toroid for car amp smps

[ssanmor]

Hello, Eva.
Nice to hear from you.
Of course the ESP article can be improved a lot. As I said, it was written 2 years ago, and now there is a lot more of useful info over there that could be included.

If you want to collaborate we could add value to the project 89, to make it more useful for DIYers. For example, a extensive description of transformer winding could be very useful.

I am open to your comments.

Best regards,
Sergio

[soundbag]

Now that Eva and Sergio agreed to modify project 89 which I happen to be interested as well.

Pls, can you make the mains version, for people who dont live in a car.

I will be glad, to see that pls.


SOUNDBAG.

[lachko973]

MAINS POWERED SMPS of about 1KW or more is my purpose. I have build the driver schematics of it but of course my problem is the power transformer (no core available ). I'l order one from some manifacturer if I could'n get some good one in my country in the near furure.

My thought was that I'l be at least very very thankfull if some of the gurus of SMPS like Eva (without giving offence to others who are here in the forum ) like to design something like @ T H I S @

I'd like to see as more as possible designs to get more fresh ideas of the whole project.
Thanks in advance !

Best regards to EVA and all other SMPS designers !

[fr0st]

A mains powered SMPS is alot difference to a comparitively simple 12v DC one. I wouldn't reccommend it to anyone that hasn't played with 12v beforehand.
ssanmor - I made rod's/your SMPS and it worked brilliantly. I managed to run 2 P3a's of the one supply without any problems.
For the australian reader of the artical it may be of interest to include the windings I used on a core obtainable from Jaycar (Australian Ratshack). I used 5.5 primary turns and I had no problems. I originally used 4 turns and it drew excessive current.
Sorry to be off topic but...
did you ever build your MAX4295 Class D amp?

Regards Matt

[jackinnj]

Originally posted by fr0st
A mains powered SMPS is alot difference to a comparitively simple 12v DC one. I wouldn't reccommend it to anyone that hasn't played with 12v beforehand.

and Matt, I go back to my point of months ago -- anyone who wants to learn will make it very easy by getting the Velleman PWM controller kit -- because it can be repurposed to a MOSFET gate driver using the SG3525 -- and you can twiddle the frequency and compensation scheme, soft-start etc to your heart's content.

with regard to a "mains" or off-line switcher -- the front-end operates at a multiple of the mains voltage -- which gets into the realm of electrical potentials most dangerous for human accidents -- and things really can fly apart --

[jackinnj]

attached is a GIF of the off-line switcher using the On-Semi MC34166 chip and MJE13005 transistors -- with the transformer information clearly laid out -- the MC34166 is a 3 amp 40V max switcher -- the idea is to use the MJE13005s as a self-starting multivibrator, then use the controller chips on the secondaries in their various configurations. On-Semi has taken the time to provide PCB patterns for the stepdown converters on the product PDF:

http://www.onsemi.com/pub/Collateral/MC34166-D.PDF

[B.I.G]

hello can i use a single SG3525 to drive four mos-fet`s each pair driving his own toroid ??? i would then put rectifiers for each toroid and then paralel them

the toroid`s would be identical of course

[Eva]

You can drive as much MOSFETs as you want from a single SG3525 as long as the maximum drive current specified for that IC is not exceeded [+-500mA I think]

Gate drive current is limited by gate resistor value. The most conservative way to calculate maximum gate drive current is to divide 12V by the equivalent resistance of the amount of gate resistors connected in paralell to each output, but looking at the waveforms with an oscilloscope reveals that actual peak current drawn is usually 2 to 4 times lower than expected due to IC rise and fall times

Obviously, MOSFET rise and fall times are highly depend on the magnitude of the gate drive current. Increasing the amount of MOSFETS driven from a single IC also requires increasing the value of gate resistors to keep the IC working below its drive capability, so adding more devices without buffering the output of the IC means slower switching

There is always a compromise buy you will need to build a prototype to evaluate it

[B.I.G]

10x for the reply .... it`s my first smps ... i have had a few attemps but failed before

i just finished my experimental smps it`s point to point uses sg3525 and 2xirfz48n .

the idle curent is about 250mA is that good ?? the mos-fet`s are cold with no load and no heatsinks !

[jackinnj]

Originally posted by B.I.G
10x for the reply .... it`s my first smps ... i have had a few attemps but failed before

i just finished my experimental smps it`s point to point uses sg3525 and 2xirfz48n .

the idle curent is about 250mA is that good ?? the mos-fet`s are cold with no load and no heatsinks !

groannnnnn....(:

don't use P2P for SMPS -- things which seem small in the 3D world are enormous when you consider that large currents switching in very small periods of time face enormous impedances. layout, groundplane, nanohenries of inductance are all critical, even with low switching speeds. it's the transition which kills.