Active pre-charge/inrush control
- Thread starter boo
- Start date
Hard to say what happened, but 22s will be dangerously close to the voltage rating of a 4110 hot off the charger. You may need to use a FET with a higher voltage rating.
You may also want to make sure the zener didn't blow for some reason. You can test it with meter in diode check mode. It should test like a regular diode.
You may also want to make sure the zener didn't blow for some reason. You can test it with meter in diode check mode. It should test like a regular diode.
Crap crap crap.
I used my last 4 fets to build the original schematic, and added a zener to be sure.
Worked exactly 1 time. Turn on , no turn off
My controller works with IRFB 4110, so I figured they would be ok for the switch too. Still have 7.6V headroom...
I was so proud of a neat looking job, but when it doesn't work...
I used my last 4 fets to build the original schematic, and added a zener to be sure.
Worked exactly 1 time. Turn on , no turn off
My controller works with IRFB 4110, so I figured they would be ok for the switch too. Still have 7.6V headroom...
I was so proud of a neat looking job, but when it doesn't work...
Which schematic are you using?
Are you sure the capacitor is non-polar?
Are you sure the capacitor is non-polar?
The first time I used your schematic, which worked like 5 times, but the ramp-up time was very slow (4 seconds).
But it worked, and it was good enough for me. Only tested with wheels of the ground, I hadn't ridden the bike yet.
The photo I oisted now was my second attempt, which was the original schematic posted by jeremy, but with an added zener.
That only worked half a time, it went up and never came down.
So now I just bridged the negative wires, and am back at the starting point - arching connectors.
Perhaps a 12V relay fed by 3S would do the trick?
But it worked, and it was good enough for me. Only tested with wheels of the ground, I hadn't ridden the bike yet.
The photo I oisted now was my second attempt, which was the original schematic posted by jeremy, but with an added zener.
That only worked half a time, it went up and never came down.
So now I just bridged the negative wires, and am back at the starting point - arching connectors.
Perhaps a 12V relay fed by 3S would do the trick?
why not use a 3-wire switch and shut gate to ground in off position? This will turn off the switch instant (cap will discharge).
Check all parts once again.
it also could be possible some FETs got damaged due to too hot soldering..
Check all parts once again.
it also could be possible some FETs got damaged due to too hot soldering..
That is a possibility, I soldered it quite hot to get the drain wire soldered to the main body.
Could also be that the 4110's are too close to their Vmax, but the ones in the controller work perfectly...
As said before, the schematics by fechter did work for me in the beginning, but rampup time was very slow, which stresses the fets even more.
Fechter, should I change out something for 22S? Will go with 4115's to be sure next time, but perhaps the R/C value needs to change for this voltage rating?
Could also be that the 4110's are too close to their Vmax, but the ones in the controller work perfectly...
As said before, the schematics by fechter did work for me in the beginning, but rampup time was very slow, which stresses the fets even more.
Fechter, should I change out something for 22S? Will go with 4115's to be sure next time, but perhaps the R/C value needs to change for this voltage rating?
fechter said:
AFAIK, only elco's are polar, this is a Tantalum cap.
http://be.mouser.com/ProductDetail/...=sGAEpiMZZMtZ1n0r9vR22aMUkmzj83C0lMPmkWiFUfQ=
Gregory
100 kW
wannesd said:
Oh no no no. You might want to check that wannesd! Well spotted Richard.
A 1 second ramp up time should be long enough. Increasing the capacitor size will slow it down. The slower it is, the less current there will be during precharge. You don't want it too slow, as it can be annoying and if you hit the throttle before it's done charging, it could blow the FETs.
If you can't find non-polar capacitors, you can use two polarized ones in series, back to back (negative to negative or positive to positive).
The basic circuit should be OK as plenty of people have tried it. Essentially the same as Alan's version here:
http://endless-sphere.com/forums/viewtopic.php?f=14&t=54225
Don't use Jeremy's earlier circuit. Plenty of people have blown FETs with that one.
If you can't find non-polar capacitors, you can use two polarized ones in series, back to back (negative to negative or positive to positive).
The basic circuit should be OK as plenty of people have tried it. Essentially the same as Alan's version here:
http://endless-sphere.com/forums/viewtopic.php?f=14&t=54225
Don't use Jeremy's earlier circuit. Plenty of people have blown FETs with that one.
Ok, 2 questions:
If I series the caps, the Farads will decrease, like resistors in parallel. So for 1µF, I'd need 2 x2µF in series, right?
Second, could the 4 second ramp up time be caused by using a polarized cap? If not, what size cap would I need?
If I series the caps, the Farads will decrease, like resistors in parallel. So for 1µF, I'd need 2 x2µF in series, right?
Second, could the 4 second ramp up time be caused by using a polarized cap? If not, what size cap would I need?
In the back-to-back mode, I'm not sure about the caps needing to be doubled. With non-polar caps, yes placing two in series cuts the capacitance in half, but with two polarized caps back to back, only one is really working in each direction, so I think it stays the same. I'll have to test that someday.
You can also tweak the charging time by changing the resistor that pulls up on the gate. If it's too slow, use a lower value of resistance.
You can also tweak the charging time by changing the resistor that pulls up on the gate. If it's too slow, use a lower value of resistance.
Another probably pointless and/or bad idea, but what if you run a "power good" line from the unit to the controller's ebrake or enable input, so that it is not possible to activate the controller until it's charged the caps up to full?fechter said:
I guess i would add more complicated circuitry to it, as a disadvantage.
But not *too* complicated: Probably need a comparator between battery side and controller side, if positive output during unequal "disabled" state, use open-collector transistor pulldown to hold ebrake closed?
geostar1024
1 µW
- Joined
- Dec 26, 2013
- Messages
- 1
I've modified fechter's design to include an automatic LVC, via a voltage divider that is compared to a 12V zener and shorts the mosfet gates to ground when the comparator changes state. The voltage divider is tuned for a low voltage of 48V (for a 16S LiFePO4 pack). At least according to the circuit simulator I've been using, it should work (and has a turn-on time of about 900ms) . . . .
Diamondback
10 kW
that circuit sounds a lot like kfong's intelligent smart switch described here.....
http://www.endless-sphere.com/forums/viewtopic.php?f=31&t=32135
it has the anti spark, and LVC, and it also features a timer to shut down power in the even that the rider forgets to turn off or unplug.
i have one set for 12S lipo on my trike, and i love it.
Jason.
http://www.endless-sphere.com/forums/viewtopic.php?f=31&t=32135
it has the anti spark, and LVC, and it also features a timer to shut down power in the even that the rider forgets to turn off or unplug.
i have one set for 12S lipo on my trike, and i love it.
Jason.
whereswally606
100 kW
I have a few 20n60s5 MOSFETs 6 of them. Would they work in this circuit for preventing the spark in a 13s limn 48v battery. Trying to be cheap and scavenge them from some server psu s I have
Those are 600V, 190 mOhms rated. That's a very high on resistance for this application. Depending on your maximum discharge current, they might get very hot.
JackB
10 W
I'm a bit confused why such an elaborate circuit it used..old style precharge cars just two switches,
first switch used a lightbulb, when it goes out, turn on the main power switch that bypass the lightbulb.
The lightbulb is a highly integrated light-emitting resistor..
first switch used a lightbulb, when it goes out, turn on the main power switch that bypass the lightbulb.
The lightbulb is a highly integrated light-emitting resistor..
izeman
1 GW
because it's more convenient and trouble free as well as idiot proof. no procedure to follow and it can be connected all the time. i got mine connected over the winter and couldn't see any battery voltage drop because of it.JackB said:
whereswally606
100 kW
Thanks for the reply fechter, I saw the spec sheet but didn't really know why they were unsuitable just guessed that they might be. Thanks for confirming this though.
dawrench
1 W
I built Fechter`s Automatic precharge 3 circuit using 5x 4110 fets. Been using it on a 48volt pack - 54volts HOC with no problem. When I tried to use this with my 18S pack - 75.2 volts - It did not work as normal. Sparks on connection of pack to controller. With the key ON I got full voltage. With the key OFF it was as though the key was on but 1.5 volts less. Bike ran just fine. I rode over 2 kilometers with the key OFF before I noticed. When I put the circuit back on the 48 volt pack it works exactly the way it should.
Any ideas?
Any ideas?
Are you using the version that grounds the gates in the off position? If so, it would be hard for the FETs to turn on. Check the wiring on that side of the switch or try your 18s pack, switch off, and measure the gate voltage against the source (two outer legs on a FET). If it's not zero, there's a problem with the switch wiring.
If it is zero and you still get output, that would be strange, like some kind of FET leakage.
If it is zero and you still get output, that would be strange, like some kind of FET leakage.
dawrench
1 W
I will check. Does the voltage rating of the switch matter?
dawrench
1 W
Yes, it is the version that brings the gates down to B-
Sure wish I could buy one of these that would switch 100V. A kit would work for me if it came with a bare circuit board.
I'm sick and tired of having to fix my large DC contactor every time some idiot flips the main switch without using the pre-charge button.
I'm sick and tired of having to fix my large DC contactor every time some idiot flips the main switch without using the pre-charge button.
Similar threads
