Restore an old BK 1604 Isolation Transformer

[rmel]

Hi All,

I bought a BK model 1604 isolation transformer a few years back, and I am only now getting a chance to look at it.  The unit looks really old, and I'm not sure the transformer inside is safe to use.  I plan to replace any wiring that seems old/worn, but I'm not sure about that transformer inside.  It's looking a bit burnt but could it just be from age?

I'm not sure I want to go through the effort of rewinding it, so I did a quick search at digikey and found this:

   It's a115V-115V transformer, 150VA, is about $45 US, plus tax and shipping (https://www.digikey.com/products/en?mpart=N-73A&v=237)

Any thoughts on whether I should replace the transformer or try to restore it?

Thanks

[Ian.M]

Assuming there's no obvious insulation damage, the output socket still grips plugs firmly and the breaker still trips at close to 1.25A, what's to restore?  It either passes 500V DC HiPot tests from primary to secondary, primary to ground and secondary to ground, or it doesn't.  If the insulation resistances are questionable post them here for comment.  If it  fails, the transformer inside it must be replaced or the whole unit condemned as B.E.R.  Rewinding (by a qualified professional) wouldn't be economic for such a simple low VA transformer.
 

[rmel]

I can check the insulation for damage and the output socket, but I'm at a loss for how to do HiPot tests or insulation resistances.

I guess I got a little scared by the look of the transformer.  This makes me think of the high voltage tests I've seen on Dave's videos. 

What's the safest way to test the circuit breaker?  I guess I would want to put a 2AMP fuse on the primary side and then plug in a cord with the hot and neutral wires shorted?  I'll have to think about this a bit..

[Ian.M]

Do *NOT*  test the breaker by shorting or otherwise overloading the isolating transformer.  Instead test it by passing a controlled current through just the breaker.  If you are careful to insulate the mains plug there's no need to remove it from the isolating transformer circuit to test it.

Do you have a Variac?   If so it makes testing the circuit breaker *MUCH* easier.  Simply feed an ordinary step-down transformer with a low voltage output rated for >2.5A out from the Variac, and use its low voltage output to feed a series circuit of a power resistor of a few ohms (with a high enough wattage rating to handle 2.5A), a DMM on its 10A AC range and the breaker you want to test.  Ramp up the Variac slowly from zero till you reach 1.25A through the breaker and wait for at least ten minutes to confirm it doesn't trip, then quickly ramp up to 2.5A and confirm it trips in under 20 seconds.

If you don't have a Variac and a suitable low voltage transformer, you could hook up a 300W to 500W 120V mains powered resistive load (e.g. some type of heater - not with a fan or electronic control) and a light dimmer to control the power and thus the RMS current, and use that to test the breaker.

To do a HiPot test you need a high voltage insulation tester, basically a <1mA current limited high voltage supply in series with a surge protected microAmmeter.  They are often calibrated in resistance but may read leakage current directly. If you've got access to one, follow its instruction manual, and don't forget to select a suitable test voltage as you want to test the insulation integrity not over-stress it to breakdown and resulting destruction.  Although its possible to improvise a DIY one, mucking around with 500V or higher DC isn't novice friendly so I am *NOT* going to give any suggestions or instructions.   

[ArthurDent]

I don't see any obvious problems with the transformer or the wiring. Lots of transformers are painted to keep everything in place to prevent loose laminations from vibrating and causing a 60 Hz hum. Transformers are low on my list of suspects in any piece of electronic equipment. Here is a photo of a high end transformer that has black paint on it except where it chipped off on the top. On your transformer you can see where there was a piece of tape that was painted over then finally removed so the part number would show.

If it doesn't smell burned then it is probably okay. Power it and plug an incandescent 60W-100W lamp in the output and leave it on for some time and see how hot the transformer gets. It is normal for it to heat up some but you should be able to hold you finger on it without getting burned.

[TheMG]

It's probably just an older style transformer, they were built that way, and this one has been painted black.

Much much more likely the transformer would fail with turn-to-turn shorts (causing it to draw excess current and heat up) long before it ever developed a primary-to-secondary short, by which point there would have been significant amount of magic smoke pouring out of the transformer.

Some simple tests you can do:

1) Plug it in and make sure it doesn't draw excessive current all on its own. There will be a bit of current due to magnetizing the transformer and some losses but it shouldn't be anything significant.

2) Check voltage at output.

3) Check voltage from both output terminals to ground. There may be some small voltage due to capacitance between the windings if the transformer isn't screened (a grounded layer between the windings), but this voltage should drop off quickly when adding 100k resistor to ground. In other words, if the isolation is good there should be virtually no current available between the output terminals and ground.

4) Plug in a load and make sure it's able to supply it correctly.

If in doubt about the breaker, replace it. They're cheap enough.

Other than that, make sure the power cord is in good condition and all connections are secure.

[WattsThat]

It’s not very usable with a two wire socket so...

The only thing I would do would be to replace that 1960’s two pole receptacle with a proper 3 wire grounded receptacle. It’s a shame the silk screening is so close, you could turn the new receptacle 90 degrees, that may avoid the legends.

Footnote: yes, you could get any transformer color you wanted back then, so long as it was black (the same was said by Henry Ford about the model A).

[rmel]

Do *NOT*  test the breaker by shorting or otherwise overloading the isolating transformer.  Instead test it by passing a controlled current through just the breaker.  If you are careful to insulate the mains plug there's no need to remove it from the isolating transformer circuit to test it.

Do you have a Variac?   If so it makes testing the circuit breaker *MUCH* easier.  Simply feed an ordinary step-down transformer with a low voltage output rated for >2.5A out from the Variac, and use its low voltage output to feed a series circuit of a power resistor of a few ohms (with a high enough wattage rating to handle 2.5A), a DMM on its 10A AC range and the breaker you want to test.  Ramp up the Variac slowly from zero till you reach 1.25A through the breaker and wait for at least ten minutes to confirm it doesn't trip, then quickly ramp up to 2.5A and confirm it trips in under 20 seconds.

If you don't have a Variac and a suitable low voltage transformer, you could hook up a 300W to 500W 120V mains powered resistive load (e.g. some type of heater - not with a fan or electronic control) and a light dimmer to control the power and thus the RMS current, and use that to test the breaker.

To do a HiPot test you need a high voltage insulation tester, basically a <1mA current limited high voltage supply in series with a surge protected microAmmeter.  They are often calibrated in resistance but may read leakage current directly. If you've got access to one, follow its instruction manual, and don't forget to select a suitable test voltage as you want to test the insulation integrity not over-stress it to breakdown and resulting destruction.  Although its possible to improvise a DIY one, mucking around with 500V or higher DC isn't novice friendly so I am *NOT* going to give any suggestions or instructions.   

Ok, thanks.  I wasn't feeling quite comfortable about shorting the wires either and probably would have done it.  And passing current through the breaker only is a good idea!  I don't have a variac however, so I'll have to see about a resistive load.

[rmel]

I don't see any obvious problems with the transformer or the wiring. Lots of transformers are painted to keep everything in place to prevent loose laminations from vibrating and causing a 60 Hz hum. Transformers are low on my list of suspects in any piece of electronic equipment. Here is a photo of a high end transformer that has black paint on it except where it chipped off on the top. On your transformer you can see where there was a piece of tape that was painted over then finally removed so the part number would show.

If it doesn't smell burned then it is probably okay. Power it and plug an incandescent 60W-100W lamp in the output and leave it on for some time and see how hot the transformer gets. It is normal for it to heat up some but you should be able to hold you finger on it without getting burned.

Thanks, that does look very similar to the one in the unit.  I'll try testing it with a 100W bulb and see how it does.

Poking around at the wires, they all feel in pretty good condition.  I was a bit concerned about the cable that plugs into the wall socket when I first looked at it, but when I handle it, it feels in good condition too.

[rmel]

It's probably just an older style transformer, they were built that way, and this one has been painted black.

Much much more likely the transformer would fail with turn-to-turn shorts (causing it to draw excess current and heat up) long before it ever developed a primary-to-secondary short, by which point there would have been significant amount of magic smoke pouring out of the transformer.

Some simple tests you can do:

1) Plug it in and make sure it doesn't draw excessive current all on its own. There will be a bit of current due to magnetizing the transformer and some losses but it shouldn't be anything significant.

2) Check voltage at output.

3) Check voltage from both output terminals to ground. There may be some small voltage due to capacitance between the windings if the transformer isn't screened (a grounded layer between the windings), but this voltage should drop off quickly when adding 100k resistor to ground. In other words, if the isolation is good there should be virtually no current available between the output terminals and ground.

4) Plug in a load and make sure it's able to supply it correctly.

If in doubt about the breaker, replace it. They're cheap enough.

Other than that, make sure the power cord is in good condition and all connections are secure.

I plugged it in and the draw from the wall is about 0.42A, according to my Kill-A-Wall, as I don't have a AC clamp meter.  I didn't expect it to be that high. Is that normal?


Checking the voltages using my BK2709B:

- Across the output terminals: 131VAC (voltage at the wall is 124.2VAC)
- Left output terminal and case: 49.6VAC
- Right output terminal and case: 131VAC

I didn't expect such high voltages between the outputs and ground.  I assume this is what you meant by the voltage due to the capacitance, but should it be this high?  I tried connecting a 100k resistor to the case and measuring from each output terminal to the resistor, but I get the same voltages.

In current mode, I measured 47uA from the left and 17uA from the right, which I guess is what you meant by there's virtually no current.

As for the breaker, I guess it works, as I accidentally tripped in when I forgot to switch back from resistance to volts.  The meter still seems ok.  Doh!

[rmel]

It’s not very usable with a two wire socket so...

The only thing I would do would be to replace that 1960’s two pole receptacle with a proper 3 wire grounded receptacle. It’s a shame the silk screening is so close, you could turn the new receptacle 90 degrees, that may avoid the legends.

Footnote: yes, you could get any transformer color you wanted back then, so long as it was black (the same was said by Henry Ford about the model A).

That hadn't occurred to me at this point.  But if I replace the two prong receptacle with a 3 prong one, where would I connect the ground to?  Just leave it floating?

[bdunham7]

The no-load current is likely reactive, what does the Kill-a-Watt say for Watts and Power Factor?

The voltages you measure are due in part to the input impedance of your DMM.  Try just putting the red lead into each socket without connecting the black one to anything.  You'll still get a reading.

A 100K resistor in series won't change anything, your meter impedance is much, much higher.  You should put the resistor from each terminal to ground and then measure across it.

Your 'leakage currents' are normal for this type of transformer and probably don't represent leakage from the line side, but rather leakage from the secondary to the case.  I would suggest the HiPot testing if you are going to rely on it.  HiPot = High Potential and is just a test done with a "Megger" or insulation tester, basically a high-value Ohmmeter that uses a high test voltage.  The one I have gives you options of 250, 500 or 1000VDC.  Your leakage may vary with temperature and humidity.

Did you check your fuses? 

[rmel]

The no-load current is likely reactive, what does the Kill-a-Watt say for Watts and Power Factor?

The voltages you measure are due in part to the input impedance of your DMM.  Try just putting the red lead into each socket without connecting the black one to anything.  You'll still get a reading.

A 100K resistor in series won't change anything, your meter impedance is much, much higher.  You should put the resistor from each terminal to ground and then measure across it.

Your 'leakage currents' are normal for this type of transformer and probably don't represent leakage from the line side, but rather leakage from the secondary to the case.  I would suggest the HiPot testing if you are going to rely on it.  HiPot = High Potential and is just a test done with a "Megger" or insulation tester, basically a high-value Ohmmeter that uses a high test voltage.  The one I have gives you options of 250, 500 or 1000VDC.  Your leakage may vary with temperature and humidity.

Did you check your fuses? 

The Kill-a-watt reads 8 Watts and 0.18 power factor -- I still need to do further reading on PF... 

I tried measuring a 100k across each terminal to ground, and now I get about 65VAC.

And I didn't check my fuses yet.  The meter still seems to work on all the ranges I tested.   

I did a quick google for an insulation tester and found this:  https://www.amazon.com/gp/product/B07BGSPH8P?pf_rd_r=23H3FFB1CKP9TETD74AQ&pf_rd_p=edaba0ee-c2fe-4124-9f5d-b31d6b1bfbee

Is that what you're referring to as a "megger"?

[bdunham7]

Yes, that's the one.  Good for checking old HV capacitors as well.

65V thru 100K = 650uA, seems a bit high.  Did you confirm the value of the resistor with the meter?

[rmel]

Yes, that's the one.  Good for checking old HV capacitors as well.

65V thru 100K = 650uA, seems a bit high.  Did you confirm the value of the resistor with the meter?

Yeah, I did use the meter, but after seeing your question, tried two more meters and they are also reading about the same.  The lowest value I got from a third meter was 62 VAC.

Then I realized why I might be getting these readings.  I had wired up the resistors to a terminal block so that I could measure both output terminals without having wires sticking out randomly.  So both resistors were wired to ground at the same time.



When I removed the red wires and connected only one at a time, I get much lower values across the resistors:  4.75 VAC and 1.68 VAC

So, now I think I understand what's happening!

When I connected the two 100k resistors to ground at the same time, since almost 0 current should be flowing to ground, I essentially connected a 200k resistor across the two terminals of the isolation transformer, like this:




So, calculating the current in this case:

130 V / 200K = 650 uA

And that's the 650 uA that I was measuring the first time.  I just learned something!

Thanks!

[WattsThat]

It’s not very usable with a two wire socket so...

The only thing I would do would be to replace that 1960’s two pole receptacle with a proper 3 wire grounded receptacle. It’s a shame the silk screening is so close, you could turn the new receptacle 90 degrees, that may avoid the legends.

Footnote: yes, you could get any transformer color you wanted back then, so long as it was black (the same was said by Henry Ford about the model A).

That hadn't occurred to me at this point.  But if I replace the two prong receptacle with a 3 prong one, where would I connect the ground to?  Just leave it floating?

Metal boxes left floating is how people get electrocuted.

Exposed metal of any mains powered device should always be grounded for safety. You don’t want to come in contact with a metal box that has become energized because it wasn’t properly grounded. Dave has covered proper chassis grounding in most of his mains supplied device tear-downs showing how some devices are downright lethal due to the lack of or improper grounding.

The device in question has a proper three wire supply plug/cord set so the metal chassis should be grounded via that ground wire for personnel safety. Any paint should have been removed such that the supply ground connection has a good, low impedance connection the the chassis/frame. The ground from the three wire outlet socket should go to the same grounding point, typically using a ring tongue crimp lug.

That transformer dates back to a time when tube/value radios had one side of the line cord set connected to the metal chassis of the radio. That metal chassis was inside a insulated box of some kind with plastic/Bakelite knobs insulating anyone operating the device. In order to service one of those devices, you had to remove the chassis from the housing to work on it, thereby creating a potentially dangerous situation because if the plug wasn’t oriented correctly, the chassis would be at mains potential. Using the transformer floated the chassis of the device under test, reducing the hazard for the person working on the unit under repair.

Have you watched Dave’s video on how to not blow up your scope?

https://youtube.com/watch?v=xaELqAo4kkQ

[rmel]

Metal boxes left floating is how people get electrocuted.

Exposed metal of any mains powered device should always be grounded for safety. You don’t want to come in contact with a metal box that has become energized because it wasn’t properly grounded. Dave has covered proper chassis grounding in most of his mains supplied device tear-downs showing how some devices are downright lethal due to the lack of or improper grounding.

The device in question has a proper three wire supply plug/cord set so the metal chassis should be grounded via that ground wire for personnel safety. Any paint should have been removed such that the supply ground connection has a good, low impedance connection the the chassis/frame. The ground from the three wire outlet socket should go to the same grounding point, typically using a ring tongue crimp lug.

That transformer dates back to a time when tube/value radios had one side of the line cord set connected to the metal chassis of the radio. That metal chassis was inside a insulated box of some kind with plastic/Bakelite knobs insulating anyone operating the device. In order to service one of those devices, you had to remove the chassis from the housing to work on it, thereby creating a potentially dangerous situation because if the plug wasn’t oriented correctly, the chassis would be at mains potential. Using the transformer floated the chassis of the device under test, reducing the hazard for the person working on the unit under repair.

Have you watched Dave’s video on how to not blow up your scope?

https://youtube.com/watch?v=xaELqAo4kkQ

Thanks for the explanation.  I thought I did understand the issues of grounding, but now I'm not sure.  I thought the point of using an isolation transformer was to avoid creating a ground loop that could blow up your scope or shock you.

I will go back and watch Dave's video(s) again.

[bdunham7]

If you install a 3-prong plug, tie the ground to the case at the same point the incoming ground is connected to the case.  If you are testing 2-prong non-grounded device, then nothing has changed.  If you connect a 3-prong grounded device, then the ground is passed through, but the power is still isolated and not referenced to ground, except by the small leakage you saw--unless the device itself has leakage.  You can also use a 'cheater plug' for the purpose of testing the ground leakage, which is useful for diagnosing things that trip GFCIs.  But be careful out there!