izeman
1 GW
of coursefechter said:
my mistake. thanks
Active pre-charge/inrush control
- Thread starter boo
- Start date
my mistake. thanks
izeman
1 GW
i changed 1m to 2m to 4.7m and as you said, time to fully open gate doubled every time. you can now see the voltage climb up to 12v on the dmm. it takes probably 2-3s to from zero to 12v now.
but there is still a small "plop" when you connect the battery. not as big as w/o the circuit but still not really spark free. my other circuit (the one alanb made pcbs for, and which has a bit more parts) does it absolutely spark free.
is this the expected behaviour?
but there is still a small "plop" when you connect the battery. not as big as w/o the circuit but still not really spark free. my other circuit (the one alanb made pcbs for, and which has a bit more parts) does it absolutely spark free.
is this the expected behaviour?
There should be no spark at all. Seems like the FET gate may have some residual charge.
Maybe try putting R1 back to 1M and place a 1M or larger resistor across the Zener diode to drain the gates.
Maybe try putting R1 back to 1M and place a 1M or larger resistor across the Zener diode to drain the gates.
izeman
1 GW
it's getting really strange. i check, double checked and tripple checked the circuit - it's a real simple one. when i connect it to the controller, and the add the battery connection i get a spark. WITHOUT voltage on the gate. if i check voltage to the controller i can see 42v there. HOW can that be???
there is no short between drain and source. i checked that as well and used brand new fets.
there is no short between drain and source. i checked that as well and used brand new fets.
Maybe try placing a jumper across the Zener to keep the gates at zero and see if that stops the spark.
izeman
1 GW
this is doomed. i removed ALL components. only drain and source connected. connected to battery, and i can see 42v at the output. there was NO cap, NO resistor. NOTHING connected to gate.
i measured the fets: 0,45v between D and S and 0 in other direction measured in diode testing mode.
measured resistance between D and S and got 2.8mOhm and OC.
this sounds quite ok. and i compared it to several other FETs i have lying around.
i measured the fets: 0,45v between D and S and 0 in other direction measured in diode testing mode.
measured resistance between D and S and got 2.8mOhm and OC.
this sounds quite ok. and i compared it to several other FETs i have lying around.
alsmith
100 kW
ok. so where are you measuring on what circuit? Is it with or without the controller in line?
izeman
1 GW
Regarding my last post : all measurements were taken w/o anything connected to it. Just the plain fets.
The voltage measurements were taken with and without controller connected, resulting in the same values.
The voltage measurements were taken with and without controller connected, resulting in the same values.
izeman
1 GW
i tested several fets. and they all show the same. i will build the whole thing with one new FET now to try again.madin88 said:
as well i will pull down gate to GND. we'll see
FET gates can hold a charge for a long time. Definitely short them to the source. Flux residue can be conductive enough to turn them on.
Also, I know you checked the circuit a zillion times, but just be sure you don't have the battery/load connections reversed. It would act like a diode in that direction.
One other thing is when measuring with a voltmeter, any FET leakage may appear like they are on when they are really off. Maybe try a 10k resistor or a light bulb for a load when testing. Again, flux residue can cause this.
Also, I know you checked the circuit a zillion times, but just be sure you don't have the battery/load connections reversed. It would act like a diode in that direction.
One other thing is when measuring with a voltmeter, any FET leakage may appear like they are on when they are really off. Maybe try a 10k resistor or a light bulb for a load when testing. Again, flux residue can cause this.
izeman
1 GW
when i short gate to source there's no spark at all. unfortunately this doesn't solve my problem as i wanted to use the circuit w/o a seperate switch. and such is needed to switch gate between high and low.
It needs a resistor from the gate to the source. Something like 1M or equal to R1. If you do this and clean the board really well, it should prevent the gates from picking up a charge before they're supposed to.
This is an oversight in the original schematic and I should update it. FET gates are extremely sensitive and can build up enough charge to turn on with picoampres of current. Just humidity could do that.
This is an oversight in the original schematic and I should update it. FET gates are extremely sensitive and can build up enough charge to turn on with picoampres of current. Just humidity could do that.
izeman
1 GW
I noticed that. If I just added ground wires to end and src and no other part, touching the gate with my finger made the fet switch.fechter said:
izeman
1 GW
i now built the circuit with the 1M between G and S. but that doesn't change a thing.
but clamping G to S when pluggin in and then switching the circuit on works flawlessly. so i cut a whole into my frame to add a 2-way switch and will do it that way.
but clamping G to S when pluggin in and then switching the circuit on works flawlessly. so i cut a whole into my frame to add a 2-way switch and will do it that way.
madin88 said:
Correct. You need about 10v on the gate for full turn-on, so you need a minimum 20v pack voltage in this configuration.
Anveshsameer
1 mW
Thats a cool pack for a battery. I was using a small precharge resistor across the operating relay which should suppress the inrush current and thus protecting the capacitors.Gregory said:
By the way what is the bag you used for the battery casing i wanna give it a try!
izeman
1 GW
unfortunately i couldn't find a 2-way switch with the correct form factor, so i have to use an on/off switch. and now i can't connect gate to ground to discharge the circuit. voltage goes down REALLY slow. takes around 10s from 45v to 15v.
turn on time is around 2s. any idea how i could improve the turn off time? or doesn't it matter anyway?
here's a short video. if you listen closely you can hear the switch click @1s, 9s, 14s and 18s.
[youtube]lEzjk7WhUYE[/youtube]
turn on time is around 2s. any idea how i could improve the turn off time? or doesn't it matter anyway?
here's a short video. if you listen closely you can hear the switch click @1s, 9s, 14s and 18s.
[youtube]lEzjk7WhUYE[/youtube]
You could try using the switch to short the gate to source and leave the 1M resistor always on. When the switch is on (gates shorted) the output will be off but the 1M resistor will drain 1uA per volt on standby. This is a very small amount of drain and would be OK for years.
izeman
1 GW
Interesting idea. I got a 48v 20Ah pack. So i guess 50uA is acceptable. On the other hand : does the slow down ramp do any harm?
I like it.fechter said:
I've taken the liberty to update the original draft as 3b to reflect the added 1M resistor and alternate switch connection per izeman/fechter recent efforts:
Switch is optional - omit for automatic-only operation.
Alan B
100 GW
This will quickly open the FETs, but the capacitance in the controller will still take a long time to drain out, so the CA may still complain. We ran into this issue before with similar circuits, a possibly better approach is for the switch to supply power to the CA and the gates, and thus when switched off the CA loses power quickly. Anytime the CA is connected across the controllers large capacitor bank there is a long drain out period.
izeman
1 GW
what about the 1M resistor between gate and source?teklektik said:
http://www.endless-sphere.com/forums/viewtopic.php?f=3&t=40142&p=1007023#p1000332
Similar threads
