Looking for power supply...

[David Henderson]

I have a DC motor that I'd like to use for a project, but I need an appropriate power supply. The motor was pulled from a treadmill.

Here's the specs listed on the motor:
Pacific Scientific
Catalog Listing: SR3628-5172-2
HP: 1.7
RPM: 3200
Volts: 90DC
Amps: 19.3
Duty: CONT.

I've currently got the motor connected the power supply from the treadmill-- it's a complicated affair with a ribbon cable out to the control/display panel and a speed sensor that causes everything to shut down if the detected speed doesn't match the expected speed based on voltage. I'd like to replace it with a simple, variable speed controller for the motor.

I've looked at eBay and a couple of the standard suppliers (digikey, mouser, etc.). I can't find a power supply that provides anywhere near that kind of voltage and amperage. Anyone have any suggestions?

(Btw, I put a multi-meter on the power supply outputs. When I crank the speed up as high as I can before things error out and shut down, the meter reads 90+ DC volts and 1-3 amps. That is under no load.)

 

[Bluejets]

I doubt you are going to find a "simple variable speed controller" for that kind of voltage and current.

If you outline your project perhaps there may be an easier alternative.

 

[David Henderson]

I plan on using the motor to drive a shop-built belt grinder. Its not an uncommon build... there's a number of them outlined in various YouTube videos.

 

[David Henderson]

It might be worth stating that it's common to find 1.5 to 2 hp motors on a grinder of this kind/size.

I know that finding even a constant PSU that provides 90v @ 19A is unlikely... that's about 1800w. That's one reason I measured the amperage. As I said, it was under no load, but it was only drawing a few amps-- nowhere near 19.

 

[Bluejets]

Seems to me that using the original matching power unit would be the best approach.

If it is just a matter of simulating a signal for underspeed then as long as you know or can find out the type of sensor arrangement, something can be fairly easily hacked up to give the unit the reqired info.

As I say, it depends on the sensor details, not only type but any pulse detail, how many pulses the control expects to see etc........... etc.

With regard to the motor current, it will be an unknown at this stage as to just what current it will draw with your new installation.

 

[David Henderson]

On using the original power unit-- That was my original plan, but the complexity of the display/controls panel gave me pause. The thing is big, so I don't really want to have to use the display panel. There's a 16 pin ribbon cable between power board and display, so at first glance it probably won't be a simple task to ditch the display/control panel and replace it with a couple of momentary switches for speed up/down.

But I am always up for a good challenge.

Here's the speed sensor:

Looks like a pretty straight forward optical sensor. In the treadmill, it was set up to bounce ir off of black/shiny lines on the flywheel. In this kind of setup, how fast is the rise/fall of the signal from this sensor? Is the signal pulsing with each change from black to shiny, or is it smoother than than?

 

[David Henderson]

Here's a couple more pictures... one of the power board and one of a simple schematic for the treadmill's wiring.

In the diagram, the "LOWER PCA" is the power board pictured in the second photo. The blue & brown at the bottom of the photo are AC power in from master switch/wall. The red and black on middle right are power out to the motor (through an inductor and capacitor as shown in the diagram.)

In the photo, the lift assembly isn't plugged in (I don't need it for the project), thus the open P2 connector. The 24pin connector on top left is the ribbon to the display/control panel. The connector in top right (w/ white, blue, red, black wires) is the connection to the speed sensor pictured in the previous post.

 

[Bluejets]

I'd be using whatever I could including any mechanics associated with the speed sensor.
As for the display unit I would assume there would be some arrangement for user input.
So even if you think it's not needed for your application, I'd be leaving it connected at least until you can proove otherwise with a test run.