Eectric go kart

[ituaku]

hi,
i really need help on h bridge protection circuit for my electric go kart..i am using a 48V 215A rated voltage and current brushed dc motor. The peak current of the motor is 400A. i decided to buy a ready made h bridge from semikron which is the SK 150 MHK 055 T SEMIKRON, leading manufacturer of diode thyristor power semiconductor modules, (IGBT, Mosfet, chips, drivers, rectifier,inverter, converter assemblies). I wanted to use a zener diode but i dont know how to connect it to the H-bridge and how to choose the right diode.

pls somebody reply...im really stuck...

 

[duffy]

The zeners won't protect you from overcurrent, sorry. They will only protect from voltage spikes, and it looks like the module already has protection diodes. Can you see the diodes (4 of them, pointing up) on that crappy little diagram on that page? Those are the freewheeling diodes. The zeners would go in parallel - but they won't protect from the 240A module from the 400A motor!!!

But, you could put 2 of these modules in parallel. That would be 480A. Run separate motor leads to each - these will work as resistors to balance any voltage differential.

 

[ituaku]

the 400A current will conduct for like a few seconds just to start the motor and after that the motor will just use 200A current...on the spec sheet, it shows that the module can go up to 500A current at Vgs 10V....thats y i thought of just using 1 module only...so does it mean i can just connect the module straight to my motor? I ask someone about whether the module have safety features n he replied this "A voltage headroom is required to allow for voltage overshoots which result from the circuit loop inductances. This is a factor of how good you layout and design is but 55V may not be enough headroom (check)". I dont understand what he means n i cant get in contact with him. hope someone can explain what this means.....another question, do u know how am i suppose to drive the hbridge? i tried looking for IC...so far im not sure which is suitable for this module...so any recommendations?

 

[duffy]

What, are you talking about that 500A on the "output characteristics" curve? Don't go by that (it has an 80µS spec anyway) go by Id and Idm. Idm is pulsed, and THAT's for less than 1ms, not "a few seconds".

Your friend is throwing the term "headroom" around in a way that's highly suspect. He's probably talking about the drain-source voltage breakdown, in which case he needs to say that. There are always voltage spikes driving an inductor, and brush type motors are especially bad. Adding a snubber (a resistor in series with a cap) across the motor leads will reduce what the diodes miss. They didn't spec the turn-on time, but they ought to be zero recovery schottky's.

Ugh. That spec sheet is AWFUL!!! Get the pdf and zoom in on the little schematic. Looks like they also put an MOV in it for you. That's that little symbol at the top of the schematic. This kills voltage spikes, too. Problem is, MOV's aren't polarized, and they have a + and - by it. WTF? Another thing - I don't see how the numbers on the schematic correspond to the anything on the mechanical diagrams! How is someone supposed to hook it up?

Question - you say your motor is rated for 215A. Is that the "stall" current, or is the stall current the 400A number?

 

[Hero999]

How long do you want the battery to last for?

What size battery are you using?

Do not use a flooded lead acid battery, use a gel cell or glass absorbed mat. Flooded batteries will leak acid everywhere if you have an accident.

I take it the current will be much lower when not accelerating, have you worked out the current required when travelling at a constant speed, i.e 30mph?

Work out what size cables you need and the voltage drop. Depending on the ambient temperature, how you've got the cables arranged and the temperature rating of the insulation, the cables might need to be as thick as 12mm (just under half an inch)!

I would also recommend using a higher voltage to lower the current, but I appreciate electrical safety would be more of an issue so it might not be practical.

 

[ituaku]

duffy: The stall curent is 400A that 215A is the rated current. from the diagram i can see that i have to connect pins 5,8,9,10 altogether to +dc and pins 19,20,21,22,23,24 to -dc. the rest of the pins will be connected to the bridge. I will try putting a snubber. Is there any way to measure the voltage spikes so i know how big the capacitor i need to use? do u know if i can simulate all this first. i dont want to damage any of the components.

hero999: i am using sealed lead acid 12V 50A. ill use 4 of it to get the 48V. i want it to last about 10 minutes. about the current at constant speed, how do you calculate it? i dont know the drag force of the kart.

 

[Hero999]

I'm not a mechanical engineer so I don't know how to calculate drag and therefore the current taken by the motor at a certain speed. It will depend on the aerodynamics, wind speed, tyres and track conditions.

All I know is that the batteries probably aren't large enough to supply 215A for 10 minutes. The capacity of a battery tends to go down as the power drawn increases, the 50Ah rating will only be valid at low currents, at 215A the capacity is likely to be not much more than 25Ah.

If the motor draws much less than 215A, about 100A then it'll probably be all right.

 

[ituaku]

hero999: im actually planning for a maximum of 20mph (slow but as long as it moves it would be fine lol) since the chassis i have wont be able to load heavy batteries. well it used to be a fuel kart so there's not much space available and the spaces are fixed as well. Anyways for the time being guess ill use the battery n see how long itll last but then again i have to complete the hbridge, controller before any actions can take place. i can foresee my long journey *sigh*

 

[Hero999]

The amount of power you have sounds way overkill for only 20mph. Your go kart will have the same amount of power as a four stroke 125cc engine which is typical for a small motor cycle.

I suppose it depends on the size of the cart and the terrain of the track. If you want to take it off road and climb steep hills then it might need the extra power.

215A at 48V is 10.32kW which is 8.265kW at the shaft assuming 80% efficiency. This translates to 11hp which is enough to do 60mph given good tyres, aerodynamics and appropriate gearing, you could probably add another 20mph if you streamline it enough.

If it's not too heavy your go kart will probably get to 20mph within a second. Make sure your gearing is strong enough to handle the torque and the breaks, tyres and chassis are man enough to handle the speed, acceleration and deceleration.

 

[duffy]

Stall current = start current. You will see a significant part of that 400A during the acceleration phase. I'm afraid that driver module just isn't rated that high.

On a more positive note, to seriously get 400A off a 48V battery you need a total circuit impedance - including the battery - of just .12Ω. That's a tough number to hit. When you pick a battery, look at the specs on it and add up all the contact resistances (going to need a low-ohm meter and kelvin clips), the motor winding resistance (lowest fixed-shaft reading), the cables and whatever. Might be high enough that 400A becomes a myth.

 

[ituaku]

hero999: at first i was planning for higher speed, but there's no space to put the extra batteries so i just cut down the speed. I bought the motor without considering the battery. so im kinda stuck with it. n its hella expensive.

duffy: i tried looking aroung for higher spec module-all i see is dead end. well maybe ill just connect them in series like u suggest before. btw does that means i have to build 2 drivers for the modules?? is there a chance both of them fail to conduct at the same time?

 

[duffy]

Not series, parallel, like I said.

 

[ituaku]

my bad.

Btw, how do u calculate the correct value for the snubber? Do I have to take measurement from the motor first??

 

[duffy]

You actually take the measurements for the snubber last, oddly enough.

I'm afraid snubber design is part engineering, part black magic. For your application, it's just a resistor in series with a cap across the motor drive - sounds simple, huh? What it's doing is killing a high frequency inductive spike that originates from a complicated interaction of things, reactances that include the lead wires and distributed capacitance in the windings, switch times from the drivers and diodes and motor commutator.

Simulators are useless for this, you have to get in there and measure it. You need a scope. You start off with a low drive voltage and crank it up until you see dangerous spikes. Your snubber RC time constant needs to be a small fraction of your minimum "on" time of the pulse width modulator or you are snubbing the PWM. The resistor is "burning off" the energy in the spike, the capacitor is storing it.

And just to make things more complicated, there are invisible spikes that are just as deadly. Current spikes. You can get a feel for them by measuring the voltage developed across a shunt resistor... but that small resistance affects the high current spike. A current probe is better - if you have one, and if it's not the moral equivalent of a snap-on ferrite.

So you damp it as close as you can to stay inside the SOA curves (which are missing from that lousy spec sheet!) without introducing too much capacitive load to the driver, which sends up the current spikes.

On a brighter note, you might not even need one. That MOV may be designed to deal with the voltage spikes. When you put the 2 drivers in parallel, the peak current spike it can handle is a thousand amps - just for 80 microseconds, but that's in the range of these current spikes. That's why they give you that number, it isn't for motor stall and start.

Inquire with the manufacturer about that "polarized" MOV, see if it was intended to go on the drive's output, or supply. Is there an SOA curve, or are we just supposed to guess? Are there numbers stamped in the plastic that aren't on the spec sheet? Right now I can't even see what pin goes to what!

 

[audioguru]

I think you are copying the circuit in our battery-powered buses.
Most of these new buses break down.

 

[Leftyretro]

I hope you aren't using normal sealed lead acid car type batteries for this application. Deep discharge batteries are required or you will have very short battery life which can get expensive in a hurry. Unfortunately for your application deep discharge batteries are even heavier and of course cost more initially.

Lefty

 

[ituaku]

audioguru: What do u mean i copied your circuit?

Duffy: thanks for the explaination.

Lefty: well teh battery charged up to an hour. its odyssey battery. as you said the charger is really expensive.

Btw the hbridge consists of N-channels from the spec sheet they gave. I dont really understand about mosfet (never learned it before) so im having problems to drive it. i read that im supposed to increase the supply by 12V to open the gate. Is it how it is supposed to be? so since i just got 48V i need level translator, am i right?

 

[audioguru]

audioguru: What do u mean i copied your circuit?

Our battery-powered busses frequently break down.
Your circuit also frequently breaks down.
The same circuit?

 

[Hero999]

For minimum voltage loss the MOSFETs need a gate voltage of 10V to turn on fully.

Unfortunately the high side MOSFET gates need to be 10V above the positive supply to turn on fully so you either need a +58V rail or bootstrapping capacitors which only work at AC.

 

[ituaku]

audioguru:

how do i do the 58V rail. can you be more detail.