The Spark and precharging the controller

[Madness]

Evening all

I have seen alot of chat on here about the spark made when connecting the battery, it killing connectors/welding switches ect.

While at work today i had a thought about using high power transistors to make a auto pre-charge system? Could this be done or am i barking up a tree that's going to catch fire and burn?

My thought was using a pic micro controller to, on power up pre-charge the brushless controller and then use some high power transistors to "switch" to high current, this could even be done on something like a 555 timer chip i think.

Now i know the transistors would use power as well and need some kind of cooling, but for the low powered of us (upto 1KW) would this work?

 

[Jeremy Harris]

It's far, far simpler to just use a simple pre-charge resistor.

Jeremy

 

[Ypedal]

http://endless-sphere.com/forums/viewtopic.php?f=16&t=7994

 

[torqueon]

Yeah those caps dumping power, due scab up connectors bad.
Currently I have come up with a simple solution, while building a pack . I choose to use lacthing relays, meaning once powered on or off they no longer need power to hold them open or closed. The way it works is the coil, takes the intail hit, by the time the contactor has closed the caps are bleed off, and the power to the coll is removed!
When you turn off the unit, hit the coil with + than it opens the contact and again removes power from coil
Assuming your switching the - side. Think about it. when off the Neg terminal is at + potentail, when on its negative
So all you need is a latching relay and a spdt switch.
So far my bench testing has work very well

 

[dogman dan]

Hmm, replacing the andersons every 2-3 years is working fine for me, at 48v.

 

[itchynackers]

What sized andersons? I initially bought a load of the 45A ones, but have come to realize they are horrible for anything higher than 40A, and for multiple disconnects (the spring gets weak, poor connection). I've been trying the 4mm bullets, so far so good.

 

[TylerDurden]

Yeah those caps dumping power, due scab up connectors bad.
Currently I have come up with a simple solution, while building a pack for my bike. I choose to use latching relays, meaning once powered on or off they no longer need power to hold them open or closed. The way it works is the coil, takes the intail hit, by the time the contactor has closed the caps are bleed off, and the power to the coll is removed!
When you turn off the unit, hit the coil with + than it opens the contact and again removes power from coil
Assuming your switching the - side. Think about it. when off the Neg terminal is at + potentail, when on its negative
So all you need is a latching relay and a spdt moment switch.
So far my bench testing has work very well

If interested I can post schematic Hope this helps

Not sure that is a solution. Usually, the controller caps are empty and draw big current as they charge when the batteries are first connected.

As far as the relay/contactor consumption, the power required to hold a relay closed is usually minor compared to operating a motor.

 

[Madness]

It's far, far simpler to just use a simple pre-charge resistor.

Jeremy

Very true, as much as a love simple (and for most things its the only way i do stuff), i love to look at other options as its how i tend to learn random stuff.

To that end i have bought some of these:
http://www.maplin.co.uk/Module.aspx?ModuleNo=22997

At 55V and 62A i think these might have just what i need.

If it works or not, i have learnt something!

 

[snowranger]

Hmm, replacing the andersons every 2-3 years is working fine for me, at 48v.

With the frequency my Ping was cutting out on me, i.e. 2-3 times a day, my andersons did not stand a chance.

I'm now using a 20 watt ceramic resistor to precharge, then add in the main + connection. My controller is switchable on/off, so the precharge is hopefully seldom used.

 

[grindz145]

Evening all

I have seen alot of chat on here about the spark made when connecting the battery, it killing connectors/welding switches ect.

While at work today i had a thought about using high power transistors to make a auto pre-charge system? Could this be done or am i barking up a tree that's going to catch fire and burn?

My thought was using a pic micro controller to, on power up pre-charge the brushless controller and then use some high power transistors to "switch" to high current, this could even be done on something like a 555 timer chip i think.

Now i know the transistors would use power as well and need some kind of cooling, but for the low powered of us (upto 1KW) would this work?

What stinks is that you're adding more silicon in the current path which means loss. We have enough silicon to switch I almost think a precharge resistor and a delay relay would be a better solution. But of course I'm thinking high power....

 

[torqueon]

I have learned much here, I am suprised that no one has tried a setup simalar to mine posted above it seams to work well on the bench and bike. I am sure refinements are needed. I may be missing somthing here, if so talk.
The idea is to use the capasitor demand to pull the self latching relay closed once closed, power is on. To go off hit with the other side of supply and the unit is off !
Mind you I am dealing with a brushed controller, most here are using brushless, from what i read the problem is the same.
Bench setup is easy check it out.

 

[icecube57]

I pondered all day on what was the simplest way to precharge the controller. Then I thought about making a plug that splits off the main plug and using it as a sacrificial plug to take the spark damage. Then plug in the regular battery harness. Not only will the plugs be identical but they will share the load technically.

 

[itchynackers]

I may be stating the obvious, but I just tested my pre-charge resistor. I left it plugged in for 14 hours, and the lipo pack voltage dropped from 77.4 to 76.6. I'd recommend not leaving this plugged in, especially if you plan on not riding for a few weeks.