Future Kit PWM DC Motor Speed Control Model FA804 12vdc, 20w.

[Trnsrus]

I connect the subject device to the fixed DC output of an old MRC powerpack and a test track with a recent Atlas N scale analog locomotive. The low speed operation was fine except for prominent motor hum, even with the locomotive stopped. I measured 1.3 volts across the rails with the potentiometer turned all the way down. Can anyone suggest a way to reduce the hum and "stopped" track voltage? I scanned the schematic and circuit description but am not allowed to post images.

[Rasputen]

Is the schematic and description posted somewhere?  All I can find is a picture of the assembled kit.
The PWM frequency is probably too low.

You need 25 posts in order to include images on this forum.

[Maletrain]

Some old MRC power packs do not provide smooth DC, and we have one at the club that is not turning off with the power knob turned down to its stop.  It is putting 0.2 VDC and 2.3 VAC on the rails when turned to zero speed.  (For Peteski - yes, that was measured with the Fluke 36, not the cheap multimeters that were giving strange AC readings with DC switching power supplies and even batteries.)  We first noticed that because it was causing a loco to buzz when it was stopped at zero speed on the MRC controller knob.

So, my first suggestion is to try a better DC power source. Eight or nine D cells in series should be enough for a reasonable test, if you don't have a 12 volt switching DC supply.

I am working on a similar controller, and it seems to provide smooth operation without buzzing at low speed.  So, I am thinking that your buzzing is coming from the MRC power supply.

PS: The PWM module I am using is this: https://www.ebay.com/itm/234341862535, with a "motor shield" to protect it from over-current (https://www.ebay.com/itm/202049250851) and a diode to protect that from reverse polarity.  [See reply #10 for explanation of crossed-out link and replacement link.]

[Trnsrus]

Thanks for replies. I posted this on TrainBoard, including schematic and description: https://www.trainboard.com/highball/index.php?threads/future-kit-pwm-dc-motor-speed-control-model-fa804-12vdc-20w.156642/

[Maletrain]

That link would have been a good idea, if the browsers would let me see it.  They all tell me it is insecure, and I have not found a way to view that site since its certificate expired.

[mmagliaro]

EDITS: I fixed this up when I realized I confused your FA804 with the one Maletrain linked from eBay.
But most of what I originally said is still correct...

The data sheet says this thing oscillates at only 264 Hz.  That's going to make a DC motor buzz buzz buzz.

AND.. I simulated the whole thing by putting the schematic into a circuit simulator, and I get 236.6 Hz, so I don't think that 264 is a mistake.  The data sheet you scanned in says 1.5A maximum.   That is believable since they are using a TO-220 driver transistor
and since it is always fully on or fully off it doesn't have to dissipate much heat.

At the 50% position on the control pot in my simulator, I get dang spot on a 50% duty cycle square wave at 236.6 Hz.

And yes, the circuit will leak at least an average 0.5v at "off" because even if it can fully push its output transistor on, the duty cycle never reaches 100%.

You can change C1 and C2 from .01uF to 1/10th of that, or .001uF, which will get you 10x on the frequency, or about 2.3 kHz.  If  you go down to 150 pF, you can get around 15 kHz, which should work pretty well and get rid of your buzz.

Something else I notice about this circuit, that you may or may not care about one day.  It works by raising the ground reference up and down.  In other words, one rail will always be at 12v, and the opposite rail will go up and down between 0 and 12.  When both rails are at 12v, the train will be stopped because there is zero differential between them.  That's all fine for controlling the motor.  But it might cause unexpected problems one day if you ever connect any other electronics to the rails that assumes that both rails are at 0 when the train is stopped.

[peteski]

Some old MRC power packs do not provide smooth DC, and we have one at the club that is not turning off with the power knob turned down to its stop.  It is putting 0.2 VDC and 2.3 VAC on the rails when turned to zero speed.  (For Peteski - yes, that was measured with the Fluke 36, not the cheap multimeters that were giving strange AC readings with DC switching power supplies and even batteries.)  We first noticed that because it was causing a loco to buzz when it was stopped at zero speed on the MRC controller knob.

The discussion you alluded to earlier was about measurement AC component on a DC output of a buck-type switching power supply.  It is different than measurement the AC component of some voltage which operates at very low frequencies like 60Hz or possibly 120Hz.  Average multimeter is designed to correctly handle such low frequencies.

Max,
It is really relevant which side is considered as a common, when there is is no common between the device being powered, or the additional devices connected to the rails?  As long as each device is only connected to either or both rails rails (no other connection between them) then there is no "common".  If you think about it, any DC throttle powering track will have its "common" (or ground/negative) on either rail depending on the position of the direction switch (which simply flips the polarity of the voltage supplied to the track).

[Trnsrus]

...You can change C1 and C2 from .01uF to 1/10th of that, or .001uF,...If  you go down to 150 pF, you can get around 15 kHz, which should work pretty well and get rid of your buzz...

...It works by raising the ground reference up and down...  When both rails are at 12v, the train will be stopped because there is zero differential between them...  But it might cause unexpected problems one day if you ever connect any other electronics to the rails that assumes that both rails are at 0 when the train is stopped.

Hi Max,

Thanks for the suggestions. I've ordered 150 pF caps and will report results after I install them.
What kind of electronics and problems can be affected and caused when connected to the output of this controller?

[mmagliaro]

It really doesn't matter which rail is called "common".  What is odd here is that both rails will have a positive voltage on them relative
to ground.  Consider the case where the train is stopped.  Both rails are at +12 relative to ground.  If you were to connect a lamp between either rail and ground, it would light up.   With a "normal" DC controller, both rails would be at ground and any load connected between either rail and ground would not light.   
I'm trying to imagine how this throttle could work in a common-rail system where the other throttles don't work this way.  How would an engine bridge across a gap from one throttle's block to another?

Trnsus: In answer to your question about "what kinds of problems", the common-rail connection with other throttles is the only one I
can think of at the moment.  I just don't like things that introduce an unnecessary quirk.  It strikes me that with an operating frequency of only 264 Hz, and since it openly states that it has a duty cycle between 1% and 99%, (no full "off") this thing was never conceived to be a model train controller.

I got so curious about this that I actually ordered one of these kits to play with it.   I do have a reservation about having to lower those capacitors down to 150 pF.  That's a very small capacitance, and it might make the whole circuit subject to stray capacitances and make it unstable.  I'd imagine that just handling the circuit board with values that small could make it change its frequency a lot.  I'll know more when I get the kit and play with it.

[mmagliaro]

Other options besides just replacing those caps would be to change the pot and resistors, so you can use higher values for C1 and C2.  There are lots of value combinations you could use to change the R/C timing to get a better frequency with stable and commonly available values of parts.   I'll play with it more when I get it.  No matter what, it will be impossible to get a true zero off with this circuit.  By its very design, it is controlling the duty cycle of the PWM with a pot, so you can't get full 0% to 100%.  The best you can hope for is to play with the pot and fixed resistors and hope you can get the "off" position to be low enough that you don't see any loco creep.

[Maletrain]

WARNING:

The link I posted to the "motor shield" was from the second batch I ordered from a different eBay seller, AND THEY ARE DEFECTIVE.  They will only auto-reset, and only at a 1 second delay.  The jumper does not work to make then manual reset (with the on-board button) and the pot to adjust the reset delay also has no effect. 

The ones I bought first do work fine. That link is https://www.ebay.com/itm/224126511915 .

I crossed out the link in the earlier post, with a note to read this one.

[peteski]

It really doesn't matter which rail is called "common".  What is odd here is that both rails will have a positive voltage on them relative
to ground.  Consider the case where the train is stopped.  Both rails are at +12 relative to ground.  If you were to connect a lamp between either rail and ground, it would light up.   With a "normal" DC controller, both rails would be at ground and any load connected between either rail and ground would not light.   

Where exactly is that ground Max?  With each rail connected to one of the throttle outputs, there is no "ground" reference.  Since the throttle's output is "floating" in reference to anything else, who's to tell what voltage the rails are at?  As long as the voltage difference between the rails is zero Volts between the rails, all is ok.  I guess I just don't understand this ground thing.  It is almost like you are describing that throttle circuit as a simulation running on a computer where you had to define a ground.  In real life the entire throttle is floating - no ground connections.

Even if you were probing that throttle's output using an oscilloscope (which does reference the earth ground of its AC power supply), since the throttle is powered by a totally floating power source (or it should be like other consumer electronics using a 2-prong AC power cord), you would have to connect one of the throttle's outputs to the scope's ground, making that output a ground reference. But either of the outputs can be used, so connected one of the ways, the throttle's output would make the waveform look like a conventional throttle output (pulses would look positive referenced to the other, now grounded at the scope lead).

The model on the tracks or other devices connected would not care which track is pulsing, and which is steady.

[mmagliaro]

Pete, I'm thinking about a dual-cab commonrail setup, with two throttle circuits of different types.  One like this PWM circuit, and one where the circuit raises the "+" output up and down above its "Gnd" terminal.  And you power them from a single power supply Like below. 
Now, there is a common reference between the two, but because of the different output schemes on their driver transistors, throttle 2 is shorting out throttle 1.    I think if they had separate power supplies, everything would be floating as you say.

[peteski]

Sure, if both throttles shared a common power supply, that would be a problem (unless the modeler did not use common rail configuration).  But in majority of multi-cab DC configurations, all the cabs used are identical. They could then use a common-positive for the common rail configuration.

I guess you are just providing this heads-up as a precaution if someone wants to use this throttle in a multi-cab setup while using a common DC power supply, and wants to use a common-rail wiring.

[Trnsrus]

Other options ...would be to change the pot and resistors, so you can use higher values for C1 and C2.  There are lots of value combinations you could use to change the R/C timing to get a better frequency with stable and commonly available values of parts.   I'll play with it more when I get it...

Max,

I operate with DCC and use the old MRC powerpacks only for testing non-equipped locomotives. I wanted to use a PWM control because the old powerpack rheostat output produces a high start speed and cannot be turned down enough to test run at a crawl speed. The voltage remaining on the rails at stop is not a problem since it is not enough to move the loco, but Is the hum created by the PWM harmful to the motor?

I have gotten the 150 pF capacitors and will install them if your test is successful.  If it is necessary to replace the pot and some other parts, it may be better fot mè to just buy another controller. It looks like many inexpensive models are available.