Suggestions for an isolation transformer

[Gregg]

Now I got some questions about this isolation transformer.   If I have a DUT hooked to the isolation transformer, and someone where to touch something hot, like a charged capactitor on the anode side, they wouldn't get shocked, because their is no path to a common ground through the person, right?   But if they where to touch the anode and cathode at the same time, then they would get shocked?   The current would flow through the capacitor, through them, and back through the capacitor, correct?

The best way to wrap your head around what is happening with your isolation transformer is to think in terms of complete circuits; mainly that YOU don’t want to become part of any circuit.
The originating circuit is the primary of the isolation transformer, but if it is truly isolated, it is completely separate from the secondary.  The source of the primary, here, is the transformer somewhere outside your house that takes high voltage and converts it to 115/230 volts that in turn goes to your house via the meter and main disconnect.  The primary has the neutral grounded and therefore there is potential from the line to anything grounded like your body if you are in contact with ground.  (This is a little over simplified, but it will suffice for this example.)  The line and neutral connected the primary of your isolation transformer forms a complete circuit back to the transformer outside your house.  The key is to think in terms of “complete circuit”.

The secondary of your isolation transformer can be considered a source that is separated from the primary and has virtually no voltage in reference to ground or the primary; it is forming a separate circuit or you could think of it as the source for a separate circuit. 
(There may be some induced or capacitive voltage between this isolated secondary and ground that you can read with a high impedance digital meter but if you put something like a 1K resistor across where you are probing it will drop to an insignificant figure because there is virtually no current available.)
The secondary of the isolation transformer has no connection with the primary unless you ground one leg, then the purpose of isolation kind of goes away. 

When you power your DUT with the isolated source but it has ground planes or a chassis that should be grounded, all sorts of weird things can happen.  If you visualize every possible circuit that can be associated with your DUT and your test equipment and take steps to mitigate disaster, you’ll be off to a good start.  No guarantees, you are on your own, think three times before probing and never touch two parts of a mains circuit at the same time.  At least gloves and a face shield are highly recommended for beginners.  Check with your multimeter before sticking your probe on something may save lots of grief and smoke.

[Gregg]

I don't understand why the fuse blew when I switched the hair dryer to the higher current mode of operation.   Why did the NTC not open up?   The higher current should have caused a higher temp, and I thought that was the purpose, to prevent too high of a current from going out the isolation transformer outputs.

Fuses are NOT current limiters; they are there to prevent the wiring form over heating from too much current.  The hair dryer on high drew more current than your fuse would allow and thereby saved your transformer.

(Edit for clarity): The hair dryer on high tried to pull more current than your fuse would allow and the fuse blew saving your transformer.

[ArthurDent]

Here is the description on line that should clear things up.

"With an NTC thermistor, when the temperature increases, resistance decreases. ... A PTC thermistor works a little differently. When temperature increases, the resistance increases, and when temperature decreases, resistance decreases. This type of thermistor is generally used as a fuse."

So the NTC is used to limit inrush current the first few seconds after power is supplied to a transformer by being high resistance when cold and low resistance when hot. They only need to be used for a few seconds at power up so this is why a relay circuit is often used to bypass them once the power has been on for a few seconds. The NTC can then cool down and be ready to do its job again.

The PTC is just the opposite, when the current gets too great, the resistance increases quite a bit and reduces or disconnects power to the load. Most return to low resistance when they cool down.

[Cliff Matthews]

I think you have the idea. If someone (I wouldn't know why..) was touching an earth ground while touching something inside the DUT you are right, there is no low resistance path.

Q- At 30-100w each time you observed some rise in output voltage over time (this is where NTC got closer to ideal temp/resistance and had less voltage drop). Did you see this with the hair-dryer on low heat 500w a bit quicker?

What you observed so far, is this model boosts mains by 4v at no load and is quite reliant on the 4-5v drop of the NTC. Do you see why clamping a relay across the NTC might have been a bad idea? Then output for light loads could be over 130v

*edit - For those that have not looked at the data sheet, these NTC's are part of the design to regulate the transformers output at light loads..  At heavier loads these NTC's will drop to a negligible resistance (under 0.1 ohm).

Those NTCs could act like a soft fuse, a resettable one, right?

NTC's are also called Varistor's - They act like a variable resistor transitioning into a low resistance wire at ~105c

[Spork Schivago]

Now I got some questions about this isolation transformer.   If I have a DUT hooked to the isolation transformer, and someone where to touch something hot, like a charged capactitor on the anode side, they wouldn't get shocked, because their is no path to a common ground through the person, right?   But if they where to touch the anode and cathode at the same time, then they would get shocked?   The current would flow through the capacitor, through them, and back through the capacitor, correct?

The best way to wrap your head around what is happening with your isolation transformer is to think in terms of complete circuits; mainly that YOU don’t want to become part of any circuit.
The originating circuit is the primary of the isolation transformer, but if it is truly isolated, it is completely separate from the secondary.  The source of the primary, here, is the transformer somewhere outside your house that takes high voltage and converts it to 115/230 volts that in turn goes to your house via the meter and main disconnect.  The primary has the neutral grounded and therefore there is potential from the line to anything grounded like your body if you are in contact with ground.  (This is a little over simplified, but it will suffice for this example.)  The line and neutral connected the primary of your isolation transformer forms a complete circuit back to the transformer outside your house.  The key is to think in terms of “complete circuit”.

The secondary of your isolation transformer can be considered a source that is separated from the primary and has virtually no voltage in reference to ground or the primary; it is forming a separate circuit or you could think of it as the source for a separate circuit. 
(There may be some induced or capacitive voltage between this isolated secondary and ground that you can read with a high impedance digital meter but if you put something like a 1K resistor across where you are probing it will drop to an insignificant figure because there is virtually no current available.)
The secondary of the isolation transformer has no connection with the primary unless you ground one leg, then the purpose of isolation kind of goes away. 

When you power your DUT with the isolated source but it has ground planes or a chassis that should be grounded, all sorts of weird things can happen.  If you visualize every possible circuit that can be associated with your DUT and your test equipment and take steps to mitigate disaster, you’ll be off to a good start.  No guarantees, you are on your own, think three times before probing and never touch two parts of a mains circuit at the same time.  At least gloves and a face shield are highly recommended for beginners.  Check with your multimeter before sticking your probe on something may save lots of grief and smoke.

I am not so much worried about me.   I am careful and do safe stuff now, like check capacitors, even if there's bleeder resistors, stuff like that.   I was hoping I could use this isolation transformer to slowly introduce my young daughter into electronics, when she gets a bit older, with it being a little safer.   I realize that even if there was no potential for shock, it would be a horrible idea to not teach her safe practices.   Sometimes, it's really hard to repair electronics now because I spend a lot of time with my daughter.   It would be nice to be able to work on some of this stuff with her around, but I would never do anything to put her in danger.

I want to start learning more about how to use my oscilloscope and power supplies.   I think this is where the isolation transformer will come in handy for the time being.   I don't want to accidently blow up my scope because I tried measuring something on the primary side of a SMPS on accident, or something dumb like that.

[Spork Schivago]

I don't understand why the fuse blew when I switched the hair dryer to the higher current mode of operation.   Why did the NTC not open up?   The higher current should have caused a higher temp, and I thought that was the purpose, to prevent too high of a current from going out the isolation transformer outputs.

Fuses are NOT current limiters; they are there to prevent the wiring form over heating from too much current.  The hair dryer on high drew more current than your fuse would allow and thereby saved your transformer.

(Edit for clarity): The hair dryer on high tried to pull more current than your fuse would allow and the fuse blew saving your transformer.

So the transformer would have tried providing that current and overheated and died?   Wouldn't the NTC though limit the amount of current the DUT was trying to draw and protect the transformer?   

[Spork Schivago]

Here is the description on line that should clear things up.

"With an NTC thermistor, when the temperature increases, resistance decreases. ... A PTC thermistor works a little differently. When temperature increases, the resistance increases, and when temperature decreases, resistance decreases. This type of thermistor is generally used as a fuse."

So the NTC is used to limit inrush current the first few seconds after power is supplied to a transformer by being high resistance when cold and low resistance when hot. They only need to be used for a few seconds at power up so this is why a relay circuit is often used to bypass them once the power has been on for a few seconds. The NTC can then cool down and be ready to do its job again.

The PTC is just the opposite, when the current gets too great, the resistance increases quite a bit and reduces or disconnects power to the load. Most return to low resistance when they cool down.

I confused the two again, the NTC and the PTC.   I was thinking the NTC where a PolyFuse.   I see where I went wrong now.   Thank you!!!!!!!

[Spork Schivago]

I think you have the idea. If someone (I wouldn't know why..) was touching an earth ground while touching something inside the DUT you are right, there is no low resistance path.

Q- At 30-100w each time you observed some rise in output voltage over time (this is where NTC got closer to ideal temp/resistance and had less voltage drop). Did you see this with the hair-dryer on low heat 500w a bit quicker?

What you observed so far, is this model boosts mains by 4v at no load and is quite reliant on the 4-5v drop of the NTC. Do you see why clamping a relay across the NTC might have been a bad idea? Then output for light loads could be over 130v

*edit - For those that have not looked at the data sheet, these NTC's are part of the design to regulate the transformers output at light loads..  At heavier loads these NTC's will drop to a negligible resistance (under 0.1 ohm).

Those NTCs could act like a soft fuse, a resettable one, right?

NTC's are also called Varistor's - They act like a variable resistor transitioning into a low resistance wire at ~105c

Yes!   I did, and I figured the voltage change was from the NTC heating up.   I did notice that it wasn't so noticable with the hair dryer though.   In my mind, I thought it was just reaching optimal temperature after running for a while, and providing a more stable resistance, I did not realize the DUT was affecting the flucuations in voltage.   Thank you!!!!   And now I understand why the relay would be a bad idea.  That's pretty smart, how they designed this to work with those NTCs.   Pretty simple, but effective.

[Spork Schivago]

Is there a way to modify the isolation transformer so instead of blowing fuses if a DUT tried to draw too much current, it would temporarily shut down until the load was removed or the short / whatever was fixed?

I thought maybe a PolyFuse, but I think they would be a really bad idea here because I don't think they make one with the proper paremeters, such as the Vmax or Imax being acceptable to just replace the fast acting fuses with a PTC.   I also worry that maybe they wouldn't open fast enough.   I was thinking maybe with some additional parts, I could use a PTC or something.   I feel I might blow a fuse more times than I'd like to replace them, using this to help repair electronics.

Why would a ground plane or a grounded chassis on the DUT cause problems with an isolation transformer?   To the isolation transformer, doesn't it just appear that the secondary side of the transformer is the primary power source?   The begining and end of the circuit?

**EDIT:  I think I made a mistake, thinking I was done with the modifications of this isolation transformer.   I unhooked the grounds from the outlet side and from the chassis side, however, I really think I should have the grounds connected to each other, but still floating.   This way, let's say two people are working on two seperate devices, and both devices have a short.   One to line, one to ground.   If they don't realize it and touch each other, while touching their DUT, they could be shocked, perhaps fatally.   So I think I should daisy chain the output grounds together, and leave them floating (just connected to each other).

[Cliff Matthews]

IMO, grounds from each DUT should float. An 8-amp breaker on the output should work. https://www.amazon.com/AMP-32VDC-250VAC-CIRCUIT-BREAKER/dp/B00OGL7GMO/ref=sr_1_3?keywords=8+Amp+Circuit+Breaker

Now "if 2 people touch each other, while touching their shorted DUT's" I don't know what say.. (sounds kinky for starter's and second, one DUT would require neutral to touch the chassis, while the other would have to have the line touching the chassis..)

OK then, It sounds rare but you can put up a warning sign that says "DANGER - NO TOUCHING IN MY SHOP!!" 

[Spork Schivago]

IMO, grounds from each DUT should float. An 8-amp breaker on the output should work. https://www.amazon.com/AMP-32VDC-250VAC-CIRCUIT-BREAKER/dp/B00OGL7GMO/ref=sr_1_3?keywords=8+Amp+Circuit+Breaker

Now "if 2 people touch each other, while touching their shorted DUT's" I don't know what say.. (sounds kinky for starter's and second, one DUT would require neutral to touch the chassis, while the other would have to have the line touching the chassis..)

OK then, It sounds rare but you can put up a warning sign that says "DANGER - NO TOUCHING IN MY SHOP!!" 

Yes, it would be rare, but would there be any downfalls for having the output grounds connected?   The various devices would then have a ground at the same potential.   Is there any reason I wouldn't want to tie them together?   If there's no downfalls, even though it'd be a very rare circumstance that would need to occur for someone to get hurt, I'd rather take the extra precaution.

Wouldn't I need two circuit breakers?   I was thinking one like this:
https://www.amazon.com/dp/B004QMYXME/ref=sspa_dk_detail_0?psc=1&pd_rd_i=B004QMYXME&pd_rd_w=Sp2ZL&pf_rd_p=733540df-430d-45cd-9525-21bc15b0e6cc&pd_rd_wg=YAzN3&pf_rd_r=K9E4PA5W4FF9PPEPE7SZ&pd_rd_r=2a9f89bd-571c-11e9-8fee-2b0771c23e42

One rated for 5 amp, the other rated for the 10 amp?   Instead of having the lines go to the fuses, I just run them to the breakers instead?

[Cliff Matthews]

I see where you're going. You want to put the breaker on the primary side? (replacing the fuses?).

Not a good idea physically, since they are part of the IEC input assembly, and not a good idea electrically to eliminate them since they still should be used as a fail-safe. FWIW, the 5 amp fuse will only be used if red the primary 230v is selected as input, and it's worth noting the fuses are unrelated to what output voltage you select with the switch on the right side.

The breaker I mentioned was for the secondary (output) and would have been 9 amps, but I'm not sure those are easy to get.

[ArthurDent]

"Wouldn't I need two circuit breakers?   I was thinking one like this:

Make sure if you use breakers they are rated for the proper voltage. The rating on the one you showed is 32VDC for auto/RV use.

Also think about how long an overload will take to trip the breaker. Thermal type breakers like the one shown in the link or on most power strips can take some time with a moderate overload to trip. If you've ever used an electric space heater on a power strip, you may have had it trip after several minutes, then you had to wait a few minutes for it to cool down so it can be reset.

A fast acting fuse or a thermo magnetic trip breaker might be better.

https://thegrid.rexel.com/en-us/knowledge/product-faqs/w/wiki/323/how-circuit-breakers-combine-thermal-and-magnetic-protection

[soldar]

I can only upload 2MB of pictures at one time, so this will be in multiple posts.

You could reduce the pixel size and file size as a consideration to those of us with slow connections. There is no need to post such huge sizes.

[Gregg]

Fuses have a wonderful property of taking time to replace; time that one can reflect on why the fuse went open circuit.  Breakers are too easy to quickly reset without thinking; I’ve see bad results of just resetting a breaker without first carefully thinking of all the ramifications.
For circuit breaker over-current protection I suggest making up some 2 pole low current breakers to put on the output.  There are many DIN rail mount ones available like this: https://www.ebay.com/itm/MERLIN-GERIN-C60N-2-POLE-6-AMP-CIRCUIT-BREAKER/173373376160  Two pole breakers are preferred for the output since there is no common reference.  Many of the pop-out breakers like Cliff mentioned will not trip against finger pressure (if you continue to hold the button in after it tripped, it will stay connected). 
I have made up several boxes with three pole breakers like the one mentioned above for when I used to attach test equipment direct to 480 volt 3 phase bus bars.  I have them rated for ½ amp through 4 amp; they are very versatile and probably saved my butt more than once.
I wouldn’t worry too much about the breaker opening fast to protect the transformer, the fuse on the input should do that job quite well.  Transformers are very robust and will take fairly large overloads for a short time, heat being a major factor in insulation breakdown; that is why things like welders have a short interval duty cycle to get the ratings they advertise. As long as you don’t overheat your transformer and let it cool down after an overload, it will be OK.

[Cliff Matthews]

I can only upload 2MB of pictures at one time, so this will be in multiple posts.

You could reduce the pixel size and file size as a consideration to those of us with slow connections. There is no need to post such huge sizes.

Yes, IrfanView image compression was already suggested in #59
( https://www.eevblog.com/forum/beginners/suggestions-for-an-isolation-transformer/msg2315613/#msg2315613 )

[soldar]

Yes, IrfanView image compression was already suggested in #59

Sorry I missed that. I heartily support the recommendation for Irfanview which is what I use. I wish there were a native Linux version although it does run well under (or is it "over"?) WINE.

[Red Squirrel]

I've had good luck finding transformers in the 1kva and even higher on ebay.  I just put in actual numbers like "1000va transformer" (or 1kva etc) in the search then sort by "price + shipping".

I ended up impulse buying a 3000va variac (not isolated) and a 1500va transformer (isolated) a while back.  I have not gotten around to it yet but my goal is to build a basic AC power supply using both together.   The transformer is a 120v to 480v so I feed it with the variac and can get any voltage between around 1-2v to around 500v.  Typically will use in 120v mode as isolation transformer but having options for lower/higher voltages is nice too.

If I feel like getting fancy I might drive the variac with a stepper motor and arduino and add some basic programming features.

[Cliff Matthews]

Hmm.. rectumfried and filterated that's 700v @ ~3 amps!   Test that Arduino code, it could be your last compile!

[jerryk]

I just purchased a Triplite IS500HG  with the intent on using it as an isolation transformer.  This thread popped up right the same day UPS dropped it at my door so I have been following it and trying to make sure I do things right when it comes to removing grounds on the output side.

I am going to disconnect both output grounds the feed the outlets on the secondary side and leave the case grounded back to mains.  There is one more ground wire that runs from the main case grounding post to the secondary side of the transformer.  I don't understand what it does.  It goes inside the winding wrap but as far as I can tell it does not pass any sort of ground reference to the neutral or the transformer core.  Do I leave it connected to the case ground that goes back to mains ground?  What is it's purpose?

I hope I'm not intruding on this thread with my question and appreciate any help.

Thanks - Jerry

[Cliff Matthews]

Mains ground only needs to connect to the case and the transformers electrostatic shield (mentioned in #49). Have you measured any resistance from the primary and mains ground to the secondary outputs?

Interesting PCB just below the breaker with 3 caps only, I see silk-screen text pointing to 6 unpopulated NTC's..

[jerryk]

Thanks Cliff for the info on the electrostatic shield.  I definitely missed that in my following of this thread.

Between mains ground and line and neutral on the output sockets reads open circuit.  In it's original condition as shipped the ground is passed straight to the ground on the output side so that read .2 ohms (no resistance).  I have disconnected both of those grounds and left the case and electrostatic shield grounded to the case which carries back to mains ground.

I assume that the overload circuit protection is in the switch at 4.5amps. I'm not sure if that is an NTC (for inrush current limiting discussed earlier?) in the close up photo of the primary windings. I have no idea what the unpopulated area of the filter cap board would be for but am all ears for ideas.

Here are better views of the circuit board and the rest of the unit.

Thanks for the Help.

Jerry   

[Cliff Matthews]

You're good to go if you lifted the output ground. No need for mods since turn-on surge will not be the same as the OP's bigger 1kVA unit. The PCB seems to be on the output side and has jumpers (likely for another model) it's also not too clear which symbols would have been either NTC or MOV on the PCB.. no matter, those caps will do an OK job of shunting out HF crap.

FWIW, the 4.5a protector is not an NTC, they don't work like that. An NTC never goes open-circuit unless you've blown it apart by exceeding its rated current. As more current is passed, it's resistance goes down until it's negligible (and has almost no effect on the circuit passing current through it. 

[jerryk]

FWIW, the 4.5a protector is not an NTC, they don't work like that. An NTC never goes open-circuit unless you've blown it apart by exceeding its rated current. As more current is passed, it's resistance goes down until it's negligible (and has almost no effect on the circuit passing current through it.

The photo below shows what I thought might be an NTC.  It's stuffed between the windings and the core and I was curious if not an NTC what is it?

With regards to 4.5a overload protection, since there is no separate circuit breaker in the unit that the on/off switch pictured above was a switch/breaker combo.  I could not find any datasheet on that switch.

Thanks for great thread - Jerry

[Bratster]

FWIW, the 4.5a protector is not an NTC, they don't work like that. An NTC never goes open-circuit unless you've blown it apart by exceeding its rated current. As more current is passed, it's resistance goes down until it's negligible (and has almost no effect on the circuit passing current through it.

The photo below shows what I thought might be an NTC.  It's stuffed between the windings and the core and I was curious if not an NTC what is it?

With regards to 4.5a overload protection, since there is no separate circuit breaker in the unit that the on/off switch pictured above was a switch/breaker combo.  I could not find any datasheet on that switch.

Thanks for great thread - Jerry

Looks kind of like it might be a thermal fuse?

Sent from my Pixel 2 XL using Tapatalk