Added:
4 connectors for thermistors to FET gates for precharge and discharge, and power resistors
Comparators, and an optoisolator for thermistor circuit
2 FETs on the isolating DC-DC to turn off the isolated side through firmware (normally on/closed)
2 connectors and circuitry for fans to cool motor controllers
LED on discharge circuit
Additional inputs to MCU for data logging
Changed:
Power resistor values
Additional thermistor output to shut off precharge (aside from precharge complete)
Components added to library:
RES 1K OHM 15W 1% TO126
RES 5K OHM 25W 1% TO220 (Since discharge FET overheated last comp, this is a pretty conservative resistance value. Can be lower.)
IC MUX 1 Circuit IC Switch 2:1 3Ohm SC-70-6 (not used)
Footprints added to library:
RES TO-126-2
RES TO-220-2
IC MUX 1 Circuit IC Switch 2:1 3Ohm SC-70-6 (not used)
New functionality:
Precharge will stop if any of the precharge/discharge resistors or FETs overheat (around 90 C). Firmware will get a signal of which part is overheating.
Ability to shut of isolated DC/DC, turning off isolated side of board. Will prevent precharge and allow discharge without shutting off entire board.
***
Upon further review, this was already possible through PA9, which enabled vbat_precharge_sw to the isolated DC/DC.
The circuit to shut off isolated DC/DC added isn't entirely useless. The PA9 enable pin enables vbat_precharge_sw. I used that to power a lot of the non-isolated stuff like overtemp comparators and fans. So we still have functionality in keeping those on to log with the MCU.
I think it also powers the contactor sense so it'd be nice to have that on whenever the board is on.
After further discussion, it is decided that two load switches would be the better alternative.
***
Dialogue:
Switch all ultrafit connectors to microfit.
Could you recheck the resistor values for a 95% precharge - or was that not the number we are going for?
Check precharge resistors for a fast precharge time.
The precharge is controlled through the precharge_sw load switch, from the PA9_Precharge_EN pin on the controller board as well.
Why hold Reset of the SR Latch LOW for 8.4ms?
Why is R50 1/4W?
Why 2.2uF 100V capacitors?
The paralleled precharge FETs should have resistors on their gates individually, not paralleled. This can cause oscillations, but is usually not an issue unless switching them on/off rapidly in a motor controller of similar.
Why is R17 a 1% resistor when the others in those voltage dividers are 0.1%?
I believe our DC-DC converter is set to 13.5V. Is there another isolated DC-DC converter that has a higher voltage rating? The XPPower one says 13.2V max. This is probably not a huge issue though.Currently, any thermistor overtemp will turn off the precharge and notify the controller board (but will not activate the discharge) This can be left up to firmware. We need to decide what to do if the discharge is overheating - ideas? Do we leave the high voltage on the motor controller, or do we keep discharge active and hope it does not burn?
___
The 95% precharge should work out. It's unchanged from last rev, but I'll check it when I get back. (This is about the voltage dividers going to the comparator right?)
The holding reset on the SR latch I believe is to reset the SR condition since each time the board is turned on since it's in a race condition. Taiping did mention I could replace it with 2 FETs or something but I'll have to look into it.
R50 was 1/4W bc i was just looking for resistor values to fit the divider. I don't think the power should matter. I believe the same is true for R17 as well.
For the rest, those were unchanged from the previous rev but I'll take a look.We can leave the thermistor overtemp up to firmware but just not populating the opto-isolator that sends that signal to the precharge FETs. The MCU should be getting signals when a comparator senses overtemping.For the discharge overheat, i believe the problem was that the discharge FET got wet, so hopefully that doesnt happen again. If it overheats normally, we can't really just stop discharge by turning on the board but leaving the isolated side off so it doesn't go into precharge.