Active pre-charge/inrush control

[Arlo1]

OK, updated schematic below. With the fuse in this position, if there is a short in the controller or overload, the fuse will blow and prevent the contactor from closing.

Fetcher... IM still lost on this.. I can whip it together and it would likely work because I trust you but I would not know why. So What I don't see is how the signal to the gate on q2 changes after q1 to make it turn on. I mean it looks to me Q1 and Q2 will just turn on at the same time.
Edit: Is it because it charges through D2 it takes longer to turn on? And the Cap C1 needs to charge before there is enough current to flow though D2 to Q2 as well the negative side of the cap is being pulled lower as the caps in the controller charge so they need to charge then C1 will fill then the current will flow though D2 to Q2?

 

[fechter]

When you turn it on, the gate voltage of Q1 climbs up to the threshold and stays there (right around 3.0V) until the controller caps get fully charged, then it starts going up again until it reaches 12v. The voltage drop across D2 prevents Q2 from coming on until Q1 is done charging. I tested this part of it.

If you look at a graph of the gate voltage vs. time, it has a flat spot for the duration of precharging.

Q1 and Q2 need to be the same type to ensure the threshold voltages are the same.

 

[Arlo1]

When you turn it on, the gate voltage of Q1 climbs up to the threshold and stays there (right around 3.0V) until the controller caps get fully charged, then it starts going up again until it reaches 12v. The voltage drop across D2 prevents Q2 from coming on until Q1 is done charging. I tested this part of it.

If you look at a graph of the gate voltage vs. time, it has a flat spot for the duration of precharging.

Q1 and Q2 need to be the same type to ensure the threshold voltages are the same.

Awesome. I think I had it understood the correct way. I will try to build this soon. I have to look at my inventory and see what I can whip together.
I will be driving some things with a 12v inverter so I might try to run the contactor off of that or I might use a resistor. But as I type this I realise how easy the 12v supply would be to use for this and I might not need the Zener ZD1 right? Because all it is there for is to stop the voltage from going to hi for the gates right?

 

[fechter]

You still need the zener diode, as the gate has to be driven from the full pack voltage through the resistor. This controls the precharge current. You could use a dc-dc between Q2 and the relay coil.

 

[Arlo1]

Ok so put the 84 v input from below the switch connected to the input of the dc-dc then the 12v output from the dc-dc to the positive of the contacter coil and because the - from the dc-dc is contactor is common - its just conneced to battery negative and all else is the same.

 

[fechter]

I would place the dc-dc input where the contactor coil is now, then wire the coil to the output of the dc-dc.

 

[ShreddinPB]

Hey guys, I am planning on building one or two of these.. I got these from digi-key
IRFB4110PBF-ND 3.57700 35.77 T
MOSFET N-CH 100V 120A TO-220AB
HTSUS: 8541.29.0095 ECCN: EAR99

I want to order some copper flat bar to connect things up.. I have ordered from onlinemetals.com before and was wondering which of these size bars would be good? Is the smallest one fine?
https://www.onlinemetals.com/merchant.cfm?id=253&step=2&top_cat=87

 

[izeman]

Don't use copper.. Go for aluminium. Its almost as heat conductive but much easier to drill and cut threads into.

 

[ShreddinPB]

Don't use copper.. Go for aluminium. Its almost as heat conductive but much easier to drill and cut threads into.

And it transfers current as well? Why is everyone using copper in this thread then?
Looking for learning

 

[izeman]

Don't use copper.. Go for aluminium. Its almost as heat conductive but much easier to drill and cut threads into.

And it transfers current as well? Why is everyone using copper in this thread then?
Looking for learning

no copper is better. by a factor of 1.5. but al is about 4 times better than fe. so still pretty good. and there are no big distances so it won't realy matter.
if you want to go the cu way - do it. if you can't get cu with the right dimensions try al.

 

[ShreddinPB]

Is this cap good for this application?
I have everything else here I think
http://www.mouser.com/Search/Produc...DDvirtualkey64700000virtualkey647-UMP1H010MDD

 

[izeman]

Is this cap good for this application?
I have everything else here I think
http://www.mouser.com/Search/Produc...DDvirtualkey64700000virtualkey647-UMP1H010MDD

any cap with the right capacity and voltage should do it. you didn't write which schematic you want to build, not how much voltage, and not how much amps you need. this is essential data

 

[fechter]

That cap would work for a system less than 50V. I prefer MLCC types because they have lower leakage and will last practically forever.
Here's a 100V 1uF MLCC: http://www.mouser.com/ProductDetail...GAEpiMZZMsh%2b1woXyUXj4dXoCGhcWKpsGj2VCq83Ew=
810-FK24X7S2A105K 1: $0.45

 

[ShreddinPB]

Is this cap good for this application?
I have everything else here I think
http://www.mouser.com/Search/Produc...DDvirtualkey64700000virtualkey647-UMP1H010MDD

any cap with the right capacity and voltage should do it. you didn't write which schematic you want to build, not how much voltage, and not how much amps you need. this is essential data

Oops!
My trike right now is 48V LifePo4, but am going to drop it to 36V.
At 48V I am running a 1000 watt motor, at 36v it will be 750 watts.

This is the one I am planning on since I have a keyswitch


Is there any negative effect using the 100V cap linked to here? Will it just take longer to turn on?

 

[ShreddinPB]

Is this cap good for this application?
I have everything else here I think
http://www.mouser.com/Search/Produc...DDvirtualkey64700000virtualkey647-UMP1H010MDD

any cap with the right capacity and voltage should do it. you didn't write which schematic you want to build, not how much voltage, and not how much amps you need. this is essential data

On another note, I also have a 48V 3000 watt motor custom build I am going to do.. what size cap would that require?
I got 10 of the mosfets
MOSFET N-CH 100V 120A TO-220AB
how many of those should I use in each installation?

 

[izeman]

Is this cap good for this application?
I have everything else here I think
http://www.mouser.com/Search/Produc...DDvirtualkey64700000virtualkey647-UMP1H010MDD

any cap with the right capacity and voltage should do it. you didn't write which schematic you want to build, not how much voltage, and not how much amps you need. this is essential data

On another note, I also have a 48V 3000 watt motor custom build I am going to do.. what size cap would that require?
I got 10 of the mosfets
MOSFET N-CH 100V 120A TO-220AB
how many of those should I use in each installation?

answering your question from the last post: YES you can use the 100v cap as well. 100v is the MAX voltage. so it wil handle 36v or 48v fine. you could order different values of the cap and try them out. the higher the value the longer the switching process should take.
the mosfet data you're posting is no data that helps. you have to post the type. eg irfb4110 or 3077 or something.
for your 1000w 36v application i would use IRFB3077 which is good for 75v 120a with an internal resistance of 2.8mohm which is quite low.
connected to a small heatsink of eg 20x5x10mm al or co you will not need more than 2 of them parallel. one should is enough, but than it needs proper cooling.
for my 80v and 60a bike (which is 4x the power what you ask) i have 4x irfb4110 and they don't even get warm.

 

[ShreddinPB]

answering your question from the last post: YES you can use the 100v cap as well. 100v is the MAX voltage. so it wil handle 36v or 48v fine. you could order different values of the cap and try them out. the higher the value the longer the switching process should take.
the mosfet data you're posting is no data that helps. you have to post the type. eg irfb4110 or 3077 or something.
for your 1000w 36v application i would use IRFB3077 which is good for 75v 120a with an internal resistance of 2.8mohm which is quite low.
connected to a small heatsink of eg 20x5x10mm al or co you will not need more than 2 of them parallel. one should is enough, but than it needs proper cooling.
for my 80v and 60a bike (which is 4x the power what you ask) i have 4x irfb4110 and they don't even get warm.

Sorry, I only copied one line of the description.. the rest is here
IRFB4110PBF-ND 3.57700 35.77 T
MOSFET N-CH 100V 120A TO-220AB

So, if I am understanding, these should be good on either the 48v 1000w or 3000w system. (should be good on all systems I think? but there is a better one to use on the 36v system)
I am however having trouble figuring out how many to use..
the 48V 3000W motor would be 62.5A right? so I can use 4x the irfb4110's that I already have and that should be fine.
the 48V 1000W motor would be 20.8A right? so I can use 2x the irfb4110's that I already have and that should be fine.

I will order some of the smaller ones for the 36V system. And attempt to customize the cap size also

These caps for the 36V system
http://www.mouser.com/ProductDetail...=sGAEpiMZZMuAYrNc52CMZGTZkznjTP8ywHVPWe4QhgU=
the fet for the 36V system
http://www.mouser.com/ProductDetail...EpiMZZMshyDBzk1/Wi5%2bqVgN3%2bWS8VfDFl6SHiKM=

 

[fechter]

Sorry, I only copied one line of the description.. the rest is here
IRFB4110PBF-ND 3.57700 35.77 T
MOSFET N-CH 100V 120A TO-220AB

So, if I am understanding, these should be good on either the 48v 1000w or 3000w system. (should be good on all systems I think? but there is a better one to use on the 36v system)
I am however having trouble figuring out how many to use..
the 48V 3000W motor would be 62.5A right? so I can use 4x the irfb4110's that I already have and that should be fine.
the 48V 1000W motor would be 20.8A right? so I can use 2x the irfb4110's that I already have and that should be fine.

I will order some of the smaller ones for the 36V system. And attempt to customize the cap size also

Yes, those will be good for either voltage. For rough approximation, I usually figure about 20A max per FET (for 4110s), which will make about 1.5W of heat for each one. This is roughly the free air dissipation for that package size. A small heat sink will be helpful either way, unless these are located where they will get lots of air flow.

A 100V rated cap will work at any voltage lower than that. The precharge time is a function of the cap value and the system voltage. I doesn't matter much what FETs are used. 1uF at 48v will have about a 1 second charge time. This should keep even a single FET well within the safe operating area during precharge. The 100V MLCCs I spec'd above are not much more expensive than the 50V rated ones and will be great if you want to run at either voltage. Keep in mind a 48v battery might be well over 50v hot off the charger.

 

[ShreddinPB]

Yes, those will be good for either voltage. For rough approximation, I usually figure about 20A max per FET (for 4110s), which will make about 1.5W of heat for each one. This is roughly the free air dissipation for that package size. A small heat sink will be helpful either way, unless these are located where they will get lots of air flow.

A 100V rated cap will work at any voltage lower than that. The precharge time is a function of the cap value and the system voltage. I doesn't matter much what FETs are used. 1uF at 48v will have about a 1 second charge time. This should keep even a single FET well within the safe operating area during precharge. The 100V MLCCs I spec'd above are not much more expensive than the 50V rated ones and will be great if you want to run at either voltage. Keep in mind a 48v battery might be well over 50v hot off the charger.

So, the caps came in, they are MUCH smaller than the ones in the stencil on page 6. Is this just because they are MLCC?

 

[fechter]

So, the caps came in, they are MUCH smaller than the ones in the stencil on page 6. Is this just because they are MLCC?

Yes. The size I put on the board layout would be big enough for an electrolytic or polystyrene cap. MLCCs are much more compact. They turn out to be about as cheap as the other types too, and much more heat resistant.

 

[Arlo1]

Shit I have to go re read this whole thread to see what I need for q1 and q2. 4115s won't work will they because they are npn and I need pnp? I want to use something rated over 100 v but if I can't for now its not a big deal. I just blew two dc-dc supplies one is fixed and the other I'm not sure the problem.

Fetcher what do I need for q1 and q2?

 

[Arlo1]

OK I got it working with 4115s for q1 and q2. My problem is the 12v supply off the DC/DC supply also powers the controller drivers and all 12v stuff in the controller.
So when I hook the battery up (using a power supply for testing) it causes the CA to read full voltage and the contactor pulses a few times. If I disconnect the 12v supply to the controller its fine then I can connect the 12v supply after words how to I get around this? A second pre-charge for the controller supply?
As well when I connect the battery voltage + to the pre-charge it causes a spark because the DC/DC is also charging its caps its not a big spark but it could cause a switch to wear out fast.

 

[Njay]

Pre-charge/inrush is a PITA do implement "afterwards". I have a "double relay" system on a mtorcycle DC/DC converter. Pre-charge must be done without load on the converter, otherwise it won't pre-charge enough to prevent sparking. One relay controls the converter's input, and only then the other at the converter's output will turn on.

 

[fechter]

OK I got it working with 4115s for q1 and q2. My problem is the 12v supply off the DC/DC supply also powers the controller drivers and all 12v stuff in the controller.
So when I hook the battery up (using a power supply for testing) it causes the CA to read full voltage and the contactor pulses a few times. If I disconnect the 12v supply to the controller its fine then I can connect the 12v supply after words how to I get around this? A second pre-charge for the controller supply?
As well when I connect the battery voltage + to the pre-charge it causes a spark because the DC/DC is also charging its caps its not a big spark but it could cause a switch to wear out fast.

4115s should work fine for Q1 and Q2.

It should handle the 12v coverter, but possibly it won't work off your power supply. It should not start pulsing. That sounds like the voltage is dropping out when the contactor pulls in.

Double check the wiring too. There should not be any spark when you engage the switch. The switch goes to a 1M resistor and the contactor FET should be off. Bad Q2? You could get a spark if you turned the switch off, then back on again before the contactor drops out.

If the wiring going from the circuit to the contactor and battery was too small, it could possibly have enough drop to cause problems.

I'll re-post the schematic, since it's back a few pages:

EDITED: wrong schematic... see next page.

 

[Arlo1]

Fetcher the 12 volt wire going to the controller has battery voltage on it untill I hook it to the 12 v output from the dc/dc converter. So this is also where my contactor + is hooked so before the dc/dc turns on even with the switch off it pulses the contactor. To fix this I need a relay between the 12v supply and the controller 12v input wire.

The spark is because where I put the dc/dc is in place of the contactor you show. So when I turn the switch on it sends 84v to Tue caps in the dc/dc causing a spark.