3 Isolated AC Outputs from 1 AC Source

I wouldn't normally ask this question, but since by your own admission you are very amatuer....

Are you sure you need AC power? If this were for DC power, it would be a much simpler solution. What is the power supply going to be used for?

** Groper Alert !!

** Small PCB transformers have poor "regulation" - ie the output voltage varies with load current by up to 25% from no load to full load.

Only YOU know what the application requires in the way of "equal" voltage matching.

I suspect that 1:1 transformers rated at 120 volts AC and only a few VA are hard to find.

...... Phil

Thanks for your replies!

exact 1:1 voltage matching is not important.. as long as it's failry linear, its fine.

If its easier to isolate in DC, thats perfectly acceptable. I can isolate before DC is turned into AC... So basically, take three isolated DC Supplies and then convert those to AC...

I do need AC at the end of all this, but since i'm converting 4.5V DC to 0-90V AC anyway, it's no problem to isolate in the DC Section. I always thought it was easier to isolate AC Supplies than DC, but if i'm mistaken, PLEASE do correct me. At the end of the day, I just need roughly 0-90V DC on three outputs, where current (for the most part) will not flow between them.

Thanks Again, Jeff

Why not? What's it for?

do need AC at the end of all this, but since i'm converting 4.5V DC to 0-90V AC anyway, it's no problem to isolate in the DC Section. I always thought it was easier to isolate AC Supplies than DC, but if i'm

mistaken, PLEASE do correct me. At the end of the day, I just need roughly 0-90V DC on three outputs, where current (for the most part) will not flow between them.

Mis Type... I Need roughly 0-90V AC not DC... but I can easily convert

3 DC Feeds to AC - so thats no big deal.

Jeff

** Desperate TOP POSTING Groper Alert !!!!

** Uninformative reply.

** Like hell it is.

** Err - what is the frequency and what is the wave form like ???

YOU are being far too anal with simple facts.

....... Phil

Unimportant.. Was simply stating that it doesn't really matter if the input voltage is exactly equal to output voltage if a transformer is used.

I wasn't saying that it is. It was my understanding that it's far easier to isolate AC, I said that only because another poster said that it would be "Simple if it were AC"

32khz Frequency and a nearly square wave-form. Definately not sine-wave power.

I'm sorry if i'm not providing enough facts, but I'm trying. I will provide any information necessary if you tell me what I need to tell you.

** Desperate TOP POSTING Groper Alert !!!!

** Another uninformative reply.

** I never said YOU said it.

I was warning YOU of bad advice.

** Christ almighty !!!

YOU should have mentioned ** THAT ** info straight off !!!!

** Very.

** Describe your darn application.

( This will be good ..... )

....... Phil

[snip]

Translation: I have a controllable AC supply and want to split that up into 3 isolated outputs.

Translation: I'm looking for isolation transformers or a 1:1:1:1 that can handle 90VACrms.

Translation: I don't want a big honking transformer just something that has a 20mA to 25mA rating.

Translation: What are the limitations of an isolation transformer? And is this the only way?

Translation: How do you search for isolation transformers?

One way to best understand the newbie language is to block out everything you know :) D from BC

Yes. In fact you cannot isolate DC from DC.

I would look for cheap transformers. Sometimes it is cheaper to convert 110 to 12 and then back to 110 or similar - two cheap transformers may be cheaper than one specific one.

"Homer J Simpson"

** So you missed the bit about it being a 32kHz square wave ......

........ Phil

For 32 kHz you can probably use an audio transformer. Your power requirement is 90 V at 25 mA or a little over 2 watts. However, at such high frequency, you might get away with a 400 mW transformer such as Mouser (Xicon) 42TU019 which is 10K CT to 600 CT. This is about 4:1 ratio. There is also 42TU012 which is 10K CT to 5K CT, or 1.4:1. They have high winding resistance of 100 to 600 ohms, so regulation will be poor. 42TU016 is 600 CT to 600 CT with 65 yo 55 ohm windings. It might not saturate at 32 kHz. Modem transformers are similar, so you might try one out of an old modem.

A power transformer might work at 32 kHz (but not very efficiently). A 6 volt winding at 60 Hz will handle 600 volts at 6 kHz. Probably a few dozen turns of wire on an iron toroid core would handle 90 VAC at that frequency. A ferrite core would probably be better. It would be a good learning exercise to try some of these ideas and see what happens. It's difficult to provide a perfect solution without knowing exactly what you are trying to do.

I also don't know why you say it is so easy to convert DC to AC. It is certainly not as trivial as the reverse.

Paul

Not really - using each primary of a dual-primary 230/110v transformer would do it, although winding-to-winding isolation would be less than normal.

"Mike Harrison"

** You have both asserted and contradicted your own utterly fallacious point in one sentence.

A " dual primary" transformer is NOT a " 1: 1 isolation" transformer.

Especially true in low VA ratings like 2 or 3 VA.

Think - "regulation factor " ......

If you can think at all.....

...... Phil

The additional information that you want to run at 32kHz is vitally improtant.

I don't like your chances of buying a suitable transformer with four identical windings, but you coulld probably wind something suitable on a decent-sized RM core - the bigger RM cores (like RM-8 and above) offer formers with a respectable numbr of pins - you'd need at least eight, more if you started inserting grounded screen between you windings.

You can buy RM12/I-3C90 half-cores from Farnell (order code305-6855) and a single section 12-pin former (order code178-513) plus the two retaining clips/earth tag (order code 433-135) that you need to clamp the two half cores around the former (after you have wound the coils onto it) and fix the core assembly to the circuit board.

The details are shown in the RM12/I data sheet at

which includes a lot of information about cores and formers that you can't buy in small quantities.

The core that you can buy offers 5.6uH inductance per root turn. One hundred turns per winding gives you a peak magnetising current of about

16mA, which should be okay.

If you need help in working out the winding details, e-mail me - my address is real.

If you seriously want identical voltages from all three windings, widing all four windings with a bundle of four twisted wires - the unloaded voltages will then be equal to one part in 10^7.

It won't be quite the same as the voltage you apply to the driven winding, due to the resistive drop as the magnetisng current rises and falls - to deal with that you'd need a fifth winding through which you'd drive a magnetising current to give the desired voltage across one of the four bundled windings.

Hope this helps.

--
Bill Sloman, Nijmegen

** What did you just guess the OPs's "completely isolated " meant in technical terms ?

Do even imagine HE knows ??

Is that wise ????

...... Phil

If size and weight are not issues. but price is. You could get 3 pairs of transformers. Each rated at 110V input to any voltage output (lets say 36V AC, for example). Then connect the 36V AC winding of one transformer to the 36V AC winding of another. So the voltage is 110V input, 36 V inter-stage, 110V output.

"Speedskater"

** ROTFL ....

The f****it OP now tells us he needs a 32kHz square wave to pass through !!!

Any ordinary 50/60 Hz tranny will need a DAMN BIG dose of salts tor pass THAT !!

....... Phil