Electric Unicycle Project - Power MOSFET Question

Thread Starter

joe426

Joined Aug 11, 2009
39
I'm working on a charging circuit for a self-balancing unicycle I'm building and had a few questions about power MOSFETs (specifically, I'm thinking of using the FQP27P06 - 60v, 27A max).

Questions:
  1. Are there any consequences of leaving a MOSFET "on" for long periods of time (~2 hours)?
  2. Can you run current "backwards" through a MOSFET (i.e. drain to source when charging the batteries)?
The max current while discharging the batteries through the MOSFETs is 25A. Max current while charging the batteries is 1A.

Here's a preliminary design for the charging circuit:



The 3mm terminals are where 22v LiPo batteries will be attached.

Thanks!
 

thatoneguy

Joined Feb 19, 2009
6,359
As long as the MOSFETs are driven completely on, they can stay on within their current rating indefinitely, as long as they have an adequately sized heatsink on them so they do not overheat.

For charging, you need to connect directly to the battery pack from the charger.

LiPo Batteries are VERY picky about charging and keeping a balanced charge between cells, in addition to temperature monitoring. They aren't like NiMH cells where you can put voltage on them and wait until they are full, doing that with LiPo makes fire.

I'd suggest getting a dedicated charger for the LiPo battery pack, or at least a charge control IC with temp and current sense inputs to control the charge rate, preferably with cell balancing as well.
 

wayneh

Joined Sep 9, 2010
17,496
[*]Can you run current "backwards" through a MOSFET (i.e. drain to source when charging the batteries)?
I'll be curious to see what others say, but my understanding is that yes, the body diode of a typical MOSFET will conduct in the "reverse" direction to the way we usually hook them up, with source at lower voltage than drain. If source voltage is HIGHer than drain, current flows the body diode. This is how a MOSFET can be turned into an ideal diode in active rectification. First you you use arrange the MOSFETS as a rectifier relying only on the body diodes. Then you use comparators to detect voltage across each body diode, and turn on the FETs when appropriate. This eliminates the voltage drop across the diode - a big source of power loss at low voltage - and allows much larger currents to flow through the MOSFET.
 

daveyjones97

Joined Oct 21, 2011
46
im an ameteur, but interested in this sort of thing. would it possible to add capacitors to the start of charge circuit so you could capture as much energy as possible going down hills or big stops without resorting to braking methods that dont harvest waste energy? i have no idea what size or rating you'd need but i think that would help battery management and run times.
 

shortbus

Joined Sep 30, 2009
10,045
While not a question you asked and no one so far has brought up, it is common and aids reliability, to derate a mosfets current. Your 25A is real close to the mosfets 27A max rating. Its better to go with a mosfet with a higher value of amperage. But thats your call.
 

T.Jackson

Joined Nov 22, 2011
328
As long as the MOSFETs are driven completely on.
Yeah that's right. Believe it or not, an audio amplifier will run hotter when the volume is at mid stick as opposed to being at full throttle.

This is because only little 'e' in the junction is posing an opposition to current flow.

Can someone explain this better, expression is not my best.
 

crutschow

Joined Mar 14, 2008
34,280
The main reason to go with a larger MOSFET is that it has a lower "on" resistance and is thus more efficient (which dissipates less power that has to be carried away by the heat sink).
 

T.Jackson

Joined Nov 22, 2011
328
MOSFET is expensive. BJT is cheap. Depends what it is interfacing to I guess.

FQP27P06 seems cheap though at < $2 for 1pcs.
 
Last edited:

SgtWookie

Joined Jul 17, 2007
22,230
Check out Chuck McManis' page on power MOSFET ratings:
http://mcmanis.com/chuck/robotics/projects/esc2/FET-power.html
...and bookmark that page. Anyone considering using power MOSFETs for a project should read that, unless they like smelling smoke.

To keep a BJT in saturation, you need to supply the base with 1/10th of the desired Ic. Base current is sort of like a "tax" - you need to supply current in order to get 10x more current; as bjt's are current controlled devices.

On the other hand, power MOSFETs are voltage controlled. The gates are more or less like little capacitors. For N-ch MOSFETs, you charge the gate to 10v using the source terminal as a reference point (called Vgs) to turn it ON (5v for logic level) and Vgs=0v to turn it OFF. MOSFET datasheets have a threshold Vgs; if Vgs is at or closer to zero than the specified Vgs then the MOSFET is considered OFF.

P-ch is basically the same, except Vgs is expressed in negative voltages.

Once the gate is charged or discharged, it doesn't require power to maintain them at that voltage; as the gate is extremely high impedance.
 

strantor

Joined Oct 3, 2010
6,782
I'll be curious to see what others say, but my understanding is that yes, the body diode of a typical MOSFET will conduct in the "reverse" direction to the way we usually hook them up, with source at lower voltage than drain. If source voltage is HIGHer than drain, current flows the body diode. This is how a MOSFET can be turned into an ideal diode in active rectification. First you you use arrange the MOSFETS as a rectifier relying only on the body diodes. Then you use comparators to detect voltage across each body diode, and turn on the FETs when appropriate. This eliminates the voltage drop across the diode - a big source of power loss at low voltage - and allows much larger currents to flow through the MOSFET.
I've never confirmed it, but the understanding I got from reading several manufaturer's app notes, is that when a mosfet is ON, current can flow either way through the junction (D to S or S to D), independent of the body diode; its not directional like a BJT. When it is off, current can flow from D to S, but only through the body diode.
 

thatoneguy

Joined Feb 19, 2009
6,359
I've never confirmed it, but the understanding I got from reading several manufaturer's app notes, is that when a mosfet is ON, current can flow either way through the junction (D to S or S to D), independent of the body diode; its not directional like a BJT. When it is off, current can flow from D to S, but only through the body diode.
This is what I've got for an answer as well, was too worried about the burning batteries in the first reply.

Off the body diode will act as a rectifier (conducting) while OFF, while ON, it acts like a low resistance resistor. Some SSRs (Solid State Relays) use MOSFETs instead of Triacs for faster switching.

For switching low frequency AC from a logic circuit an optoisolated Triac driving a high current capable Triac is the suggested method unless one really knows what exactly they are doing.
 

Thread Starter

joe426

Joined Aug 11, 2009
39
I'd suggest getting a dedicated charger for the LiPo battery pack, or at least a charge control IC with temp and current sense inputs to control the charge rate, preferably with cell balancing as well.
I was planning on charging via a dedicated charger, yes. If I have a dedicated charger, could I still charge via the MOSFETs?

Your 25A is real close to the mosfets 27A max rating. Its better to go with a mosfet with a higher value of amperage. But thats your call.
Good point, I'll go with one rated for a higher current. Thanks.

Current controlled voltage source (BJT) Vs a voltaged controlled current source (MOSFET)

Why use a MOSFET?
There is no reason in particular. Should I use a BJT instead? Would that be more efficient/better in some way?

Check out Chuck McManis' page on power MOSFET ratings:
I'll check it out, thanks.
 

thatoneguy

Joined Feb 19, 2009
6,359
i would suggest connecting directly to the battery rather than going through the MOSFETs. Going through them, you won't know for sure they are on or off, and there are losses in a MOSFET that will confuse the charger and possibly cause it to behave badly.
 

Thread Starter

joe426

Joined Aug 11, 2009
39
i would suggest connecting directly to the battery rather than going through the MOSFETs. Going through them, you won't know for sure they are on or off, and there are losses in a MOSFET that will confuse the charger and possibly cause it to behave badly.
Alright, thank you.
 

wayneh

Joined Sep 9, 2010
17,496
Correction: from S to D
Glad to hear that we all share an understanding of how MOSFETs behave. The only hesitation I had was due to a recent exchange here, where russpatterson did not observe conduction via the body diode. But I may have misunderstood his experiment.
 

strantor

Joined Oct 3, 2010
6,782
Glad to hear that we all share an understanding of how MOSFETs behave. The only hesitation I had was due to a recent exchange here, where russpatterson did not observe conduction via the body diode. But I may have misunderstood his experiment.
Yeah I remember reading some mumbo-jumbo about minority carriers and majority carriers bein the reason why MOSFETS are bidirectional but I didn't get it. I just skipped to the end where they summarize in terms that are closer to laymans terms.
 

Thread Starter

joe426

Joined Aug 11, 2009
39
One question which wasn't answered above (and I'm still interested in knowing the answer) is: Should I use a BJT instead of MOSFETs? Would that be more efficient/better in some way?"
 

strantor

Joined Oct 3, 2010
6,782
One question which wasn't answered above (and I'm still interested in knowing the answer) is: Should I use a BJT instead of MOSFETs? Would that be more efficient/better in some way?"
no. BJTs need to drain current through the base in order to drain current through collector & emitter. MOSFETs don't. They just need voltage. theoretically, you could connect a 9V battery to a mosfet gate and it would conduct indefinately, because the mosfet doesn't drain current through the gate.
 
Top