I currently use a IRLB3036PBF-ND mosfet to switch on a load fed from 2x Li-Po packs in Series.
I use the same setup to fire 2 x Li-Po packs in parallel.

My current setup is a
15k resistor from gate to source.
(-) batt. tied to source
(+) batt. tied to the laod
(+) batt tied to Off-MOM switch then to Gate

To the best I understand this is a N chan Mosfet meaning it switches the Neg of the power source?

OK here is the big question.
If a N chan switches negative could I use a combination N and P with a multi pole switch and add a 2 position switch that would allow me to change from Series Firing to Parallel Firing by just throwing a switch instead of having to reconfigure my battery?

I use the IRLB3036PBF-ND because it has never failed me and can handle the 200 watt load and fire @ 2.5volts.

I can picture this working in my head with 4 or 5 fets wired up correctly but I have no idea what to get as far as a fet the will switch the pos. A simple crude schematic and Digi-key part number would be killer.

TY in advance to anyone willing to help.

 

To the best I understand this is a N chan Mosfet meaning it switches the Neg of the power source.

Wrong, as far as I can tell. (They are much more versatile than that.) The N means stuff about the internal structure, but the part you need to know is that N turns this one off. What? It's already off until I give it positive on the gate!

Yeah, right. This one is an enhancement type. You have to enhance its conductivity with positive from gate to source. That still means N is in the "off" direction.
There are also depletion types, normally on, and they need to be given (N) negative voltage on their gate to turn them off.

For the purposes of beginners, N-channel merely means Negative voltage at the gate is the "off" direction.

Clear as mud?

 

@WoodNMetalArt
From my understanding of what you want, will FUNCTION as the drawing below. Parallel to the left, series on the right.

You want to do it with MOSFETS and a small switch because your current is so high that switches or relays would be quite heave and big. A mosfet allows a small switch to be used.

I agree with your approach. Note that you will need three MOSFETS. You must also be careful not to activate adjacent MOSFETS at the same time. You need switches that "break before make" contact. Also, you should use some high value resistors to tie gate to source to prevent them from floating (remaining in their previous state) during the switching process.

I'll leave the exact mosfet design up to you or the next posters..

You will need N-type for far left switch, n- or p-type for middle (depending on where you tie the gate) and p-type for right switch.

You may (will?) need some ground reference resistors for the middle mosfet, depending how it is designed and where your switch connects. Just be careful, it looks much simpler with switches than with MOSFETS - at least from my initial inspection.

 

Yes you nailed it. I use the 3036 to switch a 40 amp DC load. In one configuration I parallel two 4.2v Li-Po batteries for the mah boost. There are times I need more power and run the same setup in series for 8.4v. Currently I have to change my battery configuration.

To combat the adjacent firing I could use a 3 position switch with the center position as an OFF.

in my current setup I am switching the negative on and off

 

Ok so I came up with this. It may have some errors because I had to do it in paint as I don't know how to use the schematic simulators.

The Red Circled area is whats stumping me. I assume I need a MosFet or something to switch on and off the POS to connect in parallel and separate in series.

 

Anyone able to assist me on how to or what to use to switch in this positive side of the battery?

 

If you want to using a h-bridge for motor, you have to using p-Mosfet for the upper bridge.
You could google : mosfet h-bridge for motor driver.

 

@ScottWang
You'll have to read the first post. He wants to flip his batteries from a serial to parallel configuration.

@WoodNMetalArt
How many volts is each battery pack?
What is the total current draw for series and parallel connections?

 

The circuit as below, but I haven't tested yet, you could go to digikey to find the p ch mosfet for your need.

 

@ScottWang
You'll have to read the first post. He wants to flip his batteries from a serial to parallel configuration.

Thank you GopherT.
I was dizzy when I posted, I was also feel very strange during that time, why you not give him a suggestion as I did, but I still tried to using the similar way as H-bridge to think and solve the problem, but the result is quite different.

 

Thank you GopherT.
I was dizzy when I posted, I was also feel very strange during that time, why you not give him a suggestion as I did, but I still tried to using the similar way as H-bridge to think and solve the problem, but the result is quite different.

Because I could describe it with some simple switches but, once I started with the MOSFETS, it made my head hurt so I stopped. I have a half-done drawing. Thank you for pulling that together. Hopefully he returns to see and use your nice work.

I don't think the thread starter had any idea it would be so complicated to do this.

 

Because I could describe it with some simple switches but, once I started with the MOSFETS, it made my head hurt so I stopped. I have a half-done drawing. Thank you for pulling that together. Hopefully he returns to see and use your nice work.

I don't think the thread starter had any idea it would be so complicated to do this.

Comparing to his 4 mosfets circuit, mine is just plus one 2 bjts and some resistors.

 

I don't think the thread starter had any idea it would be so complicated to do this.

Agreed. Unfortunately, even the post #9 circuit can't do the job. The problem is the inherent body diode of FET2, which allows a large current to flow via FET1, short-circuiting V2 when switch Sw1 is in the position shown. Adding a series diode to prevent this would be only a partial cure, because it would have to handle 40A and would inevitably drop a significant voltage. Back-to-back FETs might be an answer, but gating those would be even more complicated.

 

Agreed. Unfortunately, even the post #9 circuit can't do the job. The problem is the inherent body diode of FET2, which allows a large current to flow via FET1, short-circuiting V2 when switch Sw1 is in the position shown. Adding a series diode to prevent this would be only a partial cure, because it would have to handle 40A and would inevitably drop a significant voltage. Back-to-back FETs might be an answer, but gating those would be even more complicated.

Alec_t, thanks for pointed out my mistake.
I met the same problem before and used the diode as you mentioned, and that is just a light current compared to here, and another problem is the voltage drop from diode may affecting the load.

Another way to solve the problem that you said was back to back, although I knew that method, but I didn't try it, because I'm afraid of it could be damaged the mosfet, have you try it in the real world?

 

the voltage drop from diode may affecting the load

It does indeed. Even if a Schottky diode is used the voltage drop prevents the two batteries sharing the current load in parallel mode: V2 provides all the current.
I haven't tried back-to-back FETs in the real world.

 

ok so a best I understand a P-Fet is like the N-Fet I use currently? Still 10-15k source to drain and applying power to gate to "switch it on" but the N switches neg with pos gate and the P switches Pos with a grounded gate?

The batteries are 4.2v
Load on 4.2v is .2 ohms
Load on 8.4v is 1 ohm

if this gets overly complicated I could just go with a DC-DC off the shelf.

 

ok so a best I understand a P-Fet is like the N-Fet I use currently? Still 10-15k source to drain and applying power to gate to "switch it on" but the N switches neg with pos gate and the P switches Pos with a grounded gate?

The Rgs resistor was designed to turn off the mosfet, and the bjts was designed to turn on the mosfet, there are just as electronics switch.

I was rechecked the circuit and found it needs one more diode to keep the V2 and V1 have a balance voltage, and when the current too big, the voltage drop of diode will be more as 1.0V or 1.1V, you could choosing the best way to do for yourself, I just offering another option, the light gray is the in parallel for batteries, the circuit as below:

 

For Scott"s circuit in Post #14:
To avoid the diode drop add an extra P-MOSFET in series with FET2 (source-to-source). This works because FETS conduct equally well in both directions when biased ON.
Connect R4 between the junction of the two sources to the two gates connected in parallel.
Remove the diode.

The batteries may still experience some imbalance in current when connected in parallel, due to the differences in ON resistance between FET3 and the two FETs in series.
The normal internal battery resistance should help reduce that.

 

For Scott"s circuit in Post #14:
To avoid the diode drop add an extra P-MOSFET in series with FET2 (source-to-source). This works because FETS conduct equally well in both directions when biased ON.
Connect R4 between the junction of the two sources to the two gates connected in parallel.
Remove the diode.

The batteries may still experience some imbalance in current when connected in parallel, due to the differences in ON resistance between FET3 and the two FETs in series.
The normal internal battery resistance should help reduce that.

The new circuit as below, I haven't try this before, only saw the circuits, although still a little worry, but I rechecked some other big current applications, so it could be fine, I should try it someday.

 

I will give this a try next time I order from DigiKey.

In Parallel it is 4.2 volts pulling 70 amps
In Series it will be 8.4 Pulling roughly 100 amps.

My Nch mosfets should work but I see you have added 3 Pch.
Can you point me in the right direction for one rated comparable to the Nch (IRLB3034PBF-ND)
I can wiggle with the amp rating down to maybe 35-50 amp range but even then I seam to only find surface mounts rated well enough. I'm not upto task to be soldering tinny tinny surface mounts.