New ClassD project Starting from 0

Status
Not open for further replies.
Eva said:
You have to do the layout, and do it properly, tame the resonances, etc...

It's like MMIC design - the layout IS the design. I see a bunch of perf board and I doubt that will go over very big when it's scaled up to full power. You can build a 50 watt class D amp out of just about anything. Add zeroes and the fun starts - things that didn't matter suddenly do.
 
Uffffffff

Uffffff, as I can see, EVERYBODY, tells me ONLY about PCB design,and feedback.

I HAVE UNDERSTAND IT, I will go to a PROFESSIONAL COMPANY to design,project and realise that PCB for me, course I will pay for that, because I'm not intending to do myself PCB.


But to go there, I must go with the final schematic....


The only thing (help) that I asked here is an OPINION, maybe a small draving,ABOUT how to REPLACE all PMos transistors with N-Mos ones.

I know I cannot modify thet part of schematic,and I need an advice how to do to use Nmos for all output stage.

I know and I have some designs with IR2110 driving NMos, but is NOT IN FULLBRIDGE MODE, that I intend to keep here.

Looks like we spoken different languages, because I asked for something , and I always receive opinions about PCB designing anf feedback...

I will put again what I need...
 
luka said:
I can also tell you PCB will be hardest, it can kill project if not tone good

About IR, you must basicly use 2 half bridges and put them together. HOW THAT, some drawings will be helpfull.. Use RCD with inverters for easy dead time adj. HOW that too?
Again some drawings will be helpfull
That's about it


That's what I need to be helped....


An externally hosted image should be here but it was not working when we last tested it.
 
For dead time circuit, check my amp on this forum or my site

I used inverter > resistor + diod > C
Resistor was trimer, so I could set dead time
What I would do now is after C use another buffer, so that IR would get only squere wave. Also about R and C values you can read HERE and HERE

About how too put them together, I think it is wery much like bridging 2 amp, 2 half bridges, but only here you can use only one input circuit/modulator... also output filter in a bit diffrent, 2 LC filters(2 amps) but with addition of another bank of capacitors across load I think

In the end it doesn't matter if you are building full or half bridge amp, P mos can and will be replaced with N with use of IR..

I don't know... about if you will be able to make it work as it should, since I think this is your first amp... and full bridge is essentially two half bdrige amps, and even one can be hard to make, trust me, you can't just throw elements up...it will work, but not for long or with much problems
 
Something

Something like that, is OK?

Common ground will be the same for all stages.

Oscillator,input,feedback,comparator (LM361) will be supplied from SEPARATE +/- 12Vcc,referenced to Common GND, and IR2110 will receive their +15Vcc and +5Vcc,also from another separate +15 and +5Vcc,referenced to Common GND.

The +5Vcc will be obtained with an 7805 from +15Vcc of IR2110 supply.

That means, power supply will have one +/- 12Vcc outputs respect to GND .

Another +15Vcc output respect to GND,from where I will derive+5Vcc.

The power part of +120,,,160Vcc, +Vcc and Common GND

+/-12Vcc out GND will have common with Common GND, also +15Vcc GND will meet Common GND in same point.

That means I will have +/- 12Vcc,+15Vcc,+5Vcc, all galvanic isolates but with their GND connected to Common GND.

Here is the ideea:

An externally hosted image should be here but it was not working when we last tested it.




Any other comments or suggestion are welcomed!
 
I think you could use only 15v for IR, no need for 5v
Dead time circuit is ok, but remove 47R, you need fastest turn off as possible

R should be as small as possible and C as big as possible, not saying this is wrong, but just to say this is due to noise pick up to be lower

Aslo no need for 18V zeners, since you will have 15v regulated I assume
 
Yes

Yes, it's working perfectly!:smash: :devilr:

Small modifications of tensions on pins 10 and 14 of LM 361.

On those small heatsinks, taken from an faulty PC power supply (very small) temperature not exceed 45-47 celsius degrees, after 40 minutes of benching on 4 ohms 300W speaker,supplied at 55Vdc,achieving on speaker 37-39VRMS till clipping.

In idle they rest 2-3 degrees more than room temperature.

In idle I have 13 miliamperses consumed by IRF540, the output residual of 125 Khz, measured on bridged output decreased from 138milivolts to 33 milivolts, and DC offset is under 10 milivolts!:bigeyes:
The coils are cold cold cold!😎 :angel:

I'm very very satisfied about it, and next weel I will try with 2 IRFP260N in parallel for each mosfet , and I will grow powersupply to 100-125-150Vcc.
After that, I will give a try with optos and TC4xxx mosfet drivers.

Those are the dimensions of heatsink that I used, and they are in centimeters.

Opinions?

An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.


I'm estimating it's randament is around 94%
 
What

What someone could say about those Mos-Fets?

Running at 125 Khz, no more, no less.

Supplied at max 170Vcc,equivalent of +/- 85Vdc

Evident in full bridge mode.

http://ixdev.ixys.com/DataSheet/99118.pdf

or

http://www.farnell.com/datasheets/307949.pdf

or this one looking to be the best,excluding capacitance:

http://www.irf.com/product-info/datasheets/data/irfp4668pbf.pdf

or that:

http://ixdev.ixys.com/DataSheet/99297.pdf

or:

http://www.irf.com/product-info/datasheets/data/IRFB4110PBF.pdf

I'm waiting for your fair opinions, regarding advantages and disadvantages of each one.
 
DjLeco!

If you want to use it on 170 V, then how could you choose IRFB4110, IXTK180N15P, IRL1004? :bigeyes:

IXTQ100N25P is more-or-less usable, but it's diode is quite slow (causing a lot of switching loss and EMI), and I bet it's very expensive.

IRFP4668 is good, if you watch it by itself, but you can run into some problems if you drive it with IR2110. With optocoupled driver it can be good.

In one of the nearby topics, you can find some preferred devices.
 
luka said:

and your supply should be 85v not 170, since you are runing full bridge

I wil suppli'em with 125 between 150, even 180Vcc in fullbridge mode.

I want to have more than 1500W/8 ohm 2000+ on 4 ohm, parraleling 2 mosfets trying into 2 ohms load...

And the mosfet winner is?
For 160Vcc singlerail supply in fullbridge mode, drived by IR2110, or even with TC44xx buffer?


:hot:

For me seems to be IRFP4668, his 10700pF capacitance is rated at 1Mhz, I will run it at 125 Khz wich is 0,125Mhz.

Any other recomandations, the ones that you have used?

Pafi said:
DjLeco!

If you want to use it on 170 V, then how could you choose IRFB4110, IXTK180N15P, IRL1004?


Sorry with that IRL, it will be intended for car SMPS from 12Volts :clown: :clown: [/QUOTE]
 
First

First I will give a try with IRFP260N drived by IR2110, at 100Vcc.

Is ok to use ES1D (15nsec/200Volts/1A) diodes,between D-S pins for fast recovery?


http://www.fairchildsemi.com/ds/ES/ES1B.pdf

Also I have alot of MURA160T3 fast diodes too.


Is enough at 1A, or I must use bigeer amperage type?

Wich are the difference between Microchip drivers TC4421 and 4422, one is inverting and other is non inverting?

http://ww1.microchip.com/downloads/en/DeviceDoc/21420d.pdf

Could be used as mosfet drivers,drived by 6N137 optos?
By example TC4422 Low side ant TC4421 High side?

Or how else?
I have 20 from each TC's and I want to use them.

Also I have alot ot TC4426, dual mosfet driver.

Are usefull too?

http://www.datasheetcatalog.org/datasheet/microchip/21422b.pdf

And also alot of MC34151 (originals that was used in Crown BCA series amplifier, but seems to be inverting ones..

http://www.datasheetcatalog.org/datasheet/on_semiconductor/MC34151-D.PDF

Waiting for some advices, pease!


An externally hosted image should be here but it was not working when we last tested it.
 
Ok

Ok,I will buy some 4227's , thank's for advice, but no one told me a formula to calculate Snubber's value (resistor+capacitor) wich will be on D-S pins of mosfets, and a formula to calculate output coil and outpul capacitor.

Please need some assistance solving that problem!
 
no one told me a formula to calculate Snubber's value

It depends on the parasitic elements, PCB layout, and have to be determined by experimentation.

a formula to calculate output coil and outpul capacitor.

It depends on the requirements and the reachable parts and materials, but it can't be transformed into a simple formula. Since you haven't specified any requirements related to output filter, and we don't know your sources, we can't help you.
 
Pafi said:


It depends on the parasitic elements, PCB layout, and have to be determined by experimentation.



It depends on the requirements and the reachable parts and materials, but it can't be transformed into a simple formula. Since you haven't specified any requirements related to output filter, and we don't know your sources, we can't help you.

The output impedance will be minimum 2 ohm, maximum 4 ohms.

The final stage will be capable to drive 2 ohm loads.

And other, I don't know, just the triangle freq will be 125 Khz.

Tell me exactly what parameters you need.
 
Status
Not open for further replies.