Topics

576 HV transformer problems - something to ponder

Ed Breya
 

I was web searching on some chemistry stuff, and stumbled upon the following paragraph, which led me to make a possible connection to the 576 bad (brown epoxy) transformer issue, which I've been thinking about lately.

"Red phosphorus can be used as a very effective flame retardant, especially in thermoplastics (e.g. polyamide) and thermosets (e.g. epoxy resins or polyurethanes). The flame retarding effect is based on the formation of polyphosphoric acid. Together with the organic polymer material, this acid creates a char which prevents the propagation of the flames. The safety risks associated with phosphine generation and friction sensitivity of red phosphorus can be effectively reduced by stabilization and micro-encapsulation. For easier handling, red phosphorus is often used in form of dispersions or masterbatches in various carrier systems. However, for electronic/electrical systems, red phosphorus flame retardant has been effectively banned by major OEMs due to its tendency to induce premature failures. There have been two issues over the years: the first was red phosphorus in epoxy molding compounds inducing elevated leakage current in semiconductor devices[5] and the second was acceleration of hydrolysis reactions in PBT insulating material.[6]"

It is within this link: https://en.wikipedia.org/wiki/Allotropes_of_phosphorus#Black_phosphorus

Interestingly, what triggered my unconscious was the red-P sample pictured immediately above the paragraph - out of nowhere, my brain suddenly said "that looks just like the brown epoxy on the 576 transformer," which I had scrutinized a few days ago. Then I scanned the first sentence of the paragraph, and it all connected.

There's a bunch of info especially in refs [5] and [6].

On further searching, I found all sorts of interesting stuff on the issues of flame retardants and electronics.

This all does nothing to fix our existing problems with patching up 576s and other old gear with the problem, but it does shed some light on a possible fundamental cause. In other words, the failure is not likely due to the epoxy resin characteristics per se, but an added flame retardant, whether red-P or something else. It also indicates possible causes of other problems in all sorts of places, that we may encounter in electronic gear - even modern stuff.

Ed

Chuck Harris
 

Hi Ed,

I think the root of the problem has finally been found. I had
always wondered why the epoxy that is in contact with the ferrite
is clear and brittle, but the epoxy that is in contact with nothing
but insulation is semi opaque, and rubbery.

That there is fire retardant isn't really surprising, as having all
that combustible beeswax/paraffin wax in the older scopes, must have
looked a lot like a by design igniter. When they wanted to make a
cleaner HV box (no wax), and higher top operating temperature, fire
must have come into their thoughts.

It doesn't help us one bit in fixing the problem, but it may explain
how it blind-sided tektronix.

-Chuck Harris

Ed Breya via Groups.Io wrote:

I was web searching on some chemistry stuff, and stumbled upon the following paragraph, which led me to make a possible connection to the 576 bad (brown epoxy) transformer issue, which I've been thinking about lately.

"Red phosphorus can be used as a very effective flame retardant, especially in thermoplastics (e.g. polyamide) and thermosets (e.g. epoxy resins or polyurethanes). The flame retarding effect is based on the formation of polyphosphoric acid. Together with the organic polymer material, this acid creates a char which prevents the propagation of the flames. The safety risks associated with phosphine generation and friction sensitivity of red phosphorus can be effectively reduced by stabilization and micro-encapsulation. For easier handling, red phosphorus is often used in form of dispersions or masterbatches in various carrier systems. However, for electronic/electrical systems, red phosphorus flame retardant has been effectively banned by major OEMs due to its tendency to induce premature failures. There have been two issues over the years: the first was red phosphorus in epoxy molding compounds inducing elevated leakage current in semiconductor devices[5] and the second was acceleration of hydrolysis reactions in PBT insulating material.[6]"

It is within this link: https://en.wikipedia.org/wiki/Allotropes_of_phosphorus#Black_phosphorus

Interestingly, what triggered my unconscious was the red-P sample pictured immediately above the paragraph - out of nowhere, my brain suddenly said "that looks just like the brown epoxy on the 576 transformer," which I had scrutinized a few days ago. Then I scanned the first sentence of the paragraph, and it all connected.

There's a bunch of info especially in refs [5] and [6].

On further searching, I found all sorts of interesting stuff on the issues of flame retardants and electronics.

This all does nothing to fix our existing problems with patching up 576s and other old gear with the problem, but it does shed some light on a possible fundamental cause. In other words, the failure is not likely due to the epoxy resin characteristics per se, but an added flame retardant, whether red-P or something else. It also indicates possible causes of other problems in all sorts of places, that we may encounter in electronic gear - even modern stuff.

Ed

Mlynch001
 

Ed,

Great Piece of information. It helps to at least understand the failure mechanism that is involved.

--
Michael Lynch
Dardanelle, AR

greenboxmaven
 

Do these transformers fail by burning out or by performing very poorly?
Is there any solvent that can dissolve the bad epoxy coating or turn it into goo that can be picked and cleaned off without damaging the winding? This is not the only time I have seen and heard of an insulating addiitive, varnish, or binder in insulation testing fine with DC but becoming conductive at RF frequencies and loading down a circuit. I have seen it in other equipment, and it could explain why a popular military surplus radio receiver (BC-348) built by one manufacturer outperforms another, even though the circuit is identical.

Bruce Gentry, KA2IVY

On 9/3/19 7:47 AM, Chuck Harris wrote:
Hi Ed,

I think the root of the problem has finally been found. I had
always wondered why the epoxy that is in contact with the ferrite
is clear and brittle, but the epoxy that is in contact with nothing
but insulation is semi opaque, and rubbery.

That there is fire retardant isn't really surprising, as having all
that combustible beeswax/paraffin wax in the older scopes, must have
looked a lot like a by design igniter. When they wanted to make a
cleaner HV box (no wax), and higher top operating temperature, fire
must have come into their thoughts.

It doesn't help us one bit in fixing the problem, but it may explain
how it blind-sided tektronix.

-Chuck Harris

Ed Breya via Groups.Io wrote:
I was web searching on some chemistry stuff, and stumbled upon the following paragraph, which led me to make a possible connection to the 576 bad (brown epoxy) transformer issue, which I've been thinking about lately.

"Red phosphorus can be used as a very effective flame retardant, especially in thermoplastics (e.g. polyamide) and thermosets (e.g. epoxy resins or polyurethanes). The flame retarding effect is based on the formation of polyphosphoric acid. Together with the organic polymer material, this acid creates a char which prevents the propagation of the flames. The safety risks associated with phosphine generation and friction sensitivity of red phosphorus can be effectively reduced by stabilization and micro-encapsulation. For easier handling, red phosphorus is often used in form of dispersions or masterbatches in various carrier systems. However, for electronic/electrical systems, red phosphorus flame retardant has been effectively banned by major OEMs due to its tendency to induce premature failures. There have been two issues over the years: the first was red phosphorus in epoxy molding compounds inducing elevated leakage current in semiconductor devices[5] and the second was acceleration of hydrolysis reactions in PBT insulating material.[6]"

It is within this link: https://en.wikipedia.org/wiki/Allotropes_of_phosphorus#Black_phosphorus

Interestingly, what triggered my unconscious was the red-P sample pictured immediately above the paragraph - out of nowhere, my brain suddenly said "that looks just like the brown epoxy on the 576 transformer," which I had scrutinized a few days ago. Then I scanned the first sentence of the paragraph, and it all connected.

There's a bunch of info especially in refs [5] and [6].

On further searching, I found all sorts of interesting stuff on the issues of flame retardants and electronics.

This all does nothing to fix our existing problems with patching up 576s and other old gear with the problem, but it does shed some light on a possible fundamental cause. In other words, the failure is not likely due to the epoxy resin characteristics per se, but an added flame retardant, whether red-P or something else. It also indicates possible causes of other problems in all sorts of places, that we may encounter in electronic gear - even modern stuff.

Ed

Chuck Harris
 

The epoxy changes and makes an already lossy situation even
worse.

The epoxy was used to make a "less messy" transformer, for
use in higher temperature situations, than the older beeswax/
paraffin wax dipped transformers.

The epoxy worked, but even at the outset was more lossy than
they beeswax/paraffin transformers. Over time, the epoxy seems
to take on water and become so lossy that the HV supplies stop
oscillating, and parts burn up.

Eventually, even drying the transformer won't cure the problem,
and the transformer enters a mode where the heat created by the
loss drives the core into an electrically lossy region, and you
get a --works when cold, but quits when hot--, mode of operation.

There is nothing I know of that can dissolve the epoxy without
dissolving the insulation on the copper wire.

-Chuck Harris

greenboxmaven via Groups.Io wrote:

Do these transformers fail by burning out or by performing very poorly?
Is there any solvent that can dissolve the bad epoxy coating or turn it into goo that
can be picked and cleaned off without damaging the winding? This is not the only time
I have seen and heard of an insulating addiitive, varnish, or binder in insulation
testing fine with DC but becoming conductive at RF frequencies and loading down a
circuit. I have seen it in other equipment, and it could explain why a popular
military surplus radio receiver (BC-348) built by one manufacturer outperforms
another, even though the circuit is identical.

Bruce Gentry, KA2IVY

greenboxmaven
 

I was doing some cleaning up in my shop and came across something from a Tektronix 454 scope that resonates with this transformer discussion. I think "contaminated" epoxy may have caused another problem in a scope high voltage power supply. The 454 had several problems that were easy to find and remedy. It still would not give anything more than a very pale glow on the screen. The second anode voltage was only about 2,000 volts. Disconnecting the transformer from the tripler circuit gave a husky corona arc to a screwdriver, so I began testing the diodes and condensers in the multiplier circuit. As I reconnected each stage, the voltage incresed. Finally I reconnected the 500 PFD "doorknob" condenser,and the voltage plummited. Having worked on television receivers decades ago, I had several replacement condensers on hand, and installing one brought the scope to life with a dazzling trace. One of the NuVistors in the preamp was barely alive, so that channel could not be brought onto the screen. A replacement got the scope working very well. I took the condenser over to a friend who has a Hipot tester, and it was fine until the voltage reached about 2,000 volts. The condenser is rated at 20KV, and the voltage on it in the scope is 12KV. The bad condenser is epoxy molded, the 60 year old replacement is tan high voltage bakelite. I wonder if Tektronix specified or accepted an epoxy formula that goes bad?

Bruce Gentry, KA2IVY

On 9/3/19 9:23 AM, Chuck Harris wrote:
The epoxy changes and makes an already lossy situation even
worse.


-Chuck Harris



John Ferguson
 

Bruce, what a delight to see a reference to condensers.

Cheers,

John Ferguson, AI4TO

On 9/4/19 3:35 PM, greenboxmaven via Groups.Io wrote:
 ....As I reconnected each stage, the voltage incresed. Finally I reconnected the 500 PFD "doorknob" condenser,and the voltage plummited. Having worked on television receivers decades ago, I had several replacement condensers on hand, and installing one brought the scope to life with a dazzling trace. ....

      Bruce Gentry, KA2IVY


Dave Wise
 

Fascinating diagnosis, Bruce. You give us another hard-to-spot failure mode to be alert to. It brings home the most basic test of any component, "Does it work in the circuit?".

Dave Wise

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of John Ferguson via Groups.Io
Sent: Wednesday, September 04, 2019 12:41 PM
To: TekScopes@groups.io
Subject: Re: [TekScopes] 576 HV transformer problems - something to ponder

Bruce, what a delight to see a reference to condensers.

Cheers,

John Ferguson, AI4TO

On 9/4/19 3:35 PM, greenboxmaven via Groups.Io wrote:
 ....As I reconnected each stage, the voltage incresed. Finally I
reconnected the 500 PFD "doorknob" condenser,and the voltage
plummited. Having worked on television receivers decades ago, I had
several replacement condensers on hand, and installing one brought the
scope to life with a dazzling trace. ....

      Bruce Gentry, KA2IVY



greenboxmaven
 

I am strange in many ways, I think condenser is easier to say than capacitor, and is the term used by older broadcast engineering and industrial electronics and control texts. I also like to say kilo or mega cycles, it describes something that is happening. I do prefer the European name for those devices that glow and control current, valve is much more descriptive than tube. And, I often refer to a CRT as a jug. At 67 years old, things get imprinted and if they work as well or better, why not?

Bruce Gentry, KA2IVY

On 9/4/19 3:41 PM, John Ferguson via Groups.Io wrote:
Bruce, what a delight to see a reference to condensers.

Cheers,

John Ferguson, AI4TO

On 9/4/19 3:35 PM, greenboxmaven via Groups.Io wrote:
....As I reconnected each stage, the voltage incresed. Finally I reconnected the 500 PFD "doorknob" condenser,and the voltage plummited. Having worked on television receivers decades ago, I had several replacement condensers on hand, and installing one brought the scope to life with a dazzling trace. ....

Bruce Gentry, KA2IVY





John Ferguson
 

at 76, I grew up with condensers - that's what Dad, an EE, called them.

On 9/4/19 4:34 PM, greenboxmaven via Groups.Io wrote:
I am strange in many ways, I think condenser is easier to say than capacitor, and is the term used by older broadcast engineering and industrial electronics and control texts.  I also like to say kilo or mega cycles, it describes something that is happening. I do prefer the European name for those devices that glow and control current, valve is much more descriptive than tube. And, I often refer to a CRT as a jug. At 67 years old, things get imprinted and if they work as well or better, why not?

         Bruce Gentry, KA2IVY


On 9/4/19 3:41 PM, John Ferguson via Groups.Io wrote:
Bruce, what a delight to see a reference to condensers.

Cheers,

John Ferguson, AI4TO

On 9/4/19 3:35 PM, greenboxmaven via Groups.Io wrote:
 ....As I reconnected each stage, the voltage incresed. Finally I reconnected the 500 PFD "doorknob" condenser,and the voltage plummited. Having worked on television receivers decades ago, I had several replacement condensers on hand, and installing one brought the scope to life with a dazzling trace. ....

      Bruce Gentry, KA2IVY

Glenn Little
 

Was the cap red?
I had a 453 MOD703H with no HV.
Found the red doorknob 500 pF HV filter cap had arced internally (saw the blackened area through the translucent red epoxy).
Replaced the cap with a ceramic 500 pF doorknob cap and the scope has worked fine since.
This was about 20 years ago.

Glenn

On 9/4/2019 3:35 PM, greenboxmaven via Groups.Io wrote:
I was doing some cleaning up in my shop and came across something?? from a Tektronix 454 scope that resonates with this transformer discussion. I think "contaminated" epoxy?? may have caused another problem in a scope high voltage power supply. The 454 had several problems that were easy to find and remedy. It still would not give anything more than a very pale glow on the screen. The second anode voltage was only about 2,000 volts. Disconnecting the transformer from the tripler circuit gave a husky corona arc to a screwdriver, so I began testing the diodes and condensers in the multiplier circuit. As I reconnected each stage, the voltage incresed. Finally I reconnected the 500 PFD "doorknob" condenser,and the voltage plummited. Having worked on television receivers decades ago, I had several replacement condensers on hand, and installing one brought the scope to life with a dazzling trace. One of the NuVistors in the preamp was barely alive, so that channel could not be brought onto the screen. A replacement got the scope working very well. I took the condenser over to a friend who has a Hipot tester, and it was fine until the voltage reached about 2,000 volts. The condenser is rated at 20KV, and the voltage on it in the scope is 12KV. The bad condenser is epoxy molded, the 60 year old replacement is tan high voltage bakelite. I wonder if Tektronix specified or accepted an epoxy formula that goes bad?

?????????? Bruce Gentry, KA2IVY


On 9/3/19 9:23 AM, Chuck Harris wrote:
The epoxy changes and makes an already lossy situation even
worse.


-Chuck Harris






--
-----------------------------------------------------------------------
Glenn Little ARRL Technical Specialist QCWA LM 28417
Amateur Callsign: WB4UIV wb4uiv@... AMSAT LM 2178
QTH: Goose Creek, SC USA (EM92xx) USSVI LM NRA LM SBE ARRL TAPR
"It is not the class of license that the Amateur holds but the class
of the Amateur that holds the license"

greenboxmaven
 

It was not a pure brilliant red, sort of a translucent reddish brown. I will guess different manufacturers or batches of the condensers could vary considerably in color, especially if they are translucent. I never thought to douse the lights and see if the flashover was visible.

Bruce, KA2IVY

On 9/4/19 11:59 PM, Glenn Little wrote:
Was the cap red?
I had a 453 MOD703H with no HV.
Found the red doorknob 500 pF HV filter cap had arced internally (saw the blackened area through the translucent red epoxy).
Replaced the cap with a ceramic 500 pF doorknob cap and the scope has worked fine since.
This was about 20 years ago.

Glenn

On 9/4/2019 3:35 PM, greenboxmaven via Groups.Io wrote:
I was doing some cleaning up in my shop and came across something?? from a Tektronix 454 scope that resonates with this transformer discussion. I think "contaminated" epoxy?? may have caused another problem in a scope high voltage power supply. The 454 had several problems that were easy to find and remedy. It still would not give anything more than a very pale glow on the screen. The second anode voltage was only about 2,000 volts. Disconnecting the transformer from the tripler circuit gave a husky corona arc to a screwdriver, so I began testing the diodes and condensers in the multiplier circuit. As I reconnected each stage, the voltage incresed. Finally I reconnected the 500 PFD "doorknob" condenser,and the voltage plummited. Having worked on television receivers decades ago, I had several replacement condensers on hand, and installing one brought the scope to life with a dazzling trace. One of the NuVistors in the preamp was barely alive, so that channel could not be brought onto the screen. A replacement got the scope working very well. I took the condenser over to a friend who has a Hipot tester, and it was fine until the voltage reached about 2,000 volts. The condenser is rated at 20KV, and the voltage on it in the scope is 12KV. The bad condenser is epoxy molded, the 60 year old replacement is tan high voltage bakelite. I wonder if Tektronix specified or accepted an epoxy formula that goes bad?

?????????? Bruce Gentry, KA2IVY


On 9/3/19 9:23 AM, Chuck Harris wrote:
The epoxy changes and makes an already lossy situation even
worse.


-Chuck Harris







Jean-Paul
 

Hello all, I have been following this most interesting thread.

As a retired magnetics designer and manufacturer, virtually all commercial firms are fabricating mass production in China. The set-up is costly and runs of less than 1000 difficult to get a quote. Your experience with Sun is typical.

Some HV transformers need special core shapes, custom bobbins, litz or very fine guage wire and/or a special winding machine. So not easy to design, tool or build.

Finally the impregnated secondaires needed a vacuum oven to pull air bubbles out of the epoxy, plastic or varnish liquid before it cures solid. Any voids or air bubbles remaining will eventually ionize and lead to carbon tracking and failures.

In mid 1980s it took us 1 yr to design and to debug the potting process and materials for a 12 kV custom avionics CRT module with transformer and voltage multiplier! We made 1000s but always got some DOA after potting, yields increased from 50% to 98% after the procedure was optimized.

Just the rumblings of an old retired EE!

Bon courage,

Jon in Paris

(also posted in other HV trsf thread)