Question for those using pwm dc motor controllers

fluidpowerpro · September 28, 2023, 12:56:12 PM

Currently I use a pwm motor control to run the winch that provides up and down on my mill head. It works ok but I am seeing a pattern of failures about every 50 hrs.
I have had 3 failures.
It always fails in the down direction, never up.
I have an engineer friend that is quite knowledgeable on these things and he feels that the issue is that in the down direction, there is inertia. The inertia causes a spike that is fed back into the controller that over time, wipes it out.
In the up direction, there is no inertia to deal with so that's why it never fails in that direction.
The controller I'm using is rated for 40 amps continuous and 60 amps max. It can be purchased on line for less than $20. (I know...you get what you pay for). I always keep a spare and it only takes about 20 minutes to replace, so no big deal, but would rather have something more reliable.
I have already purchased a larger controller that is rated to 60/100 amps, but it is larger, so I'll need to redo everything and mount in a larger enclosure. This will be a good project for this winter.
Before I do this I would like to quiz others on if they have had similar issues. What was the fix. Can anyone recommend a good durable controller.

Don P · September 28, 2023, 01:43:45 PM

I think your friend is right. I watched this and another video from this guy last night where he gets into that spike and how to protect or use it. Here he is showing a diode as protection in the first minute or so.
How to use an IGBT chopper to control DC motors - YouTube

Ianab · September 28, 2023, 01:54:30 PM

Diode protection makes sense in that scenario. As you cut power to the moving motor it will act as a generator, pumping volts back into the control circuit. It probably has some protection built in, but it would seem not heavy duty enough in the cheap controller. That's probably the first failure point, and the controller soon fries.

fluidpowerpro · September 28, 2023, 02:02:04 PM

The IGBT looks like just what I need. I'll consider one when I redo my box.
Thank you.

Don P · September 28, 2023, 07:51:37 PM

Everything is related somewhere

.
What he shows in the next video on IGBT boost converters is a simplified version of how a Pruis boosts battery voltage up to as high as I think 600 motor volts. These circuits time, capture and multiply that spike to boost the voltage. The circuit he is using in the end is basically what is under the hood of a Prius. A big IGBT, a large capacitor and what they call a reactor, an inductor coil.

Inductor and boost converter basics and IGBT boost converter - YouTube

Edit:
The analogy that runs through my mind is, water hammer is a problem in home plumbing, and it is also what makes a hydraulic ram work.

kelLOGg · September 29, 2023, 08:02:33 AM

FPP,
I use one on forward/reverse on the sawhead where there is no back-feed due to gravity. For up/down I use only 12 vdc (with no pwm speed controller) which gives constant speed. Why do you use variable speed for up/down? It is not a long travel distance and seems like overkill. What am I missing here?

jpassardi · September 30, 2023, 08:32:03 AM

I use that same PWM for both carriage travel and up/down without issue. For up/down I use a dump truck tarp motor which has worm gear reduction so it stops pretty much as soon as I cut power. I also reduce speed as I reach my set point with the dial (don't flip it off at full current).

Kellogg: having variable on the up/down allows you to creep up to your cut height and not overshoot it.

kelLOGg · September 30, 2023, 11:52:33 AM

OK. My up/down came as a manual crank. I added a tarp motor but still have use of the crank for fine adjustment - up or down - so that suffices in place of a pwm in my case. If I didn't have the crank I understand the necessity for a pwm. Thanks.

fluidpowerpro · September 30, 2023, 07:10:05 PM

Quote from: kelLOGg on September 29, 2023, 08:02:33 AM
FPP,
I use one on forward/reverse on the sawhead where there is no back-feed due to gravity. For up/down I use only 12 vdc (with no pwm speed controller) which gives constant speed. Why do you use variable speed for up/down? It is not a long travel distance and seems like overkill. What am I missing here?

As jp said. I need to be able to adjust the speed to creep up on my desired set point. Without it the winch is way too fast. I like having adjustable speed because I can go fast when I'm going up.

jpassardi · October 01, 2023, 08:16:17 AM

FPP: Just thinking - On mine the power down is also less apt to get induced current after cutting power because the LT15 has a lift assist gas strut/cable mechanism. Obviously, this reduces the torque input requirement of the lift motor.

mike dee · October 01, 2023, 08:45:13 AM

You might find this company's products interesting and usefull
https://teknic.com/

kelLOGg · October 01, 2023, 12:07:25 PM

My pwm is from SouthWest Motion products. I opted for it because it is hermetically sealed. Even though my mill is under a shed I didn't want humidity and blowing rain to damage it.

RetiredTech · October 03, 2023, 09:46:29 PM

I'm not sure what I'm missing here. I can see where there could be a feedback or spike in something like an electric vehicle. But how can you get such a spike with a winch circuit? Once the relay opens to stop the winch there is no path back to the pwm controller.

Ianab · October 03, 2023, 10:12:39 PM

That would be right if you used a relay in the circuit. But then you would have to ensure the relay opened before the motor stopped. I think with smaller motors, STOP is simply when the controller supplies zero volts. Even a sudden change in direction could generate a back EMF?

It's basically the same idea as putting a diode across the coil of a relay. When you power down the relay it generates a voltage spike, and the diode effectively dumps the current harmlessly.

If you expect a back EMF or voltage spike, then you can plan the circuit to handle it. It's the unexpected ones that fry things >:(

bushhog920 · October 04, 2023, 06:44:42 AM

Fanuc robots had a resistor in the drive to disapate induced voltage on deceleration. Sounds like you may be having this problem

Don P · October 04, 2023, 07:34:58 AM

The little controller I've been playing with is like Ian describes, stop is just turned down all the way on the electronics. Reverse is a switch that could be thrown under full power or any point in between.

I think a spike is created any time there is a pulse interrupted, a change, as in pulse width modulation. Every pulse change creates a spike of some level which depends on current, voltage and inductance in the circuit. A winding like a coil or motor has a lot of potential but it is better to keep all wiring as short as possible with high frequency delicate components to keep from boosting that spike.

The old tractor and trucks here, cranking with the coil wire dangling probably isn't going to hurt a thing. The first time I did that with the Honda I bought a new distributor.

There are also snubber capacitors in some circuits. How that is different than the ones he is using to boost the spike in the vid above is beyond me.

fluidpowerpro · October 04, 2023, 05:31:32 PM

Quote from: RetiredTech on October 03, 2023, 09:46:29 PM
I'm not sure what I'm missing here. I can see where there could be a feedback or spike in something like an electric vehicle. But how can you get such a spike with a winch circuit? Once the relay opens to stop the winch there is no path back to the pwm controller.

There is no relay like a traditional winch control. All functions are handled by the controller which has the relays built in.
When they fail I can hear the relay clicking so I know it's getting my command. I just get no output to the motor.

RetiredTech · October 04, 2023, 06:08:43 PM

I see now. I thought you had used the original winch relays for forward and reverse and then used the pwm just to contol the speed of the motor. Wiring for the relays would need to be changed a little bit, but wouldn't that solve the feedback problem?

Peter Drouin · October 04, 2023, 07:46:42 PM

Redneck way put the head all the way down hook a bungee or garage door spring to the head and pull up, get all you can without lifting the head up. Hook the other end to the frame on top.
Would slow the head coming down some. ;)

Satamax · October 06, 2023, 11:02:25 AM

Guys.

Sorry to hijack.

Two daft questions, could a PWM driver drive an older magnetic field dc motor ? I mean, i would excite the field with 195 volts as per required. And drive the 270vdc with the pwm.

Where could i find a 300vdc pwm driver rater for 2 to 3 kw ?

Thanks guys.

Don P · October 08, 2023, 12:50:30 PM

I don't know much about it and was hoping one of the electronics guys would answer but for high power, high frequency switching we've been talking about IGBT's as opposed to MOSFETs. My understanding is MOSFETs are faster, cheaper and more efficient, IGBT's are fast enough and can handle higher current and voltage. This is just some random searching for info;
Power Semiconductors - Introduction to IGBT (Power Modules) | Fuji Electric

I think just about any PWM controller that is within matching specs for output to switch the insulated gate, "IG", it will throw the relay, a bipolar transistor "BT" and pass or stop the current from the dc supply. The rapid on/off regulates the output voltage to the motor.

yet another video, this one explains what the PWM controller is doing and also explains the back EMF spike and diode protection.
Motor speed controller tutorial - PWM how to build - YouTube

I've also seen people suggesting using an IGBT of much higher voltage rating than what appears necessary.

Don P · October 09, 2023, 05:54:08 AM

This is about the same but a little more junkyard than the circuit above.
A 555 timer IC is creating the pulse, a 10k ohm potentiometer is driving the frequency and modulating the duty cycle, the pulse width. It is not perfect but can control the little fan from almost stopped to almost full throttle. A little switching of the minor components could dial that in.

I don't have a freewheeling diode on it yet to let the motor spin down without sending a spike back at the 555 chip.

The output signal from the 555 is running the motor, the orange wire running from the chip to the red motor lead. It could be triggering a MOSFET or IGBT transistor and switching more current if the load is greater.