variable speed control for 12 volt fan

[JohnEd]

Guys,

I use a auto radiator cooling fan for ventalation and cooking exhaust.  At 12v it ROARS but it evacuates the coach in a New York minute.  Close the door and your ears pop.  I have an ignition dropping resistor in series with it that is switched but it is either slow or super sonic.  I shy away from a reostat because it consumes as much power as the max draw all the time.  The advent of the SCR allowed speed control with a small, cool running device in 115 v circuits.  Are these available for 12 and 24 volt systems?  I have had this thing installed since 1990 and when I am living in the coach it runs slow for 24/7 and at higher speeds during some of the day.  Real happy with the longivity.  Speed controls for 15 amp devices???  Anybody!

Thanks,

John

[DrivingMissLazy]

Guys,

I use a auto radiator cooling fan for ventalation and cooking exhaust.  At 12v it ROARS but it evacuates the coach in a New York minute.  Close the door and your ears pop.  I have an ignition dropping resistor in series with it that is switched but it is either slow or super sonic.  I shy away from a reostat because it consumes as much power as the max draw all the time.  The advent of the SCR allowed speed control with a small, cool running device in 115 v circuits.  Are these available for 12 and 24 volt systems?  I have had this thing installed since 1990 and when I am living in the coach it runs slow for 24/7 and at higher speeds during some of the day.  Real happy with the longivity.  Speed controls for 15 amp devices???  Anybody!

Thanks,

John

The ignition dropping resistor is the same as a rheostat which is just a variable dropping resistor. They work exactly the same except one is variable and one is fixed. NAPA should have a rheostat that will solve your problem. As I recall they are about 25 watts and 3-400 ohms.
Richard

[Nick Badame Refrig/ACC]

Hi John,

I used theese dimmers in my bus to dimm my 12v puck lights in the ceiling. They can handle 70watts each.

They can easly do a 12v fan...

Sailor Sam's is one of my favorite places for bus lighting needs.
http://www.sailorsams.com/mall/dimmer_d1201.asp

Good Luck
Nick-

[boogiethecat]

Type "dc pwm control" into ebay and you'll find exactly what you want.   PWM is similar to scr technology but works on DC, and doesn't get hot and waste energy like rheostats will.

[JohnEd]

Richard,

I have the dropping resistor installed and it lacks control.....either roar or coasts.  Thanks though.


Nick,

Thanks for that bgreat reference.  I will get those unless i can do better on ebay.


Boggie,

You are the man.  I didn't have a clue what i was shopping for in terms of a "term".  I'll try ebay right now.

Thanks

John

[Sojourner]

Bottom line never use under rated DC speed control for your application. If your vent fan has amps rate on label...such as a example...5 amps at full speed then you need at least 2/3 or higher amperage rating variable control for starting current (5 by 3 x 2 + 5 = 8.3). You can add reversible DPDT switch with equal amperage or higher diode connected to both variable's output and installed switch before motor.
DC speed control w/on-off:
http://www.zaneinc.com/user/Ds-ams-l.pdf

Or my choice that I have use for many years on electronic projects. It never fail as long as you use diode rated for your needs...double voltage & amperage rating to control power surge via switching & motor starting. It usually cost less:
http://www.cpemma.co.uk/sdiodes.html

All the above will work on dc light system but it non-peak-surging....meaning you can use diode rated at 10% higher whatever it required.

[Tim Strommen]

I agree with Boogie - "PWM" is the way to go.

PWM send full voltage pulses of 12 or 24 (what ever your system is at) but varies the "on-time" to the load to give it a net "current regulation" effect (if you imagine that a motor running at 12-volts 100% of the time is drawing 5Amps over time, a motor given 12-volts for 20% of the time would draw 1Amp over time).  This is the way most new devices work - as motor especially operate better when given their design voltage.  This is also how LEDs are being dimmed these days (since they need an appropriate voltage AND current to put out the right color/ammount of light).

Most PWM devices these days are using MOSFETs which only have a dropping voltage of about .25-volts, which really gives merrit to the efficiency of these controls.

Cheers!

-Tim

P.S. Below is a simple Motor Controller concept (for those who want to do it themselves).  This is a single low-side MOSFET wich only needs a 5-12volt input signal to turn the load (in this case a motor) on or off.  By putting in a Pulse-Width-Modulation signal (either variable frequency or variable duty cycle) on the "pulse" line, you can vary the speed of the motor or dim a light.  I like low-side because the heat sink on most N-Channel MOSFETs ends up being electrical ground as well (preventing shorts). The source of said pulse can be anything from a 555-timer circuit, or a micro processor, or my favorite a saw-wave generator with an analog voltage from a potentiometer going into a comparator to generate a fixed-frequency/variable duty-cycle PWM signal (I can put up a schematic if enough requests come down...).

Oh - and remember to use a fuse (not shown in the drawing below!)-T

[Stan]

My experience with pwm controls on DC motors was RF noise so I just used a MOSfet with a pot in to vary the input signal. The other drawback to any solidstate method is you cannot get full speed. It is surprising how much speed that last volt (dropped across the device) increases.

[JohnEd]

Well, remember that old...."Too soon old, too late smart"?  I am still on board with that one.  I went to ebay andlocated a few of those pwm's for sale/bid.  I found one that owuld handle 12 amps and was no bigger than a quarter and in "kit" form no less.  Lucky me and cheap.  Reading everything I could find about PWM's TODAY, and after winning the auction, I learned MORE.  It seems that dc motors don't work well with "high frequency switched dc" as the moror windings are basicallly coil inductors and tend to block higher(?) frequencies than DC.  Sojourners steer to ZANE INC or ZANEINC.COM sells DC speed controlers specially built for DC motor speed control.  The discussions I have read indicate that DC motors can be most effectively controlled with a switching freq of 400 to 800 cps.  My pwm is running 4.5KHrz and that is modest as these devices go.  It is specified as a DC motor controler but further research indicates that might be a really small motor.  Mine will always be usefull for controling interior lights so all is not lost.

Should my device fail, Sojourners recommendation will certainly work.  Thanks, Guy.

To all, thank you for you info and help.

John

[Tim Strommen]

Stan,

EMI/RFI from PWM drives mostly comes from the "switch", the "line" and the "load".  With proper decoupling at the switch supply (like a cap between the +supply and -return path) and a choke on the supply line - a lot of potential EMI/RFI can be avoided.  For an application like PWM control on a motor things as simple as having a supply and return wire for the motor (not grounding the motor straight to the bus chassis), and twisting the wires will help a lot more too.  Then, placing a 1nF and 10nF ceramic "disc" capacitor right at the motor terminals (or the brushes if you can get to them) to be controlled to cut down any noise from the brushes (again, if applicable) will also help.

As for the PWM itself, a common way around a single-frequency radiation (especially at about 50% duty cycle when a PWM looks like a square-wave) - is to use "spread-spectrum" PWM, where the duty cycle is selected by the user, but the PWM frequency for each cycle is randomized to lessen a specific frequency peak.  This is where a microprocessor would show its strengths as a signal generator for a MOSFET control.

As for losses with a MOSFET (solid-state switch) my favorite N-Channel MOSFET, the International Rectifier "IRF3805" drops about 33mV with a 10-Amp load - which means you're more likely to suffer "significant voltage loss" from the transmission lines to the load than the "switch".


JohnEd,

Don't worry too much about an "experiment", we all make mistakes .  Anything between 75-80kHz should be okay for a motor control.  If you have the ability, try to locate the control itself very close to the motor being controlled.  This will help to reduce the potential for emissions (EMI/RFI), and is a popular approach to robotic "stepper" controllers nowadays... (that is - motor integrated control, with remote management).

Cheers!

-Tim

[Rick Brown]

The PWM circuit Tim posted requires a reverse diode across the motor to absorb field collapse energy.  The posted circuit will work fine for a resistive load, but an inductive load stores energy in its field and when the current is interrupted the field collapses and induces a reverse voltage over the motor.  That voltage is felt by the transisor and field energy dissipated there.  Most transistors designed for this type of work can withstand this when switched occasionally, but 900 times a second will burn them up pretty quick.
-RickBrown in Reno, NV

[DrivingMissLazy]

Gee, after reading all this, I think I will just go down to NAPA and buy a rheostat and suffer the ten watts or so of loss. If you can not afford to lose one amp of power then you have much more serious problems. LOL
Richard

[Sojourner]

Tim Strommen...I always look forward to your knowledgeable interest in electronic projects and that is a great asset to me and others on this board. However I believe there are some non- electronic bus nuts looking for simpler design that is easy to put together without doing the fine tuning to work best. I am not as much up to date & knowledgeable as you. I have also done many experimental circuit & adjustment till is tuned for the application. If they want to use what you suggest is fine....they will learn more & get the fun of doing it to say "I did it"....Great! I believe what I have suggested this diode-in-series fan speed changer is very effective & reliable for this application. I am sure you agree this simpler system will work on this large RV fan motor unless they want a variable control.

For those who want to use "series diode" voltage step down design and from 1/3 speed to full...read on.
This design will not cause sine wave noise but by motor only. Each silicon diode has voltage drop of .7 volts. So each stage of series diode will reduce another .7v and so on. There is no waste of energy and cooler then PWM which is variable but this is in stepping down voltage system. This is good enough for large vent fan.
If you can solder wire with rosin core @ 63% Tin & 37% Lead...solder.
Then you can build this. If not but want to.... http://www.teamnovak.com/tech_info/how_to/solder/index.html

Parts list for series diode voltage reducer:

12p Rotary Switch #105-14571 @.....................$4.76
11 pc. 15a 1200v Diode #747-DSEP12-12A @ $1.78 ea
1pc 20amp slow blow fuse   
1pc rectangular metal electrical box
1pc rectangular extension for above box.
1pc round or square metal cover plate.
The above electronic parts are list in this link: http://www.mouser.com/
If you can find better source at lowest price...Good for you...let us know if you will.

12 positions rotary switch will start at 7.7 volts lower than full voltage. Label 1 to 12 on cover. #1 is lowest speed to #12 at full speed.

This will work at any voltage up to diode limit in volts. Above list of diodes is 1200v. In other words 24v will work but needs to double diode in series which will be 1.4v between each drop stage.
Follow the "B" schematic only this will be 12 positions instead of 6 and much lower current system. Make "R1" a 20 amps fuse.
http://www.cpemma.co.uk/sdiodes.html

For those non-electronic geeks....read this in simple terms about diode: http://www.allaboutcircuits.com/vol_3/chpt_3/1.html

I am sure you may still have questions...feel free to post. If I can't answer, I know Tom Strommen can...our electronic Geek!

FWIW

Sojourn for Christ, Jerry

[JohnEd]

Jerry,

It never dawned on me that I could get a voltage drop across diodes without heat.  Great application/design.  I need a box and knob anyway.  I think I will build one of these for every fan in the coach.....7 each including the destratification squirrel.

Thank you very much.

Richard,

I thought like you do.  Whats an amp to get excited about.  Seems my ignition dropping ceramic resistor gets hot enough to have your skin stick to it if touched.  TOASTY!  I guess that fan draws 15 amps or so and running at slow speed 12.5 volts are being dropped across that 8 ohm resistor.  I never measured it of course but if temps are any indication...well!  For Bonnie work it was a waste of power.  That and the problem with my liking to fondle "cool" resistors.  Thanks for your help.  Your comments and logic are good as far as I can see.

John

[Rick Brown]

*
It never dawned on me that I could get a voltage drop across diodes without heat
*
You cannot.  Assume your motor is drawing 1 Amp that flows through a .7 Volt drop over a diode.  Mr. Ohm once said Power = Amps * Volts.  That is .7 Watt per diode.
-RickBrown in Reno, NV