Suggestions for an isolation transformer

[L_Euler]

I know you went ahead and bought one, but I've made a couple really simple, but effective isolation transformers by just hooking a pair of identical standard transformers back to back. For instance, 120VAC to primary of transformer 1 with secondarys of both transformers connected to each other gets you 120vac isolated at the primary of transformer 2 as long as you don't use the ground or neutral wire.

That's what a lot of us do in a pinch, but No-Load to Full-Load regulation is about 10% for one transformer. AFAIK, that's 120v noload and 108v full load, so double that for back-to-back and full load output could drop below 100v.. not to mention Isolation transformers have a lower capacitive coupling for better/safer DUT isolation.

Agreed.  However, if you know enough to understand you need an isolation transformer, you probably know enough to cope with the limitations of this method.

[SoundTech-LG]

Not sure I would suggest this, but the ebay used price, and free shipping were hard to resist. Suggestions on the primary 240V to 208V drop are welcome. My solution is to use another transformer with a 30 amp 24 volt winding in series with the primary. It will still be a bit high. I had it apart. Forgot to snap photos of the huge toroid it contains. Hopefully I'll do that soon.

https://www.amazon.com/Tripp-Lite-SU5000XFMRT2U-Rackmount-Transformer/dp/B0007DGFEU

[Spork Schivago]

5a is for 230v and 10a for 120v - Hey, do we get any pics of the belly of this beast?

Yes sir, I had them backwards, I've been real tired.   I will open it up very soon and take pictures.   Since it was dropped in shipping, I ran some basic tests and am a little confused.   I have it set to 115VAC input, 115VAC output.   I hook up a C13 / C14 cord to the output and put my multimeter, in volts mode, across the two terminals of the plug.   I read roughly 125VAC.    I take one of the probes from the DMM and hook it to the neutral hole on the plug and put the red probe on the big metal bolt on the top and read roughly 7VAC.

I take the red probe and put it on the side nuts of the isolation transformer and read roughly 47VAC.   Is this normal and safe?    Or is there a chance something is making that chassis hot?

Thanks!

[Spork Schivago]

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

[Spork Schivago]

Here's some more.more.

[Spork Schivago]

The final two.

[Wolfgang]

... looks very similar to the transformer I used. I Just added the inrush current limiter, fuses, an output voltage switch (115/1230), and two analog AC panel meters, plus a few hefty sockets - job done. Never let me down so far.

[Cliff Matthews]

Hmm, those SG-44 inrush limiters are rated at 8 amps.. which is 960 watts. That's pushing it a bit, I'd think. Still, how often will anyone push these to the limit? If it ever stops working and the fuse is not to blame, one of these will be cracked open and the other one will be your spare if you never use the 240v input.
The 125v you measured is unloaded but still OK (1v under USA limit of +/- 5%). It might be educational, for those following along, to know what you measure for various loads up to a hair-dryer on low heat setting (medium and high may blow it's fuse). I think the 7v is just a floating average from minimal capacitive coupling. To be sure it's prepared for isolation use, either L or N on any of the output terminals should measure high resistance to any contacts on the primary side of the transformer. It's time to unplug and do some resistance checks.. Nice purchase!

*edit Varistor SG-44 PDF attached

[bson]

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

The input and output mains earths both look connected to the chassis.  I wouldn't trust this transformer to isolate any 3-prong device without carefully determining what it does with its earth lead.  If the goal is noise isolation it's perfect the way it is - you want a shared earth.  But if it's to float you should consider removing the output side earth wiring since it prevents effectively floating the outputs.  Or maybe leave it on one terminal and clearly label it to give you a choice.

[Cliff Matthews]

Yup, that was mentioned with an optional GND-lift switch in #13 (or he has the option to modify his DUT bench power cords with ground clipped). Regarding picture uploads to the forum, Dave has mentioned a free utility called "IrfanView" before and I can confirm this is not only 10x faster than the Windows viewer, but it also allows a lot of sharpness detailing and compression as well. It's got to be one of the best freeware offerings from Europe I've ever used, hands down.

[Spork Schivago]

Would the SG-26 or SG-32 be better choices?

Code: [Select]

SG-26
-----
R@25°C: 5.0Ω
R Tolerance (±%): 15
Imax: 12 Amps
RImax: 0.060Ω
Surge Rating: 100 Joules

SG-32
-----
R@25°C: 4.0Ω
R Tolerance (±%): 20
Imax: 14 Amps
RImax: 0.050Ω
Surge Rating: 100 Joules

SG-44
-----
R@25°C: 5.0Ω
R Tolerance (±%): 20
Imax: 8 Amps
RImax: 0.050Ω
Surge Rating: 40 Joules


[Spork Schivago]

I got memory problems, and totally forgot about post #13, but yeah, I think I am going to remove the grounds that go to the outputs.

For the resistance check, just measure from L or N on the output to all the wires that don't go to output or to the grounds connected on the chassis?   Thanks!

**EDIT:

My Craftsman 82357 is an older meter, but it's been fairly good to me and when I used to have the manual, had good specs (although, I haven't had the manual in years and cannot locate one).   Regardless, it shows OL (overload), or infinite resistance, when measuring from the L and N on the output side to any of the other wires coming out of the transformer, including the ground wires.

So now, I just need to remove the grounds that go from the chassis to the outputs, maybe upgrade the thermistors, replace the one fuse with a 5A fuse, and then build the soft-start circuitry thinga-ma-jig.   Can someone help me find parts for that circuit and show me where I wire everything up?   The relays and NTCs?   Thank you!

[Spork Schivago]

If I used that 12-volt transformer, what would be the purpose of that?   I am not sure I understand enough to build the soft-start stuff myself yet.   I do not want to attempt to do something I don't fully understand, for fear of getting hurt, or hurting someone.

[Cliff Matthews]

I had a look for a manual for a Craftsman 82357, and aside from usual Russian bait-link-and-switch for virus, found nothing. I'll just attach a photo so if anyone here has the PDF, they can help out.

As for mods, hold off until you take 5 minutes to check the output voltage with both a 60 or 100 watt lamp and a hair dryer (on low heat) over a period of 1 minute each. This way you can observe any voltage change as the limiter heats up.

Why? It's unwise to change perceived faults in a design before you truly observe the original. I'm thinking that with a relay engaged, this transformer may output >126v to light loads (in other words, it may depend on the limiter a bit for light loads).

*edit - while your at it, you should measure the mains too (input line voltage), it may be the reason you're getting 125v out. 

[Cliff Matthews]

If I used that 12-volt transformer, what would be the purpose of that?   I am not sure I understand enough to build the soft-start stuff myself yet.   I do not want to attempt to do something I don't fully understand, for fear of getting hurt, or hurting someone.

You are correct, if you're not up to that, don't do it. The 12v transformer's only function was to power the 555 and hence the relay after a 2 second delay. But after reading your 'stock' output voltages, it appears the limiter(s) do have a useful function at low currents, so I think you should hold off any mods (with the exception of removing the previously mentioned output grounding).

As for changing the limiters, the SG-26 may be OK but the SG-32 is likely not OK (just stay with the stock, or ask the MFG about why they used an under-rated part.. after you get your mounting brackets..). Once again, you got a good deal.

[Spork Schivago]

Sears used to have the manual for that DMM, but they no longer have it.   I even tried the WayBack Machine, but no luck, unfortunetly.

I will hold off on the mods, minus the ground lifting.   I was thinking of purchasing something like what is included in the attachments here, one of those ground lugs, if I could find the correct size.   I would use epoxy or something to attach it to the case, and isolate those output grounds.   Then, if I ever wanted the output grounds hooked up, I could just run a jumper wire from the new ground lug things to the old posts.

I cannot figure out another safer way to lift those ground wires.   I could wrap them in electrical tape, but I would worry the tape might eventually fall off...

I have a special light socket that has a hot and a neutral wire.   I can use that to test the 60 watt and 100 watt bulbs, assuming I can find 60 watt and 100 watt bulbs.   How do I test the voltage with the hair dryer hooked up though?   Where could I put my probes without damaging the hair dryer?   I don't think I can put them inside the transformer.   My wife would kill me if I tore apart her hair dryer...


**EDIT:

The lifting of the grounds was easier than I originally thought.   I removed the nut for each one of them, but where they hooked up to the output receptacles, they were just slid on, they weren't soldered like the hots and neutrals.   Last night, I had just taken a quick glance and seen the hots and neutrals were soldered and must have assumed the grounds were as well.   So now I removed all the grounds going to the output recepatcles.   I left the ground hooked up that goes to the input power receptacle, and I left the primary side ground hooked up.   Now, no grounds go to the output receptacles and they are floating!

Thanks!

[Spork Schivago]

If I used that 12-volt transformer, what would be the purpose of that?   I am not sure I understand enough to build the soft-start stuff myself yet.   I do not want to attempt to do something I don't fully understand, for fear of getting hurt, or hurting someone.

...
As for changing the limiters, the SG-26 may be OK but the SG-32 is likely not OK (just stay with the stock, or ask the MFG about why they used an under-rated part.. after you get your mounting brackets..). Once again, you got a good deal.

Now, why would the SG-32 not be okay?   I will stay with stock for now and ask the manufacturer, but I want to learn, and to me, the SG-32, I thought would have been better, because it could handle up to 14 amps.   So I'm missing something with these NTCs still...

[Gregg]

I cannot figure out another safer way to lift those ground wires.   I could wrap them in electrical tape, but I would worry the tape might eventually fall off...

For the ground wires, put the ones for your isolated outlets to a binder post.  Leave the input ground bonded to the case and run a wire to another binder post adjacent to the isolated ground,  Then you can easily jump between the binder posts for non-isolated ground or make a plug in jumper with banana plugs if you switch the grounding scheme often.  Same idea for meter connection, just install a pair of the flush type sleeve capable meter jacks in the case and bob's your uncle. 

[Spork Schivago]

I cannot figure out another safer way to lift those ground wires.   I could wrap them in electrical tape, but I would worry the tape might eventually fall off...

For the ground wires, put the ones for your isolated outlets to a binder post.  Leave the input ground bonded to the case and run a wire to another binder post adjacent to the isolated ground,  Then you can easily jump between the binder posts for non-isolated ground or make a plug in jumper with banana plugs if you switch the grounding scheme often.  Same idea for meter connection, just install a pair of the flush type sleeve capable meter jacks in the case and bob's your uncle. 

What do you mean by the meter connection?   Are you talking about modifying this isolation transformer so it's a bit like my Rigol Programmable Power Supply, where I can take probes, plug them into the banana jacks, and use them to power some circuit?

For what it's worth, I found a 5-amp 250V fuse down in my basement.   It's physically larger than the LittleFuse that was in there, but it fits.   I don't have any of those LittleFuse's to my knowledge, but it would be nice to get a 5-amp 250V LittleFuse one of these days.   The fuse door / input selector switch door closes a little harder because of the larger fuse.   The fuse holder appears to be designed for either type (the LittleFuse or the larger ones).

[Cliff Matthews]

NTC resistance curves are in the data sheet. The original is a 5 ohm (@25c) 0.6 inch disc that will drop to <1 ohm >80c (see RT/R25 curve A) while passing ~1amp.. so with transfo losses, that's roughly a 100 watt lamp on the output. The SG-32 NTC is 4 ohms so when cold, the secondary side may produce >126v (over US limit). The SG-32 is also larger 0.9 inches so it can dissipate more power before it reaches its "sub-ohmic" operating point while possibly melting that nylon barrier strip. I think only 1 NTC is used at 120v-in while both are used at 240v-in - that's why I said you'll have a spare. Hope that helps!

[Spork Schivago]

Code: [Select]

Mains Input Voltage: 122.5VAC
60 Watt Incandescent Bulb for Load: 124.1VAC - 124.5VAC
No Load Output: 126.5VAC

[Multi-Watt Bulb]
30 Watt: 124.8VAC - 125.2VAC
70 Watt: 123.7VAC - 124.2VAC
100 Watt: 122.7VAC - 123.1VAC

I had trouble finding the incandescent bulbs and need to hook up a three-way touch switch between the three-way bulb and the isolation transformer first before I can get the power output of the 100 watt bulb.

**EDIT:  I added the measurements for the 30 / 70 / 100 watt bulb.   Here are the measurements, in VAC, for those powers.

[Gregg]

Spork,
Just put a pair of the banana jacks like these https://www.ebay.com/itm/10-Pcs-4mm-Safety-Protection-Plug-Binding-Post-Banana-Jack-Black-Red-W5H1/173105429056 somewhere in the transformer case connected to output so that you can easily connect your meter.  You could put fuses and/or a double pole switch if you want.

[Spork Schivago]

I don't see how I could connect the binding posts though.   Anything connected to the case will be connected to ground.   Unless you guys mean drill holes through what I call the front, where the output receptacles are, and hook them to the transformer.   That could come in handy.

I have a programmable DC power supply, but no AC, and that could be really handy, even though it isn't programmable.   Having a ~120VAC / ~240VAC transformer case that accepts that common bannan plug would come in handy...and I think I know what you mean now.   I can hook the DMM up real quick like while I have a device plugged in to measure the load easy like.

[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?

**EDIT:

Hairdryer on low -- Hot showed 119.7VAC - 120.9VAC.   High, as preditected, broke the transformer.   I will see if I blew a fuse or if time off will fix it.   Those NTCs could act like a soft fuse, a resettable one, right?   Or no?

[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.