Alan B
100 GW
Hi JRH, I can't reply to your private message you have disabled them. 
eBike Master Switch Design
- Thread starter Alan B
- Start date
Alan B
100 GW
Parts are being ordered for the test board, according to the board tester.
Alan B
100 GW
This is a quick phone pic of the OSH Park board. Purple with gold plating on a budget!
powersupply
100 W
Wow, looks cool!
Am sure it will perform as cool....
Am sure it will perform as cool....
mwkeefer
1 MW
Alan,
Id like to order 2 of those PCBs if you still have them available for sale, shipping would be to 19086 EMS.
Thanks in advance!
Regards,
Mike
Id like to order 2 of those PCBs if you still have them available for sale, shipping would be to 19086 EMS.
Thanks in advance!
Regards,
Mike
Alan B
100 GW
Hi Mike, we're in the testing phase right now, these have not been tested. So far we have one tester, do you want to be another? I can provide you one bare board, you provide parts, build it up and report on how it works. Then we'll make any changes and have more made. I only have three boards in this first batch.
mwkeefer
1 MW
Count me in !
-Mike
-Mike
powersupply
100 W
Hey Alan,
how hard would it be to implement a basic short circuit protection?
A sense resistor and a comparator to turn the mosfets off?
Or a quad OP/comparator for additional overvoltage/reverse?
THAT I would buy immediately.
how hard would it be to implement a basic short circuit protection?
A sense resistor and a comparator to turn the mosfets off?
Or a quad OP/comparator for additional overvoltage/reverse?
THAT I would buy immediately.
izeman
1 GW
hi migueralliart, i guess you're referring to my project which unfortunately never came to live because of too less time. building them was way too much work so i'm happy to see that there is a new project now which seems easier to build.migueralliart said:nicobie said:We really need a switch like this. I hope yours works!
I had high hopes for Method's switch, but it seem that he has lost interest in everything.
There was also an attempt from someone else but it never got to the point where he could sell them.
Subscribed!
still feel free to look at "my" switch design, maybe something to re-use in it.
http://endless-sphere.com/forums/viewtopic.php?t=49026#p723441
btw: i run 4x 4110 in parallel and it doesn't get above ambient at 4500w peak (78v 60a).
in case those fets become warmer than wanted one could make a small copper/aluminium bar above all fets lying on the GND metal base using the holes to screw it to the fets and pcb all together.
izeman
1 GW
hi alanAlan B said:
i'd like to test those as well. i have all parts in my workshop and the experience to solder them properly and test them. pm me if interested
johnrobholmes
10 MW
I'll be helping with this project as well. We will be offering built units after prototypes are tested, for those not wanting to build their own. I will also build up Alans unit for testing since he is time strapped but being so generous as to generate the board files.
Alan B
100 GW
Thanks folks, now all three test boards are spoken for, they should go out in the next day or two.
Then we'll hopefully get some quick feedback and see what changes to make for a larger run. I'm not riding right now so we need to get these tested in real conditions. I may also make some 2 device variants of these boards which will be sufficient for many use-cases.
Then we'll hopefully get some quick feedback and see what changes to make for a larger run. I'm not riding right now so we need to get these tested in real conditions. I may also make some 2 device variants of these boards which will be sufficient for many use-cases.
Alan B
100 GW
izeman said:hi migueralliart, i guess you're referring to my project which unfortunately never came to live because of too less time. building them was way too much work so i'm happy to see that there is a new project now which seems easier to build.migueralliart said:nicobie said:We really need a switch like this. I hope yours works!
I had high hopes for Method's switch, but it seem that he has lost interest in everything.
There was also an attempt from someone else but it never got to the point where he could sell them.
Subscribed!
still feel free to look at "my" switch design, maybe something to re-use in it.
http://endless-sphere.com/forums/viewtopic.php?t=49026#p723441
btw: i run 4x 4110 in parallel and it doesn't get above ambient at 4500w peak (78v 60a).
in case those fets become warmer than wanted one could make a small copper/aluminium bar above all fets lying on the GND metal base using the holes to screw it to the fets and pcb all together.
I reviewed your thread izeman, very nice! I like those chunks of copper! Those switches will handle a LOT of power!
Hopefully these little PC boards will handle enough power, and be really easy to make.
On the next batch I will add an LVC connector, so if you have an opto output LVC detection system it can override this master switch and shut power down. It is probably better to shut down the throttle rather than kill power, but this will give us that option in case it is needed.
Alan B
100 GW
Test Board Status
Board #1 went out today.
Board #2 is packed, will go out most likely tomorrow.
Board #3 negotiating details, will go out soon.
Board #1 went out today.
Board #2 is packed, will go out most likely tomorrow.
Board #3 negotiating details, will go out soon.
Alan B
100 GW
I see now that I committed boards to four testers, my mistake. Apologies for the confusion. I only have three boards now, want to get some feedback before cutting the next set. I'll PM the one I forgot about.
What do we want to test?
1) Does it work as in turn on and off the load? No smoke. (Is the gate voltage rising to the zener level?) We don't want to put a heavy load on it until we are sure that the gates are saturated at 10-15V.
2) Does it ramp? Do the component values seem to be good choices?
3) At what current levels does it not overheat? (Depends partially on FET)
4) At what current levels does it overheat?
As soon as we determine that it works we can make improvements
PLEASE NOTE - DANGEROUS VOLTAGES ARE PRESENT ON THIS BOARD!
Whether it is on or off there are both Battery positive and negative on this board. When checking temperatures either use a noncontact thermometer, or insure that power is removed before probing with fingers!!!!!
What do we want to test?
1) Does it work as in turn on and off the load? No smoke. (Is the gate voltage rising to the zener level?) We don't want to put a heavy load on it until we are sure that the gates are saturated at 10-15V.
2) Does it ramp? Do the component values seem to be good choices?
3) At what current levels does it not overheat? (Depends partially on FET)
4) At what current levels does it overheat?
As soon as we determine that it works we can make improvements
PLEASE NOTE - DANGEROUS VOLTAGES ARE PRESENT ON THIS BOARD!
Whether it is on or off there are both Battery positive and negative on this board. When checking temperatures either use a noncontact thermometer, or insure that power is removed before probing with fingers!!!!!
Alan B
100 GW
Note that the BOM and picture of this first board are near the bottom of page 1 of this thread. I'll update the BOM there as it evolves.
powersupply
100 W
Seeing the excitment here, curious how your BMS will be received....
Might have gotten lost, just to add to the wishlist of the next generation of these (where you want to add a LVC cutout):
Short circuit protection!
Might have gotten lost, just to add to the wishlist of the next generation of these (where you want to add a LVC cutout):
Short circuit protection!
Alan B
100 GW
powersupply said:
I saw your posting when I was unable to respond, and didn't notice it again. Apologies for that. Seem to be doing too many things at once lately.
That might be possible in a more complex deluxe version. It would have to be able to sense the current and turn off the FETs very quickly which is not an easy task. I'm curious where your short circuits are coming from? The other thought is that the controller should really do that, it has all the right elements, unless it is the thing that shorted. Depending where the short is, the current can rise so quickly it is hard to protect. The bike should really have a fuse to protect against shorts. We do all have master fuses, right.
powersupply
100 W
The shorts are due to user error. I read that this happens all the time, even to experienced people.
One has to listen to them and plan for such events. I mean this is a hobby envioment
Fuses are REALLY tough to get, they need to be high V and high A and ultrafast to avoid total damage.
In fact, I have not found any fuse which can disconnect at 100V and 40A within the time of the pulsed drain current of (one) FET from the controller.
Some are for aviation and more expensive than the whole bike.
And all of them are one time fuses.
I'd rather spend the money for a few high quality FETs in your unit to protect the many more in the controller/BMS.
It also makes sense from a recycling point of view.
Some resistors and a transistor or comparator is all it takes it seems.
Sure, there is the melting type fuse which avoids disaster, but it cannot avoid the replace-the-12-FETs-disaster.
I agree the controllers would have everything to protect themselves, but they don't and modding them is harder than implementing it externally.
It would be really great if you could at least give it a thought.
One has to listen to them and plan for such events. I mean this is a hobby envioment
Fuses are REALLY tough to get, they need to be high V and high A and ultrafast to avoid total damage.
In fact, I have not found any fuse which can disconnect at 100V and 40A within the time of the pulsed drain current of (one) FET from the controller.
Some are for aviation and more expensive than the whole bike.
And all of them are one time fuses.
I'd rather spend the money for a few high quality FETs in your unit to protect the many more in the controller/BMS.
It also makes sense from a recycling point of view.
Some resistors and a transistor or comparator is all it takes it seems.
Sure, there is the melting type fuse which avoids disaster, but it cannot avoid the replace-the-12-FETs-disaster.
I agree the controllers would have everything to protect themselves, but they don't and modding them is harder than implementing it externally.
It would be really great if you could at least give it a thought.
Alan B
100 GW
You are right about the fuse, it protects the wiring and battery, but not the FETs.
FETs are very hard to protect. There is a big capacitor bank in the controller that may be sufficient to blow the FETs in the controller when the motor shorts.
I'll give it some thought but short protection is a tall order. I already plan to add a jack for remote turn-off of this switch board, so one possibility is a separate circuit that detects problems and turns off this switch that can be added externally.
FETs are very hard to protect. There is a big capacitor bank in the controller that may be sufficient to blow the FETs in the controller when the motor shorts.
I'll give it some thought but short protection is a tall order. I already plan to add a jack for remote turn-off of this switch board, so one possibility is a separate circuit that detects problems and turns off this switch that can be added externally.
powersupply
100 W
Sorry, didn't mean shorts between motor-controller, those are out of the scope and reach of your unit of course.
Rather shorts after the battery or reversed connections.
Those seem to be the most common catastrophic failures from what I read
I produced a short like that myself, a big flash and a plasma later the 4mm connectors were vaporized.
The fast acting fuse was gone, too and too late.
It was a $6 spare fuse for the Fluke.
And how many controllers have blown due to reversed power.
If your unit would be in line, it could prevent that.
Thanks for your consideration.
Rather shorts after the battery or reversed connections.
Those seem to be the most common catastrophic failures from what I read
I produced a short like that myself, a big flash and a plasma later the 4mm connectors were vaporized.
The fast acting fuse was gone, too and too late.
It was a $6 spare fuse for the Fluke.
And how many controllers have blown due to reversed power.
If your unit would be in line, it could prevent that.
Thanks for your consideration.
Alan B
100 GW
Since this design ramps up the voltage, if the controller was hooked up backwards the fuse would blow at much lower voltage and current, and perhaps that would be sufficient for the controller to survive. With full voltage from the battery to a backwards controller the small inductance and resistances of the supply are going to cause current to rise much faster and deposit a lot more energy while the fuse is deciding to blow. It is hard to say if the controller would survive with the soft-starting master switch board, but the chances are much better than with a direct battery/fuse connection.
johnrobholmes
10 MW
The FET body diodes would pass the current if it were hooked in reverse, another bank of FETs in series would be needed. To get the master switch back to the original resistance, it would quadruple the number of FETs used.
Alan B
100 GW
I think he meant the motor controller was reversed polarity, rather than this master switch. John's correct, if this master switch is reversed input/output wise it will conduct all the time through the FET body diodes, and cease to be a switch.
I have a design for the charge control switch, and it has dual FETs for control in both directions. That is not a big problem since our chargers don't handle the same currents as the motor does. Except maybe Luke's chargers!
I have a design for the charge control switch, and it has dual FETs for control in both directions. That is not a big problem since our chargers don't handle the same currents as the motor does. Except maybe Luke's chargers!
powersupply
100 W
Yes, I meant if the motor controller was connected to the battery with swirched polarity.
It's just things I read here, never happened to me, but who knows, it could.
I only had a short which could have been avoided/stopped with such a smart short circuit switch. Because the arc kept the plasma going until the connector was used up or the gap became too long, or the actual fuse finally blew.
A smart switch could have at least prevented that.
And another smart switch at the entrance to the controller would prevent damage if one hookes up its plus to minus etc.
I would solder these right close into the path of the battery or the controller / BMS.
No catastrophic events anymore.
It's just things I read here, never happened to me, but who knows, it could.
I only had a short which could have been avoided/stopped with such a smart short circuit switch. Because the arc kept the plasma going until the connector was used up or the gap became too long, or the actual fuse finally blew.
A smart switch could have at least prevented that.
And another smart switch at the entrance to the controller would prevent damage if one hookes up its plus to minus etc.
I would solder these right close into the path of the battery or the controller / BMS.
No catastrophic events anymore.
Similar threads
