[SOLVED] FS Anti-Spark is trolling me (turning back on magically)

How-Tos & Esk8 mechanics Esk8 Mechanics
#1

Hey everyone I have a super strange problem. My Flipsky Smart anti spark (the one with push to start) is turning back on a few seconds after I turnend it off.

I have measured the voltage at the Vesc side and after I turn it off there is still 3,5v left that’s slowly discharging 0,1v per second. After some seconds the anti spark turns back on. My theory is that the push to start feature detects the changing voltage and turns back on.

I need something that discharges the capacitors quickly after I turn it off.
Right now the 5V devices drain it quickly but when they turn off at 3,5v its to slow.

If you can bear my terrible englisch you can watch my detailed explanation video

@Gamer43 you designed it maybe you know what’s going on.

5 Likes
#2

Yes this is a known issue, if you have enough capacitance on the VESC side this will happen. I assume you have 4 4.12 FSESCs hooked up to it which would cause this.
Edit: I see its only 2 hooked up and it still happens, I’ve ran it with 2 and it didn’t happen but I can see it being possible… hopefully Gamer43 can help.

2 Likes
#3

I have 2x 4.12 and 2x 4.20 but only the two 4.12 are enough to cause the Problem. The 4.12 have 2040 uF each so quite a bit more than the usual Vesc. I discused removing some here.

Ideally I would like to keep them and install another solution.

My Total capacity is under the maximum 6000uF stated on the FlipSky website.

I have a lot of components so if there is a known solution I could probably built it.

#4

I guess if everything fails I could make a circuit that shorts the capacitors with a small resistor for 10 seconds after 5v goes out.

3 Likes
#5

You might be able to have the resistor there all the time, similar to what they do to discharge some caps for safety with certain circuits.

2 Likes
#6

what value would be appropriate in your opinion? The standard ones can dissipate like 0.5W so at minimum it should be 5000Ohm when its full at 50.4v

I also don’t want to disrupt the push to start behavior.

#7

So 6000uF is the maximum(with some headroom) that if exceed the MOSFETs will fail.

The sensitivity for auto-turn on can vary, especially if you get caps whose capacitance tolerance is on the upper +20%.

If you do not wish to remove a capacitor then you will need a separate bleed circuit to discharge the caps in a timely matter.

As shown by the 4.20, you really do not need that much capacitance.

I apologize for the inconvenience.

6 Likes
#8

Does the anti spark actively discharge the capacitors or does it rely on self discharge anyway?

#9

Relies on self discharge

I would’ve needed costly power rated components to discharge the caps which would’ve made the thing 2x bigger and cost 2x as much.

#10

I just started to put random resistors parallel with the vesc and it seems like a 20k ohm resistor removes the problem. Ist not the most elegant solution but it’s okay if I don’t want do design a whole new circuit.

5 Likes
#11

I don’t quite understand why it would be high power? The receiver and stuff drain the voltage very fast to under 5V and for the rest you only really need a 1A FET with a resistor to discharge the rest. Or am I missing something?

#12

Sorry, I don’t have a good recommendation as it can depend on timing requirements in the AS module.

I would make the timing the priority though, how fast it needs to discharge, and then select a resistance and power rating (for the resistor) that can do the discharge at that speed. Or reduce the capacitance.

Are you sure you have a 20k resistor? That will take a couple of minutes to discharge a couple thousand uF: Capacitor charge and discharge calculator | MustCalculate

#13

I just randomly tested 10k 20k and 50k. 10k and 20k helped and 50k didn’t. I guess im right on the edge of discharging fast enough. Ideally I want the passive resistor to be as big as possible to not waste unnecessary power.

Yes I have measured it. I guess it only needs to help a bit from 3,5V to 2V then its in the safe zone.

Here is a video of me testing it.

now its almost 5:00 in the morning. This stupid stuff can keep you up for days.

4 Likes
#14

1A at 5V is 5W.

5W will blow up most SMD resistors.

#15

But wouldn’t it only be a super short burst? There is not much energy saved when it turns on at like 4v. But even 0.25A would probably already help enough. The default passive discharge is just very slow.

#16

Short bursts of energy is what kills MOSFETS.

The issue is the component needs to absorb all the energy stored. There is not enough time for it to dissipate into the environment.

1 Like
#17

Agreed.
IMO what you said about choosing a resistor that is a bit lower than what seems to work is the right thing to do…perhaps 10k?

If you will be leaving the resistor connected to the VESC input then at 50V it would be about 1/4W of power being dissipated in the resistor so perhaps two 1W 20k resistors in parallel to spread the heat?

1 Like
#18

After turning it off, although the battery is not powered, there is still power on the motherboard,

#19

But the caps aren’t kept charged after it’s powered down, are they?

#20

The whole vesc is disconnected when it’s powered down. So the caps are not powered by anything. There may be a tiny leak current but the self discharge is faster.

2 Likes