Figuring out why a simple pre-amp would go incendiary!

Prostheta · Post by **Prostheta** » 11 Jul 2011, 07:18

Hi everyone. A bit of a conundrum which I'm hoping to shed some light on. I am hoping to reproduce a small batch of pre-amps for early 70s/80s Aria basses which - given their age - have a certain number of failed units. The rub is, the pre-amp is nothing more than a balanced unity gain buffer made from two op-amps (JRC4558N/LM4558N, no tube scream voodoo, please) on a balanced 18v supply from 2x PP3s. Not entirely sure of one cap value and the actual IC id due to epoxy problems (hammer removal, not chem).

The circuit "BB noisekiller" was encapsulated in a relatively big block of epoxy and general reports of death include the pickup dying at the same time. I find this somewhat coincidental. I've considered that the volume of epoxy might have heated the components during curing, however I would have thought quicker failure this this would result. Additionally I have had more than one report of the BB literally smoking and threatening to become an epoxy incendiary device! The circuit has a 10-pin Molex connector and looks like this:

Track side circuit sketch:

Schematic:

It fits into the wider instrument circuit as so (linked offsite, not my content):
The "enjoy the gig with" schematic

Whether this failure rate is acceptable and normal, I am not sure. Something somewhere is causing them to die! Perhaps accidental battery reversal damage, one battery dying before the other, heat dissipation issues....? If there is a crucial design flaw, I would like to avoid it instead of reproducing it! I can't seem to put my finger on it though. Puzzling.

TIA

DrNomis · Post by **DrNomis** » 11 Jul 2011, 09:50

Prostheta wrote:Hi everyone. A bit of a conundrum which I'm hoping to shed some light on. I am hoping to reproduce a small batch of pre-amps for early 70s/80s Aria basses which - given their age - have a certain number of failed units. The rub is, the pre-amp is nothing more than a balanced unity gain buffer made from two op-amps (JRC4558N/LM4558N, no tube scream voodoo, please) on a balanced 18v supply from 2x PP3s. Not entirely sure of one cap value and the actual IC id due to epoxy problems (hammer removal, not chem).

The circuit "BB noisekiller" was encapsulated in a relatively big block of epoxy and general reports of death include the pickup dying at the same time. I find this somewhat coincidental. I've considered that the volume of epoxy might have heated the components during curing, however I would have thought quicker failure this this would result. Additionally I have had more than one report of the BB literally smoking and threatening to become an epoxy incendiary device! The circuit has a 10-pin Molex connector and looks like this:

Track side circuit sketch:

Schematic:

It fits into the wider instrument circuit as so (linked offsite, not my content):
The "enjoy the gig with" schematic

Whether this failure rate is acceptable and normal, I am not sure. Something somewhere is causing them to die! Perhaps accidental battery reversal damage, one battery dying before the other, heat dissipation issues....? If there is a crucial design flaw, I would like to avoid it instead of reproducing it! I can't seem to put my finger on it though. Puzzling.

TIA

Do you know exactly what component is smoking?, is it possible to narrow it down to one component?, if so, my bet is that there is too much current flowing through the component...hope that helps...

Prostheta · Post by **Prostheta** » 11 Jul 2011, 09:56

Not sure. It wasn't mine, and it was encapsulated in epoxy. I might be able to get ahold of it to find out. I agree about the excessive current. Do Tantalum caps go closed circuit if they fail? I wonder if one became resistive and started acting like a heater....it would have the full force of a battery behind it too!

DrNomis · Post by **DrNomis** » 11 Jul 2011, 10:07

Prostheta wrote:Not sure. It wasn't mine, and it was encapsulated in epoxy. I might be able to get ahold of it to find out. I agree about the excessive current. Do Tantalum caps go closed circuit if they fail? I wonder if one became resistive and started acting like a heater....it would have the full force of a battery behind it too!

Yeah, I would suspect the Tanalum caps, I seem to recall someone telling me a story about how they had some Tanalum caps literally burst into flames, you could try replacing the Tantalum caps with Aluminium Electrolytic types, or Low-ESR types, provided that they have an adequate working-voltage rating, it's possible that the Tantalum caps were inadvertently put in the circuit the wrong way around, and therefore went leaky, personally I think encapsulating the whole thing in epoxy would have exasperated things with regards to heat dissipation, which would have contributed to the Tantalum caps going incendiary...

Prostheta · Post by **Prostheta** » 11 Jul 2011, 10:20

Interesting. Would the Tantalums take some time to do that despite the reversed polarity? The circuit functioned perfectly for some time til failure. I can see why there are in the design, and how it could still function without one being aligned correctly.

If I get ahold of the BB in question, I'll certainly concentrate on those first.

DrNomis · Post by **DrNomis** » 11 Jul 2011, 10:37

Prostheta wrote:Interesting. Would the Tantalums take some time to do that despite the reversed polarity? The circuit functioned perfectly for some time til failure. I can see why there are in the design, and how it could still function without one being aligned correctly.

If I get ahold of the BB in question, I'll certainly concentrate on those first.

It's possible that the Tantalums would take some time to do that, Tantalums are constructed differently to standard Electrolytics, in Electrolytics, there's an electrolytic paste which is smeared onto one of the "plates" which is formed by a long legnth of aluminium foil, this electrolytic paste "forms" the capacitor when an electric current is applied to the cap, when a capacitor is put into a circuit backwards, the forming of the capacitor deteriorates, and the capacitor eventually fails and goes leaky, to be able to pack a large amount of capacitance into a small space, manufacturers started using Tantalum instead of an electrolytic paste, Tantalum caps are supposed to have low-leakage when put into circuit the correct way around, but will eventually go leaky if put into a circuit the wrong way around, I think the Tantalum caps made in the 70's were more prone to failure than ones made in later years, since manufacturers were still learning how to perfect them....

Prostheta · Post by **Prostheta** » 11 Jul 2011, 10:57

Very informative, thanks. I wonder if a Tantalum cap going leaky would have the knock-on effect of damaging a pickup....I don't see how. This is possibly a spurious false positive attributing two problems to one issue.

Is there a consensus on the "handedness" of Tantalums, ie. "the positive facing side is always on the left of the writing"? I had a few difficulties with identifying component values when physically degooping the rebaked epoxy, including print coming off the polyfilm which I intuit as having been 1uF-4.7uF by its size. Ohwait....if a Tantalum truly did fail, I doubt it would have much print left anyway.

DrNomis · Post by **DrNomis** » 11 Jul 2011, 11:22

Prostheta wrote:Very informative, thanks. I wonder if a Tantalum cap going leaky would have the knock-on effect of damaging a pickup....I don't see how. This is possibly a spurious false positive attributing two problems to one issue.

Is there a consensus on the "handedness" of Tantalums, ie. "the positive facing side is always on the left of the writing"? I had a few difficulties with identifying component values when physically degooping the rebaked epoxy, including print coming off the polyfilm which I intuit as having been 1uF-4.7uF by its size. Ohwait....if a Tantalum truly did fail, I doubt it would have much print left anyway.

The only cenario I can think of for a way that a Tantalum cap could possibly damage a pickup, is if it went leaky and allowed DC from the battery to flow directly through the pickup, this would cause the pickup to go open-circuit as it tried to pass the full current from the battery, assuming that the battery is a 9V type, usually the manufacturers use two methods of indicating the polarity of a Tantalum cap, they place a "+" sign near the cap's positive terminal, and they make the positive terminal longer than the negative terminal, in my Jaycar Electronics catalogue, there is a picture of six different Tantalum caps, one has the "+" sign to the left of the marked capacitance value, the other 5 are a bit unclear, but I would say that I agree with you that the "+" sign is to the left...

The working-voltage rating of a Tantalum cap, or, any cap for that matter, is basically how many volts the dielectric insulator, between the two plates of the cap, can withstand before being "punctured" by that voltage and developing a short-circuit between the two plates....

mictester · Post by **mictester** » 11 Jul 2011, 11:25

Prostheta wrote:Not sure. It wasn't mine, and it was encapsulated in epoxy. I might be able to get ahold of it to find out. I agree about the excessive current. Do Tantalum caps go closed circuit if they fail? I wonder if one became resistive and started acting like a heater....it would have the full force of a battery behind it too!

Yes. Tantalum caps can go short.

I once had a batch of 10V rated tantalums that were marked 35V and couldn't work out why they kept exploding on a 28V supply! I went through four of them (they'd work for a while, then explode like a match head!), and then decided to check them using a variable supply. They were fine up to (about) 14 Volts, so I concluded that they must be 10V rated ones. They were returned to the supplier who confirmed that they were "faulty", and shipped me a new batch, similarly marked, but slightly fatter! The replacements were fine.

Prostheta · Post by **Prostheta** » 11 Jul 2011, 11:29

Can't entirely take it as read though! I remember they were "left-handed" when I unearthed them. I can't see how the battery could dump through the pickup coil. There are those 470k resistors in the nearest feasible path. At most a warm pickup and a string deadening magnet field generated! An anti sustainer.

Edit due to new post: sounds like ditching tantalums for cans is the way forward. I was always under the impression that Tantalums were pretty good gear.

DrNomis · Post by **DrNomis** » 11 Jul 2011, 11:40

Prostheta wrote:Can't entirely take it as read though! I remember they were "left-handed" when I unearthed them. I can't see how the battery could dump through the pickup coil. There are those 470k resistors in the nearest feasible path. At most a warm pickup and a string deadening magnet field generated! An anti sustainer.

So, maybe the potting compound itself went conductive?, but then again, if it did, I'd imagine that the current passed would be very small, like a miliamp at the most....

So, have you checked the pickups with a multimeter?, if not, try doing that to see whether the pickup is open, or short-circuited, the other way I can see that may have caused the pickup to fail, is that somewhere on the winding, the enamelled copper wire could have been eaten through by corrosion, and this could have happened at the same time when the Tantalum caps smoked....

DrNomis · Post by **DrNomis** » 11 Jul 2011, 11:44

Prostheta wrote:Can't entirely take it as read though! I remember they were "left-handed" when I unearthed them. I can't see how the battery could dump through the pickup coil. There are those 470k resistors in the nearest feasible path. At most a warm pickup and a string deadening magnet field generated! An anti sustainer.

Edit due to new post: sounds like ditching tantalums for cans is the way forward. I was always under the impression that Tantalums were pretty good gear.

Usually Tantalum caps are pretty reliable nowdays, since the technology has been perfected, but I think it wouldn't hurt to replace the Tantalum caps with Aluminium Electrolytics....

DrNomis · Post by **DrNomis** » 11 Jul 2011, 11:46

mictester wrote:
Prostheta wrote:Not sure. It wasn't mine, and it was encapsulated in epoxy. I might be able to get ahold of it to find out. I agree about the excessive current. Do Tantalum caps go closed circuit if they fail? I wonder if one became resistive and started acting like a heater....it would have the full force of a battery behind it too!
Yes. Tantalum caps can go short.

I once had a batch of 10V rated tantalums that were marked 35V and couldn't work out why they kept exploding on a 28V supply! I went through four of them (they'd work for a while, then explode like a match head!), and then decided to check them using a variable supply. They were fine up to (about) 14 Volts, so I concluded that they must be 10V rated ones. They were returned to the supplier who confirmed that they were "faulty", and shipped me a new batch, similarly marked, but slightly fatter! The replacements were fine.

Mictester, thanks for that, good to know that I was correct in my thinking...cheers...

Prostheta · Post by **Prostheta** » 11 Jul 2011, 11:52

DrNomis wrote:
Prostheta wrote:Can't entirely take it as read though! I remember they were "left-handed" when I unearthed them. I can't see how the battery could dump through the pickup coil. There are those 470k resistors in the nearest feasible path. At most a warm pickup and a string deadening magnet field generated! An anti sustainer.

So, maybe the potting compound itself went conductive?, but then again, if it did, I'd imagine that the current passed would be very small, like a miliamp at the most....

So, have you checked the pickups with a multimeter?, if not, try doing that to see whether the pickup is open, or short-circuited, the other way I can see that may have caused the pickup to fail, is that somewhere on the winding, the enamelled copper wire could have been eaten through by corrosion, and this could have happened at the same time when the Tantalum caps smoked....

I haven't got any pickups on hand which failed during these circumstances. They were also epoxy potted however! Not entirely convenient. I did dissect one last month, and the coil to lead connection looked less than perfect. Otherwise, very good condition and construction. My failed pickup was impact damaged, so no clues were forthcoming.

DrNomis · Post by **DrNomis** » 11 Jul 2011, 11:59

Apart from that, I can't really think of much else off the top of my head, but deffinitely try changing out the Tantalum caps, and see if that stops the smoking, etc.....

Prostheta · Post by **Prostheta** » 11 Jul 2011, 12:04

My smoking isn't the problem! I don't want to make replacement preamps with the same doomsday device inbuilt

deltafred · Post by **deltafred** » 11 Jul 2011, 12:29

Is there any possibility of plugging the Molex connector on the wrong way round?

That would put 9v straight across the pickup and the 2 opamp + inputs which it might survive depending upon what return path the ground connection provides.

Without building one up and connecting it up wrongly it is pretty much impossible to say what voltage and polarity the tants would see.

Just a thought

Prostheta · Post by **Prostheta** » 11 Jul 2011, 12:43

The Molex connectors are a pair; one five-way and a three-way from the pickup. I doubt reversal would be the cause, as even non-engineers could easily attribute cause and effect there! Failure is sudden and unexpected, with this problem dating back to the eighties. The common potential issue sounds like the caps, however the pickup part of the equation is unknown. It probably isn't part of the same equation, and dual failure is likely anecdotal and/or misattribution. I am happy to consider them as seperate failures.

The only other common attribute would be the epoxy of course.

DrNomis · Post by **DrNomis** » 11 Jul 2011, 12:44

deltafred wrote:Is there any possibility of plugging the Molex connector on the wrong way round?

That would put 9v straight across the pickup and the 2 opamp + inputs which it might survive depending upon what return path the ground connection provides.

Without building one up and connecting it up wrongly it is pretty much impossible to say what voltage and polarity the tants would see.

Just a thought

Very good point there....

Prostheta · Post by **Prostheta** » 11 Jul 2011, 12:49

I'll scope my spare loom out at home. On the droid again.