Wed, 23 May 2012 18:19:37 +1000, "Phil Allison"
put finger to keyboard and composed:

Try this, might actually be a picture.

http://s17.postimage.org/3o3gjr0z3/pa_cp_amplifier.jpg


** Blank again.

I can see it, but the resolution is crap.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

On Wed, 23 May 2012 18:38:53 +1000, Franc Zabkar
put finger to keyboard and composed:

On Wed, 23 May 2012 18:19:37 +1000, "Phil Allison"
put finger to keyboard and composed:

Try this, might actually be a picture.

http://s17.postimage.org/3o3gjr0z3/pa_cp_amplifier.jpg


** Blank again.

I can see it, but the resolution is crap.

This is better:
http://s18.postimage.org/vktthz2yv/pa_cp_amplifier.jpg

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

"Franc Zabkar"

** Blank again.

I can see it, but the resolution is crap.

This is better:
http://s18.postimage.org/vktthz2yv/pa_cp_amplifier.jpg



** Blank for me.

Sure you do not have to join up with this site first ?



..... Phil

"Franc Zabkar" wrote in message
...
On Wed, 23 May 2012 18:19:37 +1000, "Phil Allison"
put finger to keyboard and composed:

Try this, might actually be a picture.

http://s17.postimage.org/3o3gjr0z3/pa_cp_amplifier.jpg


** Blank again.

I can see it, but the resolution is crap.

- Franc Zabkar


The resolution things a bit weird - if you click on my first post of the
HTML link, it does look crap, but when you zoom in and zoom back out again
(Windows 7 zooming), it magically reverts to full res, as I uploaded it.

The jpeg link seems permanently crippled at low res though.


I tried a few sites last night trying to post this schematic, but many were
full of crap - links that took you where you didn't want to go etc.

Why is something so easy so difficult?

Cuh.


Gareth.

On Wed, 23 May 2012 18:44:58 +1000, "Phil Allison"
put finger to keyboard and composed:


"Franc Zabkar"

** Blank again.

I can see it, but the resolution is crap.

This is better:
http://s18.postimage.org/vktthz2yv/pa_cp_amplifier.jpg



** Blank for me.

Sure you do not have to join up with this site first ?

Nope. I use Opera.

However, if I use OffByOne (which doesn't support Javascript), I don't
see anything and I get a "No connection to ..." message.

Also, Firefox starts to display the image box, but then reverts to a
blank screen and a new URL of
http://postimage.org/image/4zramf0lh/full/

Anyway, I got sick of farting around and uploaded it to my own web
space (why can't everybody do this?):
http://www.users.on.net/~fzabkar/tem...ath_CP1000.jpg

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

On Wed, 23 May 2012 09:50:31 +0100, "Gareth Magennis"
put finger to keyboard and composed:

Why is something so easy so difficult?

Why not upload it to a site that specialises in hosting free service
manuals, eg ElektroTanya or eserviceinfo.com, rather than some
annoying, ad infested image/file hosting site?

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

"Franc Zabkar"

This is better:
http://s18.postimage.org/vktthz2yv/pa_cp_amplifier.jpg


** Blank for me.

Sure you do not have to join up with this site first ?

Nope. I use Opera.

However, if I use OffByOne (which doesn't support Javascript), I don't
see anything and I get a "No connection to ..." message.

Also, Firefox starts to display the image box, but then reverts to a
blank screen and a new URL of
http://postimage.org/image/4zramf0lh/full/

Anyway, I got sick of farting around and uploaded it to my own web
space (why can't everybody do this?):
http://www.users.on.net/~fzabkar/tem...ath_CP1000.jpg


** OK, the original ( low res) link will not open but works if I save it
first.

Same goes for the one you provided above - which is high quality.

Many thanks.



..... Phil

"Franc Zabkar" wrote in message
...
On Wed, 23 May 2012 09:50:31 +0100, "Gareth Magennis"
put finger to keyboard and composed:

Why is something so easy so difficult?

Why not upload it to a site that specialises in hosting free service
manuals, eg ElektroTanya or eserviceinfo.com, rather than some
annoying, ad infested image/file hosting site?

- Franc Zabkar
--


Didn't think of eserviceinfo, which I use a fair bit, but will do that next
time.
I thought this would just work, or most would be able to access ABSE
instead.

Thanks for taking the trouble to put it on your site.


Cheers,

Gareth.

Phil Allison wrote:

"Gareth Magennis"

The problem is around the two Lateral Mosfets Q16 and Q21 (EXC10N20 /
10P20).

Another major problem is actually getting these Mosfets - you can't, as
far as I can tell, and Allen and Heath no longer supply parts,

** Semelab BUZ901P and BUX906P are the same devices made in the same
factory.

http://uk.farnell.com/semelab/buz901...83?Ntt=buz901p

http://uk.farnell.com/semelab/buz906...85?Ntt=buz906p

Can you post that schem somewhere other than ABSE ??

Like most here, I have no access.

..... Phil


You can access ABSE at this URL, but someone flooded the group with a
couple gigabytes of outdated software and the post may not be there.
Also, it can take days for a message to appear, but it's better than
nothing.

http://www.usenet-replayer.com/cgi/content/vsq?days=1&groups=alt.binaries.schematics.electron ic



--
You can't have a sense of humor, if you have no sense.

"Phil Allison"

http://www.users.on.net/~fzabkar/tem...ath_CP1000.jpg


** In the above schem, 2.7 ohm resistors R35 and R36 are each bridged by the
B-E junctions of three power transistors and their associated 0.15 ohm
emitter resistors. It is therefore not possible for the 2.7ohms to burn if
the connections are good.

In normal operation the standing voltage across each 2.7 ohm is about 0.4
volts - ie not enough to bias the output transistors into conduction. So
the two mosfets drive the load at low power and the other devices come in to
assist above about 100mW into 8 ohms.

Is that schem really correct ???

Hard to believe there are no zeners protecting the gates of the mosfets -
if the output is shorted, the gates can be overdriven and the mosfets
destroyed.

Also hard to believe the rail fuses connect DC to the *whole* amplifier -
cos if one rail fuse blows, the output looks bound to swing over to the
other rail resulting in relay and speaker destruction.

If instead, the rail fuses were only in the supplies to the 6 power
transistors and DC is applied permanently to the rest of the amplifier -
then burnt R35 and R36 resistors and dead mosfets are easily explained. It
happens whenever fuses blow and these parts then have to carry all the
current to the load.

Finally, 1 kohm resistors R30 and R 33 can burn if the associated mosfet
fails - cos the dead mosfet will have a low resistance from Gate to Source
so Q15 and Q20 will drive current through them to the load.


..... Phil

"Phil Allison" wrote in message
...

"Phil Allison"

http://www.users.on.net/~fzabkar/tem...ath_CP1000.jpg


** In the above schem, 2.7 ohm resistors R35 and R36 are each bridged by
the B-E junctions of three power transistors and their associated 0.15 ohm
emitter resistors. It is therefore not possible for the 2.7ohms to burn if
the connections are good.

In normal operation the standing voltage across each 2.7 ohm is about 0.4
volts - ie not enough to bias the output transistors into conduction. So
the two mosfets drive the load at low power and the other devices come in
to assist above about 100mW into 8 ohms.

Is that schem really correct ???

Hard to believe there are no zeners protecting the gates of the mosfets -
if the output is shorted, the gates can be overdriven and the mosfets
destroyed.

Also hard to believe the rail fuses connect DC to the *whole* amplifier -
cos if one rail fuse blows, the output looks bound to swing over to the
other rail resulting in relay and speaker destruction.

If instead, the rail fuses were only in the supplies to the 6 power
transistors and DC is applied permanently to the rest of the amplifier -
then burnt R35 and R36 resistors and dead mosfets are easily explained.
It happens whenever fuses blow and these parts then have to carry all the
current to the load.

Finally, 1 kohm resistors R30 and R 33 can burn if the associated mosfet
fails - cos the dead mosfet will have a low resistance from Gate to
Source so Q15 and Q20 will drive current through them to the load.


.... Phil





Thanks very much for the analysis.

Changed Mosfets thursday, but has now blown 4 out of 6 output devices on
power up. (no load)

Been away since then, will tackle this very carefully again wednesday.


The 2 fuses had not blown by the way, the front panel limit LED was lit
though. They are blown now.



It looks from your analysis and the schematic, that the amp should power up
into its quiescent state with no 2SC/2SA devices present at all. (with
R38/41 in circuit)
Is this correct?
Is this wise or helpful?!


Gareth.

"Gareth Magennis"

Changed Mosfets thursday, but has now blown 4 out of 6 output devices on
power up. (no load)


** You might have swapped the N and P channel mosfets.

FFS !!!

Always power up gradually with a variac and small value rail fuses after
repairing a blown up power amplifier.

Monitor the AC supply current OR the DC supply current at the same time too.

It looks from your analysis and the schematic, that the amp should power
up
into its quiescent state with no 2SC/2SA devices present at all.

** Yep.

Now answer my questions:

Are the DC rail fuses connected as shown in the schem ?

Are there really NO zeners on the mosfet gates ?


.... Phil

"Phil Allison" wrote in message
...

"Gareth Magennis"

Changed Mosfets thursday, but has now blown 4 out of 6 output devices on
power up. (no load)


** You might have swapped the N and P channel mosfets.

FFS !!!

Always power up gradually with a variac and small value rail fuses after
repairing a blown up power amplifier.

Monitor the AC supply current OR the DC supply current at the same time
too.

It looks from your analysis and the schematic, that the amp should power
up
into its quiescent state with no 2SC/2SA devices present at all.

** Yep.

Now answer my questions:

Are the DC rail fuses connected as shown in the schem ?

Are there really NO zeners on the mosfet gates ?


... Phil



1. The devices were the right way round. I marked the heatsink before
removing the originals.
2. The amp was carefully powered up on a variac and 100W series lightbulb.
Relays clicked and no limit LED this time on front panel.
3. Prior to that I had used a +-30v bench supply on the PCB, no current
draw at all.
4. With lightbulb disabled, amp then blew a few seconds after power up.
Yes, stupidly I did not fit 500mA fuses as I usually do. That'll teach me
not to be so cocky. Doh!

The original fuses were intact and did NOT blow on power up - just that
sides limit LED lit.
They are blown now.

No Mosfet resistors are burnt this time, the Mosfets again read good on my
tester, but that seemingly is meaningless.


I will check out your other questions later.



Cheers,

Gareth.

"Phil Allison" wrote in message
...

"Gareth Magennis"

Changed Mosfets thursday, but has now blown 4 out of 6 output devices on
power up. (no load)


** You might have swapped the N and P channel mosfets.

FFS !!!

Always power up gradually with a variac and small value rail fuses after
repairing a blown up power amplifier.

Monitor the AC supply current OR the DC supply current at the same time
too.

It looks from your analysis and the schematic, that the amp should power
up
into its quiescent state with no 2SC/2SA devices present at all.

** Yep.

Now answer my questions:

Are the DC rail fuses connected as shown in the schem ?

Are there really NO zeners on the mosfet gates ?


... Phil



DC rails are as per the schematic.

There are no gate zeners.



Cheers,

Gareth.