Repair job Story

[Phil Allison]

Phil Allison wrote:
------------------------

Hi to all my loyal followers,

The term "bounce" is applied when an item that has been recently repaired is brought back for a similar or repeat failure.

Mostly it's a false alarm, a new problem or just once in a while the same failure involving the new parts. This is the worst scenario and for me very rare.

One that comes to mind was a MusicMan hybrid guitar amp - the RD50. The circuit is all solid state except for a pair of 6L6GC tubes for output.

https://www.thetubestore.com/lib/thetubestore/schematics/MusicMan/Musicman-RD50A-Schematic.pdf

One came to me with the input op-amp ( IC1) having it's output stuck to a DC rail. Fitted a new TL072 and away it went, good as new. Unusual failure, but shit happens sometimes.

It was back in a week with the new IC blown, just like before.

Another TL072 fixed it again, but I smelt a big fat rat.

Since the op-amp interfaced with the external world of musicians and their assorted paraphernalia - it must had failed due to abuse, twice.

But what exactly ? To stop the nonsense, I fitted 1N4148 diodes from pin 3 to each rail as a precaution. Walter C. Jung's "IC Op-Amp Cookbook " gave support for fact that such inputs were vulnerable.

The customer was not happy, he insisted on speaking with me and was very aggressive. I had understood he was a school teacher but it turned out he was in fact a *Taekwando* instructor !!

He expected the new repair to be covered by warranty, but I was not budging.

He insisted that guitar amps were built very tough to take all sorts of abuse.

I assured him he was mistaken: "I make my living because that's not true" - I stated firmly.

After extracting a few details from him, it all became clear.

Any guess what seemingly innocent act he had performed done, twice ?

-------------------------------------------------------------------

** No-one here is ever gonna get this - it is just too darn simple.

On both occasions, when the input TL072 failed, the owner was neither at home nor a gig. He was using an upper floor room of a neighbour's very run down 120 year old terrace house for band rehearsal.

His amp made an unusual buzzing all the time, put down to dodgy power wiring.

The amp failed to proceed from the get go, after switching on & connecting his wall wart powered wha-wha pedal with a lead already plugged at the amp end.

The amp was not earthed, so floating about 150VAC above ground due to stray C inside the AC tranny. So that sort of voltage was transferred momentarily to pin 3 of the TL072 via 10kohms as the plug tip touched the pedal.

About 20 mA peak - just enough to be fatal.

Other models of MusicMan amps I checked have protection diodes fitted or much greater input resistances.

The sod of an owner was aware the room and it's ancient wiring had a lot to do with the failures, but kept it secret until I really pushed him.

On the grounds that it was worth a try to get free repair and mod.

Wot an utter ass.

Never saw him or his MM amp again, the mod worked.


..... Phil

 

[Lasse Langwadt Christense]

lørdag den 9. november 2019 kl. 06.08.58 UTC+1 skrev Phil Allison:

Phil Allison wrote:
------------------------

Hi to all my loyal followers,

The term "bounce" is applied when an item that has been recently repaired is brought back for a similar or repeat failure.

Mostly it's a false alarm, a new problem or just once in a while the same failure involving the new parts. This is the worst scenario and for me very rare.

One that comes to mind was a MusicMan hybrid guitar amp - the RD50. The circuit is all solid state except for a pair of 6L6GC tubes for output.

https://www.thetubestore.com/lib/thetubestore/schematics/MusicMan/Musicman-RD50A-Schematic.pdf

One came to me with the input op-amp ( IC1) having it's output stuck to a DC rail. Fitted a new TL072 and away it went, good as new. Unusual failure, but shit happens sometimes.

It was back in a week with the new IC blown, just like before.

Another TL072 fixed it again, but I smelt a big fat rat.

Since the op-amp interfaced with the external world of musicians and their assorted paraphernalia - it must had failed due to abuse, twice.

But what exactly ? To stop the nonsense, I fitted 1N4148 diodes from pin 3 to each rail as a precaution. Walter C. Jung's "IC Op-Amp Cookbook " gave support for fact that such inputs were vulnerable.

The customer was not happy, he insisted on speaking with me and was very aggressive. I had understood he was a school teacher but it turned out he was in fact a *Taekwando* instructor !!

He expected the new repair to be covered by warranty, but I was not budging.

He insisted that guitar amps were built very tough to take all sorts of abuse.

I assured him he was mistaken: "I make my living because that's not true" - I stated firmly.

After extracting a few details from him, it all became clear.

Any guess what seemingly innocent act he had performed done, twice ?


-------------------------------------------------------------------

** No-one here is ever gonna get this - it is just too darn simple.

On both occasions, when the input TL072 failed, the owner was neither at home nor a gig. He was using an upper floor room of a neighbour's very run down 120 year old terrace house for band rehearsal.

His amp made an unusual buzzing all the time, put down to dodgy power wiring.

The amp failed to proceed from the get go, after switching on & connecting his wall wart powered wha-wha pedal with a lead already plugged at the amp end.

The amp was not earthed, so floating about 150VAC above ground due to stray C inside the AC tranny. So that sort of voltage was transferred momentarily to pin 3 of the TL072 via 10kohms as the plug tip touched the pedal.

About 20 mA peak - just enough to be fatal.

Other models of MusicMan amps I checked have protection diodes fitted or much greater input resistances.

The sod of an owner was aware the room and it's ancient wiring had a lot to do with the failures, but kept it secret until I really pushed him.

On the grounds that it was worth a try to get free repair and mod.

Wot an utter ass.

Never saw him or his MM amp again, the mod worked.

here few things have a ground connection so with the input filter y capacitors in switch mode supplies most things are floating at 1/2 mains

makes for some surprises when people plug a computer into a TV grounded through the antenna cable

 

[Lasse Langwadt Christense]

lørdag den 9. november 2019 kl. 20.49.04 UTC+1 skrev Phil Allison:

Lasse Langwadt Christensen wrote:

----------------------------------


Wot an utter ass.

Never saw him or his MM amp again, the mod worked.



here few things have a ground connection so with the input filter
y capacitors in switch mode supplies most things are floating at 1/2 mains


** Same here for any Class 2 piece of electronics, particularly wall warts since the iron transformer kind were banned. Yes really.

Yamaha sold their EMX series of large powered desks into Australia as Class 2 items - a ridiculous idea as they were bloody lethal in practice.

Fitted 3 core leads to a few to stop singers getting nasty shocks on their lips.

The whole desk and everything attached could go live with a singe wiring error in a power plug.

When you are a self employed tech, the buck stops with you regarding electrical safety. Stages are dangerous environments and the rule is " make it safe or disable it"

The latter tends to annoy customers ...

here even a lot of stuff that should have ground doesn't, because most many outlets doesn't have a ground and if they do the schuko plug most devices come with doesn't connect to ground in a Danish outlet with out an adapter

 

[Phil Allison]

Lasse Langwadt Christensen wrote:

----------------------------------

Wot an utter ass.

Never saw him or his MM amp again, the mod worked.



here few things have a ground connection so with the input filter
y capacitors in switch mode supplies most things are floating at 1/2 mains

** Same here for any Class 2 piece of electronics, particularly wall warts since the iron transformer kind were banned. Yes really.

Yamaha sold their EMX series of large powered desks into Australia as Class 2 items - a ridiculous idea as they were bloody lethal in practice.

Fitted 3 core leads to a few to stop singers getting nasty shocks on their lips.

The whole desk and everything attached could go live with a singe wiring error in a power plug.

When you are a self employed tech, the buck stops with you regarding electrical safety. Stages are dangerous environments and the rule is " make it safe or disable it"

The latter tends to annoy customers ...


...... Phil

 

On Fri, 8 Nov 2019 21:08:54 -0800 (PST), Phil Allison
<pallison49@gmail.com> wrote:

Phil Allison wrote:
------------------------

Hi to all my loyal followers,

The term "bounce" is applied when an item that has been recently repaired is brought back for a similar or repeat failure.

Mostly it's a false alarm, a new problem or just once in a while the same failure involving the new parts. This is the worst scenario and for me very rare.

One that comes to mind was a MusicMan hybrid guitar amp - the RD50. The circuit is all solid state except for a pair of 6L6GC tubes for output.

https://www.thetubestore.com/lib/thetubestore/schematics/MusicMan/Musicman-RD50A-Schematic.pdf

One came to me with the input op-amp ( IC1) having it's output stuck to a DC rail. Fitted a new TL072 and away it went, good as new. Unusual failure, but shit happens sometimes.

It was back in a week with the new IC blown, just like before.

Another TL072 fixed it again, but I smelt a big fat rat.

Since the op-amp interfaced with the external world of musicians and their assorted paraphernalia - it must had failed due to abuse, twice.

But what exactly ? To stop the nonsense, I fitted 1N4148 diodes from pin 3 to each rail as a precaution. Walter C. Jung's "IC Op-Amp Cookbook " gave support for fact that such inputs were vulnerable.

The customer was not happy, he insisted on speaking with me and was very aggressive. I had understood he was a school teacher but it turned out he was in fact a *Taekwando* instructor !!

He expected the new repair to be covered by warranty, but I was not budging.

He insisted that guitar amps were built very tough to take all sorts of abuse.

I assured him he was mistaken: "I make my living because that's not true" - I stated firmly.

After extracting a few details from him, it all became clear.

Any guess what seemingly innocent act he had performed done, twice ?


-------------------------------------------------------------------

** No-one here is ever gonna get this - it is just too darn simple.

On both occasions, when the input TL072 failed, the owner was neither at home nor a gig. He was using an upper floor room of a neighbour's very run down 120 year old terrace house for band rehearsal.

His amp made an unusual buzzing all the time, put down to dodgy power wiring.

The amp failed to proceed from the get go, after switching on & connecting his wall wart powered wha-wha pedal with a lead already plugged at the amp end.

Isn't the wall wart a class 2 device with low capacitance between
primary and secondary ?

>The amp was not earthed, so floating about 150VAC above ground due to stray C inside the AC tranny.

The stray capacitance would be in the order of 10 - 100 pF not capable
of storing much charge.

Was the amplifier a proper class 1 device with proper mains filter
Y-capacitors ?

So that sort of voltage was transferred momentarily to pin 3 of the TL072 via 10kohms as the plug tip touched the pedal.

About 20 mA peak - just enough to be fatal.

I very much doubt that transformer stray capacitances alone could
generate such currents.

It sounds like _two_ class 1 devices (amplifier and e.g. desktop
computer) with mains filter Y-capacitors were connected to separate
ungrounded class 0 sockets and due to Y-capacitor tolerances, the
chassis potential was different, not exactly Vmains/2.On this case,
quite large currents could flow in the signal conductor. The busing
sound also suggests that another class 1 device was involved.

When powering multiple class 1 devices in a room with only ungrounded
class 0 sockets available, use a grounded class 1 extension cord with
multiple sockets. The whole system will float close to Vmains/2 but
the chassis potential on all devices is the same.

 

[Robert Baer]

Lasse Langwadt Christensen wrote:

lørdag den 9. november 2019 kl. 06.08.58 UTC+1 skrev Phil Allison:
Phil Allison wrote:
------------------------

Hi to all my loyal followers,

The term "bounce" is applied when an item that has been recently repaired is brought back for a similar or repeat failure.

Mostly it's a false alarm, a new problem or just once in a while the same failure involving the new parts. This is the worst scenario and for me very rare.

One that comes to mind was a MusicMan hybrid guitar amp - the RD50. The circuit is all solid state except for a pair of 6L6GC tubes for output.

https://www.thetubestore.com/lib/thetubestore/schematics/MusicMan/Musicman-RD50A-Schematic.pdf

One came to me with the input op-amp ( IC1) having it's output stuck to a DC rail. Fitted a new TL072 and away it went, good as new. Unusual failure, but shit happens sometimes.

It was back in a week with the new IC blown, just like before.

Another TL072 fixed it again, but I smelt a big fat rat.

Since the op-amp interfaced with the external world of musicians and their assorted paraphernalia - it must had failed due to abuse, twice.

But what exactly ? To stop the nonsense, I fitted 1N4148 diodes from pin 3 to each rail as a precaution. Walter C. Jung's "IC Op-Amp Cookbook " gave support for fact that such inputs were vulnerable.

The customer was not happy, he insisted on speaking with me and was very aggressive. I had understood he was a school teacher but it turned out he was in fact a *Taekwando* instructor !!

He expected the new repair to be covered by warranty, but I was not budging.

He insisted that guitar amps were built very tough to take all sorts of abuse.

I assured him he was mistaken: "I make my living because that's not true" - I stated firmly.

After extracting a few details from him, it all became clear.

Any guess what seemingly innocent act he had performed done, twice ?


-------------------------------------------------------------------

** No-one here is ever gonna get this - it is just too darn simple.

On both occasions, when the input TL072 failed, the owner was neither at home nor a gig. He was using an upper floor room of a neighbour's very run down 120 year old terrace house for band rehearsal.

His amp made an unusual buzzing all the time, put down to dodgy power wiring.

The amp failed to proceed from the get go, after switching on & connecting his wall wart powered wha-wha pedal with a lead already plugged at the amp end.

The amp was not earthed, so floating about 150VAC above ground due to stray C inside the AC tranny. So that sort of voltage was transferred momentarily to pin 3 of the TL072 via 10kohms as the plug tip touched the pedal.

About 20 mA peak - just enough to be fatal.

Other models of MusicMan amps I checked have protection diodes fitted or much greater input resistances.

The sod of an owner was aware the room and it's ancient wiring had a lot to do with the failures, but kept it secret until I really pushed him.

On the grounds that it was worth a try to get free repair and mod.

Wot an utter ass.

Never saw him or his MM amp again, the mod worked.


here few things have a ground connection so with the input filter y capacitors in switch mode supplies most things are floating at 1/2 mains

makes for some surprises when people plug a computer into a TV grounded through the antenna cable

All you "need" is a source that is 120V on the low side, 240V on the
high side and a leaky cap to the metal chassis.
Can one be half dead? (leaky cap was from low side to chassis)

Bad wiring was due to fact the building was partly re-wired numerous
times: change/upgrade from 50Hz to 60Hz, added in 2-phase 90 degree for
running some motors, reuse of that wiring for partial 120/240 use (the
real cause of the above ignored problem), and a number of upgrades to
allow more power.
Did not help there were no docs for this, just memory "oh yea, we...".

Palo Alto High School.

 

[Phil Allison]

upsid...@downunder.com wrote:

-----------------------------

The amp was not earthed, so floating about 150VAC above ground due to
stray C inside the AC tranny.

The stray capacitance would be in the order of 10 - 100 pF not capable
of storing much charge.

** Try 250 to 400pF. The tranny is large and with overwound primary and secondary.

I very much doubt that transformer stray capacitances alone could
generate such currents.

** The peak current available is several amps.


> It sounds like _two_ class 1 devices

** AFAIK, one MM amp amplifier with a dodgy AC power socket - loose parts, wiring and corroded contacts.

Very likely the neutral was bad, so the only connection was to active at sometimes So chassis then floating at 240VAC.

FYI

spikes from a 120VAC supply, via a 400pF cap, will damage the B-E junction of a BC108 or similar. A few strikes halves the Hfe.

An unprotected, tiny JFET gate at a TLO72 input is vasty more fragile.



...... Phil

 

On Sat, 9 Nov 2019 13:58:05 -0800 (PST), Phil Allison
<pallison49@gmail.com> wrote:

upsid...@downunder.com wrote:

-----------------------------

The amp was not earthed, so floating about 150VAC above ground due to
stray C inside the AC tranny.

The stray capacitance would be in the order of 10 - 100 pF not capable
of storing much charge.

** Try 250 to 400pF. The tranny is large and with overwound primary and secondary.

The capacitive reactance at 50 Hz is 8 Mohms.

I very much doubt that transformer stray capacitances alone could
generate such currents.

** The peak current available is several amps.

At 240 V 50 Hz the current through the 400 pF would be 30 uA, which
would hardly damage any components.

In practice, the wall wart capacitance would be much less and that
capacitance would be in series with the 400 pF, so the total
capacitance would be even less.

The maximum size of an Y-capacitor is 22 nF, so the leakage would be
up to 1.5 mA, so this could cause some harm to sensitive electronics.
But if the only path to ground is through the stray capacitance of the
wall wart, the total series capacitance would be below 100 pF.

It sounds like _two_ class 1 devices

** AFAIK, one MM amp amplifier with a dodgy AC power socket - loose parts, wiring and corroded contacts.

Even with two class 1 devices with maximum Y-capacitors connected to
class sockets, the current would be about 1 mA.worst case.

Very likely the neutral was bad, so the only connection was to active at sometimes So chassis then floating at 240VAC.

FYI

spikes from a 120VAC supply, via a 400pF cap, will damage the B-E junction of a BC108 or similar. A few strikes halves the Hfe.

I you assume 1.5-2.5 kV non-repetive spikes could indeed do some
damage, you had to assume that the spike occurs just when plugging in
the tip of the plug is connected before the ring is connected through
the cable shield. This might happen, but is highly unlikely.

>An unprotected, tiny JFET gate at a TLO72 input is vasty more fragile.

Wasn't the input stage inverting so there is a series resistance
limiting any large currents. Never heard that TL072 would have been
more fragile than early (C/P/N)MOS logic chips.

..... Phil

 

[Phil Allison]

upsid...@downunder.com wrote:

-----------------------------

** Try 250 to 400pF. The tranny is large and with overwound
primary and secondary.

The capacitive reactance at 50 Hz is 8 Mohms.

** True, but also 100% irrelevant.

Only the stored energy at 340 volt peak maters.

Whether actually stored or just about to be in the next few uS.

** The peak current available is several amps.

At 240 V 50 Hz the current through the 400 pF would be 30 uA, which
would hardly damage any components.

** However, the peak current is only limited by series resistance.

Charge a cap, short it out - how many amps, peak ?

The maximum size of an Y-capacitor is 22 nF,

** Red herring, not mentioned and not involved.


> Even with two class 1 devices with maximum Y-capacitors

FFS dickwad - there are NO such fucking caps.

spikes from a 120VAC supply, via a 400pF cap, will damage the B-E
junction of a BC108 or similar. A few strikes halves the Hfe.

An unprotected, tiny JFET gate at a TLO72 input is vasty more fragile.

Wasn't the input stage inverting

** No.

Fuck off you retarded idiot.


..... Phil