New heat sinks on the bridges

Posted on September 1, 2014


Njay: I’ve built the new heat sinks for one of the bridges. It’s got a lot more air surface than the previous ones. The downside is that they won’t fit in the old boxes any more.

Vnevoa: Cool, that’s good to hear. Never mind the boxes, I’m going to revisit the whole fixation to the Trike anyway.

Njay: I’ve tested the bridge with around 12~15A continuous phase current and it looks good, no extreme warm-ups anywhere. I’m using a stalled motor to ramp up the current. But the current on the battery side is low, like 2~5A, due to the transformer effect of PWM power transfer.

stalled motor

Motor rigged for stall current. The bridge electronics have to limit the current, or something will blow.

Vnevoa: Niiiice! 🙂 What’s the upper limit on the current? How far can we go?

Njay:Well, that depends on the time… if it’s short, we can pull a lot more than the 15 Amps. I’ve tried with 19A for a minute or so, it does get warm but it handles it. I think we can trust it.

Vnevoa:  Great. I’m going to have push the bridges all the way to the max in my testing, before the Faire. If it’s going to burn, it should burn before the Faire, not in front of everyone there. 🙂

Njay: Sure, push away, full throttle. 🙂

Vnevoa: How do you regulate the max current?

Njay: There’s a potentiometer on the bottom of the bridge, it’s a reference voltage for a comparator connected to the micro-controller. The comparator trips an external interrupt and the interrupt routine will switch off the power when the bridge is doing the active duty part of the PWM. So when you hit the max current you’ll hear the motor growl and vibrate, but it keeps moving.

Njay: Now let me show you the build process for the heat sinks…

heatsink plan

Plan for the new aluminium heat sink, made with 2 different extruded profiles.

heatsink mask

Paper plan with 1-to-1 real size markings of the holes to drill. The critical markings are the 3 for the transistors

heatsink drilling

Drilling the backing plates, two at a time (stacked), on my spiffy new drill jig.

heatsink threaded holes

Drilled holes. Notice that I’ve gone to the trouble of threading the 3 holes that secure the MOSFETs. This should help during assembly and keep all the transistors better glued to the plate.

heatsink assembled

Assembled heat sink, almost ready to attach to the bridge.

final bridge

Final assembly, bridge with new heat sinks.

Vnevoa: Yeeeeahhh… that’s very cool tech (pun intended). 🙂 Okay, one down, one to go.

Njay: Yep, we’ve got to build another pair of sinks for the second bridge. But I don’t have the time… it took me more time than I wanted to make these (less sleeping time for me today!).  If you can save me that time, it’s great. It’s your turn. You cut and drill and then give me the parts back so I can put the thermal paste and assemble the whole thing.

Vnevoa: Yeah! Finally I get to do something. 😀 What, I’m only going to drill, nothing more?…

Njay: Dude, it took me a long time just to cut and drill and thread. If you can subtract that work from my life, it’ll be great. Pay attention to these points:

  • avoid damaging the surfaces
  • the mosfet pins are more fragile than they look, handle with a lot of care
  • the holes for the FETs and the middle hole are all asymmetrically positioned
  • when in doubt, use the 3mm drill
  • drill 2 plates together
  • remove the metal shavings from all the holes and cuts.

Vnevoa: Yes daddy. 😉 Anything else?

Njay:Yeah, after the bridges are assembled with the new sinks, you’ll need to fix them more rigidly, they’re too flimsy right now. The large heat sink will easily break the MOSFET pins if you don’t fixate the sinks somewhere.

Vnevoa: Got it. Don’t worry, I’ll screw them on to some part of the Trike.

Njay: Make it detachable, for easy maintenance.

Vnevoa: Obviously. So, we’re now at exactly one week from the end of the Maker Faire registration, and I have to make a decision about going there or not. It’s pointless to register if we don’t have a chance of getting the Trike running. Are there news from the I²C chips?

Njay: More or less, they should be arriving tomorrow or close to that.

Vnevoa: Cool. Because, if we find out that your bridges can’t be used for some reason, I can always try to use the MD03s I’ve got. In any case, your add-on board for I²C insulation is critical to stabilise the Trike and protect all the digital circuits. We have to get it working in this remaining week, or bust.

Njay: Yep, I think we shouldn’t have a problem there, the add-ons are assembled and just waiting for the chips. I did however make a mistake when I interpreted the chip’s datasheet, and the pins are reversed on my board. But it’s okay, we can solder the chip upside down and all will be well. 🙂

Vnevoa: Hehehe. Thanks to symmetric designs.

Njay:Yeah, we lucked out on this one. 🙂

(to be continued…)Lisbon_MMF_logos_Banner_big