Offline full-bridge SMPS....need help

[xeclipse]

Hi Chas1, I wanted to modify a pc power supply for amp testing and low voltage source for testing Luka's smps. I removed all the secondaries and left the primary winding. There were 17 turns then the secondaries then another 17 turns.
Here's how I rewound the former:
17 turns primary, 5+5 turns secondaries (for +/-30 v), then the remaining 17 turns primary.
I got +/-25v output. The main former was heating within a few seconds with no load (can't hold the back of my finger on the core for more than 5 secs). There was also a whine, barely audible, from the former.
So I increased the secondary turns to 40 then 45 then 50; but the first primary winding(17 turns) remaining the same, only changing the second primary winding turns.
The core was still heating up but taking longer, about 15 mins with 50 turns.
What could be the problem? I may not have noticed the heating before because the pc psu had a cooling fan.
I will appreciate your assistance.
Regards,Xeclipse.

[chas1]

Need more info:

What core set are you using or what is the power output of supply you are modifying, in any case you should recalculate the turns needed along with Bmax based on fsw you are using,core could be saturating.Also you need to load the power supply for at least 10% of required output.A schematic would help.Rewinding the transformer should follow 1/2 primary,complete secondary then the rest of primary,secondary should be bifilar and primary halfs should have equal amount of turns and tape used for isolation.Make sure you remove shields in PC transformer.Most PC transformers have 40 turns in primary 20,20... an last but not least check waveforms.

[xeclipse]

Originally Posted by chas1

Need more info:

What core set are you using or what is the power output of supply you are modifying,...

Its the same core that was in the psu, etd29. The power output was 215 watts.

Originally Posted by chas1

... in any case you should recalculate these turns needed along with Bmax based on fsw you are using,core could be saturating.

I haven't changed the fsw.

Originally Posted by chas1

Also you need to load the power supply for at least 10% of required output.A schematic would help.

Yes, there were 2w resistors to gnd at psu outputs, I removed them. I don't have a schematic, the only modification I did was turn ratios keeping the original 34t primary.(17t first then whole secondary then 17t primary).

Originally Posted by chas1

... primary halfs should have equal amount of turns and tape used for isolation.Make sure you remove shields in PC transformer.Most PC transformers have 40 turns in primary 20,20... an last but not least check waveforms.

Could the different number of turns in the primary halves be the problem? Because when I increased the primary turns, I only increased to second half, ie:17t primary, whole secondary, then 23t.
Yes I removed the shields and taped the half primary before winding the secondaries then taped the secondaries before winding the second half primary.
Unfortunately, I don't have a scope.
Thanks for the reply.

[stewin]

hi guys . can this ultrafast igbt IRG4PC50U be used in this circuit (lukas smps ver 6) with increasing the freq .to give an output of 3kwts ?

[Workhorse]

Originally Posted by chas1

I would like to second Luka's comment, and I would also like to thank all who shared in this thread.Info about my website is cedlabs.com and at this moment under construction because I had a lot of LLC converter notes that needed tidying up along with changing the theme.My plan is to have complete designs of Fullbridge offline,Halfbridge offline,Two switch forward as well as LLC info.All will have complete simulations using Ltspice along with detail explanations of closing the loop and zener feedback.I will allow others to post there designs as long as they are complete if they want.
chas1

Chas1,

Any idea, whose website is this

K6 700 Watt Audio Amplifier and Switching Power Supply

http://www.a-and-t-labs.com/K6_Sw_Amp/Supplyboard.jpg

[clocktower]

Originally Posted by Workhorse

Chas1,

Any idea, whose website is this

K6 700 Watt Audio Amplifier and Switching Power Supply

http://www.a-and-t-labs.com/K6_Sw_Amp/Supplyboard.jpg

[chas1]

The designer of the supply is Reinhard Metz and Myzil Boyce and they are also
the owners of the website a-and-t-labs.com and Reinhard Metz is a member of this forum
and has a thread in support of this supply.You can find the schematic,parts lists as well
as a complete discription of the supply on their website.The design was published in the now defunct Radio Electronics magazine which later became Electronics Now.The design was to power a 700watt amplifier that they designed.

[Workhorse]

Originally Posted by chas1

The designer of the supply is Reinhard Metz and Myzil Boyce and they are also
the owners of the website a-and-t-labs.com and Reinhard Metz is a member of this forum
and has a thread in support of this supply.You can find the schematic,parts lists as well
as a complete discription of the supply on their website.The design was published in the now defunct Radio Electronics magazine which later became Electronics Now.The design was to power a 700watt amplifier that they designed.

Thanxz for the info.

[chas1]

follow up

This is the url

K6 700 Watt Audio Amplifier and Switching Power Supply

I constucted this supply with some modifications and used it to power a 2kW Crown for a test
There were some small issue's but I reduced the weight from about 150 lbs to about 60.I still have
the dual transformers and rectifier assembly from the Crown as a reminder what switcher's really do
in high power amps.

[stewin]

more threads on the above smps

K6 Switching Power Supply

K6- Switching power supply

A and T labs K6 SMPS for Class D amp use