Greetings,
I'm planning to make a couple of amps with TAS5630. "On the side", for one reason or another, i've been collecting and scavenging an assortment of old and/or broken computer PSU's, and as such, i've put together quite a stock of toroidal cores. Judging by their color coding, mostly mix -26 and -52 (according to Micrometals color codes).
I'll be running the amps in 2xBTL per each chip. Now, considering that some board-space-saving wouldn't hurt, my question is, is there some way to wind a coupled BTL output inductor, without blowing anything up?
I've got the "mini ring-core calculator" and Micrometals' inductor design/analysis program, to help.
I'm planning to make a couple of amps with TAS5630. "On the side", for one reason or another, i've been collecting and scavenging an assortment of old and/or broken computer PSU's, and as such, i've put together quite a stock of toroidal cores. Judging by their color coding, mostly mix -26 and -52 (according to Micrometals color codes).
I'll be running the amps in 2xBTL per each chip. Now, considering that some board-space-saving wouldn't hurt, my question is, is there some way to wind a coupled BTL output inductor, without blowing anything up?
I've got the "mini ring-core calculator" and Micrometals' inductor design/analysis program, to help.
Also, it is needed to use this chip in AD modulation mode then and feed its inputs with strongly differential signal, as I know...
Very good drawings and explanations! 🙂
However I had no success to use coupled inductor in BTL configuration with BD modulation mode earlier... It was working only with AD modulation for me. And the inductor was wound slightly different, than in your drawings. You have shown the 'common mode' inductor (input currents are always in phase), whereas for AD modulation I have used the 'differential mode' inductor (input currents are out of phase).
Please correct, if I am wrong. 🙂
Oooooh, yeah 
Many thanks for the enlightenment 😱
Am i right with my hunch about the core dimensioning, though?
Many thanks for the enlightenment 😱Am i right with my hunch about the core dimensioning, though?
Thanx! In BD modulation the magnetizing current have an important role, and to maintain this magnetizing current you need to use a relatively low permeability core for this inductor. Maybe this wasn't low enough in your case. I used it only with BD modulator, and it worked well with 2 stacked AL=60 iron powder core. I think an EE (or RM) core (with huge airgap) would be better, but I haven't try it yet, however I've wound it already.
Input currents are out of phase in my case also, but there is a common mode component too at switching freq. This common mode current would be huge with separated inductors, because of the low switching freq, this is one of the reasons I used this kind of filter. With AD modulation higher fsw is needed, this because the ripple current is not too high.
Khron!
Probably it's BD in bridge mode.
You don't have to worry about flux, but it's not that easy. In this inductor the most important thing is the leakage inductance, and this is very hard to calculate, it depends on the winding geometry too. It's better to simply experiment, measure inductance with one end shorted ! Hint: you will need much more turns then usual. In return you will get a much lower core loss.
Pafi, the TAS5630 can work in either AD or BD modulation, in the 2xBTL configuration.
The top sketch in your posted pic resembles how the windings look on common-mode inductors, on SMPS inputs, so i'm guessing the bottom sketch is the differential-mode winding? So that's how a coupled inductor for a bridged amp should be wound? Or... Which type of modulation should be configured, for which type of winding? 😕
And by "one end shorted", i guess you mean with one of the windings shorted, on a "double-wound" toroid, right? (Better safe than sorry 🙂 )
The top sketch in your posted pic resembles how the windings look on common-mode inductors, on SMPS inputs, so i'm guessing the bottom sketch is the differential-mode winding? So that's how a coupled inductor for a bridged amp should be wound? Or... Which type of modulation should be configured, for which type of winding? 😕
And by "one end shorted", i guess you mean with one of the windings shorted, on a "double-wound" toroid, right? (Better safe than sorry 🙂 )
Note that in the bottom sketch (AD) the flux is always contained within the core, while on the top sketch (BD) the flux becomes progressively spread everywhere (as in a pair of air core inductors) as output voltage becomes higher or lower than 0V.
A-haaaaa, now i get it! That looks like L = leakage L to the audio signal, but as Lnominal to the carrier. The handwriting wasn't 100% readable, but now it's all so much clearer, and it makes sense 😀 😱
Many, many thanks!
Many, many thanks!
No, you should short two input wires while measuring on output wires.
Eva!
No, this is not that one. Check current directions! Both drawings show BD modulation, but in different aspect (upper: normal BD signal current, lower: common mode content of BD).
I still cannot understand, how it is possible to use coupled inductors in output filter with BD modulation 😕 At non zero modulation index, both half-bridges in BD modulation mode will produce the pulses, which will be different in length. This means, at some time point one half-bridge will already change its output state, whereas other half-bridge not yet. As the output inductors are coupled, we will get an pulse transformer between half-bridges outputs. And if the leakage inductance is not high enough, then this will produce huge currents, I think, as the pulses on the transformer windings are out of phase...
I have tried to do this with T106-2 core earlier, with 12+12 windings on it, wounded like Pafi has shown. I did get huge current consumption and excessive overcurrent protection triggering. Since this, I am thinking that it is not applicable for BD modulation 🙂
I have tried to do this with T106-2 core earlier, with 12+12 windings on it, wounded like Pafi has shown. I did get huge current consumption and excessive overcurrent protection triggering. Since this, I am thinking that it is not applicable for BD modulation 🙂
I think 81bas right,not only BD mode but alos AD mode can not use common iron core.
Maybe someone influenced by TAS5613 EVM schematic,http://focus.ti.com/general/docs/lit/getliterature.tsp?literatureNumber=slou286&fileType=pdf
I have try it when i gOt tas5613PHD days ago,but no success.IC always protect.
Maybe someone influenced by TAS5613 EVM schematic,http://focus.ti.com/general/docs/lit/getliterature.tsp?literatureNumber=slou286&fileType=pdf
I have try it when i gOt tas5613PHD days ago,but no success.IC always protect.
it worked for me with TAS5630 in AD modulation mode, but it is needed to make sure, that inputs are feed with strongly differential signal (to make both half-bridges switching synchronously).
A dual inductor will work in a full-bridge AD mode amplifier,and inductors designed for this are sold by Coilcraft (for example), but are likely to be an EMC nightmare as any common-mode noise from the switching edges will pass straight-through. You would need a separate common-mode dual choke to keep the RF emissions down.
Yes, if it is not high enough. But as I said, "you will need much more turns then usual". I tought my picture (and my post) told how it works. I've already told leakage is the main inductance in this operation.
Of course that's not enough! That's less then 2 uH even in common mode, it's next to nothing!
I'm so said! I tought my amplifiers work, but now you showed me they don't.
The linked method (wich is completely different from mine) can work (with AD modulation), but it needs a high Q core, and type -26 iron powder is definitely not a high Q material!
- Status
- Not open for further replies.