Anti spark question

[ak136]

Hello,

I am thinking about ordering xt90 anti spark plugs although I am not sure do I really need them

I want my bms and controller to last as much as possible

When I am turning my bike on I first plug in the battery connector then press the switch on the battery
I am thinking is this still causing a large current spike? I only have a cheapo multimeter so I dont think I would be able to catch such fast transients

 

[gromike]

Sounds like you don't need anti-spark connectors. If you plug the battery connectors together and you get a spark, then maybe you do. As you have a battery switch, I don't think you need an anti-spark connector. Size the connector for your amperage,
Though I must admit that I recently bought a bag of XT-90 anti-spark connectors. Gotta plan for the future, I guess.

 

[ak136]

Sounds like you don't need anti-spark connectors. If you plug the battery connectors together and you get a spark, then maybe you do. As you have a battery switch, I don't think you need an anti-spark connector. Size the connector for your amperage,
Though I must admit that I recently bought a bag of XT-90 anti-spark connectors. Gotta plan for the future, I guess.

It would be great if someone had an osciloscope or something like that to measure the current when powering via switch on the battery.

Because if there is a large current spike then still you will end up having damage except now not on the battery connector rather on the bms mosfts and controller caps

 

[Chalo]

It would be great if someone had an osciloscope or something like that to measure the current when powering via switch on the battery.

Because if there is a large current spike then still you will end up having damage except now not on the battery connector rather on the bms mosfts and controller caps

Because the spark is a feature of pretty much all controllers, I think the engineers who designed them aren't as concerned about it as you are. My advice is keep an eye on the condition of the plug, and otherwise don't worry about it. Maybe switch to Anderson SB50 connectors, which are also designed to tolerate the spark.

 

[ak136]

Because the spark is a feature of pretty much all controllers, I think the engineers who designed them aren't as concerned about it as you are. My advice is keep an eye on the condition of the plug, and otherwise don't worry about it. Maybe switch to Anderson SB50 connectors, which are also designed to tolerate the spark.

What about engineers being forced to design parts for planned obsolescence? I wouldnt be suprised if they knew the exact number of high current precharge spikes in order to blow the capacitors inside the controller

And to add last week I had a controller that wouldnt turn on, after opening one of the capacitors blew inside, I replaced it with a new one and now it works normally

 

[Chalo]

What about engineers being forced to design parts for planned obsolescence?

All of them? Seems far-fetched to me.

 

[E-HP]

It would be great if someone had an osciloscope or something like that to measure the current when powering via switch on the battery.

Why do you need to know that? Just take control of the situation and make a pre charge circuit with a pushbutton and resistor. Size the resistor to whatever makes you feel better. I agree with Chalo. Plus your current procedure doesn’t require one, as bromike pointed out.

 

[ak136]

Why do you need to know that? Just take control of the situation and make a pre charge circuit with a pushbutton and resistor. Size the resistor to whatever makes you feel better. I agree with Chalo. Plus your current procedure doesn’t require one, as bromike pointed out.

I would like to know that to prevent future unecessary damage

Taking control and building my own precharge circuit would be unnecessary as an anti spark plug would be far simpler

 

[ak136]

All of them? Seems far-fetched to me.

Not all but as I am mostly working with kt controllers I think that might be the case for them

 

[E-HP]

I would like to know that to prevent future unecessary damage

Taking control and building my own precharge circuit would be unnecessary as an anti spark plug would be far simpler

The anti spark is ONLY protecting the connector contacts.

 

[docw009]

Back in 2015, I was worried about the spark on my XT60's. I bought the anti-spark XT90 and I also made this anti park cirucit.


The switch was pretty crude. I just took a household 110V SPST switch and soldered a resistor across it, plus Xt60's. After a year, I set it aside, as I found it unwieldy to use. I recall Chalo telling me it wasn't a problem as my batteries had on/off switches. I quit using it. I don't use it on my batteries w/o switches either,

It's not a bad idea for protection, but I;ve not seen a controller fail, except for water ingress.

 

[E-HP]

Back in 2015, I was worried about the spark on my XT60's. I bought the anti-spark XT90 and I also made this anti park cirucit.
View attachment 364680

The switch was pretty crude. I just took a household 110V SPST switch and soldered a resistor across it, plus Xt60's. After a year, I set it aside, as I found it unwieldy to use. I recall Chalo telling me it wasn't a problem as my batteries had on/off switches. I quit using it. I don't use it on my batteries w/o switches either,

It's not a bad idea for protection, but I;ve not seen a controller fail, except for water ingress.

They are a pain. I'm currently using a 100W 60 ohm resistor in my antispark. I don't think it offers anything more than the 40A AC breaker "switch" I had been using with thousands of actuations, and the tear down showed the contacts had no damage. I'm going back to that, since it's simple, less to go wrong, and doesn't require a two step process like the precharge circuit or the XT90 antisparks. Plus, the XT90 antispark dies after a few uses at 72V, so becomes a normal XT90.

 

[amberwolf]

It would be great if someone had an osciloscope or something like that to measure the current when powering via switch on the battery.

You normally use an ammeter, wattmeter, multimeter, or other current-measuring device with a peak hold feature for that, not an oscilloscope.

The latter is used to see the waveform of different things, so you could use it to see exactly what a current spike looked like, but there is not much value in that for this situation.

 

[DanGT86]

My understanding is that the mosfets in things like BMS are not damaged by inrush due to lack of precharge. I had recently wondered about the same thing. I assume because their entire purpose is high speed switching of large currents that there is effectively no spark jumping a gap.

I believe the main function of precharge is to protect mechanical switches and connectors from spark damage.

Anti-spark XT90s require that you carefully make the contact before fully seating the connector. Its really easy to go to fast and I believe I have heard a spark when that happens. It would be a huge pain to gently hold it in place while you turn the battery on before fully seating the connector. XT90s also take a lot of force to connect and disconnect and I find them super annoying to connect and disconnect every ride. As mentioned above I don't think they are recommended for high voltage like 72v packs.

I have standardized all of my controllers and batteries with 2 power connectors. One is the main XT90 and the other is the precharge which is smaller wires with a 50-100 ohm resistor and anderson connectors. I connect the anderson first for precharge then connect the XT90 for main power. If you must run precharge on your setup I would recommend something like that instead of trying to hold an XT90AS half way in while you hit your on switch.

Whatever precharge you come up with will have to be the only thing connected when you hit your on switch.

 

[amberwolf]

My understanding is that the mosfets in things like BMS are not damaged by inrush due to lack of precharge. I had recently wondered about the same thing. I assume because their entire purpose is high speed switching of large currents that there is effectively no spark jumping a gap.

Correct--the spark damage is almost always confined to the tips of the connectors as they first touch.

The "only" time FETs in BMS (or controller) would be damaged is:
--voltage spikes generated by current spikes, where voltage exceeds Vdsmax.
--current spikes thru DS that exceed Idsmax or that cause excessive heating within the die before the casing can disssipate that to whatever heatsinking has been provided.

Neither one of those is likely to happen from a battery-plugin/unplug event (which is all the antispark connectors would help with).

In a really high votlage system, the arcing could potentially create enough RF to damage FETs (or votlage regulators) that are directly on the battery bus, depending on the kind and amount of filtration on the bus. Not very common--it's more likely to happen from a poor battery, fuse, or phase connection that has frequent or even continuous arcing.


Properly-sized, non-counterfeit / non-garbage capacitors aren't likely to be damaged from the inrush for the average ebike system. In a high voltage system (well above the typical 48-52v nominal types) the current is sustained longer and can be higher, and the heating from the current flow has more chance of damaging the caps--but even this is something that would have to be cycled every time you use it to be much of a worry.

(A really high voltage system (couple hundred volts+) has more chance of such damage, but those are even less likely to be problematic caps or improperly sized/rated (few cheap / etc controllers out there at these voltages)


As an example...I use a cheap battery disconnect switch to turn my 52v system on SB Cruiser on and off, cycled at least twice a day, nearly every day, with no precharge setup (no antispark).

Both the charger *and* the controllers*** and the Cycle Analyst and an old 15v wallwart used as a DC-DC are directly connected to that switch, so all are cycled with whatever current spike happens to each during power-on. I don't use a BMS. None of them has had FET or cap failures from the years of switching.

***Presently it's a pair of Phaserunners; previously have used various mismatched cheap controllers, some of which have had their issues but not from this.

 

[E-HP]

It's ironic that the capacitors that create the spark that the OP is concerned with, are the very components that protect to rest of the controller's electronics from transients. The XT90 is just a connector.

 

[DanGT86]

I noticed in the VESC world the vendors all seem to caution that an anti spark switch must be used. Are those particularly vulnerable?

I always figured when a spark jumped the connector gap it would be a momentary extreme high voltage event. I don't really understand why it just seemed like something electricity would do. I associate sparks ionizing air with high voltage.

So during the brief moment of the spark is the controller seeing voltages hundreds or thousands of times its recommended max?

Feels like that would be bad for anything with delicate electronics.

 

[amberwolf]

I noticed in the VESC world the vendors all seem to caution that an anti spark switch must be used. Are those particularly vulnerable?

I always figured when a spark jumped the connector gap it would be a momentary extreme high voltage event. I don't really understand why it just seemed like something electricity would do. I associate sparks ionizing air with high voltage.

A spark isn't by itself a high voltage event, it's a high current event, enough to turn the air into plasma.

You only need enough voltage to jump whatever the gap is to start the process. The smaller the gap, the lower the voltage required. IIRC it's something like three thousand volts per mm? (open air). I'm not sure what the minimum breakdown votlage of air is, but there's some point at which it isn't "linear" from that.


LIghtning is a high voltage event because of the huge distances involved, but with connector / switch / contact sparks, it can be a very tiny voltage. Watch a little 3volt brushed motor inside when connected to that voltage, and you'll still see sparks at the brushes (brush arc); it's still only 3v...just an extremely tiny gap.

So during the brief moment of the spark is the controller seeing voltages hundreds or thousands of times its recommended max?

No, the voltage doesn't rise above the actual voltage. The capacitance will cause voltage to rise "smoothly" on the actual circuits, according to the amount of capacitance, and the resistance between the voltage source and the capacitors.

The RF generated by the arcing can induce voltages into circuits along the conductor paths, especially where there is a lot of inductance in the paths to various parts. If the parts aren't able to suppress this (built-in TVS, etc) then it can damage them, but most stuff has some form of TVS in it to protect against ESD, and for most very short RF events it's enough. Long or continuous arcing (like with poor connections and high current draws, not precharge/connection issues) could cause enough RF long enough to destroy the TVS and then reach the actual electronics on the die, and THEN you get POOF.




If there is a lot of inductance in a circuit, and you *cut* the current, then the collapse of the fields in the inductors can cause voltage spikes above what had been present. This can happen when a BMS shuts off to protect the battery from overcharge on a long downhill during regen when the battery was almost full already (or has some internal cell / etc problem with high resistance). The current in the motor coils ceases and collapses the induced magnetic fields, which spikes a voltage that can be high enough to blow the FETs in the controller. Not a common problem, but it happens.

 

[DogDipstick]

Correct--the spark damage is almost always confined to the tips of the connectors as they first touch.

The "only" time FETs in BMS (or controller) would be damaged is:
--voltage spikes generated by current spikes, where voltage exceeds Vdsmax.
--current spikes thru DS that exceed Idsmax or that cause excessive heating within the die before the casing can disssipate that to whatever heatsink

NO

This is " incorrect"

and this ( diagram) is the mechanism for damage.

BMS and cells themselves are CERTAINLY protected by an antispark, Amber.. and...

if you disagree..

Consider I did not draw this.. an EV professional did.

You should and do know this though.

 

[DogDipstick]

The anti spark is ONLY protecting the connector contacts.

I believe the main function of precharge is to protect mechanical switches and connectors from spark damage.

Antispark and precharge are two different things.

A spark can be attenuated with 5 ohms on a 72V system.. but.. that will still slam the pack, controller, BMS and cells with damaging current potentially. This is why you see longer precharge times and real resistances on a timer period with ( real professionally designed ) pre charging circuits. Ther is a calculation for that that considers the system energy, time, and resistances needed to attenuate. It is called the RC time formula.

People think "antispark" is same as "precharge" and no... i dont think so.

 

[amberwolf]

NO

This is " incorrect"

and this ( diagram) is the mechanism for damage.

BMS and cells themselves are CERTAINLY protected by an antispark, Amber.. and...

if you disagree..

Consider I did not draw this.. an EV professional did.

You should and do know this though.

I did, and I posted that...which you quoted:

--voltage spikes generated by current spikes, where voltage exceeds Vdsmax.
--current spikes thru DS that exceed Idsmax or that cause excessive heating within the die before the casing can disssipate that to whatever heatsink

The diagram shows the current spike (not sure if it is showing a negative voltage drop during the current spike, but that coudl happen too).

Unless I just don't understand what the diagram is stating / implying?

If so, could you tell me specifically what is different?