DC (treadmill) motor power supply

Thread Starter

ziqquratu

Joined Mar 29, 2017
10
Hi everyone,

Firstly, apologies if this is in the wrong forum - please move it if so.

I'm trying to construct a power supply to run some brushed DC treadmill motors that I've collected, with which I plan to drive some power tools. They're all around 220 VDC at 1.5-2 HP. I've tried to understand the control circuits that came with the treadmills themselves, and have attached three different schematics (just the power circuitry, not the control side, which is largely painted over and resistant to analysis - I assume it's just PWM control in any case) that I've been able to more-or-less trace below. These diagrams are not complete (each circuit has a choke and an EMI filter on the input, and a small inductor on the leads close to the motor, for example), but I *think* they show the most important parts of the power electronics. Also, the "0R" resistors are all high-power (>5W) resistors that are too low for my sad little multimeter to measure, so all <0.1 ohm, and my local mains is 230V at 50Hz. Further, while I'm obviously not an expert, I've been trained in working with mains voltages and so forth, and feel comfortable in my ability to deal with the hazards here.

I think I understand the basic principles behind these controllers, but am not clear on a few points and would like to clarify them before trying to roll my own driver. The plan is to control speed with an MSP430 via an optocoupler, with a hall-effect sensor to feed back the motor speed.

So, questions:

1) I assume R4 (circuit 1) and R3 (circuit 2) serve as brake resistors to slow the motor when power is shut off. But what is the purpose of R3 and R2 (circuits 1 and 2, respectively)? Likewise, what is the purpose of R1 (also high power low resistance) in circuit 3?

2) In circuit 1, there is a capacitor across the rectifier, between AC and +ve - I've absolutely no idea what role it serves. There's also R1 from +ve to -ve (can't properly see the bands or measure its resistance) - is this likely just a bleed resistor for C3?

3) In circuits 1 and 2, the transistor is driving the motor from the high side; in circuit 3 it's driven low side. I've also seen other motor control circuits driving low side. Is there likely to be a particular reason for favouring one over the other?

4) The smoothing capacitors (C3, C1 and C1, respectively) all seem woefully inadequate for the current in question. Is this likely a case of "good enough" smoothing for a low-precision application, or might there be a reason for the values chosen?

5) What are C2 and R4 doing in circuit 2? Suppressing voltage spikes?

6) A more general question about these types of motors. When the motor says that it's rated for, say, 220V - is that likely to mean 220V peak or RMS? My assumption is RMS, but I've been unable to confirm that anywhere.

I think that's about all I needed to ask at this point. I'm starting to put together a schematic for my own power supply, drawing as best I can from these ones, and I'll incorporate anything that I can learn from people here, and hopefully post a schematic for feedback after that.

Thanks very much in advance for your help!
 

Attachments

MaxHeadRoom

Joined Jul 18, 2013
28,619
220VDC is the rated voltage, and is the voltage required to operate the motor at its rated RPM, there is no 'RMS' value when working with this system.
Typically the P.S. can be at least +10% above motor rated voltage when operating with PWM.
Max.
 

Thread Starter

ziqquratu

Joined Mar 29, 2017
10
Hi Max,

Thanks for the reply. Sorry if I'm using the wrong term, but what I mean by RMS here is basically "the average of the DC ripple voltage" - given that I'm assuming the smoothing capacitors are inadequate, it seems reasonable to expect quite a large ripple in the output voltage and thus a peak voltage somewhat above the rated voltage.

Not sure if that changes your answer, but hope my meaning is clearer.
 

IamJatinah

Joined Oct 22, 2014
136
Hi everyone,

Firstly, apologies if this is in the wrong forum - please move it if so.

I'm trying to construct a power supply to run some brushed DC treadmill motors that I've collected, with which I plan to drive some power tools. They're all around 220 VDC at 1.5-2 HP. I've tried to understand the control circuits that came with the treadmills themselves, and have attached three different schematics (just the power circuitry, not the control side, which is largely painted over and resistant to analysis - I assume it's just PWM control in any case) that I've been able to more-or-less trace below. These diagrams are not complete (each circuit has a choke and an EMI filter on the input, and a small inductor on the leads close to the motor, for example), but I *think* they show the most important parts of the power electronics. Also, the "0R" resistors are all high-power (>5W) resistors that are too low for my sad little multimeter to measure, so all <0.1 ohm, and my local mains is 230V at 50Hz. Further, while I'm obviously not an expert, I've been trained in working with mains voltages and so forth, and feel comfortable in my ability to deal with the hazards here.

I think I understand the basic principles behind these controllers, but am not clear on a few points and would like to clarify them before trying to roll my own driver. The plan is to control speed with an MSP430 via an optocoupler, with a hall-effect sensor to feed back the motor speed.

So, questions:

1) I assume R4 (circuit 1) and R3 (circuit 2) serve as brake resistors to slow the motor when power is shut off. But what is the purpose of R3 and R2 (circuits 1 and 2, respectively)? Likewise, what is the purpose of R1 (also high power low resistance) in circuit 3?

2) In circuit 1, there is a capacitor across the rectifier, between AC and +ve - I've absolutely no idea what role it serves. There's also R1 from +ve to -ve (can't properly see the bands or measure its resistance) - is this likely just a bleed resistor for C3?

3) In circuits 1 and 2, the transistor is driving the motor from the high side; in circuit 3 it's driven low side. I've also seen other motor control circuits driving low side. Is there likely to be a particular reason for favouring one over the other?

4) The smoothing capacitors (C3, C1 and C1, respectively) all seem woefully inadequate for the current in question. Is this likely a case of "good enough" smoothing for a low-precision application, or might there be a reason for the values chosen?

5) What are C2 and R4 doing in circuit 2? Suppressing voltage spikes?

6) A more general question about these types of motors. When the motor says that it's rated for, say, 220V - is that likely to mean 220V peak or RMS? My assumption is RMS, but I've been unable to confirm that anywhere.

I think that's about all I needed to ask at this point. I'm starting to put together a schematic for my own power supply, drawing as best I can from these ones, and I'll incorporate anything that I can learn from people here, and hopefully post a schematic for feedback after that.

Thanks very much in advance for your help!
Hi There, Great questions!~ The A-Star unit I do know, and comes from Horizon treadmills prior to 2004. On to your questions.....
1). These large resistors are in series with the motor as drawn, and will carry total series current, which is watched by the controller.
I don't know if things are true as drawn, as large current resistors can often be in parallel, halfing resistance and doubling current ranges.
All large resistors in this question are series resistors at high wattage and are intended for current controls.
2). That capacitor across the AC Line is for noise suppression, and the resistor is a cap bleed-off resistor like you thought.
3). Not sure on the preferred high-side or low-side question, personally, I like low-side drive but either works fine as far as I have seen thru the years. I can say Cybex 425 type controllers had issues with high-side drive feedback to the PWN generator(3524, 3843) which could kill that chip.
4). Normal input capacitance for a 2hp DC PM Motor is usually about 2000uf total, and I have seen as little as 470uf and as much as 4400uf.
5). C2 and R4 are possibly not drawn correct. These may be in series as a "Snubber Circuit" which reduces switching spikes and ususally resides across the main kickback diode. Great question! Without Snubbers, a a designer could go batty during proto-testing.
6). A motor rated at 220v single phase can take up to 220v x 10%. Regulation of motor speeds are controlled by that PWM switching which supplies the motor with square-wave signals of varying duty cycle, all at peak Vbulk, resulting in yes, and average operating voltage applied to that motor. Great questions! A master PWM must "match" the expected input frequency for each different controller, as "filters" on the controllers will only allow a viable select main drive frequency. For example, a 500hz speed command will not operate a Horizon, Matrix, Icon, True, Bowflex, as each manufacturer has "selected" a sweet frequency they prefer and filter out other freq's. These drive PWM frequencies can be as slow as 10hz and as fast as 500hz for the controllers I have experience with.
Let me know if I botched anything or confused anyone ;o)
 

Thread Starter

ziqquratu

Joined Mar 29, 2017
10
Hi IamJatinah,

Thanks for the very helpful reply!

I went home last night and checked my drawings again, so let me clarify a couple of things you commented on:

1) The high power resistors are indeed connected as drawn, in series from the transistor to the motor. I did find some connections to the control circuitry that I didn't notice (i.e. didn't look for properly!) before, which supports your suggestion that they're current sensing.

5) C2 and R4 (circuit 2) are definitely connected as drawn (they are literally bridging - even share solder with - the motor lugs).

In any case, your reply has answered pretty much everything very clearly, and definitely helps me in designing my simple controller. Thank you, again, for your very informative response.
 

Thread Starter

ziqquratu

Joined Mar 29, 2017
10
Hi everyone,

I've put together a very rough-and-ready schematic and I'm going to post it now, crude as it is, for comments, as I may not have the chance to do so over the weekend. I've based it on the treadmill controller circuits I posted above, along with a motor speed controller I found in Silicon Chip magazine (I've attached a cut-down image of the schematic, but if that's a breach of copyright that's not OK here, please let me know and I'll delete it).

Please bear in mind that I've not designed anything quite like this before, and am learning many of these things as I go, so there may be some absurdities in here!

So, first off, the microcontroller - it's a placeholder at this stage. I'll worry about which pins to use when I start trying to put it on a PCB. Basically, it will read from the hall-effect sensor (which will be on a small breakout board), measure the current RPM, compare to the desired RPM (set by the pot RV2, probably in 100rpm steps; you'll have to hold down push switch SW2 to enter "set speed" mode and avoid risk of set point changes due to vibration or bumps). It will put out a PWM signal, say 2kHz, to the optocoupler; there will also be a 16x2 LCD attached, which I haven't bothered to include here. All this control circuitry will be powered by a small wall-wart, and so be isolated from the main power circuitry. I *think* the basic principles here are sound, but please let me know if I've overlooked something or done something ridiculous.

On to the high voltage side of the circuit. The varistor RV1 across the AC lines will help prevent spikes (and I'll probably reuse the EMI filters from the treadmill controllers I have as well). The NTC thermistor TH1 will limit inrush current and so "soft start" the motor. After the rectifier, the voltage will be smoothed by C3 (reusing a cap from the controllers I have), before going through brake resistor R10 to the motor, itself protected by a high-power diode D4. The motor will be low-side driven by IGBT Q4, which is protected by a snubber (R12 and C7) and goes via current sense resistor R11 to ground.

To drive the IGBT, a 15V line will extend via R2 and zener D3 (with C4 to smooth the supply and D2 to prevent C4 discharging backwards if the rectified mains dips due to insufficient smoothing). The optocoupler will drive transistors Q2 and Q3 (via a 300R current limiting resistor), which will in turn drive the IGBT (protected from spikes by the second 15V zener D5).

As a guard against extreme over-current, I've included Q1 like shown in the Silicon Chip circuit. With the values shown, a current of 48A gives 1.2V across R11. The voltage divider R8 and R9 means Q1 would see 0.6V, which would cause it to pull the optocoupler output to ground and prevent it triggering Q2.

I know I could drive the IGBT directly from the optocoupler, however I understand that a driver like this one allows me more room to increase the PWM frequency if I wish to (e.g. if I find the lower frequency is generating too much noise), and will help reduce switching losses, so for a couple of nothing components, it seems worthwhile. I'm not sure about the overcurrent protection offered by R11/Q1 - seems like a good idea, but I'm entirely open to criticism. I do want to include some means of checking the regular operating current as well - would it be OK to have the micro read the voltage across R11 directly, or is there a better/safer way (that allows me to keep the high-power side totally isolated from the micro circuit)?

Oh, and of course there will be a fuse and physical power switches - a small switch to turn on the control circuit, and a NVR switch to allow me to turn the motor on and off without turning off the controller itself (so you can, say, set the speed with the motor turned off).

Alright, so I hope that the explanation of my thinking makes sense, and I look forward to hearing the various criticisms people must have!
 

Attachments

Thread Starter

ziqquratu

Joined Mar 29, 2017
10
Hi Max,

I looked at the IR2110 (found out about it on that very blog, in fact!), and if I had one I'd probably use it, but I figure the transistor pair is probably more than good enough for this job (and I actually have them on hand!) - do you think they'll be a problem, and if so, should I simply remove them and drive directly from the optocoupler? The choice of an MSP430 is for the same reason - I snagged a handful just before TI stopped giving out free samples, so I'll use them. And I've gone with a hall sensor because it's a dust-proof solution, and woodworking tools (which I want these motors for) generate a LOT of dust! Finally, having the option to accidentally spin a blade backwards would be an unnecessary hazard, so I'm deliberately excluding any means of bi-directional drive from this controller.

Thanks again for your comments!
 
Top