Use a P-Channel Jfet as a variable resistor? / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Use a P-Channel Jfet as a variable resistor?

1 2 next

Move Thread

LAN_403

GammaRay

Mon Dec 31 2012, 07:31PM

Registered Member #5323 Joined: Fri Jun 15 2012, 02:14PM
Location:
Posts: 104

I checked the archive before posting. I'm powering a tiny 3v DC motor/fan blade with an AeroGel Super Cap (5v@5F) and would like the motor/fan to spin as long a possible between chargings. Torque and RPM is not the issue, but obtaining the maximum length of operating time is the goal even if the motor turns at a leisurely slow speed, that is okay. The motor/fan spins at an acceptable RPM all the way down to 0.4v. (Already tried a PWM approach, but the PWM circuit I have requires 5v to operate and fails at 4.5v).

I'm wondering if a P-Channel Jfet could be used as a variable resistor to limit the motor operation to a slow, yet constant speed throughout the Super Cap's 5v-0.4v discharge cycle? Sanity check requested. Thank you.

Adam Munich

Mon Dec 31 2012, 07:35PM

Registered Member #2893 Joined: Tue Jun 01 2010, 09:25PM
Location: Cali-forn. i. a.
Posts: 2242

Since a resistor would just burn any power not consumed by the motor, you're not going to get a longer run time by using any form of resistance, regardless of what the component is. All you'll have is just a motor with less torque.

Try a PWM circuit with a CMOS 555 timer and a logic-level mosfet. That should accomplish what you're after.

Myke

Mon Dec 31 2012, 08:16PM

Registered Member #540 Joined: Mon Feb 19 2007, 07:49PM
Location: MIT
Posts: 969

Gren wrote ...

Since a resistor would just burn any power not consumed by the motor, you're not going to get a longer run time by using any form of resistance, regardless of what the component is. All you'll have is just a motor with less torque.

You do get less torque due to the reduced current but that would also increase the time that it spins since the energy is drained from the cap at a slower rate. It's just not very efficient.

Gren wrote ...

Try a PWM circuit with a CMOS 555 timer and a logic-level mosfet. That should accomplish what you're after.

He did try PWM but it had problems as he described. Maybe try to use separate caps for supplying the switching control so that the lower voltage doesn't cause the logic to cut out? You could also try to use 3.3V logic things which should give you more headroom. With this, you could probably get a lot longer run times since you have more control of the rate of draining the cap. You could run the motor at the minimum acceptable RPM meaning short pulse lengths which also might improve the switching control current drain (depending on what you use for PWM).

Sulaiman

Mon Dec 31 2012, 08:21PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

A series resistor would increase run time a little by running ata lower current
it would be good to short out the resistor at low capacitor voltages
For ultimate long run-time it comes down to stored energy
0.5 x 5 x 5^2 = 62.5 Joules
being delivered to the motor at minimum reliable power, e.g.
1V x 10mA = 10mW
If you could convert all of the stored energy at 100% efficiency
Run Time = Energy/Power = 62.5/0.01 = 6,250 sec = 1.736 hours .. etc.
If you make a 'Buck' converter that works down to 1.5V input
9% of the capacitor energy will be remaining.
e.g. 91%Energy x 63%Efficiency = 1 hour @ 1V/10mA
I'd expect there to be loads of ICs and fierce commercial competition in this area.

any special reason for p-channel jFet?
this is not the first type of transistor that would have come to my mind for such an application.
an n-channel jFet such as j310 (or similar) could be used as a sort of linear regulator,
but I think that an LDO linear regulator is cheaper!.

I'd start by looking at 'Energy Harvesting' technologies. e.g.

P.S.
two (or maybe three) red led in series with the motor would be almost as good as a resistor
(IF the motor current is less than the maximum allowed led current)
shorting/switching an led out as the capacitor voltage falls would extend run time.
Simple and easy to demonstrate BUT definitely not as efficient as a switch-mode converter.

Conundrum

Tue Jan 01 2013, 02:02AM

Registered Member #96 Joined: Thu Feb 09 2006, 05:37PM
Location: CI, Earth
Posts: 4061

N channel would work equally well, presumably here he is trying to use high side switching.

I've used an LM317LZ before as a temperature sensitive fan controller and this sort of worked.

Daedronus

Tue Jan 01 2013, 10:05AM

Registered Member #2329 Joined: Tue Sept 01 2009, 08:25AM
Location:
Posts: 370

why not use the motor itself as a switch?

Dr. Slack

Wed Jan 02 2013, 05:36AM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

Important things being equal, like the same motor and the same power source, you will get longer run time from a well-designed PWM than you will from any sort of variable resistor.

However, if you don't have a well designed PWM, but one that stops working in a region where you still want it to be working, then you need a better design of PWM.

For operation at very low input voltages, a PWM will effectively have two outputs, one for its own bias supply, and one for the load. Obviously there is a startup problem if you want it to start at very low volts, but for converting a discharging capacitor, it sounds like there's initially enough voltage to make it work.

If your motor will run at 0.4v, then being able to PWM/buck from 5v to 0.4 instead of drop from 5v to 0.4 will give you 10x the run time.

GammaRay

Fri Jan 04 2013, 02:43PM

Registered Member #5323 Joined: Fri Jun 15 2012, 02:14PM
Location:
Posts: 104

Here's what I've found so far. I used a Decade resistance box to manually guadually reduce resistance from the Supercap to the motor as the Supercap discharged from 5v down to 0.4v. Although quite wasteful, it did cause the motor to operate longer by a factor of 2 by causing the motor to turn quite slowly. Alternatively, I used a buck/boost circuit to provide voltage to a PWM circuit, but alas, the combined current drain of the buck/boost and the PWM results in dismal time duration (worse time than with no circuit at all). I agree that a properly designed PWM that operates to very low voltage is my best bet. I would like to purchase an off-the-shelf & assembled PWM that will operate down to around 1-2 volts (if such a circuit exists) can anyone point me in the right direction?

Adrenaline

Fri Jan 04 2013, 03:16PM

Registered Member #235 Joined: Wed Feb 22 2006, 04:59PM
Location:
Posts: 80

I haven't researched a ton or know all of your requirements, but Linear has some devices:

Proud Mary

Fri Jan 04 2013, 03:27PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Have you looked at low voltage motor driver ICs like DRV8832?

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob