Printable View
Hi
I know the Big Nipper have used this to great effecton their robots but I am having a few difficulties on comming up with a design. Having talked to Paul at Technobots he confirmed that they are Plastic fets so they dont need to be insulated. So my Idea was simply to bolt a piece of aluminium, the correct thickness between the two rows of fets and then attach a PC heat sink to the top of that.
Has anyone any other ideas?
Regards
Ian
could you solder 1 row of fets to the bottem side of the circuit board? That would give you enough space to place 2 nice heatsinks.
We cut the PC heatsink in half and took a few fins off, thinned the heatsink base until it fit inbetween the fets and drilled 4 holes in it. Used 2 strips of ally each with 4 holes (just as washers really) and bolted them either side of the fets to squash the fets to the heatsink with lots of heatsink paste inbetween and bolted them on.
Dont surpose anyone took pictures of this at Magna?
Mark
If no pictures come forward would you consider taking one?
Regards
Ian
Got plenty of time this weekend as Ive nothing planned anymore, so will do.
Thanks Very Much mark, Much appreciated.
Regards
Ian
Ian, look at my profile. Hope this helps
Looks good.
Any idea how many amps it can safely handle now? Does the device current limit to a preset level or switch off if it overheats? I cant see either mentioned on its datasheet.
Thanks
Mark
Thanks Mark.
My only slight problem is my Pc Heat sink is has alot tinner of a base. I suppose I will have to look for one.
Thanks
Regards
Ian
The pic on Marks profile has a green automotive fuse in the holder ... this is I believe 30 Amps.
There isnt any current limiting or protection on these controllers. You do need a fuse to protect them.
Yes it is a 30 amp fuse in the picture. With a heatsink the controller should take 27amps per channel continuous,but if you only stall one motor then a track on the pcb burns out and doesnt blow the fuse, Like when my robot got jammed under the arena flipper. I am now going to put a 25amp fuse in plus a backup 30amp fuse swicthed in with a servo just in case. I am running 30volt motors at 33.6v, normally the wheels spin all the time even when pushing other robots and pulls between 5 and 10 amps per motor but when stalled would easily pull 100+amps.
Graeme - little hitter 2.
ive heatsinked my controller too, pic on profile
I have removed the FETs from my H-Bridge and used short single wire links to a new set of four higher rated FETs (75A) on a seperate heatsink. Although this is more work than fitting a heatsink to the existing FETs, it does overcome the inadequate PCB tracks that are vulnerable to damage at high motor loading beyond its designed rated specification. It also allows a more adequate heatsink to be used with less restrictions on shape and size. I have also added a Hall type current limit cct and a thermistor temperature sensor on the heatsink. If anyone decides to try this, then I would advise that all PCB to FET wire links are kept as short as possible and not bunched or provided as part of a ribbon cable. The FETs can be cut leaving just a short stubs for soldering on the wire links which should be very carefully checked to ensure that they are each connected to the correct point on the new FET bridge. Each bridge should then be capable of a max of approx 50A, assuming an adequate heatsink. Do not attempt this unless you have some previous experience in electronics construction.
clive would you be able to get a photo of it as it sounds very interesting?
Gary - I have uploaded a photo of my H-Bridge mod to my profile. The FETs are fitted to two alluminium bars, four on each bar. The thick wire link shown is the common negative rail linking the low FETs on both bridges. The high FETs are arranged on a seperate bar and are isolated with insulating pads, so that the two bars can be tied together with an alluminium bridge piece capped with a PC fan heatsink. There are no doubt more elegant ways of doing this but I used some pre-tapped FET bars that I had in my scrap box. The short wire links between the FETs and the PCB are clearly shown. The bridge power and motor screw terminal block is redundant (and could be removed), as I have taken the supplies directly directly to new termination points on the new bridge.
My post above should read that the low side FETs are insulated from the bar. The high side FETs are mounted without insulating pads.
After further testing, I have noticed a tendency for my modification to cause shoot-through, probably because I have not been able to shorten the wire links enough to prevent a timing alteration of the FET switching.
You can slow down the turn on and speed up the turn off by added another resistor and a diode in the MOSFET gate path.
Unfortunately, I am no longer able to experiment, as the 5V data rail is short cct. This is probably because of lack of opto-isolation which allowed a large spike to damage the control circuitry when the shoot-through occurred.
help!!!
I was being a bit careless and decided to attach my battery direct to the motor whilst the controller was still attached, anyway the FETs gave off lots of smoke and it no longer works at all. Can this be repaired???
Ooops, You let out the magic smoke:sad:!
Either:
a) Buy new fets
b) Put the smoke back in
Regards
A bored Dave:)
Could anyone put up a pic or two of this mod? I am about to try it myself and want to see what every one else decided to do before I start.
You have mail Joseph.
Thanks Woody! Ill have a go at that next week sometime.