Hello,
I am planning on building a power supply (posted this in another thread), consisting of 8 stages. I will not go into further details, but provide the following information: the maximum continuous DC total wattage that these 8 LM317/LM350 are going to dissipate, is around 36 watts.
In short, I am going to ensure a 6 volt difference between input and output, and the two LM350's will provide 1.5A current max each, while the 6 LM317's will provide 0.5A max each. So, total wattage on the regulators = 2*6*1.5+6*6*0.5 = 36 watts.
Now, I was thinking of mounting them directly on the aluminium chassis for cooling. The chassis dimensions could be approximated by 50cm x 15cm x 10cm, and it will be 3-5mm thick.
Do you think it would be adequate as a heatsink? The transformer powering the whole system will be a toroidal mounted inside the same chassis.
I am planning on building a power supply (posted this in another thread), consisting of 8 stages. I will not go into further details, but provide the following information: the maximum continuous DC total wattage that these 8 LM317/LM350 are going to dissipate, is around 36 watts.
In short, I am going to ensure a 6 volt difference between input and output, and the two LM350's will provide 1.5A current max each, while the 6 LM317's will provide 0.5A max each. So, total wattage on the regulators = 2*6*1.5+6*6*0.5 = 36 watts.
Now, I was thinking of mounting them directly on the aluminium chassis for cooling. The chassis dimensions could be approximated by 50cm x 15cm x 10cm, and it will be 3-5mm thick.
Do you think it would be adequate as a heatsink? The transformer powering the whole system will be a toroidal mounted inside the same chassis.
barely ...
a square horizontal 25cm x 25cm x 3mm bare aluminum plate has Rth approx. 2.5 K/W for a single heat source in it's center
the body temperature of a single 36W device in the center would therefore be at 90C above room temp
... don't know whether multiple devices distributed is better or worse but it seems that you need something like 1 K/W heat spreader with fins on the outside or backside of your box
a square horizontal 25cm x 25cm x 3mm bare aluminum plate has Rth approx. 2.5 K/W for a single heat source in it's center
the body temperature of a single 36W device in the center would therefore be at 90C above room temp
... don't know whether multiple devices distributed is better or worse but it seems that you need something like 1 K/W heat spreader with fins on the outside or backside of your box
What is better in terms of cooling as far as positioning the components is concerned?
I mean, is vertical mounting worse than typical horizontal? I have a 5W wirewound resistor that I want to place on my pcb, and positioning it vertically will save me much space. Will it also aid cooling of the resistor or the opposite?
I mean, is vertical mounting worse than typical horizontal? I have a 5W wirewound resistor that I want to place on my pcb, and positioning it vertically will save me much space. Will it also aid cooling of the resistor or the opposite?
If you refer to the consumption posted above, I have changed my plans, so that about 15W at most will be dissipated across the chassis. I will also use an aluminum corner that will be the basic sink.
Now I was thinking about mounting the 5W resistors vertically to cut the box size, but since there seems to be no other benefit I will not, after all it is not that critical.
I will use 105C capacitors. And biggest DC dissipation across each 5W is designed not to exceed 0.67W. I can touch them at such a dissipation - hot, but not untouchable for seconds (minutes) on end. So I assume that the inside temperature shall not exceed 20C above ambient.
Now I was thinking about mounting the 5W resistors vertically to cut the box size, but since there seems to be no other benefit I will not, after all it is not that critical.
I will use 105C capacitors. And biggest DC dissipation across each 5W is designed not to exceed 0.67W. I can touch them at such a dissipation - hot, but not untouchable for seconds (minutes) on end. So I assume that the inside temperature shall not exceed 20C above ambient.
As far as heat exchange is concerned, is it better to have the inside of the chassis painted black or not?
Painted only the outside matt black will be good, or must the inside also be black?
I don't know what applies to contacting metals - an aluminum corner mating with the chassis should be both left unpainted, or black color would enhance heat transfer?
Painted only the outside matt black will be good, or must the inside also be black?
I don't know what applies to contacting metals - an aluminum corner mating with the chassis should be both left unpainted, or black color would enhance heat transfer?
So no need to paint it internally. Thanks.
Yes, I was thinking about drilling some holes, around 3mm. Problem is, I may mount the toroid on the top panel "hanging", so that would not enable me to drill the top everywhere I want, so I will see how this will go. Besides, maybe nothing will get that hot.
Do you suggest drilling holes to attenuate RF, or you suggest these holes are kept small, not to ruin RF attenuation? It was not clear to me.
Yes, I was thinking about drilling some holes, around 3mm. Problem is, I may mount the toroid on the top panel "hanging", so that would not enable me to drill the top everywhere I want, so I will see how this will go. Besides, maybe nothing will get that hot.
Do you suggest drilling holes to attenuate RF, or you suggest these holes are kept small, not to ruin RF attenuation? It was not clear to me.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.