Two HW4.12 ESCs is more like 4mF. (680μF x 3 x 2)
I mean a huge capacitive load like 22mF
Fixed it. I haven’t made one yet. But it would be interesting to see.
As @Kug3lis says, any quantity of energy required to charge capacitive loads needs to pass through the mosfets. So generally speaking you would need to see how much avalanche energy a single mosfet is capable of handling; then you would calculate how much avalanche energy your capacitive load generates; this way you’d know how many mosfets are needed to cope with your particular inrush current.
1 Like
Why not go for a similar FET as the fatboy one? Looks like the single but powerful fet works quite well.
Yes its possible as its SOA is pretty big 
But don’t forget for high constant (>50A) currents you need cooling. It has higher internal resistance compared to smaller friends 
The thing is IF this layout is capable of coping with the inrush currents, then the big boy mosfet is not needed.
Since the chip that I selected is specifically designed to manage inrush currents, I believe that the effect @Kug3lis is predicting will not happen as long as you follow the manufacturer reccomendations (which is adding the ramping up cap C1). Real world testing will have the final word on this. If the six mos I choose fail, then an industrial mos should be choosen OR a buck converter must be added to replicate the precharge circuit of diebiems.
Now the comparison in terms of cost and heat generated: SOA mosfet: I believe is around 15eu, not sure though 6 NTMFS5C628NLT1G are about 6 eu at quantity 1000
6 NTMFS5C628NLT1G have 0.4mohm rdson (combined) so when 100amps are travelling through them, only 4 watts are dissipated
Then it’s fully opened. When it’s partially (charging gate capacitance it works in linear mode V*I) so 50V * 100A = 50000W) check
As this works as a high side switch gate voltage is Vds + Vgate so around 60-65V depends on IC Vg voltage.
So while Cgate chargers resistance will be huge inside mosfet. Look at the graph (those voltages are Vds + as its working in high side switch) so at 52V (Vds 50V) its resistance is 9mΩ so at like Vgate 10-20V it will be like 20-50mΩ
Even our big mosfets was blowing with long rising time so we have to find golden spot for enough current to not kill mosfet and enough time to not burnt it.
1 Like
I believe @Maxid wanted a comparison at normal operating condition (gate at 12v)
Definitely agree. Real world testing is a must to understand if the design is capable or not to handle specs.
62V 
10char
lt4356 controls n-side mosfets and uses 12v charge pump at the gate. Edit: 12v respect to vcc.
High side switching mosfet gate voltage reference is Vds not 0V in case of low side that’s why you need charge pump to lift up your source voltage to Vds + Vgate voltage 
Now let’s go back to the high-side drive. Let’s say you apply a voltage of 12V (with reference to ground) to the MOSFET gate. However, when the MOSFET is on, voltage at source is equal to +V. Let’s assume +V is +15V. Now the problem is +12V gate drive (with reference to ground) will not keep the MOSFET on. When the MOSFET is on, the MOSFET source will be at a potential of +15V. To be on, the MOSFET must have +8V VGS minimum. So, if source is at +15V, the voltage at the gate with respect to ground must be at least +23V. If source was at +300V, for example, gate drive would require a minimum of +308V with respect to ground. This is if the gate drive is referenced to ground. If you have a separate isolated power supply whose ground and the ground of the MOSFET-based circuit are isolated, then you can use that to drive the MOSFET as well.
1 Like
Thanks for explaining. Do you know if there are reliability differences between low side and high side switching?
Don’t really remember much difference, just as I have worked with auto/moto industry high side is preferred way because of ground loops/cases and etc. Low-side is the easiest to make as it doesn’t need charge pumps and etc. But if you have somewhere ground available is bit pain in the ass because you have to make sure your electronics will not get ground from anywhere else not through communications ports like CAN and etc
Good discusion happening here:
For an isolated circuit, no there is no great difference between high and low-side switching. For higher load currents, low side semiconductor switches (for example NPN transistors and N-channel MOSFETs) are often less lossy than their high-side equivalents, and so are preferred.
However, if the circuit is connected to external devices with their own power connections, this becomes blurred. If these external devices provide a connection to the same ground reference as the power supply to the circuit and you switch this in and out then the external devices will provide an alternate route to ground, your switching will be ineffective and you may end up damaging something not rated for the appropriate current along the way.
Similarly, if the external devices provide a V+ supply which is referenced to the same ground as the supply that you are switching, you can end up back-powering the positive voltage rail via the externally powered devices, again with undesirable results.
2 Likes
Getting back to the precharge discussion, I was thinking that integrating this switch directly on the vesc would have this advantage: we could use the soft-swicth to also enable the drv (and put some delay to enable lt4356). This way we could use drv buck converter to do the precharge of the caps, eliminating the inrush problem when lt4356 wakes up. What do you think?
I was more concerned with the total footprint than anything else. 6 FETs will need much more space than a single fatboy one. It’s also the reason why I don’t like the Vedder ones - they are just so huge for what they do. Until @kuglis nobody used the vertical space available (I am obviously talking about use cases where you would not need the aluminium block (in for example low power commuter boards))
They are 5 by 6 mm. So 3cm long when they are placed in line. (max 3.5cm if you space them properly)
Update. I have added the precharge function with a cheap and reliable 555 timer. I also made a first pcb draft; the soft switch function has been left out temporarely to achieve a compact design and a single side smt assembly. The whole thing will be open source after testing prototypes. I may also incorporate it into the escalate pcb.
@Maxid I layed out the design in a way that allows the user to cut off one or more mosfets without altering the functionality, like you would do with an LED strip.
7 Likes
I would reccomend independent 10-20 ohm gate resistors for each parallel FET.
2 Likes
I assumed that few fets with low gate charge would be equivalent to having a big fet; thus I only placed one resistor. A quick research revealed that I was wrong since the resistor also suppresses the gate oscillation if you have high inductance in the gate trace. Thanks for the tip.
1 Like
Well its all good if you have switching load, here we have a constant on/off so there will be not much oscillation as it will be always on
I thought the same, especially with the slow-ish turn on of the gate. But I had some really strange behavior going on with my switch which seemed to disappear when I added individual gate resistors close to each FET and things quit exploding. It isn’t much cost or real-estate anyway.