Reply with QuoteThat is very interesting trivia, thanksI have studied BJT operating modes in reverse bias situations too but I wouldn't easily dare to use them because data sheets always lack SOA information concerning 'upside down' cases.
SMD discrete MOSFETs
IRAMS10UP60A power module
Discrete through-hole MOSFETs with heatsink
SMD discrete MOSFETS + possibility to connect own custom external power stage
Controller logic only without integrated power stage. An external power stage must be used.
Some other solution (please specify)
Some bipolar transistor trivia:
1) Small-signal bipolars make excellent 6.8V temperature compensated zener diodes. Use the collector of a PNP as the 'zener' cathode, the emitter as the anode. The base must be left unconnected.
2) Using a bipolar 'upside down' gives a Vsat in the sub-millivolt range. The emitter becomes the collector, the collector becomes the emitter. Downside is a very low Hfe and a Vce of only 6VDC. If these limitations are acceptable, in some circuits they will out-perform FETs as analog switches.
Mariss
That is very interesting trivia, thanks
Well, that is what a lab is for. You start with the mfg's datasheets, then you instrument and measure for yourself to determine unpublished device characteristics. A 200V diode mfg reasonably expects his device will be used as a 200V rectifier and his datasheet reflects that expectation. It does not include details that are of interest to you because you are using the device in an unconventional way.
While on the same subject, incoming quality control should always test at least a few parts from every shipment as if they had never been qualified for use in a design. You would be surprised by the horrors this practice has unearthed over the years.
You can intentionally induce a lower trip point at higher temperatures by locating the PNP nearer the power devices than the temco diode.
Again, modulate the offset Vf by your choice of diode forward current.
Mariss
Xerxes,
Marris has also invented IIL (I2L) logic, where transistors are used "upside down". The gain is less than one, but if you operate at near zero volt swing you can pick up some speed. (if you are interested, don't just wiki, the net seems weak in this area)
Barry
my projects:
http://www.barryfish.com
Oh Yeah...
Kudos to both Marris and Xerxes,
It's guys that bolt together hardware, hammer out code, smoke fets, and "get 'er done" that I respect. That's both of you.
Congrats on being successful entrepreneurs in todays world of should-of, could-of's and why-me's.
Thanks for getting new hardware out there to the masses, and at reasonable prices.
Barry
my projects:
http://www.barryfish.com
Mariss,
Just curiois, what kind of surprises have you had with purchased components? I think any quality manufacturer tests them to ensure them to be within specs.
--
In the next drive I'm using 100V mosfets. There are three very good candidates:
IRF540N
IRF540Z
FDP3682
The FDP3682 looks very similar to IRF540Z for the price of IRF540N. Any thoughts?
BTW, I wonder why IR titles some of their MOSFETs for specific application. Some are "automotive mosfets" and some "digital audio mosfets" but both seem all the way better than their general purpose ones. Why they want limit applications of those parts by choosing very selective name?
Hi Xerxes
Great job on your ac drive!
I am in the process of finishing my bachelors degree in electrical and electronic engineering, and in the process I am doing a final year thesis project.
This project is to design and construct a DSP controller for a 10kW Brushless DC Motor (BLDC), with the guidelines that the controller should handle 300V and 400A operation. Yes I know this is much more than 10kW! Anyway, when i was choosing power devices for my project, the MOSFETs that I (briefly) considered were said to be "Automotive" types.
The difference that I gathered are that Automotive type MOSFETs are not meant for linear applications, but rather switching as in a PWM scheme. However "Digital Audio" MOSFETs would imply the same thing if they are truly meant for digital audio applications such as Class D amplifiers, but maybe they are able to operate more in the linear region than the Automotive MOSFETs.
Just a thought I Might share, it is early (or should I say LATE) morning here anyway.
Cheers,
Andy.
Thanks Andy! I think most mosfets are meant for switching and nearly nobody uses them as analog devices (BJTs are better for that).
I think IR means class D audio amps. It is very demanding switching application because frequencies are around 500 kHz and deadtimes only few nanoseconds.
See http://www.irf.com/technical-info/appnotes/an-1070.pdf
Xerxes,
Over the years, every imaginable surprise.:-) Motorola MOSTFETs that miserably failed the datasheet reverse energy ratings, X7R ceramic SMT caps from Murata, Kemet and AVX that developed thermal stress tension fractures after reflow, 1% resistors out 20% from rated value, printed circuit boards certified as 100% electrically tested not tested at all (look for the tiny dimples left by test probes on circuit pads to see if tested), Nichicon 100VDC electrolytics that fail at 24VDC, reels of TI 74HC14s that contain occasional 74HC04s, Bourns trimpots that wouldn't trim (no wiper contact), On-Shore Technology modular 5mm terminal blocks with soft copper sockets (no Beryllium, no spring. An RoHS thing), IR's new and improved IRS2104s that latch up and turn off both MOSFETs, etc. and etc.
The FDP3682 looks to be a little too fragile for my taste. 55mJ vs IRF540Ns 700mJ single pulse reverse energy rating.
The different application ratings are for devices that have certain characteristics optimized for that application. In some respects "automotive" rated parts have to meet more stringent requirements than even mil-spec parts. "Digital Audio" devices are optimized to PWM efficiently at 200kHz or more. However the LC filtering on the output is a far more benign load than motors are.
Mariss
Wow! That's crazy
Sometimes specs are difficult to compare. For example the footnote of avalanche energy spec for IRF540N says "This is a typical value at device destruction and represents operation outside rated limits". It is also tested using bit smaller current and higher inductance than the fairchild part.
How large energy rating would be enough? If you would get 5W switching losses (exaggerated) per MOSFET at 20kHz you would be having 0.25mJ avalanches max. FDP3682 also operates safely in Id=150A Vds=100V condition for 10µs so it should be short circuit safe for long enough period.
This gives a whole new meaning for "Chip Sweeper".Originally Posted by Mariss Freimanis
Jeff
Yet another option
I would opt for discrete through hole MOSFETS + possibility to connect own custom external power stage of any power. Enough power for most applications (<500W) but enables to drive higher loads too.
Why? Through hole can be reworked by a hobbist with fewer tools and a lower skill level. Ever replace an SMD chip?
Why the external power stage? If heat sink geometry will not fit your board, you can still make it work, and SMD wires to the outside world are uncommon.
Last edited by cnc2; 03-17-2008 at 11:31 PM.
any body have same project
my servo develop kit
my new servo control board
Hi
Dear tivoidethuong; I am implementing to control of an ac 3ph servo motor with FOC. I am going to buy a Dspic motion control development card.
What do you think about this subject. Can I make similar development card?
Can you send to me your development board pcb and circuits.
Thanks
my schematic not yet testing,
i think you can try dsPICDEM-MC1L 3-Phase Low Voltage Power Module to control servomotor
yes. tivoidethuong I will try Dspicdem mc-lv.
Are there usefull knowledges or files in your documents about ac 3 ph servo motor controlling boards or etc..?
Can you post to me?
Thanks for your interest
to sakarya83
i think you can try with MC1 dspic develop kit for motor control of microchip
http://www.microchip.com/stellent/id...&part=DM300020
b.r
Posting Permissions
