ETD49 Overheating

[badboy_6120]

Hi guys

I started to build 1Kw smps designed by Ludo3232 and Microsim in this forum
http://www.diysmps.com/forums/showthread.php?79-1kW-smps-project-(based-on-MicrosiM-design)

I face a bunch of problems until now and thanks to "Jagd.Panther" and "Microsim" I manage to handle them
now the problem is transformer over heating

with no load at 70Khz transformer cores temperature get to 65°c !!!!!!
Is this right???
should transformer heat at all in idle situation???

Please everyone who can help me solve this problem post what should I do ether here or at its original forum

I also have ETD59 in hand and I can replace it with this one but I don't think would solve this problem

Thank you in advance

 

[Jagd.Panther]

The overheating is caused by losses in the core, the losses depend on flux and frequency.

To put it simple: increase number of turns

That will help.

 

[badboy_6120]

The overheating is caused by losses in the core, the losses depend on flux and frequency.

To put it simple: increase number of turns

That will help.

Hello

I'm confused because Ludo3232 also used 22 turns for his transformer's primary and the temp as he said was about 55 degree for 45min in 1200W load (post 46 on page 5)

Hi all..

Smps is stable with 1200W load.. for 45 minutes tested..

- Mos 52°C without fan
- Tranformer 55°C without fan
- Out diodes 61°C with fan (but small heatsink..)

On 1200w load i have 20V drop.. is due to transformer, now is winded with 5 x 0.4 wires.. my new transformer is winded with 0.1mm wire and have less voltage drop..

 

[Jagd.Panther]

Is your core made of N27 material?

 

[badboy_6120]

Is your core made of N27 material?

I honestly don't know that because the only shop that I bought it from has no Idea about any more information about it


I did some test yesterday to prevent transformer heating
now wounded the winding on ETD59
I'll report the results

 

[badboy_6120]

I test the smps with ETD59 now and the transformer heats normal

without any load :
I test it for 1 hour and it did not heat at all ( room temperature 30°c)

with about 600W load in 1 hour the transformer's core temperature get to 52°c and 51°c for winding

How about these temperatures?

I'm still wondering how Ludo3232 claimed that its transformer's temperature in 1200W load for 45 minutes is 55°c without any fan?????

I wound the ETD59 as below
Primary: 14+14 >> 36*0.25mm
Secondary :15+15 >> 30*0.25mm

I increase the turns so the transformer won't heat in idle condition

 

[Jagd.Panther]

Hi Badboy,

Do you have any specs for the ETD59 core you use?

Anyway ETD59 has like 1.5 higher Ae & higher Al. You can either keep number of turns you used on ETD49 core or even reduce the number of turns if ETD59 is 'cold' on idle.

Ludo probably used different core material with lower losses/higher Bsat.

 

[badboy_6120]

Hi Badboy,

Do you have any specs for the ETD59 core you use?

Anyway ETD59 has like 1.5 higher Ae & higher Al. You can either keep number of turns you used on ETD49 core or even reduce the number of turns if ETD59 is 'cold' on idle.

Ludo probably used different core material with lower losses/higher Bsat.

The shops here does not know anything about these cores but their names and where they come from
The ETD49 is from china and ETD59 is from Germany (as he claimed)
I wounded the primary of ETD59 with 14 turns first (I get this number from Excellent 7300 program on another forum) but result was the same as ETD49 and the core gets hot on idle, so I increased the turn until I reach to 28 turns for primary without any cores heat on idle condition

Now my turns are :
Primary : 28 turns >> 36x0.25mm
Secondary : 16+16 turns >> 30x0.25mm

I also test the transformer in 1400w on output for more than 1 hour
Transformer winding temp =70"c
Transformer cores temp = 50"c
With a big fan working in front of it

Is my winding not enough for this much power?

 

[badboy_6120]

here's the ETD49 and ETD59 that I used in this smps

 

[Jagd.Panther]

Now my turns are :
Primary : 28 turns >> 36x0.25mm
Secondary : 16+16 turns >> 30x0.25mm

I also test the transformer in 1400w on output for more than 1 hour
Transformer winding temp =70"c
Transformer cores temp = 50"c
With a big fan working in front of it

Is my winding not enough for this much power?

Looks ok. Is there any free space to use thicker wire?

 

[Jagd.Panther]

here's the ETD49 and ETD59 that I used in this smps

View attachment 5468 View attachment 5469

Genuine ETD N97 cores made by EPCOS AG, sold by farnell . According to the label on the box they were made in Czech Republic.

 

[badboy_6120]

Looks ok. Is there any free space to use thicker wire?

Maybe a little more
I twisted the wires together with my hands and maybe if I use some kind of tool I could squeeze the wires a bit more and use more wires in parallel but I'm afraid to scratch the isolation layer of wires
The winding overheat is from primary or secondary???

 

[badboy_6120]

Genuine ETD N97 cores made by EPCOS AG, sold by farnell . According to the label on the box they were made in Czech Republic.

They look very nice
It is obvious that my Cores are fake but unfortunately they are the only ones that I could find here

 

[Jagd.Panther]

Maybe a little more
I twisted the wires together with my hands and maybe if I use some kind of tool I could squeeze the wires a bit more and use more wires in parallel but I'm afraid to scratch the isolation layer of wires

Looks good, don't try to increase wire diameter then.

The winding overheat is from primary or secondary???

Most likely both. You can either calculate current in pri/sec winding or open LTspice or other SPICE program and quickly check.

 

[Jagd.Panther]

They look very nice
It is obvious that my Cores are fake but unfortunately they are the only ones that I could find here

There are many materials out there, some are older and have higher losses, some are newer and have higher B, lower losses. If you got them for cheap then don't bother finding anything else. Use what you have.

 

[badboy_6120]

There are many materials out there, some are older and have higher losses, some are newer and have higher B, lower losses. If you got them for cheap then don't bother finding anything else. Use what you have.

Price was not my concern at all
I did not fing anything beside these anywhere here

 

[badboy_6120]

Do you have any specs for the ETD59 core you use?

Hello Jagd.Panther

I'm back after a while with some information about the ETD59 core that I used and here they are:

It's material is MF102

 

[MicrosiM]

While your SMPS is runnig at fixed duty cycly 50%, the transformer will heat.

Improving your transformer design and or core material, will not decrease the temreature in idle so much.

 

[badboy_6120]

While your SMPS is runnig at fixed duty cycly 50%, the transformer will heat.

Improving your transformer design and or core material, will not decrease the temreature in idle so much.

when I used 22turns for primary on etd59 the cores start to heat in idle condition (at 70Khz transformer cores temperature get to 65°c) so I increased the turns to 28 for primary and 32 for secondary and now there's no heating in idle but at 1400W load the windings gets really hot:
for 1 hour
Transformer winding temp =70"c
Transformer cores temp = 50"c
With a big fan working in front of it
are not my litz wires enough for this much power? (primary 36*0.25mm secondary 30*0.25mm)

Dear MicrosiM
you made smps more than these
did those smps trasnsformers heats like this??
"Ludo3232" claimed that its transformer's temperature in 1200W load for 45 minutes is 55°c without any fan

what did he do to get this result?

Thanks in advance

 

[Jagd.Panther]

Hi Badboy,

Go to farnell.com / digikey.com / element14.com / etc and order some EDT cores made of better material. Ferrite material, its permeability, losses, maximum induction, play big role in trafo design. The "better" the material the less turns you need, so you can use thicker wire and cut down copper losses.

Also you can look at toroid cores, do you have anything over 50mm in diameter in local shops?