Quote:
Originally Posted by Astro
Do any of the other prototypers/beta testers need voltage boost? If not then developing the controller with 4 IGBT capability may be the way to go and save the 3 IGBT controller boards for people who don't need the voltage boost.
|
I have not sourced an AC motor as yet. I have a couple of 5 HP, 230VAC motors and a monster 30 HP 575VAC motor that I can use to test the controller. Nothing that I'd put into a car. Shipping the 3 pair IGBT controller board works for me. I'll test with the motors that I have when I receive the controller.
Paul - I'm wondering how much headroom is required on the max voltage for the IGBTs for a safety margin. Have you scoped the AC controller on a motor yet? (Or did I miss that?) Do you see any reflected wave?
Some boring background - In my day job we have problems with reflected waves on AC controller outputs and motor terminals, especially on motor leads over 200 feet. We don't really have any 'short' cable runs that I have experience with. We have installed load reactors, dv/dt filters, motor terminators, and in one memorable case a full sine wave filter (that cost almost as much as the controller). We have not lost IGBTs, as far as I know, but we regularly fail the insulation in the first wound coil in the AC motor's stator. Which phase is not predictable. The motor rewinder says something about 'the dv/dt is too high'. After the motor is rewound, it usually lasts longer, from 4000 to 10000 operating hours ... some as high as 25000 hours. The motor rewinder adds an extra 2 layers of insulating paper to the first coil of each phase, and one extra layer to the next 3 coils. Does that make any sense?
We are using a variety of AC controller vendors, and different vintages - Rockwell Powerflex, 1336 Plus; Siemens Master CUVC; Toshiba G3 and G7; Mitsubishi A560 and A760. The IGBTs on the drives that do fail (I take them apart to see what failed) appear to be rated between 1600V and 2200V. But they are driving 575V motors, and their DC bus voltages surge to 1000VDC, and we definitely have reflected waves in the 180% range.