Quote:
Originally Posted by lutach
I can work with opto as most of my ESC are opto. I think demand is there, but once folks understand the benefits of HV, there should be more demand and I'm sure the boat guys will use it as well. The Monster will be using the SUPERSO8 (TDSON-8) package MOSFET and Infineon has a BSC079N10NS that seems real good. I wanted to try something with my HV110 that I still have to send to Castle for repair. The MOSFET that I wanted to use has a 14 week lead time and the only one available is a 100V that has a cont. 34A at 25C and 21A at 100C. It has the same SO-8 foot print. All I would need is a simple forward/brake car program if it's possible. I called you, but I think it was on the day you were out sick. I know you are very busy and I would rather have you do it to my controller than waste a good one.
|
Actually, the BSC079N10 doesn't work well at all in motor controller applications. We've been working with samples of that MOSFET for about six months and have had very serious issues getting that FET to work well in a bridge application.
However, the BSC118N10 works great, and is the FET we are using in the SHV controller. It is a little higher on resistance, but works in bridge applications. The BSC079 won't stay off on the low side due to gate capacitance ratio issues.
Now that said,
the manufacturer's continuous amp ratings on ANY FET are total BS. It's mostly specmanship between manufacturers. There is no way the BSC079N10 can take 34 amps unless it was by itself with a near infinite heat sink... That's 10 watts of waste heat at 34 amps -- would burn up almost instantly.
Just look at how many FETs we use per amp on the phases of the HV controller.
You can assume that a dissipation of about 300-400 milliwatts (.3-.4 watts) per FET (of I^2R losses - -assumes very fast turnon / turnoff) is going to be a safe maximum. So for the BSC118 (11.8 milliohms) that gives a maximum of about 10-12 amps per FET on each half-bridge leg, when WELL heat sinked. Given that the MMM has 36 FETs (6 FETs per half-bridge leg) that would only yield a 60-70 amp controller. (I leave it as an exercise to the reader to try the math on the MMM NTMFS4833 FETs, to get a valid maximum continuous current for the MMM)
The HV-110 would be a better platform with 108 output FETs (18 per phase), but it cannot dissipate as much heat per FET as the MMM.
So, I think a 72 FET controller would probably be best -- using two power boards like the MMM, with good heat transfer. That would yield a pretty bulletproof controller in a smallish form factor.
We won't even talk about the SHV-250, because it is a little big (5" x 2.5" footprint.)
Patrick del Castillo
President (And still principle engineer!)
Castle Creations
I like the silly smilies -- even though my wife thinks they are annoying...
