Feedback problem for SG2525

I have built control logic for an symmetrical flyback converter based on IC SG2525. This IC has 2 outputs to control external BUZ11, one for each primary winding. From output of transformer positive side (+40V) i have isolated feedback (voltage to freq. via transformer to freq. to voltage) to the control electronics to get good regulation but I have no feedback from negative side (-40V) and no control of negative voltage. SG2525 has only one input to control duty cycle of 2 outputs. Is there any simple way to also keep track of the negative? //Peter

The quick answer is no.

The -ve supply relies on the +ve supply for its control.

If you need really tight control of the -ve supply, I think you will have to build a separate supply.

Once you have built it, try it and see how bad (or how good) it is. Is it good enough for your requirements?

JimB

Yes, the solution I use now is 2 separate positive flybacktransformers, with their own control logic, in series to get 2x40V output so I do have good regulation of negative also. Very space demanding and not so compact. It would be nice to make use of both Out A and Out B on one SG2525 to control both primary winding with only one SG2525 but it does not look very easy to do. Thanks alot for reply. //Peter

What actually is the supply for?, it sounds like it's for an in-car amplifier?.

Yes, primary I designed this flyback converter for 2pcs Velleman K4005 2x100Wrms amplifier to use in car. Also the design of the control electronics is very universal with isolated feedback for all types of flyback converter constructions. I build the control logic on one eurocard size board and the transformer (EDT49) and switchtransistors on a separate eurocard, it is a very heavy duty transformerboard with 5 pcs. BUZ11A. I thught about that the only way to separately control one of the outputs on SG2525 should be to have some extra logic after the output signal to modify it based on feedback from negative side.

GT20D101 said:

Yes, primary I designed this flyback converter for 2pcs Velleman K4005 2x100Wrms amplifier to use in car. Also the design of the control electronics is very universal with isolated feedback for all types of flyback converter constructions. I build the control logic on one eurocard size board and the transformer (EDT49) and switchtransistors on a separate eurocard, it is a very heavy duty transformerboard with 5 pcs. BUZ11A. I thught about that the only way to separately control one of the outputs on SG2525 should be to have some extra logic after the output signal to modify it based on feedback from negative side.

Click to expand...

That's what I thought you were doing?, there's no need to monitor BOTH supplies, they come off a single centre tapped winding, so will be pretty well identical - and you can't alter them individually anyway!.

Look at any switch-mode supply, they only usually monitor ONE output voltage, including the one in your PC!.

Thats not true, maybe for use with amplifier, but not if there is different load on negative and positive side. I wanted to have a universal control working not only for equal load on both side.

GT20D101 said:

Thats not true, maybe for use with amplifier, but not if there is different load on negative and positive side. I wanted to have a universal control working not only for equal load on both side.

Click to expand...

Then build two completely seperate supplies, one for each rail - you can't have two seperately regulated supplies off a single transformer (unless you regulate on the secondary side).

But I'll repeat again! - switchmode supplies almost always (I've never seen one any different!) regulate off ONE secondary output, it's all that's required, and there's no way the supply can adjust one without adjusting all the others - as it's a single transformer all the other windings will be in proportion to that one voltage. Obviously they won't be quite as accurate as the regulated winding, but they will be accurate enough for most purposes.

Generally you regulate from the most critical winding, for a PC this would probably be the 5V winding - although more modern PC boards use 3.3V a lot as well! - I can't say I've studied any really modern PC PSU's.

Your car amplifier requirements is VERY easy to do, it's not at all critical, so doesn't require good regulation anyway!.

GT20D101 said:

Thats not true, maybe for use with amplifier, but not if there is different load on negative and positive side. I wanted to have a universal control working not only for equal load on both side.

Click to expand...

Not that I'm trying to push the reset button on your whole design, but check out linear technologies controllers. They offer some that allow regulation of both +ve and -ve rails independently. I agree that most often only one side is regulated (sometimes off of an auxiliary winding) but dual regulation controllers exist.. Off the top of my head I can't recall a part number.. maybe worth a look for you.

Optikon said:

Not that I'm trying to push the reset button on your whole design, but check out linear technologies controllers. They offer some that allow regulation of both +ve and -ve rails independently. I agree that most often only one side is regulated (sometimes off of an auxiliary winding) but dual regulation controllers exist.. Off the top of my head I can't recall a part number.. maybe worth a look for you.

Click to expand...

So how can you independently regulate two supplies on the same transformer?.

Nigel Goodwin said:

Optikon said:

Not that I'm trying to push the reset button on your whole design, but check out linear technologies controllers. They offer some that allow regulation of both +ve and -ve rails independently. I agree that most often only one side is regulated (sometimes off of an auxiliary winding) but dual regulation controllers exist.. Off the top of my head I can't recall a part number.. maybe worth a look for you.

Click to expand...

So how can you independently regulate two supplies on the same transformer?.

Click to expand...

I dont believe the regulation is truly independent for the same xformer.
But have a look at LT1683. I've used this part before in a single transformer converter with both +ve and -ve feedback.

The FB pin is the usual implementation. The NFB pin goes to an intermediate stage (negative feedback amp) which also drives the error amp. My take is that it steals regulation from +ve to give regulation to -ve.
This allows _some_ regulation on both but neither has full regulation capability. But I dont think the OP discussed how much regulation he needs. So this allows both to have some. ?? Anyone care to venture a guess as to how this part works with the NFB?? Page 9 of the datasheet shows a block diagram, they didnt discuss it much otherwise.

Optikon said:

Page 9 of the datasheet shows a block diagram, they didnt discuss it much otherwise.

Click to expand...

The text on the datasheet is quite confusing, but the block diagram makes it clear, the 'negative' input is simply inverted (to make it positive) and applied to the 'positive' feedback input. So if the positive rail goes too positive, OR the negative rail goes too negative, BOTH will be turned down. So it's not offering regulation of both negative and positive rails, just from either of them (whichever is the greater).

But, I'll repeat again, this thread doesn't need particularly good regulation, and certainly doesn't need both controlling - check every in-car amp ever made! :lol:

2 pcs. SG2525

I did not wanted to control 2 primary windings on the same core, I just wanted to have both outputs of SG2525 independently regulated. As it is now there is only one amplifier connected to the two outputs of SG2525 and I think the only way to also control the other output is to modify one of the outputsignals after leaving the chip based on a feedbacksignal. Maybe the easiest thing to do is to have 2 pcs. of SG2525 on the same board (leaving 2 outputs unused). I think 2 primary (separated) and 2 secondary (separated) windings on the same core leaves the feedback electronics with a heavy work to do as it have to compensate for flux from the other primary winding, more current cross winding 1 causes more flux in the core and must affect winding 2 raising output voltage on secondary 2 and the feedback electronics have to compensate for that. But I do not now if it really is a problem (oscillation maybe?).