ucd2k overheating

Status
Not open for further replies.
Jump to Latest
I

iryont

Member
Joined 2011
blibestmentalt.no/54981
#1
ucd2k overheating?

Hey guys

I'm having an issue with my ucd2k modules since both of them are hitting over 130C on the coil itself after some time. Most of the time they were doing nothing (no signal) except being connected to 8R speakers with serial crossover (not a parallel one though). The sound is fine and there are no distortions.

P9ii1Fx.jpg


TZcqXj5.jpg


I can understand the need of providing better heatsink according to the datasheet. However, the coil is the one which makes me worried.

How do proceed with it? Should I add a heatsink for the coil, too?
 
Last edited:
#6
iryont said:
Adjusting frequency in self oscillating class D amplifier?

I did contact Hypex. They are aware of the problem and they told me it would be good to cool down the coil with a thermal pad to a heatsink as well. I will do that then.
Click to expand...

Aware of the problem? and shipping these products world wide?

And the solution is to attach a heat sink?

Not sure if the wire inside the coil can handle 139C for too long.

At least good to know these issues about these modules 😀
 
#7
Apparently. It does look to me like a total failure - 140C on idle is not something withstandable with higher output powers.

I just hope I can cool the coil down to reasonable temperature. They did state that their amplifier has idle loss of 35W which is true - mine was drawing 215 mA across 160V swing. They didn't specify where the loss is happening 😛
 
#8
iryont said:
They are aware of the problem and they told me it would be good to cool down the coil with a thermal pad to a heatsink as well.
Click to expand...
...now I am close to buy one, just for curiosity. I am struggling to believe that they sell such... It really does not sound like Hypex.
Even if they choose high temp wires, tapings and bobbin, then still it does not appear reasonable. Most power ferrites have their lowest losses around 100°C, even considering that the ferrite losses dominate at idle and the goal was to keep copper losses low at high power - even then I am not convinced that 140°C at idle would be the right choice for smallest size.
In any case 140°C is headache for the life time of the soldering, the PCB, surrounding components and a real killer for the e-cap next to it....

P.S.
Can you measure the operating frequency and double check with Hypex?
I still cannot believe that they leave you alone with 140°C at idle.
If something went wrong with tolerances or wrong component value inserted
for a relevant component, then you can get such effects.
Typically ferrite losses go up extremely with increasing flux density which results towards lower frequencies and this effect is just partially compensated by the loss reduction towards lower frequencies. In sum a low idle frequency can lead to overheated chokes.
...or they simply got delivered chokes with really scrappy or wrong ferrite.
 
Last edited:
#9
You gotta believe in that. I have two different modules and they act the same - both of them do reach the same temperature on the coil itself.

This was the message I did receive:

Yes, we recently found that these coils run a bit hot indeed. It wouldn't hurt to apply some heat conductive filling pads between the coils and the heat sink mounted to the radiator.
Click to expand...

That's all of it. I didn't receive any other suggestions or a replacement possibility.

I'm going try to cool it down with a thermal pad to a main heatsink. The idle loss is about 35W which is exactly what they specify in their datasheet, but I didn't expect the loss to happen on the coil itself. That's not desirable and certainly the coils would die if they were used with much higher powers for a longer period of time.

I will measure the operating frequency as soon as I get some free time.
 
#11
"..heat conductive filling pads between the coils and the heat sink mounted to the radiator."

...hm... if it was intended by the design to transfer the heat with a gap filler to the heat sink, then it would be interesting to see how much the idle temp would go down, when providing a heat sink which idles at 50C or less... This might change the picture.
Is there nothing written in the describtion?
 
#12
Coils are flat from the top. It looks like they were intended to be cooled with a thermal pad. I will find out once I give it a try. However, I couldn't find any information about it within the datasheet itself, but on the other hand they do specify that "user has to meet the condition of < 95C with an external heatsink".

I will let you guys know, but first I need to receive the heatsink.
 
#13
Just another idea. I saw your thermometer, have bought a similar one for little money with a laser diode in it pointing to the object. Fact is, there is an parallel offset error of 2~3 cm due to the construction. Are you sure you are focusing onto the chokes with your thermometer?

Btw - 35W idle power is really heavy stuff!
 
Last edited:
#15
iryont said:
Coils are flat from the top. It looks like they were intended to be cooled with a thermal pad. I will find out once I give it a try. However, I couldn't find any information about it within the datasheet itself, but on the other hand they do specify that "user has to meet the condition of < 95C with an external heatsink".

I will let you guys know, but first I need to receive the heatsink.
Click to expand...

Would you pls share a fotograph of the coil(s)?
 
#17
iryont said:
Take a look at the first two pictures. You can see them since I was measuring their temperature.
Click to expand...

thanks, I see. Looks like some RM or PM core of appropriate size. I really wonder why they get so hot in idle mode. Can you measure (oscilloscope plot) inductor current by inserting an inductive current probe? This could give an idea of ripple current and possible core saturation.
 
#18
Thanks for pointing this out, I have a pair of these 😱
I have to check this out though, as soon as possible.

Mine are mounted in two of these, with a regulated psu using the top plate as
the main cooler. There is also a 120mm fan mounted at the bottom plate to
get some movement of air. They ran pretty hot without the fan so it feels
better to have that as well. Maybe I could squeeze in a laminate of thermal
pads and aluminum between the coils and the top plate. From the drawing in
the datasheet it looks like a 10-12mm distance. There are 12mm thick
thermal pads but does that really work 😕

Also squeeze in some between the coil and the heatsink.


$_12.JPG
 
Last edited:
Status
Not open for further replies.
Top Bottom