Much More Powerful Pie Now!!!

[Leslie]

Modulating your contoller in such a way only has you getting your solder gun on the board once. Just cut the old ones off.

One of the wires off the ISOTOP fet terminals for both source and drain can shunt straight to their coresponding power poles.

If you were to screw eyelets and insert there wire connectors into the core of some good thick wire all the solder eyelets solid onto the wires for the ISOTOP FETs screw conncetions I cant see how it wouldnt be an ample solution with good rewards.  Look for silver eyelets . .  Best conductors would be a refreshing approach towards these types fets.

[Leslie]

Really all you need is a path to ground to the controller and the PWM signal into the gates of your fets you can get your main source and drain from the motor and battery.

So then the power leads to your controller only need to a of a small gauge you can then controll the voltage easikly to your controller and overvolt to 72v in the future when you need to updgrade your pack. so two small wires from each fet need to come from the PCB.  Hmm 12 wires

Biasing those fets for your need will be fun. You can make them trigger precision at all voltages if you want.


Are you sure your controller is fried?  Maybe Email Tom not for a return and ask him what to do?

[Leslie]

Edit:

maybe use a small cap for sharing gate clamp voltages and from the controller to avoid messing with nothing other than putting in wires ,  This way you don't have any impedance issues over adding more modules.

Then parrallel another module to get even lower on-resistance any time you want without touching solder to the controller.

Ive actually never done this BTW, but I am getting to building it myself.

I all ready have enhancement mode modules that triger at the touch of a finger running speed controll if I press harder.  Great idea but not great with sweaty hands.

[o00scorpion00o]

Hi 317537, I think it's all a little too complicated for me lol. I Install new fet's no problem. But other than that, my knowledge of electronics is fairly limited!

I could follow pics maybe lol hint hint!

I bet one or two fet's are toast because of the fact the wheel is hard to turn, that's basically from my experience with my other controller!

I wonder did it fry because of over current or heat? I thought the controller had thermal protection?

How about 6 of these? IRFB4110PBF

I think it will be hard to mount some other kind of heatsink in there, But I think the fet's should be individually bolted to the heatsink!

[Leslie]

Sorry I got a little excited.  I do have some bits I could try out.

I got one HBS hub inductance tests are spot on and halls are operational.

1 dead old GM regen controller.

And now I need some of those fets you just posted.


IRFB4110PBF

Oh yeah, they certainly have the ability to run the motor.

Do you think I could get away with just using one gate feed off each chanell?

Ok Ill give it a try.

Ill try run my HBS off my own fet bridge from the logic of a dead GM 36v controller.

Ohhh BTW do it.  Replace the fets.

[o00scorpion00o]

I Wish I had an electronics store closer then 50 miles. Such a pain! 

Does anyone know if more speed can be got from this Magic Pie before I fix the controller and close it up again?

I think I will remove the shunt and solder some thin wire and see if it tames it down enough, as it is hard to keep the front wheel on the ground. Does anyone know if the controller has some soft start option? I see speed scale, My cable is broken and I can't fix it. it would be interesting if I could still turn the amps down via the software. I just wish I could get more speed though!

317537. I opened up the controller and found 1 dead fet! Maybe that's the reason the controller was actually still alive? Usually I would find 2 or more fried! I am still wondering was it over current or heat that killed it. If it was heat then that means the thermal protection didn't work, or maybe it hasn't got any?

[Leslie]

I Wish I had an electronics store closer then 50 miles. Such a pain!  

Does anyone know if more speed can be got from this Magic Pie before I fix the controller and close it up again?

I think I will remove the shunt and solder some thin wire and see if it tames it down enough, as it is hard to keep the front wheel on the ground. Does anyone know if the controller has some soft start option? I see speed scale, My cable is broken and I can't fix it. it would be interesting if I could still turn the amps down via the software. I just wish I could get more speed though!

317537. I opened up the controller and found 1 dead fet! Maybe that's the reason the controller was actually still alive? Usually I would find 2 or more fried! I am still wondering was it over current or heat that killed it. If it was heat then that means the thermal protection didn't work, or maybe it hasn't got any?

Ok your mod may of been a little robust, more than 50 amps.  Hold your horses if you gets it fixed and review the circuit and all its parts and what they are doing and why.

Possibly the cap did some damage when you pulled to much power through its leads and noise got into your DC power line causing the cap to draw current and switching spikes from the motor, when the fet switches off if that cap is drained low enough when the fet was on it will conduct very high current to fill back up from both sides of the input and output through the motor to the positive track.  The coils are inductive so they will give this cap what it needs and so will the battery but the lack of resistnace may be tipping towards the motor side more than the battery side with all the series cells and BMS solid state resistance allowing a less solid supply to your controller.  Hence the need for caps in the first place to filter out any noise that can come into your system..

You bypassed the shunt to that 63v cap and that shunt may of been what was limiting the that caps demand on the current back from the motor by shunting the path to ground to the said cap.


Its hard to tell we need more data.  I can simulate this in spice if I gets more data on the caps and how much current the Pie was able to draw with your mod would help too.

What are the caps there.  We were discussing whether or not your controller has the big caps like in MC 36v the picture or Bikemads MP has some behind the circuit board.  .  I assumed there was nothing under that brown cap.

Is that brown cap have anything parallel with it it or is that the only cap before the shunt?

[Leslie]

On some controllers I am seeing the polyester cap over the shunt to filter ripples out over the shunt, and the shunts always come after the main cap not before so that there is good power to the caps and the shunts limit how those caps are drained.

Placing a resistor shunt before the caps just seems to allow a lot of noise as the resitance is stopping the caps from recharging and the motor is at risk of draining the caps dry if the shunt is not providing enough current from the battery.

[o00scorpion00o]

hi 317537. here are some pics that might explain things better.

I saw on the cycle analyst where the peak current draw was around 120 amps, lol. R.I.P! 
However On the actual readings on take off, I noticed around 65 amps max. Climbing the hill was around 30-35 amps max. Once the speed picks up the current draw reduces a lot. That is why I think pulling the throttle quickly every second or two is what maybe drew excessive current on the hill, very low if any rpm, because I was very concious of doing wheelies and falling off! 

I would love to see what this motor's torque actually is at that current.

do you think then 317537 that I should have put a shunt parallel to the existing one, or maybe just added soder, or a much lesser gage wire? 

[Leslie]

Id replace the big cap you have there with a tantalum 100uf low esr cap its there to absorb spikes not provide energy to your fets.  Beef up the caps before your shunt and look around for a decent shunt bar that is rated.

I can send you some 10 amp shunts from my junk pile, replace the one thats on your controller with three 10 amp shunt bars for 30 amps and they are very thin compared to your mod.

Beef up all the tracks from you power pole to the shunt and use the shunts I send you only.

[Leslie]

hi 317537. here are some pics that might explain things better.

I saw on the cycle analyst where the peak current draw was around 120 amps, lol. R.I.P!  

 

LOL you maxed the limit of the fets man.  Well done.

[Leslie]

If you PM me with your details I will send  you the shunts this wednesday when I go into town.  They should arrive to you by next week.

I will send you them for free.

[o00scorpion00o]

Thanks 317537, Just sent you a P.M. 

Should I replace both the 470uf 63v caps? they wouldn't be bigger would they as there is not a lot of room in there. How low esr is good enough? his is as close as I can find, did you see the price???     

http://ie.farnell.com/vishay-tansitor/135d107x9060f6e3/capacitor-tant-100uf-60v-2-5ohm/dp/1618665

Also I think I will replace the fet with the same, as I will tame it down a bit as the torque is too much to handle, maybe I can adjust it with the software to my requirements.

Also where should I install the other two shunts as there is only place for 1 if I remove the old one? and would it be better to install just 1, 30 amp shunt instead?

Sorry for bombarding you with all the questions lol

[Leslie]

Thanks 317537, Just sent you a P.M. 

Should I replace both the 470uf 63v caps? they wouldn't be bigger would they as there is not a lot of room in there. How low esr is good enough? his is as close as I can find, did you see the price???     

http://ie.farnell.com/vishay-tansitor/135d107x9060f6e3/capacitor-tant-100uf-60v-2-5ohm/dp/1618665

Also I think I will replace the fet with the same, as I will tame it down a bit as the torque is too much to handle, maybe I can adjust it with the software to my requirements.

Also where should I install the other two shunts as there is only place for 1 if I remove the old one? and would it be better to install just 1, 30 amp shunt instead?

Sorry for bombarding you with all the questions lol

http://endless-sphere.com/forums/viewtopic.php?p=216474#p216474

You want to run as much capacitance as you can.
Lowering the input capacitance will increase your chance of failure.
When they cost about $1.50 I cant see any argument for running anything less than 100V 1000uF

I can fit a 100V 1000uF in a 6 fet Infineon for the input
For the rail you can fit a 100uF 100V
For the regulator you can fit a 100uF 100V

Basically the same values as stock but 100V instead of 63V

There is a trade-off between the size of the cap and the length of the leads.
I would personally rather have a 1000uF Low ESR cap with slightly longer leads than a 470uF cap with shorter leads

Capacitance is your friend.

even if you don't run a higher volt the 100v caps with high voltage spikes will be less hot and operate better from what I read.  I do have some old switchmode board laying around with some very stocky interesting looking LOW ESR 100v caps on board.

We can look at our options before I send you the shunts.

[Leslie]

Thanks 317537, Just sent you a P.M. 

I can adjust it with the software to my requirements.

Hi

Interesting thread about the shunt bypass.

And take the time to program your controller to meet the BMS and motor specs.

You were warned to do this on the first page. .

I tend to be good at seeing the warning signs because I have made almost every mistake in the book 4 times over.

However!

This is interesting.

Is the software in this case going to help your fets survive?  What AMPS was the controller set to, to begin with?  would of it made a difference to peak currents and voltages?  Maybe not hence the need for the shunt to begin with.