An argument for *not* using a contactor

[methods]

So I did a test on my Kelly KBL 12221 controller
I placed an ammeter in line with the main battery to the +B of the controller
I then took measurements with the controller OFF and ON

In both situations the leakage current was 4.7ma

If I round that up to 5ma (0.005A) the following is true:

It would take 8.33 days to burn 1 ah
Overnight (10 hours) would only burn 50mah

This being the case, I am going to make a permanent connection between my battery and controller.
There will be a 75A Andersen, but I will leave it connected.

The only counter arguments I can think of go like this:

1) Unsafe (errr... I am a guy who rides around with 9Ah 100V lipo 3 inches from my nu*s so. . .)

2) Long vacation (I would disconnect it in that case)

3) What if you park it with a dead pack? (Good argument, but I never, ever, ever run my Lipo packs down that far.)

Seems like I can simplify operation, wiring, and packaging.

To further my argument, inspection of the Kelly instructions show a 300ohm resistor permanently attached between the battery and the controller. I tested the current through this resistor and it was still 5ma

It is my guess that it is customary to use a contactor to make the vehicle more safe from accidents during storage.
Would you agree?

-methods

 

[Link]

There shouldn't be a leakage current from a contactor. It's basically a glorified eletromagnetically activated switch. It's mostly used so you can cut power completely without having to unplug something.

Oh, and that's a precharge resistor. It's to keep the contacts from deteriorating from the huge inrush current that you'd get if you suddenly fed voltage to the uncharged capacitors in the controller.

A permanent connection won't solve your problem. The controller is what's drawing power through that resistor. Just add a button in series with the resistor and remember to push it for a second before you flip on the contactor. That's how my ride's setup, anyway.

 

[Malcolm]

Link is right as usual. The precharge resistor is needed to prevent excessive arcing and possibly welding of the contactor contacts. Without it, the inrush of current when you complete the circuit will also significantly shorten the life of the main capacitors in the controller and is likely to lead to premature failure. If you do replace the contactor with Andersens you should still precharge the controller with a resistor before plugging the Andersens together.

On my motorbike I have the precharge resistor wired direct across the contactor all the time, so I don't need to worry about giving the capacitors time to charge up before closing the contactor. Like you say, the drain is very small and it would take a long time to drain the pack. I do it this way partly because I'm lazy and partly because I like to keep things simple.

I think safety is the main reason for using a contactor, but it's an important consideration for me at least. I wouldn't be happy leaving my bike unattended if it was ready to go all the time.

 

[TylerDurden]

Here's fechter's diagram:

With the keyswitch between the batt and the precharge resistor, there will be no drain.

The caps should charge before the contactor closes, even if the killswitch is closed.

 

[methods]

Sorry, you guys got me all wrong.

I know what the pre-charge resistor is (remember that I run at 100V. . .) and I know that there is no leakage due to a contactor. Currently I dont even have a contactor, I just plug two 75A andersens together

My entire point was to *eliminate* the main contactor in my new build
I never stated that the leakage was a problem. Quite the oposite what I was arguing was that it was *not* a problem.

My argument was that I did not need a main contactor because the leakage current was negligible
I was arguing to just make a hard connection
No problem to solve. Just trying to pose an argument for a permanent connection.
I mentioned the resistor to prove the point that with or without a contactor the leakage current is there
(unless you set up a switch for the pre-charge)

How did we run so far into the weeds?

-methods

 

[Link]

LOL

No problem with that, I guess. Don't know how the charger would react, but odds are it won't do anything different.

 

[rkosiorek]

i think that it is mostly an asthetic thing. for my own vehicle i can live with a huge 75A powerpole. those things are like 3" long when plugged together.

on the other hand if i was selling it to someone else, especially if i was trying to pretend that this was a commercial product i'd use the contactor.

remeber that if you get rid of the contactor you also get rid of the small size "Kill Switch" you are currently using. it can be small because it only controls the contactor coil. otherwise you're gonna need a big switch with large leads for that function. so i guess it depends on what your requirements are.

rick

 

[Ypedal]

My lack of being an engineer, and other factors, lead me to sleep better at night knowing that the 2 wires that exit my battery pack(s) are not in any way connected to anything electrical.

When i have my helmet on, gloves, and in the driveway i plug in and go. And when i step off the bike i disconect before removing my helmet.


I just can't break that habit.

 

[fechter]

The 'leakage' drain of the main capacitors in the controller can be significant over a long time. If you were going to park it for longer than a month or so, it would be good to disconnect the battery.

I've also seen more than one report of a power line spike going through the charger and blowing up the controller even though the controller was off. In one case this caused a scooter to spontaneously go full throttle while parked indoors. Something like this could start a fire.

Most chinese scooters do not have a contactor and just leave the controller always connected.

 

[TylerDurden]

How did we run so far into the weeds?

Because it's not really a wash...

Using a contactor assumes the use of a low-voltage control switch; which when appropriately placed (relative to the resistor), results in no drain.

Sure, you don't really need to use a contactor if you have a big kill-switch or a tether to unplug your pack in the event of an emergency.

(As stated above, it is wise to disconnect the pack when not operating.)

 

[Doctorbass]

Using contactor is great if you plan a REAL SAFE killswitch able to handle high current like close to 100A.

Contactor open : NO leakage current at all
Contactor closed: High current carying capability all juice needed.. no loss

It is difficult to find a switch able to support voltage as high as we used when over 48V.... and to suport high current.

Contactor is the ideal solution to me. Tyco make very small one that can habdle 500A make or brake contact at 320Vdc !
for around 85$

and they only draw 1.7W... and they have an internal presurized gas.. probably nitrogen to kill the arcing posibility.

it is impossible to have current leakage with that.

And you can energize them to power ON a controller at 100V that all caps are at at 0V inside without any plasma.. they can handle that for thousands of cycles.

if you want to preserve the life of the input capacitors of the controller and avoiding to charge them too quickly, the precharge resistor is a great solution... by using a simple push botton in serie with it to precharge them only when needed...

I plan to do that with my setup.

Every high performance EV have them.. why not every performance ebike?

Doc

 

[methods]

Ah, much better!

Thanks guys, exactly what I wanted, a set of arguments for why I *should* use a contactor.

Lets see:

* To avoid charger problems: That sounds reasonable. I usually charge with mine plugged in so no problems but I could see where there could be

* To make a more professional build: That is another good argument. One thing I noted in all the arguments was the assumption that I would want/need some sort of kill switch. I had not thought of that and currently I dont even have one. Now that I think about it, I really do need one.

* To sleep better at night: Fair enough

* Leakage over time, power spike, safety: All good arguments. I did not consider what would happen if a power spike came through the wall, through the charger, and into the controller. That is a scary thought. . .

* Need for a kill switch / tether: Ok, one more argument for having the ability to kill the power. That really is the part I did not conside.

* Finally, Doc asks “why not”: Well, I am short on space and I really like simplicity. I would really like to eliminate everything that is not needed.



Ok, after reading all the arguments for using a contactor the thing that resonates strongest is the need for a kill switch. The need to be able to just kill everything in an emergency. I absolutely agree with that.

I think I am going to go with the diagram posted by Tyler Durden that fechter drew. Thanks guys. I will also look into the contactor that Doc mentioned as the ones I have seen are HUGE!

-Patrick

 

[Doctorbass]

Here is my unfinished setup.. I need to build the low voltage section for the N-O and N-C switch to activate or cut the contactor and the precharge resistor.. but a similar way than fechter... I will use my 25-240V to 12V "converter" to loop the activated contact.

First, a precharge resistor thru a switch will rise the voltage 50 or 100V input on this converter to produce the 12V output.. then the produced 12V will energize the contactor(serie or parallel configeration).. then the closed contactor will cary the 50 or 100V to replace the precharge resistor and will cary directly the 100V to the input of the converter ... sustaining the loop: 12V that activate contactoer that activate 12V..etc.. a simple N-C push botton will allow the cut of the 12V loop and act as killswitch or power off...
In fact.. i will have two precharge resistors: one that pre-activate the loop of the contactor energized position and the second that in the same time will pre-charge the caps of the controller(.. let-say.. the controller...)

The Power ON switch will be a dual position ON-off-ON switch. with one ON that is not holded(spring loaded).. to precharge...

Here are pics of my 3 contactors for serie-parallel and 4 gauge wire... oh!... CABLE i shold say setup.. not bad for the room they take...

Oh.. and i will put my special duc tape housing on it for sure!

Doc

 

[methods]

Doc. . . I suspected that you might be a little nutty.
After seeing the latest pictures of your bike I am now sure that you are completely insane

You guys have convinced me to do a contactor setup.

So have you looked around for a more powerful motor yet? I am sure you are experiencing the same thing as me: The power limitation is not the controller or the batteries, but only the motor! I can not get my 5305 to make any more power without melting down. My only choice now is to step up the voltage another 6S to 120V for use in very short bursts. My controller and batteries can take it but I worry about the CA. I may have to modify the CA to run that high.

Has anyone ran an X5 at 120V yet?
I suspect that it will melt down which is somewhat fine by me since the damn axle threads are stripped anyway. . .

-methods

 

[Doctorbass]

Methods, that's exact in my case, the limitation is now as incredible as it appear... the motor!!.. yess. guys, we reached the limitation of the famous X5

The factor no 1 is the Heat!

that's why i dont want to go higher than 100V 100A peak.. 9-10 kilowatt is like the real maximum we can use for burst... and like 3kW continuous with an outside ambiant temp of around 20 degree C max.

To ensure to never overheat or reach a critical temp for the hall sensors and wiring varnish melting, i use my BBQ temp alarm sensor that will buzz to the determined temp i decide. I 've seted it to 135 Celsius.. so every time the alarm start, i simply lower the throttle until it stop buzzing..

Doc

 

[methods]

The Kelly has a temperature input that it will use to attenuate the output power but I have not hooked it up.

-methods

 

[fechter]

Methods, that's exact in my case, the limitation is now as incredible as it appear... the motor!!.. yess. guys, we reached the limitation of the famous X5

The factor no 1 is the Heat!

That is a wild setup Doc, Now it just needs some clear Lexan covers and some neon lights in there. I think you have enough voltage to run neon lights directly off the battery.

Forced-air cooling can about double the continuous power rating It's not so easy on a hub motor, but certainly possible. Imagine what that's going to look like Liquid N2 or CO2 injection might be another option.


I use a wimpy little 40amp surplus relay on mine. It seems to handle 100A peaks OK.

 

[methods]

I have a bunch of 12V Peltier devices from a drinking fountain
I could paste about 10 of those on the motor case
They only draw 4A each so it would only take 480W @ 40A


Seriously though, heat sinks could do wonders. My bike frame even warms up.
Too bad we cant leverage the spokes as heat sinks. . .

Has anyone tried drilling cooling vents? They could be protected with a fine mesh to stop particulate mater while still allowing a flow of air. Then little "scoops" could be added for the forced effect mentioned above.

I am considering running two motors. I am thinking of trying two lighter models. . . or at least a lighter one in the front. I am picking up a 9 continents kit for my wife's cruiser bike so maybe I will try bolting one of those on the front for a week or two. Come to think of it, if one controller is start imitate the other could be peddle first.

-methods

 

[TylerDurden]

Has anyone tried drilling cooling vents?

I hear some xlyte sidecovers are self-ventilating...

 

[Doctorbass]

hmmm... I would worrie about drilling hols in the cover of a motor like this... a simple splash of water or some sand or mud that could goes inside could grind the winding or make corosion.. some hub motor without holes on them can already have water infiltration.. Oh.. and that was discussed already about the fact that using very high power in our hub motor could just help evaporating this water.. but now i'm talking about sand and other particules...

We already know that the hall sensor are very sensitive and that the rare earth magnet around the hub can become fragmented with object inside...

well.. these discussions will certainly one day pushing me to try that famous idea of cooling it (the stator) with circulating water in cooper tubes!.. maybe by removing the 7 speed casette and just keeping the 11 tooth gear.. to make enough room to drill the exposed axel and get some inlet and outlet tubes ... hmmmm just thinking about puting 15-20kW into it make me dreaming!.. oh... another torque arm problem to solve! :lol:

Doc

 

[Link]

well.. these discussions will certainly one day pushing me to try that famous idea of cooling it (the stator) with circulating water in cooper tubes!.. maybe by removing the 7 speed casette and just keeping the 11 tooth gear.. to make enough room to drill the exposed axel and get some inlet and outlet tubes ... hmmmm just thinking about puting 15-20kW into it make me dreaming!.. oh... another torque arm problem to solve! :lol:

*coughhttp://endless-sphere.com/forums/viewtopic.php?f=3&t=7296cough*