| Electronics > Projects, Designs, and Technical Stuff |
| 3-phase inverter using Infineon CIPOS module? |
| << < (2/2) |
| Le_Bassiste:
--- Quote from: Yansi on August 29, 2019, 05:45:31 pm ---you will be surprised, when you find there are even whole microcontrollers with integrated power transistors in one package, or even microcontrollers with integrated high voltage FET/IGBT drivers in the package. --- End quote --- mind to share a source? |
| mrflibble:
--- Quote from: Yansi on August 29, 2019, 05:45:31 pm ---you will be surprised, when you find there are even whole microcontrollers with integrated power transistors in one package, or even microcontrollers with integrated high voltage FET/IGBT drivers in the package. --- End quote --- Yeah, I noticed Infineon had several arm cores, but couldn't really find what I was looking for. Then again, I did not spend hours with parametric search, so entirely possible I overlooked something. Do you have a suggestion for "jelly beans" components in this segment? You can assume I am not looking for esoteric performance. Three phase motors, 1-3 kW range, rampup, ramp down, control speed, have things not go *boom*. --- Quote from: T3sl4co1l on August 29, 2019, 05:47:14 pm ---Switching speed. Overcurrent may be internal, or it may be up to you. Depends how integrated of a module you're looking at. --- End quote --- Slower switching speed would be acceptable, within reason of course. I'm not looking for bleeding edge performance, mostly I'm looking for convenience in the form of lots of small subproblems already being solved. Ideally I'd like something that handles: * synchronous rectification, or basically single phase or 3 phase in, DC bus voltage out. 1 or 3 phase depending on use case. * variable speed. * overcurrent protection * dynamic braking * resistive braking (using external dump resistor) * easy MCU control, primarily to be able to change speed. Think lathe variable spindle speed based on work piece diameter, that sort of thing. That's about it really. Obviously there are the "Oh it would be nice if..." items, but lets keep it simple. In that regard, if it supports only 1 braking mode then that would still be doable. I realize that it would probably be cheaper to buy 3 chinese inverters, but where's the fun in that. ;D |
| Yansi:
For example STMicro has a lot of such MCUs: probably some of the Powerstep, STSpin32 series, ... etc But for a 1-3kW VFD where circuit and layout simplicity is concerned, I'd opt for any of the standard off-the-shelf IPM (such as STMicro SLLIMM module) and any suitable mcu (probably STM32F3 series, which has more than enough computational grunt and very good analog peripherals) located directly on the mains side (only isolating the necessary control interfaces - isolating digital IOs is way cheaper and simpler, than many analog signals from the power stage). Please be aware that most IPMs with integrated boot-strapped IGBT drivers are mostly 500-700V only, so suitable only for a rectified single phase DClink supply below 400V. When large power outputs (more than 500W) are required (from a single phase mains), using an active power factor correction will become more than necessary (in EU probably even mandatory, above 100W or so). Also it is interesting to note, that by using discrete power transistors and driver ICs, one may typically achieve more impressive results for less price (but not in more compact form factor, IPMs take little space). Three phase rectified mains (DClink voltages well above 500V) require typically 1200V rated IPMs, which are quite more costly and usually require external signal isolation for the high side. Resulting circuitry is not that simple. //EDIT: Synchronous rectification on a VFD input is quite useless, considering the voltages involved. You can not spare much power dissipation there, not for cheap for very sure! |
| T3sl4co1l:
Note that you get the effect of synchronous rectification from the co-pack diodes, but don't count on saving diode drops, not at high voltages. You need very expensive MOSFETs for that (~$20/ea at this power level). Tim |
| mrflibble:
Good points about the sense and nonsense of synchronous rectification. A bit of envelopeless back of the envelope calculations showed that, indeed, for my use case synchronous rectification makes no economical sense. So regular diode drop it is! After completing some background thought processes I suspect that the best way to get into this would be to get a dev board. The idea being that a known-to-work situation should get me up to speed quicker than doing everything from scratch. There's only so much time in the day... So I guess that results in a new question.... Can anyone recommend a good devkit for this? |
| Navigation |
| Message Index |
| Previous page |