I am considering stepping up to 14s. I am using a d100s vesc. Is the switch with vesc safe enough, or should I go with external switch? Will a switch rated to 12s survive? Should I step back to xt90 key? What other components need changing with 14s? Besides of course 14s bms and charger.
I forget the math but the resistor in an xt90s is small and there’s still a large inrush current hitting the esc capacitors. You won’t get a spark but it’s rough on the esc and the plugs break quickly as well. I had QS8 and QS10 plugs made with bigger resistors for higher voltage and sell them cheap. 40$ for two pairs qs8 w shipping. Message me if internet
1 Like
Just use the supplied switch.
1 Like
You don’t need to use an antispark with the D100s because it has a built in power switch to activate sleep mode on the ESC
However, there will be a slight spark when you plug it in the first time due to the caps charging up. Easiest way to avoid this is just making the discharge lead of the battery and XT-90s. Only plugging it in one time is not going to wear out the resistor and will prevent any damage from happening to either XT90
At the same time, you don’t really need to worry that much about plugging in the ESC without an antispark a single time - it may cause a tiny bit of soot to get on the contacts of the plug, but it’s not going to be a lot and likely won’t affect you.
I put XT90s on my batteries anyway though.
Thats the kind of thing I was looking for. Secret knowledge. Not just the blurb. Tricks and tips. Include also, how 14s is an advantage.
1 Like
I have heard others opt for external antispark switch, as there is some disadvantage to using the sleep inducing one. Please elaborate maximally. All the choices, all the angles.
I aint one of those others fam, I have shared my bit haha
I like using a loopkey with the anti spark plug and simple electronics and reliable on/off switch.
More voltage is a faster motor or gear it down for more torque. More power either way.
1 Like
Skyart built me a 14s12p battery,it’s fantastic. 62 lb board 230 lb rider, typical ride I gain/ loose 2400’ of elevation 26 mph average speed yields 45-50 miles of range
1 Like
Yeah, I am building a 12s7p but the hummie deck has JUST enough room to put another 2 p groups.
I am going to run 170kv motors, with 4.6 gear drives. I think I will try the 12s for a while. But the 14s 7p would be sorta like a 12s8 or maybe 9p. Range wise.
is 14s really a problem for an xt90s as an antispark? it’s only 16% higher voltage than 12s.
1 Like
12 is already too high for an xt90s. maybe someone else will do the math or I’ll find it. calculating the typical capacitance in the esc, the voltage, the resistance of the xt90s, time, and getting the inrush current…it’s not good for the capacitors.
Good question.
For sure I have blown XT90s on 20s after one or two connects, and I’ve found a lot of posts where others have the same issue.
12s doesn’t seem to kill the XT90s.
Hummie special ordered the QS8s with 15 ohm resistors instead of the 5.6 ohm that the stock QS8 and the XT90s come with, and they work fine on 20s.
Below is a simple circuit to show a battery, the pre-charge resistor and the ESC’s electrolytic capacitors. It’s not a perfect representation as there are a lot of parasitics and ceramic caps not included in this calc, but since the resistor is the bottleneck, that shouldn’t change the results much.
Before looking into the simulated power using an anti-spark like the XT90s, here is a battery to ESC connection with no inrush limiting.
Current and power levels are very high. Like 2,000A for a few microseconds high.
Zooming in shows just how fast it charges the caps. This includes the ESR of the caps, the connection resistance, wire resistance, wire inductance and a 20s4p P42A pack’s internal resistance.
A resistor can be put in series to limit that current and slow down the filling of the capacitors. This is what the XT90s does. The problem with this method is that the resistor has to dissipate a lot of power and is going to get very hot. The problem only gets worse with higher battery voltage.
Legend:
Green line: Battery voltage
Blue line: Capacitor voltage
Red Line: Current across the resistor
Dark Green: Watts lost as heat across the resistor
XT90s at 12s (this is OK)
About 500w peak is flowing through that probably 1w rated resistor for a short time.
XT90s at 14s
About 630w
XT90s at 20s (this will fail)
About 1250w
Hummie QS8 with 15ohm resistor at 14s
About 230w. The time it takes to fill is quite a bit longer though, so the resistor is running over it’s rating for longer.
Hummie QS8 with 15ohm resistor at 20s
About 450w, very similar to the 12s XT90s with a stock resistor in it.
Zooming out a bit - it takes about 50ms instead of 20ms to get down to 0.5A. Probably still fast enough that pushing it in at a normal speed would still allow the resistor to work.
Will the XT90s survive 14s? Probably. Do those mosfet based anti-spark switches address this inrush problem? No. Most are a very simple switch… and probably why they aren’t super reliable. The only one that did manage inrush, that I know of at least, was the Stormcore. Even that system occasionally failed for some reason or another.
9 Likes
Great info and pics.
Won’t the inrush current also be greatly determined by the capacitance in the esc? The more capacitance (bigger the esc) the greater the inrush?
I also have qs10 connectors with 80ohm resistors for the highest voltage (126v) vescs being made these days.
1 Like
The higher the capacitance, the longer the inrush.
The lower the cap ESR, the higher the peak inrush current (limited by the pack DC IR).
3 Likes
This is true, but the second part is only important when not using a precharge resistor.
When using a precharge resistor, the capacitor and battery pack resistance values are so small by comparison (milliohms vs ohms) that they are mostly insignificant.
If the resistor value gets too high and the ESC total capacitance is large enough, it may not charge the caps enough before it hits the main connection terminals.
That could all be measured with a scope to see how long it takes to push a plug together at normal speed, then simulated to see what the optimal resistor value would be… but as long as it’s given a bit of time to charge (~0.5s?) before being fully plugged together, then it probably doesn’t matter.
3 Likes
At >= 20S the AS150 got blown again and again with my Psychoframe (that uses 2x genuine Trampa VESC 100/250) and I replaced them by Jeti AFC 8mm Anti-Spark Connectors because AS150 gets blown at 20S.
I have used them again with my new 20S build :
I like them more than XT-S because you don’t have a key and it saves room.
2 Likes
Agreed, good point.
I was just referring to the actions of caps+pack alone as there’s no real inrush when using the resistor.
1 Like