UCD 25 watts to 1200 watts using 2 mosfets

[Pafi]

Hi coolet!

The one you see on my recent attachment, is made from 2 pairs of EFD25 core, material N87.
http://www.epcos.com/inf/80/db/fer_07/efd_25_13_9.pdf

Total cross-section: 115 mm^2 , air gap was originally 1 mm, but I increased it by 2*0.5 mm (0.5 mm inside and 0.5 mm outside). Wound with litz wire. VDD=85 V, f=330 kHz. Temperature is about 45 C.

You can use your core for smaller power, but the bigger is the better!

 

[Dzony988]

I use EE 33 pc core with gap 2x1mm and 16 turns. Sound is fine, core is cold, but wire on core are little hot? Do I wrong somewhere?

Thanks

 

[Pafi]

1: Keep wire away from the gaps!
2: Use litz wire, or stranded wire at least!

 

[Dzony988]

I varnished coli with nail polish 🙂.Temperature has dropped, now it is just lukewarm. This wit PC core was good solution.

Tell mi pafi, this core EE33 has AL(25%nH)=4300
How many uH my coil has with 2x1mm gap and 16turns? (in my case, number of turns was only experiment)

I changed 330pF in Feedback to 100pF.
I hope that now swiching freq is above 200KHz, with this coil until I find T130-2 core somehow 🙂

Thank you
Regards

 

[qsa]

maybe this schema the protection part can be in-cooperated .it uses 555 and ltv817 optocupler when it senses the load is high or there is a short circuit it disconnects ground from r10 .

have remove turn on delay from tekko and add the protection part from russian UCD 2.5

but replace LTV817s with SFH610A-1 optocoupler and BAV99 with 1n4148 because not available

The crowbar is for protecting speakers from DC and power supply RAILS if mosfet fails

check schematic, could this work ?

 

[coolet]

Hi coolet!

The one you see on my recent attachment, is made from 2 pairs of EFD25 core, material N87.
http://www.epcos.com/inf/80/db/fer_07/efd_25_13_9.pdf

Total cross-section: 115 mm^2 , air gap was originally 1 mm, but I increased it by 2*0.5 mm (0.5 mm inside and 0.5 mm outside). Wound with litz wire. VDD=85 V, f=330 kHz. Temperature is about 45 C.

You can use your core for smaller power, but the bigger is the better!

thanks pafi. can we used shut down facilty of IR2110 (pin11) when there is dc voltage on the output.

regards
coolet

 

[Pafi]

thanks pafi. can we used shut down facilty of IR2110 (pin11) when there is dc voltage on the output.

regards
coolet

Yes of course!

 

[nigelwright7557]

Yes of course!

I use an RC network through a gate into my 2113.

 

[70145]

have remove turn on delay from tekko and add the protection part from russian UCD 2.5

but replace LTV817s with SFH610A-1 optocoupler and BAV99 with 1n4148 because not available

The crowbar is for protecting speakers from DC and power supply RAILS if mosfet fails

check schematic, could this work ?

Shouldn't your feedback be connected to the inverting input of the LM311?

 

[NMOS]

Shouldn't your feedback be connected to the inverting input of the LM311?

Feedback from member for protection circuit is not big, but its very important

especially under voltage protection

 

[Dzony988]

Hi to all

I find at Farnell.com feritte cores, transportable in my contry and its cheap
Which of these cores can I use?

1) Multicomp 33RI 31.5x16x19
2) Multicomp 33RI 28x20x16
3) Epcos 25.3x14.8x10 material T37
4) Ferroxcube TN25/15/10-2E25
5) Ferroxcube TN32-19/13-3E25
6) Ferroxcube TX29/19/7.6-3E27
7) Ferroxcube TX32/19/15-3C90

http://www.farnell.com/datasheets/491631.pdf Datasheet for all Ferroxcube

http://www.farnell.com/datasheets/80884.pdf Datasheet for Epcos

Thanks

 

[Ouroboros]

Of the Ferroxcube materials you have listed, the only one which is intended for power applications is 3C90. The rest are intended for EMI suppression filters.

Search on the web for "Soft Ferrites and Accessories". This contains the full data on Ferroxcube cores and ferrite materials, and is a very useful reference.

 

[Dzony988]

You mean Ferroxcube TX32/19/15-3C90?

Will it be enough to 500 600W 4 ohms? How many turns for this core for 20uH coil?
its cheap 2.06 pounds

Thanks

 

[Ouroboros]

Yes, but this is an ungapped Ferrite toroid (not dust-iron), so it will need a gap cutting in it to prevent magnetic saturation.

The core will be (roughly) 32mm outside diameter, 19mm inside diameter and 15mm height.

if you cut a gap in it (with a diamond coated circular tile-cutter saw) so that it had an AL of 100nH, then my estimation is that it could tolerate an i^2*L of 4mJ. As an example, a 22uH coil carrying a peak current of 13A will have a peak i^2*L of 3.7mJ.

I have never tried gapping a ferrite toroid with a tile saw, so if you did that then you would have to wind a few turns on the gapped core and measure the inductance to determine the AL for the gapped core.

 

[Ouroboros]

Actually I wouldn't try to go above 400W into 4 Ohms, with this core gapped to give an AL of 100nH.

 

[Dzony988]

I must to gap that core? how much? Some other feritte core without gaping for 400 500w into 4 ohms?

 

[Dzony988]

Can I use any core from here without gaping

Farnell Export

thanks

 

[Pafi]

I must to gap that core? how much?

Summed volume of gaps should be ~100 mm^2 at least.

Some other feritte core without gaping for 400 500w into 4 ohms?

No, it doesn't exist practically.

 

[lorylaci]

Dzony988:
MUST USE GAP!
Buy two of these:
EPCOS|B66361G1000X187|FERRITE CORE, HALF, GAPPED, ETD34 | Farnell Magyarország
This will give you 2mm gap (170mm^3), wind it with 18 turns of multiple thin wire (as much as you can, but don't wind over gap), this will lead you to an inductor of 30uH with 43A saturation current. There is 94mm^2 winding space, and 50% effective space use (even if you wind it very nice), that means 2,5mm^2 wire-group, so if you do right the Rdc would be under 15mOhm. Should be pretty good over 500W 4 Ohm too...

 

[Pafi]

Laci!

I'd rather say 28 A! I prefer higher Ae, lower Aw, since core loss is always there, but audio load is rather weak.

But this is already much better than a 106-26.