|Jan 24, 2013, 08:32 PM|
Designing a FET board for a TURNIGY 100A HV 5-12S
I have a burnt out ESC that I am changing the present FETs to higher amperage FETs
I will be using 12ea IRF3205 FETs on a separate board and a large heat sink much like mjsas had done. (http://www.rcgroups.com/forums/showp...postcount=104?).
Most of the components on this board has been identified except the processor and the FET drivers. The drivers had conformal coating on them and scraping it off most of the identification came off. The ID that I can see is that it is a IRF P/N 5200 7NP or 714P. What ever it is I am trying to get a cut sheet to see If these can drive the larger IRF3205 also need the pinout.
There is no real application for this and is an experiment.
|Jan 25, 2013, 02:46 AM|
Removed controller from motherboard
The controller has 9 outputs plus V+ and V- from the batteries.
The outputs are:
1. VS phase A
2. HO phase A
3. LO phase A
4. HO phase B
5. VS phase B
6. LO phase B
7. HO phase C
8 LO phase C
9.VS phase C
After tracing everything out,the driver has a similar pin out as the IR2308. Still do not know what the original driver is.
Original driver also had a 10k resistor across pins Hin and ground, no doubt to pull the single low coming from the processor.
|Jan 25, 2013, 07:33 AM|
Australia, WA, Perth
Joined May 2007
Driver is an ISL 6700.
Yes, there should be a flyback diode up there! Not sure that caused the burnout - I can check these things because one of these ESCs is in my scrap bin too
The micro is strange. Pin 3 goes to GND, it runs on 12V... The ident has been scratched off mine too.
Are the drivers up to the task of switching those FETs?
|Feb 01, 2013, 01:32 AM|
Thanks for the info.
These chips can drive 1.25 amps and according to the FET data sheet they are designed for low on resistance. Exactly what this resistance is was not stated.
The driver chip traces goes back through a circuit as well as to the drain of the FET. This circuit has a 5.1 volt zener diode for a single going back into the processor. I wonder what the processor uses this for?
|Feb 01, 2013, 03:36 AM|
Australia, WA, Perth
Joined May 2007
Where's VS come from exactly?
The zener looks like it's just there to clamp the signal to tolerable levels for the micro. But why it needs it, depends on what VS is. Sorry if I missed something obvious!
EDIT: I missed the bleeding obvious!
I just had a poke around mine, and 2 of the 3 VSs are fed into the micro, through around 70k. I get hits on 4 pins, ranging from 66 to 80k from board connector pin to micro. I think that one that isn't connected is why it's in my parts bin It reads open circuit...
I would expect four connections: Line voltage for phase A B and C, and a false centre (average of the three, through a Y of resistors). The false centre isn't strictly needed, so may not be present on this particular ESC.
How much of the ESC schematic have you drawn up?
|Feb 01, 2013, 02:15 PM|
Joined Nov 2007
The schematic for the IRF3205 FET board is very generic and looks to be the same as your burnt up board. (I posted that schematic a few years ago.) You should be able to use it as a direct replacement for your board.
The burnt diode in your picture is very needed, it is part of the bootstrap circuit to generate the gate voltage for one set of high side drivers. Without the diode one set of high side drivers will never turn on and while the motor would run it would be running on less then three phases. Usually those diodes never burn out so you may have other problems.
The FET board uses two IFR3205 FETs which are rated at 110 amps but are really 75 amp FETs so the result should be 1.7 * 75 * 2 amps maximum (about 200 amp max). I would keep the load to about 100 amps continuous. The IRF3205 is rated at 390 amps pulse so two in parallel can pulse nearly 800 amps. Actually there might be two phases driving so the pulse could be way over 1000 amps.
The wiring has to be robust and able to carry several hundred amps with very little voltage drop, also the capacitors have to be big enough.
Note: All FET specs show current ratings for the chip, not the package. In this case the 110 amps is for the chip and 75 amps for the package. The current and power rating of the body diodes is also very important as they can sometimes carry as much current as the FETs. The spec shows a 1.3 volt forward voltage drop for the diode so lots of heat comes from these diodes.
The irf3205 FET is a cheap high voltage FET, it has sort of low input capacitance so is easy to drive.
|Feb 04, 2013, 03:10 AM|
New Zealand, Wellington, Porirua
Joined Jan 2013
The missing diode looks like its part of the charge pump the FET driver controls (along with the capacitor) to provide the higher than source pin voltage for the high side FET. Without it, it may not be able to boost the voltage high enough to fully turn the FET on, leading to increased resistance and power dissipation in the FET when it is on.
If you're going to shove on much larger FET's then I'd check their datasheet with the datasheet of the original FET's to make sure the gate capacitance is within spec of the FET driver. If the gate capacitance is too high, the FET will take longer to turn on, increasing power dissipation.
There should be a forumal to see how long the FET's will take to turn on based on the capabilities of the driver (it will effectively be modelled as a voltage source with a series resistance), the gate resistor and the gate capacitance.
|Thread||Thread Starter||Forum||Replies||Last Post|
|Sold TURNIGY Sentilon 100A HV 5-12S Speed Controller (Version 4)||dizzane||Aircraft - Electric - Power Systems (FS/W)||1||Apr 15, 2010 11:09 PM|
|Wanted Turnigy 100A HV 5-12S ESC||fly_boy99||Aircraft - Electric - Power Systems (FS/W)||0||Oct 05, 2009 05:26 PM|
|Wanted Turnigy 100A HV 5-12S ESC||fly_boy99||Aircraft - Electric - Power Systems (FS/W)||1||Sep 04, 2009 07:56 PM|
|For Sale TURNIGY Sentilon 100A HV 5-12S For sale!!||jetfan20||Aircraft - Electric - Power Systems (FS/W)||5||Jul 26, 2009 07:27 AM|
|Sold TURNIGY Sentilon 100A HV 5-12S Speed Controller||Gnarly1||Aircraft - Electric - Power Systems (FS/W)||3||May 24, 2009 06:45 PM|