LMD18245T heating

[phantomcow2]

I was wondering if what I am experiencing with these chips is normal.
I just soldered together a stepper driver which utilizes two of these chips. Chips are rated 3amps, and I am running my motors bipolar parallel at 2.8A. Well I got a piece of 1/8" thick aluminum with the proper holes drilled bolted on the two drivers as a heatsink. I did use heatsink compound.
Well this driver has a test, where 5v is applied continuously to another terminal labeled "Enable". Driver will generate pulses to run motor at 60RPM. Motor did turn fine, and smoothly I might add.

Heating is my concern. Within 3 minutes of operation, there was a good amount of heating with that heatsink. My heatsink was too hot to touch. Motor was unloaded.

Is this typical with this driver? My heatsink setup is 2x1.25", it covers both chips. When all is operating, I will have a 120mm case fan blowing on it.

[Madclicker]

Everyone says these chips run hot and you are trying to run them near limit with a makeshift heatsink. I love the smell of toast in the morning.

They should work fine if you heatsink them properly.

[jfong]

A small 2"x1.25" x .125" thick piece of AL will not do much for dissapating any heat. You need to use a larger heatsink with fins that increases the surface area. Older pentium cpu heatsinks work good if they are large enough.

I have 6 LMD18245 chips mounted to a 7"x6"x"2" heatsink from a old car audio power amplifier. All set at the maximum 3amp current. There is a 3" computer fan blowing on the heatsink and it gets barely warm even after a few hours cutting metal on the mill.

On my testbench, I have another driver with 2 LMD's poorly screwed on to a 3"x3" peice of AL angle. No heatsink compound, no fan. I use it to stress test the driver board. It gets really hot after spinning a motor at 2000rpm for 20 minutes but the chips seem to handle it pretty good. It doesn't even trigger the internal thermal shutdown protection. The chips can take alot of heat but if you want them to last, use a decent size heatsink with fins on it.

Jim

[phantomcow2]

Well. I have a lot of this 1/8 thick stock. All 1.25 wide, but 30" lengths. What if each chip had its own piece, maybe 4 inches in height? Would this be more suitable?

[jeffs555]

It is surface area that disipates the heat. That is why heat sinks have fins, to maximize the surface area. Do yourself a favor and get a good heat sink. Even an old 486 heat sink(1.7"x1.7"x.8") would have twice the surface area of your 1.25"x4" plate.

[phantomcow2]

I dont have a 486 sink to use. But how about a heatsink from an ATX psu? These are something like 3" long, 2.5 in height. But they fan out at the top, with several extruded fins.

[jeffs555]

I wasn't recommending a 486 heat sink, just illustrating that even a small finned heat sink would be better than flat aluminum. I don't know anything about the ATX heat sink, but you can measure the surface area of the fins and see how it compares. Sounds like it is worth trying.

PS I looked at the LMD18245 specs, and it looks like at max current the power dissipation could be around 10 watts. The maximum operating junction temp is spec'd at 125 deg C. There will be a junction to case differential of about 15 degC, so the maximum case temperature should be less than 110 degC. The lower the temperature the better for the life of the chip.

[lucas]

Originally Posted by phantomcow2 But how about a heatsink from an ATX psu? These are something like 3" long, 2.5 in height. But they fan out at the top, with several extruded fins.

I use something similar (from a AT PSU) but it wasn't enough without active cooling (fan's).
A bigger heatsink might help but what if next week the environimental temperature rises? IMO fan's are the best solution, big and good heatsink's aren't cheap if you have to buy them, fan's can be found everywhere...

Just my 2 cents.

[phantomcow2]

Well Like I said in my first post, I was planning to use a 120mm fan ontop of everything when this thing is actually in use. The heating I experienced was only from some testing.
I am going to give the ATX PSU a shot, should make some difference.

[phantomcow2]

I've been running this 2.8amp motor in test mode for 45 minutes with the power supply sink on the chips and voltage regulator, and an 80mm case fan blowing right at it. Big difference, the heatsink feels pretty warm, but comfortable enough for skin contact.
The heatsink now is 2.5" in height, and 3.2" wide. Base material is 3/16" thick, but at the top it fans out to 1.5" thick.
Thanks for the help all.