Dropping from 15V AC to 9V AC?

[LektroiD]

How can I reduce an AC input down to 9VAC?

I'm using a toroidal transformer which outputs 2x15VAC, This will run into an AC-DC PSU giving 12VDC at the other end to drive the main circuit. I would like to have an additional tap from the 15VAC and reduce to 9VAC for a VU meter driver board also.

What's the best way to drop from 15VAC to 9VAC?

[danadak]

Maybe I am missing something but you mention a tap, that's exactly
one of the better ways, another tap in transformer for the lower voltage.


Regards, Dana.

[LektroiD]

What I meant by tap was to use the existing 15VAC to attach to another circuit to produce 9VAC. The toroidal transformer does not have existing 9VAC taps, or I would be using that, and would not have needed to pose the question.

[IanB]

What I meant by tap was to use the existing 15VAC to attach to another circuit to produce 9VAC. The toroidal transformer does not have existing 9VAC taps, or I would be using that, and would not have needed to pose the question.

To convert one AC voltage to another you use transformers. So the "other circuit" is simply another transformer. Either a small mains to 9 V transformer (easiest) or a 15 V to 9 V transformer (unlikely to be a standard part). Alternatively you might wrap an additional winding around the toroidal transformer to produce 9 V.

[Brumby]

Alternatively you might add wrap a second winding around the toroidal transformer to produce 9 V.

This was my first thought.

As long as the total VA rating of the windings that are USED does not exceed the toroid's specification, you won't have any problems on that front.  For 9V and modest current needs, I can't see this being a difficult task.

How many turns, you ask?  Just wind half a dozen turns and measure the voltage you get.  Then apply some basic math to give you the answer.

Which direction should they be wound?  If it's going to be turned into DC, it doesn't matter.

[David Hess]

You could add a second transformer or add a winding to the existing transformer like IanB suggests.

If the load draws a relatively constant current, then a series capacitor can be used to drop the AC output voltage without without loss.

[BrianHG]

I curious as to why you cant use the 15v ac to drive the VU meter driver board?

[VEGETA]

well, either get another transformer or tap... or if your current requirement is too low, you can use a resistor divider (power resistors preferably) to drop it which works fine.

It's only problem is power rating since common resistors are 0.25 or 0.5 watts in general. However, if you have the 5 watts power resistors, then I guess it is fine. Still, another tap or transformer is kinda better for you.

If your target circuit runs off DC voltage then that is easier for you, just provide it with DC voltage enough to run it. Most likely it has an internal bridge rectifier thus 1.4v drop voltage to take care of.

[LektroiD]

I curious as to why you cant use the 15v ac to drive the VU meter driver board?

The VU driver board requires 9V AC. I'm assuming it has some 9V maximum rated components on there. I'd be happy to feed it 15V if I knew I could get away without blowing any components on the board, but then wonder why it asks specifically for 9V AC and not 9-15V AC.

[LektroiD]

well, either get another transformer or tap... or if your current requirement is too low, you can use a resistor divider (power resistors preferably) to drop it which works fine.

It's only problem is power rating since common resistors are 0.25 or 0.5 watts in general. However, if you have the 5 watts power resistors, then I guess it is fine. Still, another tap or transformer is kinda better for you.

I have no idea how to make a tap on a toroidal transformer (This is why I posted this thread in the "Beginners" forum). I'm assuming you peel off the plastic outer layer, scratch off the enamel of the winding (somewhere), and solder another wire on top of it? But how to find the point where it is a 9V? It could get rather messy and leave loads of points exposed where I've guess-scratched off the enamel... Testing each point with live mains supply into it doesn't fill me with confidence either, especially if it has several exposed points where the enamel has been scratched off.

Then there's the problem of re-wraping the plastic outer layer after I have located 9V to tap from. I guess I could use electrical tape afterwards, but I would assume there's a reason the toroidal transformer manufacturers are using that specific type of plastic outer shell. Maybe electrical tape will become slimy after a while?

If your target circuit runs off DC voltage then that is easier for you, just provide it with DC voltage enough to run it. Most likely it has an internal bridge rectifier thus 1.4v drop voltage to take care of.

The target (main) PCB accepts 15V AC, it has a 15V AC-12V DC power supply onboard. I could omit that, but looks like it is a high power supply, which I am unlikely to find a replacement PSU easily. I would prefer to use the PSU that was designed for it which is integrated on the PCB. I can't ask the designer any questions about power ratings, etc, as he died a few years ago. The project page (which was left as work in progress) is here: http://jhaible.com/legacy/vocoder/living_vocoder.html

[VEGETA]

I'm assuming you peel off the plastic outer layer, scratch off the enamel of the winding (somewhere), and solder another wire on top of it?

I haven't try it but I guess your assumption is correct since it is the way official taps are made xD.

But how to find the point where it is a 9V?

You could test them to find out but it is kinda dangerous... well, 15v is not really that dangerous.

Then there's the problem of re-wraping the plastic outer layer after I have located 9V to tap from.

Here is why you use the crappy silicon hot gun 

Just fill it all with it and it will kinda work. or just but the whole transformer into a sealed plastic enclosure and only get the wires out.

The target (main) PCB accepts 15V AC, it has a 15V AC-12V DC power supply onboard.

That is all you need, the 12v. Measure current requirement or at estimate it, and if it is below 1.5A then you can use LM317! basic LM317 will work for you and you can put a bridge rectifier before it so it will be 15vAC from transformer that goes in the full wave rectifier (4 diodes) then to LM317.

You can use 2 LM317 to give around 3A or maybe use one of these Ebay power supply boards (5$) which is a buck converter. Some of them can give 10A!

As for the power supply of the board, it is in here: http://jhaible.com/legacy/vocoder/schematics/power_supply.pdf

it uses ua723 voltage regulator and it is using an external transistor (TIP3055) which is capable of 15A. So you need to kinda estimate the current draw to know if LM317 is gonna be enough or not.

Luckily for us, the U2701 IC is a 4-Amps bridge rectifier so it is the maximum. I guess then you can use some E-bay board.

Also, you can actually feed in a DC voltage to the bridge rectifier so a 15v DC will be good enough.

[Ian.M]

Trying to tap an existing winding would be crazy. You'd have a high probability of either wrecking the transformer or making it unsafe, and even if you manage to get 9V AC, it certainly wont play well (high probability of smoke) with another bridge rectifier when the two loads either share a common ground, or are otherwise interconnected without isolation.   The only tap that is usable for multiple outputs when the winding ends feed a bridge, is an exact center tap as that directly gives you a rail that splits the bridge output in half (commonly used for balanced +/- supplies with the center tap as 0V)  However even if the winding center could be located accurately enough and safely tapped, that's no help in this case ay you need 9V AC not 7.5V AC.

You've only got three choices:
1. add a 9V AC winding as an overwind.
2. Add a separate transformer for 9V AC - either direct from mains or as a 5:3 ratio off the 15V AC secondary
3. Redesign the VU driver board PSU section to either run on a higher voltage, or to tap into rails available in the main PSU.

However if the VU meter requires a floating supply, option 3 wont work.  You'd need to study its schematic carefully to determine if its possible.

[LektroiD]

Trying to tap an existing winding would be crazy. You'd have a high probability of either wrecking the transformer or making it unsafe, and even if you manage to get 9V AC, it certainly wont play well (high probability of smoke) with another bridge rectifier when the two loads either share a common ground, or are otherwise interconnected without isolation.   The only tap that is usable for multiple outputs when the winding ends feed a bridge, is an exact center tap as that directly gives you a rail that splits the bridge output in half (commonly used for balanced +/- supplies with the center tap as 0V)  However even if the winding center could be located accurately enough and safely tapped, that's no help in this case ay you need 9V AC not 7.5V AC.

I figured this, which is why I was surprised to see it suggested so many times for a device which has a 240V AC primary at your fingertips. I will use the onboard PSU which was designed specifically for this project, it would be counter intuitive to redesign another supply which going by the notes, I'm guessing is 2A.

You've only got three choices:
1. add a 9V AC winding as an overwind.
2. Add a separate transformer for 9V AC - either direct from mains or as a 5:3 ratio off the 15V AC secondary
3. Redesign the VU driver board PSU section to either run on a higher voltage, or to tap into rails available in the main PSU.

However if the VU meter requires a floating supply, option 3 wont work.  You'd need to study its schematic carefully to determine if its possible.

Thanks for your suggestion, This was also IanB's suggestion, above, and I think after reading all the other replies above, option 2 is the cleanest way; I can use a small, low power 9V toroidal next to the recommended 15V 80W toroidal (although I'm not sure how '80W' converts into 'VA', which is what these transformers are rated in?).

I've only ever dealt with DC. I thought AC would be just as simple as DC, where you can just throw in a regulator or buck converter, or even a simple voltage divider circuit. It's all a learning curve...

[julian1]

I curious as to why you cant use the 15v ac to drive the VU meter driver board?

The VU driver board requires 9V AC. I'm assuming it has some 9V maximum rated components on there. I'd be happy to feed it 15V if I knew I could get away without blowing any components on the board, but then wonder why it asks specifically for 9V AC and not 9-15V AC.

What does the VU driver board do and how much current does it consume? It might be fine at the higher voltage, or just with some heatsinks on the linear regulator(s) if in fact it has DC regulators.

[Ian.M]

Yes. we need to see the VU board schematic or at least its PSU section to detemine whether it can run from a higher voltage or if it really needs its own isolated 9VAC supply.

If it does need its own isolated AC supply, assuming the toroidial transformer has an open center (i.e not potted). try the overwind experiment Brumby suggested, only to make the maths easy, wind TEN turns of hookup wire through the center of the toroid (not six as suggested).  Measure the voltage from that  and you can calculate how many turns are actually required for 9V AC. 

Anything under about 100 turns is no particular problem to wind. If its going to need more turns (unlikely) I'd reconsider.

Mark the two ends of the temporary winding were they cross the same edge of the toroid, unwind it and measure the length between the marks.  Calculate the length required for your 9V winding and spool off enough (with about 10% extra + 2x leadout wire length) thin (but not too fine) magnet wire onto a spool that will fit through the core e.g an empty cotton reel. 

Thread thin sleeving (e.g. stripped from 1mm solid wire) long enough for the lead length you want over the end of the magnet wire with 1" of wire emerging which should be folded back and taped, and cow hitch a 6" length of heavy unwaxed linen thread round the other end of the sleeving.  Cow hitch another linen thread round the magnet wire just past the end of the sleeving.  Temporally tape the sleeving to the outer side of the toroid, with about a 1/4" gap between the tape and the nearest linen thread.  The cow hitch should be centered on the face. Lay both ends of both linen threads along the toroid face in the direction of the proposed winding and secure the far end with tape.

Wind twenty turns with the magnet wire (or half the required turn count, whichever comes first) tight and close packed but not overlapping, then untape the linen threads, pull them as tight as possible, fold them back towards the start of the winding and tape them out of the way.  Continue winding until there's 20 turns remaining (or if less, already don't continue).  lay two 8" loops of linen thread along the outer side of the toroid, looped end away from the winding, one on the center line and the other a bit towards one edge, in the direction the winding will be coming from.  Tape the tails down on top of the winding and tape the loops down as far round the toroid as they reach stretched tight.

Carry on winding, over the top of the new linen threads till the desired turn count is reached. Temporarily tape the winding end, and power the transformer to check the new secondary voltage.  5% high is good as it will drop a bit when loaded.  Adjust the turn count if its too far off.

Cut the wire end a few inches longer than the first one. Thread a matching piece of sleeving over it, untape the loop ends of the linen thread, form cow hitches with them round a drinking straw, and thread the wire end with sleeving through the drinking straw and pull out the straw so the cow hitches are round the wire and sleeving end just like the starting end. Pull the free ends of the linen threads to tighten the cow hitches and lightly tension the wire inside the sleeving and fold back the end.

Make sure everything is snugly in place and all the linen threads have been tightened then its just a matter of locking the threads in place to permanently anchor the winding and sleeving by applying enough liquid superglue to saturate the linen threads, trim off the tails and let it all set. Remove as much as possible of any remaining temporary tape.  If you've got any electrical grade lacquer, conformal coating or varnish, now's the time to give the whole new winding a coat, and let it fully dry.

Reinforce where each sleeving end is attached with a short strip of heat resistant insulating tape (preferably cloth, mylar or kapton and definitely not PVC) running round the outer diameter of the toroid, then wind insulating tape over the winding, starting with a full turn before the winding and finishing with a full turn after it.  The sleeving with the leadout wires should emerge between turns of the tape. 

Result: a professional quality overwind 9V AC secondary.

[mariush]

if the  lower ac voltage doesn't have to be isolated from the secondary ac winding,
maybe you could put pairs of diodes in series with the output to lower the ac voltage

ex

15v AC :---[===|]---: ~14.3v AC
         \-[|===]-/

[BrianHG]

if the  lower ac voltage doesn't have to be isolated from the secondary ac winding,
maybe you could put pairs of diodes in series with the output to lower the ac voltage

ex

15v AC -[===|]---: ~14.3v AC
         \-[|===]-/

2 back to back 5.1v, 5w zener diodes would work.

[madires]

Yep, the circuit diagram for the VU meter would help. But usually the VU meter shares ground with the main circuit to have a common reference for the signal. So it isn't isolated. And it also needs DC power. I'd assume that you can simply power the VU meter by the +12V of the main circuit. Maybe you have to add a voltage regulator if there's none in the VU meter circuit.

[Ian.M]

There's the possibility the VU meter uses positive and negative rails, and if so either it needs isolated AC in to feed its two half-wave rectifiers, or it would need to be modded by removing the diodes and feeding its positive and negative rails separately.

The main PSU uses two identical secondaryy => bridge rectifier => regulator circuits to get its positive and negative regulated rails.  The two secondaries have a standing DC offset with respect to the 0V rail so Mariush and Brian's idea of a diode dropper wont work unless the VU board's signal input is transformer coupled ad its ground is isolated.

[madires]

Let's not overcomplicate things by dwelling into theoretical possible designs. We want to see the circuit diagram

[LektroiD]

I don't have the schematic as it wasn't provided, but I just looked at the board, and noticed a 7812 v-reg on there. So it looks like I'll be able to feed 12DC straight in from the PSU on the main board...

 

I feel a little stupid for not checking the PCB thoroughly before posting, but at the same time I've learned a lot about AC from the above postings. I also noticed a dented electrolytic which I'll be replacing before powering up.

[Ian.M]

Maybe keep the 7812 regulator and feed the 7812 regulator from the  unregulated input to the main PSU +12 regulator to minimise any possibility of unwanted interaction between the VU board and the main board.  If it needs more than 100mA, however that may cause the 7812 to run too hot.

If the VU board power in lineup goes (ignoring fuses): bridge rectifier => reservoir cap => 7812, either that or your idea should be OK, as long as *NOTHING* taps off the AC before the bridge and signal ground has continuity with the negative side of the bridge. 

If there is a fuse before the bridge rectifier, it would be best to remove the bridge and jumper where it was so the fuse is in the +feed and the other AC in terminal is grounded, then use the AC in terminals for DC in (mark their new functions). 

If you are going to use +12V regulated power, fuse it at 1A or any mistake could literally burn your VU board.

If you post good sharp closeup photos of both sides of the VU board, with where you plan to mod it marked, we can check your proposed mod is OK.

[LektroiD]

Like this:

In fact, I won't need to cut the traces if I'm removing the bridge rectifier

[Ian.M]

Yes, I'd simply jumper straight across the bridge on the track side.  You can leave it in situ as the diodes in it that you haven't shorted out are reverse biassed. No track mods required and easy to reverse near-invisibly if you need to make a warranty claim.

[PA0PBZ]

You can also leave the whole pcb as it is and just feed the unregulated output from the other PSU into CN6.