Hi,

I have a Trace Elliot 300W bass amp here that uses 2 pairs of BUZ901/906
mosfets. All output devices have been resoldered or replaced at some point,
and the amp now contains 2 x BUZ901 (SML) and 2 marked STUD1P.

I've repaired a few of these in my time but never come across this marking.
Is it a Trace internal number, or have these been substituted in from
elsewhere? Are they actually BUZ906?

(One BUZ901 is blown, the other 3 check out OK on a PEAK component tester).


Cheers,



Gareth.

"Gareth Magennis is a ****ING MORON "


I have a Trace Elliot 300W bass amp here


** I suppose you whole life is over if you tell us the ****ing model number
?


that uses 2 pairs of BUZ901/906 mosfets. All output devices have been
resoldered or replaced at some point, and the amp now contains 2 x BUZ901
(SML) and 2 marked STUD1P.


** The device is labelled " STUPID " ???

April 1st has gone you know.


One BUZ901 is blown,


** I suppose your whole life depends of not telling us what YOU mean by the
word "blown" ??

Lateral MOSFETS fail ( if they fail) in a unique way - not like any
other power devices.

WHAT IS IT ???




..... Phil

Gareth Magennis wrote in message
news
Hi,

I have a Trace Elliot 300W bass amp here that uses 2 pairs of BUZ901/906
mosfets. All output devices have been resoldered or replaced at some
point,
and the amp now contains 2 x BUZ901 (SML) and 2 marked STUD1P.

I've repaired a few of these in my time but never come across this
marking.
Is it a Trace internal number, or have these been substituted in from
elsewhere? Are they actually BUZ906?

(One BUZ901 is blown, the other 3 check out OK on a PEAK component
tester).


Cheers,



Gareth.




While at it toroidal mains Tx ? any inrush limitation ? no ken those
markings.
Trace Elliot GP12 SMX, 1998 300W bass amp last week PA was marked "Bipolar
Bear" on overlay and devices under spring clips, not seen , but checked out
as BCE npn & pnp devices

"Phil Allison" wrote in message
...

"Gareth Magennis is a ****ING MORON "


I have a Trace Elliot 300W bass amp here


** I suppose you whole life is over if you tell us the ****ing model
number ?


that uses 2 pairs of BUZ901/906 mosfets. All output devices have been
resoldered or replaced at some point, and the amp now contains 2 x BUZ901
(SML) and 2 marked STUD1P.


** The device is labelled " STUPID " ???

April 1st has gone you know.


One BUZ901 is blown,


** I suppose your whole life depends of not telling us what YOU mean by
the word "blown" ??

Lateral MOSFETS fail ( if they fail) in a unique way - not like any
other power devices.

WHAT IS IT ???




.... Phil






You do make me laugh, Phil.


If it matters, it is a 7215 combo, i.e. a GP7 300 Watt SM series, but I
believe all Trace Amps share the same 2 amplifier modules.
There's a Bi-polar Bear which is Bi-polar (you might like that one), or the
150/300W Mosfet board which is Mosfet. I guess the 150W one uses just 2
Mosfets and smaller PSU caps.


But the question is only about the markings on the output devices, and
whether this STUD1P device is the original Trace part or some numpty
repairman has retrofitted a dodgy part he manufactured himself in his
kitchen of took it out of a Studiomaster or summat.



Sigh.

p.s. it has been repaired at a place I know about that is well known for
dodgy repairs, and the soldering is absolutely shocking.

On Wed, 06 Apr 2011 15:45:02 +0100, Gareth Magennis wrote:

p.s. it has been repaired at a place I know about that is well known for
dodgy repairs, and the soldering is absolutely shocking.

Why not just replace all 4 with known pairs? TT manufactures
a bunch of different audio mosfet devices
http://www.semelab.com/magnatec/alfet.shtml



--
Live Fast Die Young, Leave A Pretty Corpse

"N_Cook" wrote in message
...
Gareth Magennis wrote in message
news
Hi,

I have a Trace Elliot 300W bass amp here that uses 2 pairs of BUZ901/906
mosfets. All output devices have been resoldered or replaced at some
point,
and the amp now contains 2 x BUZ901 (SML) and 2 marked STUD1P.

I've repaired a few of these in my time but never come across this
marking.
Is it a Trace internal number, or have these been substituted in from
elsewhere? Are they actually BUZ906?

(One BUZ901 is blown, the other 3 check out OK on a PEAK component
tester).


Cheers,



Gareth.




While at it toroidal mains Tx ? any inrush limitation ? no ken those
markings.
Trace Elliot GP12 SMX, 1998 300W bass amp last week PA was marked "Bipolar
Bear" on overlay and devices under spring clips, not seen , but checked
out
as BCE npn & pnp devices




Toroid, will get back to you tomorrow with the rest.



Gareth.

"Meat Plow" wrote in message
news
On Wed, 06 Apr 2011 15:45:02 +0100, Gareth Magennis wrote:

p.s. it has been repaired at a place I know about that is well known for
dodgy repairs, and the soldering is absolutely shocking.

Why not just replace all 4 with known pairs? TT manufactures
a bunch of different audio mosfet devices
http://www.semelab.com/magnatec/alfet.shtml



--
Live Fast Die Young, Leave A Pretty Corpse




Yes, that is my preferred option, and I could get them from Farnell
tomorrow.
However, that would cost me about £50 for the four, so perhaps £60 to the
customer, which is getting on for $100.


If there is any doubt what these oddly marked devices actually are I will do
just that.



Cheers,


Gareth.

Ah, I see now you meant alternatives.

Not really my policy I'm afraid. If it blew up again I wouldn't have a leg
to stand on.


Cheers,


Gareth.

On Wed, 06 Apr 2011 20:25:57 +0100, Gareth Magennis wrote:

"Meat Plow" wrote in message
news
On Wed, 06 Apr 2011 15:45:02 +0100, Gareth Magennis wrote:

p.s. it has been repaired at a place I know about that is well known
for dodgy repairs, and the soldering is absolutely shocking.

Why not just replace all 4 with known pairs? TT manufactures a bunch of
different audio mosfet devices
http://www.semelab.com/magnatec/alfet.shtml



--
Live Fast Die Young, Leave A Pretty Corpse




Yes, that is my preferred option, and I could get them from Farnell
tomorrow.
However, that would cost me about £50 for the four, so perhaps £60 to
the customer, which is getting on for $100.


If there is any doubt what these oddly marked devices actually are I
will do just that.



Cheers,


Gareth.



$100 isn't bad for a slayed output stage. Especially on previously bodged
repairs. For that combo and most others including vintage my minimum fee
is $100 before parts. I do more work on vintage than anything. People
drag in these Silvertone, Airline, Ampeg, Gibson etc.. from the late
1950's/60's all the time. If it's a small combo like a GA series
Discovery,Maestro,Explorer etc...I'll slide the $100 minimum fee unless
it needs a restoration. Last Silvertone 50 watt head I restored I charged
$200 labor and the customer never batted an eyelid.

I'd replace all 4 of those devices and take a hard look at the whole thing
probably before I ordered parts.

--
Live Fast Die Young, Leave A Pretty Corpse

"Gareth Magennis"


If there is any doubt what these oddly marked devices actually are I will
do just that.


** What does your PEAK say about the pin out and the Gate threshold
voltage ?

Should be Source on the case and about 0.5 to 1.0 volts.


..... Phil

"Phil Allison" wrote in message
...

"Gareth Magennis"


If there is any doubt what these oddly marked devices actually are I will
do just that.


** What does your PEAK say about the pin out and the Gate threshold
voltage ?

Should be Source on the case and about 0.5 to 1.0 volts.


.... Phil



Doesn't display voltage information, it either says its a "* channel mosfet"
and identifies the pins, or "no valid part found". Its a PIC based device
I believe.

I bought it purely to check Mosfets go or no go, and it works well.


Gareth.

Gareth Magennis wrote in message
news
Hi,

I have a Trace Elliot 300W bass amp here that uses 2 pairs of BUZ901/906
mosfets. All output devices have been resoldered or replaced at some
point,
and the amp now contains 2 x BUZ901 (SML) and 2 marked STUD1P.

I've repaired a few of these in my time but never come across this
marking.
Is it a Trace internal number, or have these been substituted in from
elsewhere? Are they actually BUZ906?

(One BUZ901 is blown, the other 3 check out OK on a PEAK component
tester).


Cheers,



Gareth.




perhaps obtained from these people
http://www.autogarduk.com/stock%20list%2027.htm
41 P and 274 N

What are the test parameters of a Peak?

"N_Cook" wrote in message
...
Gareth Magennis wrote in message
news
Hi,

I have a Trace Elliot 300W bass amp here that uses 2 pairs of BUZ901/906
mosfets. All output devices have been resoldered or replaced at some
point,
and the amp now contains 2 x BUZ901 (SML) and 2 marked STUD1P.

I've repaired a few of these in my time but never come across this
marking.
Is it a Trace internal number, or have these been substituted in from
elsewhere? Are they actually BUZ906?

(One BUZ901 is blown, the other 3 check out OK on a PEAK component
tester).


Cheers,



Gareth.




While at it toroidal mains Tx ? any inrush limitation ? no ken those
markings.
Trace Elliot GP12 SMX, 1998 300W bass amp last week PA was marked "Bipolar
Bear" on overlay and devices under spring clips, not seen , but checked
out
as BCE npn & pnp devices




Toroid and no inrush limit device. The Bipolar Bear uses 2SA1695 and
2SC4468 devices.



Gareth.

"Gareth Magennis"


Doesn't display voltage information, it either says its a "* channel
mosfet" and identifies the pins,


** IS THE *SOURCE* ON THE ****ING CASE OR NOT !!!

You ****ING MORON !!



......Phil

"Phil Allison" wrote in message
...

"Gareth Magennis"


Doesn't display voltage information, it either says its a "* channel
mosfet" and identifies the pins,


** IS THE *SOURCE* ON THE ****ING CASE OR NOT !!!

You ****ING MORON !!



.....Phil





Bye bye.

"N_Cook" wrote in message
...
Gareth Magennis wrote in message
news
Hi,

I have a Trace Elliot 300W bass amp here that uses 2 pairs of BUZ901/906
mosfets. All output devices have been resoldered or replaced at some
point,
and the amp now contains 2 x BUZ901 (SML) and 2 marked STUD1P.

I've repaired a few of these in my time but never come across this
marking.
Is it a Trace internal number, or have these been substituted in from
elsewhere? Are they actually BUZ906?

(One BUZ901 is blown, the other 3 check out OK on a PEAK component
tester).


Cheers,



Gareth.




perhaps obtained from these people
http://www.autogarduk.com/stock%20list%2027.htm
41 P and 274 N

What are the test parameters of a Peak?



Don't have a clue. It runs on a 9v battery.

It does claim to measure hfe, but again I don't know with which parameters.
I don't think its meant to be a precision instrument but I find it very
useful, mostly for finding dead mosfets and possible good ones, and
identifying pinouts of bipolars.

I once had terrible problems with a C-audio amp that I just couldn't get
working. They had loads of BC184 (I think) doing the bias and stuff. I
replaced the lot with BC184L, not realising that the "L" meant the pinout
was different. Took me a looong time to suss that one out.
Now I always double check with the Peak when replacing any signal
transistor, just in case.

Likewise Farnell once sold me some resistors that were a decade lower than
marked on the packaging. Again took me a while to suss out the problem, and
I now always meter EVERY resistor before soldering it in.

Cheers,

Gareth.

Gareth Magennis wrote in message
...

"N_Cook" wrote in message
...
Gareth Magennis wrote in message
news
Hi,

I have a Trace Elliot 300W bass amp here that uses 2 pairs of
BUZ901/906
mosfets. All output devices have been resoldered or replaced at some
point,
and the amp now contains 2 x BUZ901 (SML) and 2 marked STUD1P.

I've repaired a few of these in my time but never come across this
marking.
Is it a Trace internal number, or have these been substituted in from
elsewhere? Are they actually BUZ906?

(One BUZ901 is blown, the other 3 check out OK on a PEAK component
tester).


Cheers,



Gareth.




perhaps obtained from these people
http://www.autogarduk.com/stock%20list%2027.htm
41 P and 274 N

What are the test parameters of a Peak?



Don't have a clue. It runs on a 9v battery.

It does claim to measure hfe, but again I don't know with which
parameters.
I don't think its meant to be a precision instrument but I find it very
useful, mostly for finding dead mosfets and possible good ones, and
identifying pinouts of bipolars.

I once had terrible problems with a C-audio amp that I just couldn't get
working. They had loads of BC184 (I think) doing the bias and stuff. I
replaced the lot with BC184L, not realising that the "L" meant the pinout
was different. Took me a looong time to suss that one out.
Now I always double check with the Peak when replacing any signal
transistor, just in case.

Likewise Farnell once sold me some resistors that were a decade lower than
marked on the packaging. Again took me a while to suss out the problem,
and
I now always meter EVERY resistor before soldering it in.

Cheers,

Gareth.



If mains fuse is designated T3.15 amp, it wil probably need an inrush
fitting sometime in the near future , see thread here last week T-E GP12 SMX
and Studiomaster comparison

Dealt with a heap of C-audio Pulse 2 1100 last year. Stored outdoors (open
sided barn) with the thinking that plastic tarpaulin over the rack with a
pack of silica gel (as found, not activated) inside it in the autumn would
keep them ok over a GB winter.

Except for a few colateral damage burnt out 1/3W R , all the problems were
those TO92 , gone slightly ohmic presumably from damp ingress. Throws out
the bias and then amp always goes into protect.

"N_Cook" wrote in message
...
Gareth Magennis wrote in message
news
Hi,

I have a Trace Elliot 300W bass amp here that uses 2 pairs of BUZ901/906
mosfets. All output devices have been resoldered or replaced at some
point,
and the amp now contains 2 x BUZ901 (SML) and 2 marked STUD1P.

I've repaired a few of these in my time but never come across this
marking.
Is it a Trace internal number, or have these been substituted in from
elsewhere? Are they actually BUZ906?

(One BUZ901 is blown, the other 3 check out OK on a PEAK component
tester).


Cheers,



Gareth.




perhaps obtained from these people
http://www.autogarduk.com/stock%20list%2027.htm
41 P and 274 N

What are the test parameters of a Peak?




Just realised I've got the old version of the Peak. The new one does more
things than mine, like gate threshold voltage.

http://www.rapidonline.com/Tools-Fas...ource=googleps



Gareth.

"N_Cook" wrote in message
...
Gareth Magennis wrote in message
...

"N_Cook" wrote in message
...
Gareth Magennis wrote in message
news Hi,

I have a Trace Elliot 300W bass amp here that uses 2 pairs of
BUZ901/906
mosfets. All output devices have been resoldered or replaced at some
point,
and the amp now contains 2 x BUZ901 (SML) and 2 marked STUD1P.

I've repaired a few of these in my time but never come across this
marking.
Is it a Trace internal number, or have these been substituted in from
elsewhere? Are they actually BUZ906?

(One BUZ901 is blown, the other 3 check out OK on a PEAK component
tester).


Cheers,



Gareth.




perhaps obtained from these people
http://www.autogarduk.com/stock%20list%2027.htm
41 P and 274 N

What are the test parameters of a Peak?



Don't have a clue. It runs on a 9v battery.

It does claim to measure hfe, but again I don't know with which
parameters.
I don't think its meant to be a precision instrument but I find it very
useful, mostly for finding dead mosfets and possible good ones, and
identifying pinouts of bipolars.

I once had terrible problems with a C-audio amp that I just couldn't get
working. They had loads of BC184 (I think) doing the bias and stuff. I
replaced the lot with BC184L, not realising that the "L" meant the pinout
was different. Took me a looong time to suss that one out.
Now I always double check with the Peak when replacing any signal
transistor, just in case.

Likewise Farnell once sold me some resistors that were a decade lower
than
marked on the packaging. Again took me a while to suss out the problem,
and
I now always meter EVERY resistor before soldering it in.

Cheers,

Gareth.



If mains fuse is designated T3.15 amp, it wil probably need an inrush
fitting sometime in the near future , see thread here last week T-E GP12
SMX
and Studiomaster comparison


I contributed to that thread. I'll be sticking a T4 in it instead. ;-)
4 new Mosfets and labour is almost an uneconomical repair as it is.





Dealt with a heap of C-audio Pulse 2 1100 last year. Stored outdoors (open
sided barn) with the thinking that plastic tarpaulin over the rack with a
pack of silica gel (as found, not activated) inside it in the autumn would
keep them ok over a GB winter.

Except for a few colateral damage burnt out 1/3W R , all the problems were
those TO92 , gone slightly ohmic presumably from damp ingress. Throws out
the bias and then amp always goes into protect.



I looked inside 2 of these C audio pulse (nothing to do with C audio really)
and quickly put them back together again and gave them straight back. Fist
time I'd ever done anything like that.



Gareth.

On Thu, 07 Apr 2011 14:03:34 +0100, Gareth Magennis wrote:

"N_Cook" wrote in message
...
Gareth Magennis wrote in message
news
Hi,

I have a Trace Elliot 300W bass amp here that uses 2 pairs of
BUZ901/906 mosfets. All output devices have been resoldered or
replaced at some
point,
and the amp now contains 2 x BUZ901 (SML) and 2 marked STUD1P.

I've repaired a few of these in my time but never come across this
marking.
Is it a Trace internal number, or have these been substituted in from
elsewhere? Are they actually BUZ906?

(One BUZ901 is blown, the other 3 check out OK on a PEAK component
tester).


Cheers,



Gareth.




perhaps obtained from these people
http://www.autogarduk.com/stock%20list%2027.htm 41 P and 274 N

What are the test parameters of a Peak?




Just realised I've got the old version of the Peak. The new one does
more things than mine, like gate threshold voltage.

http://www.rapidonline.com/Tools-Fas...quipment/Test-
Equipment/Analysers-Testers-Counter/Atlas-DCA-Semiconductor-component-
analyser/36637/kw/85-2514?source=googleps&utm_source=googleps



Gareth.

I own a really old Sencore Cricket transitor/fet checker. You don't have
to identify BCE correctly because it has 6 pushbuttons on it. It's still
a useful tool to check leakage and gain.



--
Live Fast Die Young, Leave A Pretty Corpse

On Fri, 08 Apr 2011 10:17:06 +0100, Gareth Magennis wrote:

Interestingly the customer said that the amp went DC and fried his
speaker - the amp has no relay or DC protection. My little bit or
research so far indicates that dv/dt failure in industrial motor
applications can be down to pulling the power cable off a stalled motor,
zapping the gate/substrate connection in a nanosecond. Might this also
apply to a faulty speaker in this kind of audio amp which burns short
then eventually self disconnects?

I also understand extremely electrically noisy environments can also
induce this failure mode. The customer is daisy chaining this amp with
another one via the FX send, I'm wondering if this has any chance of
being the cause of the failure.

Slaving another amp via the send isn't likely but a failing speaker or
poorly wired/connected load might be the reason.



--
Live Fast Die Young, Leave A Pretty Corpse

"Meat Plow" wrote in message
news
On Fri, 08 Apr 2011 10:17:06 +0100, Gareth Magennis wrote:

Interestingly the customer said that the amp went DC and fried his
speaker - the amp has no relay or DC protection. My little bit or
research so far indicates that dv/dt failure in industrial motor
applications can be down to pulling the power cable off a stalled motor,
zapping the gate/substrate connection in a nanosecond. Might this also
apply to a faulty speaker in this kind of audio amp which burns short
then eventually self disconnects?

I also understand extremely electrically noisy environments can also
induce this failure mode. The customer is daisy chaining this amp with
another one via the FX send, I'm wondering if this has any chance of
being the cause of the failure.

Slaving another amp via the send isn't likely but a failing speaker or
poorly wired/connected load might be the reason.




Check this out, if just for educational purposes. Quite interesting.

http://www.4qd.co.uk/serv/MOSFETfail.html



Gareth.

On Sat, 09 Apr 2011 00:12:23 +0100, Gareth Magennis wrote:

"Meat Plow" wrote in message
news
On Fri, 08 Apr 2011 10:17:06 +0100, Gareth Magennis wrote:

Interestingly the customer said that the amp went DC and fried his
speaker - the amp has no relay or DC protection. My little bit or
research so far indicates that dv/dt failure in industrial motor
applications can be down to pulling the power cable off a stalled
motor, zapping the gate/substrate connection in a nanosecond. Might
this also apply to a faulty speaker in this kind of audio amp which
burns short then eventually self disconnects?

I also understand extremely electrically noisy environments can also
induce this failure mode. The customer is daisy chaining this amp
with another one via the FX send, I'm wondering if this has any chance
of being the cause of the failure.

Slaving another amp via the send isn't likely but a failing speaker or
poorly wired/connected load might be the reason.




Check this out, if just for educational purposes. Quite interesting.

http://www.4qd.co.uk/serv/MOSFETfail.html



Gareth.

Mostly pertaining to motor control applications. My experience is limited
to audio applications. Some similarities exist thought. Still a good read
for failure modes.



--
Live Fast Die Young, Leave A Pretty Corpse

"Gareth Magennis"

Check this out, if just for educational purposes. Quite interesting.

http://www.4qd.co.uk/serv/MOSFETfail.html


** Got nothing WHATSOEVER to do with LATERAL mosfets !!!

These are completely different animals to switching mosfets.

Originally invented by Hitachi and sold in TO3 packs as 2SK ( N ch) and 2SJ
( P ch) numbers as the perfect output devices for audio amplifiers. Power
amp makers in the UK, Australia and NZ pounced on them and for a while in
the mid 1980s, MOSFET amps were the industry standard.

Later, Hitachi supplied plastic pack ( TO3P) versions too - then stopped
making the TO3 devices completely in the late 1980s. There was shock, horror
and nashing of teeth until it was discovered that an obscure semiconductor
maker in Scotland ( of all places) was making very similar lateral devices
to Hitachi ( under licence?) - this was Semelab.

Initially, all Semelab devices had SML part numbers on them, but later this
changed to BUZ part numbers. The exact same devices are also sold under the
Exicon brand and Magnatec. Magnatec also supply dual chip versions of the
same devices - this simply doubles the max current and max power
dissipation ratings.

Hitachi lateral mosfets in plastic packs are now sold under the brand "
Renesas " .

A lateral mosfet will always have the source connected to the case or
heatsink tab, have a threshold voltage of about 0.7 volts, come in P and N
channel versions and be found in audio amplifiers and very little else.

OTOH - switching mosfets have the drain connected to the heatsink tab, are
almost all N channel types ( no genuine P ch compliments exist) have
threshold voltages of 2 to 5 volts and are found mostly in switching PSUs,
motor drives and class D amplifiers.



..... Phil

"Meat Plow" wrote in message
news
On Sat, 09 Apr 2011 00:12:23 +0100, Gareth Magennis wrote:

"Meat Plow" wrote in message
news
On Fri, 08 Apr 2011 10:17:06 +0100, Gareth Magennis wrote:

Interestingly the customer said that the amp went DC and fried his
speaker - the amp has no relay or DC protection. My little bit or
research so far indicates that dv/dt failure in industrial motor
applications can be down to pulling the power cable off a stalled
motor, zapping the gate/substrate connection in a nanosecond. Might
this also apply to a faulty speaker in this kind of audio amp which
burns short then eventually self disconnects?

I also understand extremely electrically noisy environments can also
induce this failure mode. The customer is daisy chaining this amp
with another one via the FX send, I'm wondering if this has any chance
of being the cause of the failure.

Slaving another amp via the send isn't likely but a failing speaker or
poorly wired/connected load might be the reason.




Check this out, if just for educational purposes. Quite interesting.

http://www.4qd.co.uk/serv/MOSFETfail.html



Gareth.

Mostly pertaining to motor control applications. My experience is limited
to audio applications. Some similarities exist thought. Still a good read
for failure modes.




Isn't a speaker a motor?



Gareth.

"Gareth Magennis"

Isn't a speaker a motor?


** And chaulk an cheese are the same too.

****wit.


.... Phil

"Gareth Magennis" wrote in message
...


"Meat Plow" wrote in message
news
On Sat, 09 Apr 2011 00:12:23 +0100, Gareth Magennis wrote:

"Meat Plow" wrote in message
news On Fri, 08 Apr 2011 10:17:06 +0100, Gareth Magennis wrote:

Interestingly the customer said that the amp went DC and fried his
speaker - the amp has no relay or DC protection. My little bit or
research so far indicates that dv/dt failure in industrial motor
applications can be down to pulling the power cable off a stalled
motor, zapping the gate/substrate connection in a nanosecond. Might
this also apply to a faulty speaker in this kind of audio amp which
burns short then eventually self disconnects?

I also understand extremely electrically noisy environments can also
induce this failure mode. The customer is daisy chaining this amp
with another one via the FX send, I'm wondering if this has any chance
of being the cause of the failure.

Slaving another amp via the send isn't likely but a failing speaker or
poorly wired/connected load might be the reason.




Check this out, if just for educational purposes. Quite interesting.

http://www.4qd.co.uk/serv/MOSFETfail.html



Gareth.

Mostly pertaining to motor control applications. My experience is limited
to audio applications. Some similarities exist thought. Still a good read
for failure modes.




Isn't a speaker a motor?


Maybe the "speech coil" head actuators in old 5&1/4 FH HDDs blur the
distinction a bit.

On Sat, 09 Apr 2011 00:43:51 +0100, Gareth Magennis wrote:

"Meat Plow" wrote in message
news
On Sat, 09 Apr 2011 00:12:23 +0100, Gareth Magennis wrote:

"Meat Plow" wrote in message
news On Fri, 08 Apr 2011 10:17:06 +0100, Gareth Magennis wrote:

Interestingly the customer said that the amp went DC and fried his
speaker - the amp has no relay or DC protection. My little bit or
research so far indicates that dv/dt failure in industrial motor
applications can be down to pulling the power cable off a stalled
motor, zapping the gate/substrate connection in a nanosecond. Might
this also apply to a faulty speaker in this kind of audio amp which
burns short then eventually self disconnects?

I also understand extremely electrically noisy environments can also
induce this failure mode. The customer is daisy chaining this amp
with another one via the FX send, I'm wondering if this has any
chance of being the cause of the failure.

Slaving another amp via the send isn't likely but a failing speaker
or poorly wired/connected load might be the reason.




Check this out, if just for educational purposes. Quite interesting.

http://www.4qd.co.uk/serv/MOSFETfail.html



Gareth.

Mostly pertaining to motor control applications. My experience is
limited to audio applications. Some similarities exist thought. Still a
good read for failure modes.




Isn't a speaker a motor?



Gareth.

It can be perceived as a sound motor.



--
Live Fast Die Young, Leave A Pretty Corpse

On Sat, 09 Apr 2011 09:42:02 +1000, Phil Allison wrote:


"Gareth Magennis"

Check this out, if just for educational purposes. Quite interesting.

http://www.4qd.co.uk/serv/MOSFETfail.html


** Got nothing WHATSOEVER to do with LATERAL mosfets !!!

These are completely different animals to switching mosfets.

Originally invented by Hitachi and sold in TO3 packs as 2SK ( N ch) and
2SJ ( P ch) numbers as the perfect output devices for audio amplifiers.
Power amp makers in the UK, Australia and NZ pounced on them and for a
while in the mid 1980s, MOSFET amps were the industry standard.

I cried the day Hitachi quit manufacturing the 2SK49 / 2SJ135 pairs. I
have one old SCS 350w/ch power amp that uses them. I acquired it with a
blow channel back in the late 90's. It wasn't easy back then to find
replacements. I imagine it's probably a lot easier these days to find NOS
replacements. Maybe even a sub is made today?



--
Live Fast Die Young, Leave A Pretty Corpse

"Phil Allison" wrote in message
...

"Gareth Magennis"

Isn't a speaker a motor?


** And chaulk an cheese are the same too.

****wit.


... Phil



Oh and there's me thinking a motor is a device that converts electrical
energy into mechanical energy.


****wit.

On 4/9/2011 9:06 AM Meat Plow spake thus:

On Sat, 09 Apr 2011 00:43:51 +0100, Gareth Magennis wrote:

"Meat Plow" wrote in message
news
On Sat, 09 Apr 2011 00:12:23 +0100, Gareth Magennis wrote:

"Meat Plow" wrote in message
news
On Fri, 08 Apr 2011 10:17:06 +0100, Gareth Magennis wrote:

Interestingly the customer said that the amp went DC and
fried his speaker - the amp has no relay or DC protection.
My little bit or research so far indicates that dv/dt
failure in industrial motor applications can be down to
pulling the power cable off a stalled motor, zapping the
gate/substrate connection in a nanosecond. Might this also
apply to a faulty speaker in this kind of audio amp which
burns short then eventually self disconnects?

I also understand extremely electrically noisy environments
can also induce this failure mode. The customer is daisy
chaining this amp with another one via the FX send, I'm
wondering if this has any chance of being the cause of the
failure.

Slaving another amp via the send isn't likely but a failing
speaker or poorly wired/connected load might be the reason.

Check this out, if just for educational purposes. Quite
interesting.

http://www.4qd.co.uk/serv/MOSFETfail.html

Mostly pertaining to motor control applications. My experience is
limited to audio applications. Some similarities exist thought.
Still a good read for failure modes.

Isn't a speaker a motor?

It can be perceived as a sound motor.

It *is*, in fact, a motor. Which, technically speaking, is anything that
converts electrical energy to mechanical motion. So by this definition,
even solenoids and piezo transducers are motors.

Now, it's not a motor in the conventionally-understood sense--something
that produces rotary motion--but it is, nonetheless, a perfectly good motor.


--
The current state of literacy in our advanced civilization:

yo
wassup
nuttin
wan2 hang
k
where
here
k
l8tr
by

- from Usenet (what's *that*?)

On Sat, 09 Apr 2011 13:23:48 -0700, David Nebenzahl wrote:

On 4/9/2011 9:06 AM Meat Plow spake thus:

On Sat, 09 Apr 2011 00:43:51 +0100, Gareth Magennis wrote:

"Meat Plow" wrote in message
news
On Sat, 09 Apr 2011 00:12:23 +0100, Gareth Magennis wrote:

"Meat Plow" wrote in message
news
On Fri, 08 Apr 2011 10:17:06 +0100, Gareth Magennis wrote:

Interestingly the customer said that the amp went DC and fried his
speaker - the amp has no relay or DC protection. My little bit or
research so far indicates that dv/dt failure in industrial motor
applications can be down to pulling the power cable off a stalled
motor, zapping the gate/substrate connection in a nanosecond.
Might this also apply to a faulty speaker in this kind of audio
amp which burns short then eventually self disconnects?

I also understand extremely electrically noisy environments can
also induce this failure mode. The customer is daisy chaining
this amp with another one via the FX send, I'm wondering if this
has any chance of being the cause of the failure.

Slaving another amp via the send isn't likely but a failing speaker
or poorly wired/connected load might be the reason.

Check this out, if just for educational purposes. Quite
interesting.

http://www.4qd.co.uk/serv/MOSFETfail.html

Mostly pertaining to motor control applications. My experience is
limited to audio applications. Some similarities exist thought. Still
a good read for failure modes.

Isn't a speaker a motor?

It can be perceived as a sound motor.

It *is*, in fact, a motor. Which, technically speaking, is anything that
converts electrical energy to mechanical motion. So by this definition,
even solenoids and piezo transducers are motors.

Now, it's not a motor in the conventionally-understood sense--something
that produces rotary motion--but it is, nonetheless, a perfectly good
motor.

My reference was to the failure modes of motor speed control mosfets.
They aren't the ideal comparison to the failure modes found in audio
power amps using lateral or bipolar mosfets. None the less the 4QD read
was enlightening.

Lots of things can be considered motors if you think outside the box.
Anything that conveys motion can be considered a motor.


--
Live Fast Die Young, Leave A Pretty Corpse

"Meat Plow"

I cried the day Hitachi quit manufacturing the 2SK49 / 2SJ135 pairs. I
have one old SCS 350w/ch power amp that uses them. I acquired it with a
blow channel back in the late 90's. It wasn't easy back then to find
replacements. I imagine it's probably a lot easier these days to find NOS
replacements. Maybe even a sub is made today?


** Finding NOS for those numbers is nigh impossible now.

But the Semelab BUZ900 and BUZ905 lateral mosfets are near identical.

Where devices operate in parallel, best not mix the numbers up -
particularly if there are no source ballast resistors.


..... Phil