pwm for dummies

[Mcgyver]

yet another newbie pwm plea for help. I tried to help myself but have hit a wall.

Objective: homemade, discrete pwm for el cheapo treadmill motors manually controlled via a pot. Motor is 90v dc to be used to drive a small drill press.

years ago an engineer at Unitrode kindly sent me the attached brush motor schematic and some UCC3581 samples. I tried and tied with the help and encouragement of my Unitrode benefactor to get the circuit working, but no luck.

Recently I was learning about 555's and thought I’d give it another try. Basically I replaced the 3581 part of the first circuit with a 556 astable/monostable setup as per the attached schematic (credit to http://casemods.pointofnoreturn.org/pwm/circuit2.html)

I could get it to act as a dimmer for a couple of light bulbs but as soon as it tried with the 90v motor I’d fry the fet. I was using a 100k pot and fr =120k

So my question is can anyone either 1) help understand why this didn’t work and/or suggest how to make it work or 2) scrap the whole darn thing and point to a pot controlled pwm circuit that'll work?

I've been following some of the other threads but I’m not sure how to take what’s offered there and parse stuff out for my simple manual controller. With the money and time I've spent on this I could have bought one with gold plated contacts but its become a quest to prove I’m not a hopeless idiot with electronics. help please

[pminmo]

Couple of questions. One your circuit shows an IRF540, and what looks to be 12V to operate a fan. To run a 90Vdc motor, the IRF540 won't work, it will fry. The IRF740 as shown in the first schematic will handle a 90V DC motor. I need to know exactly the part number of the fet and clarification of how you hooked up the 90V motor.

Phil

[Al_The_Man]

You did not mention what current your motor was? Also the first schematic only shows a 100µfd cap after the bridge, that needs to be quite a bit larger for a high current motor.
A cheaper approach is to use an IC that was used in another version of treadmill motors using a Universal motor controller,(AC/DC motor), the configuration used a phase controlled triac firing into a bridge which fed the motor, hence it was phase controlled with full wave pulsating DC output, there should be no reason why a DC motor would not work in this configuration as well as a universal one. There would be no need for the Large cap either.
The IC around the design is a TDA1085C, it has tach feedback option and ramp options also.
Al

[Mcgyver]

thanks Gents,

I did use the irf740. The circuit showing the IRF 740, the file “bn shcem pwm dc control” was the base for this. When I couldn’t get that circuit working (and believe me I tried) I went about figuring out how else to get pulses to the FET and be able to control the duty cycle.

Hence the 556 circuit. It was a low voltage application for a computer fan taken from the net. Having a computer fan kicking around, I use a 12v wall wart, a 7809 ic and got the circuit working.

Next I thought I could slipstream this into the irf740 circuit. The right hand side of the base circuit (the irf740 one, “bn shcem pwm dc control”), basically has power in (the connection between pin 3 and the diode), a connection to the gate, and ground. So I took the 556 circuit, made rf 120k, rdc = 100k potentiometer, the gate that went to the 540 goes to the 740 and ground was ground.

Al, I sort of followed what you were saying…is the TDA1085C commonly available? I like stuff I can pick up Active or Sayal (retail outlets) – that way as I blow them up they are easily, and with minimal fuss and embarrassment, replaced. This is for the drill head on the pillar tool I made – it spins small bits in light applications and likely never never be cnc controlled.

I wish I could give more on the motor. It says:

Triple Play Motor
Part No.: 117085
Voltage: 90 volts - DC
RPM: 6000
Rotation: CW

It says made by GS electric for Legend Sports, a product of Proform Fitness Products. It’s 3” dia. and 4” long with two leads and no place to access or replace brushes. Googleing reveals little except someone who speculates its 500 watts. Unfortunately that is all I can say about the motor. I picked it up from princess auto a few years ago for $10.

Thanks again for the support.

CNC content: bought some l297 & l298’s this week and have been hangin out at Phil’s page

[ViperTX]

Well...with all the rambling...I'm not sure what is the latest....but I suspect that when you turned off the FET the motor continued to generate a voltage that exceeded the Vds of the FET.

[Al_The_Man]

Originally Posted by Mcgyver Al, I sort of followed what you were saying…is the TDA1085C commonly available? I like stuff I can pick up Active or Sayal (retail outlets) – that way as I blow them up they are easily, and with minimal fuss and embarrassment, replaced. This is for the drill head on the pillar tool I made – it spins small bits in light applications and likely never never be cnc controlled.

The TDA1085C is stock in Digi-Key for $1.68 ea.
I will post the reverse engineered schematic of the treadmill amp using this IC this week-end hopefully.
Al

[H500]

Mcgyver: On the 556 circuit you showed, it's missing a flywheel diode across the motor. Don't forget that the load is inductive. That's probably why your fet smoked.

You need a fast recovery diode such as a MUR860. A 1n4001 won't do. You should give the fet generous heatsinking during development so that they don't fry due to circuit errors.

By the way, are you from Calgary? I didn't know there were any other cnc victims here.

[Mcgyver]

thanks for the help. Would replacing the 4001 in the circuit with the faster acting one get me on track, or is the flywheel diode something else?

H500, in TO, but don't hold that against me

[H500]

The 556 circuit you posted don't have any diode across the motor. If you're refer to the other schem, then it's the 4004 diode across the motor.

A 4004 takes a long time to turn off when the polarity reverses, so in effect, it shorts the power supply into the mosfet for a few hundred nanosecs. If the pwm speed is high enough, these high current spikes causes localize overheating and eventual failure of the fet.

[Mcgyver]

sorry, it was the 4004 circuit that i was referring to. I'll try to get a fast switching diode and try it. I think I understand what's going on and thanks for the help. the fet was very hot with a small heat sink, even when I was using light bulbs for a load, but stop working the second i tried the motor. The motor ran, but at full speed

Al, when you get a chance it would be great to check out the schematic you were referring to. If there was a tried and true schematic, I'd scrap the current 556/fet idea. thx

[Al_The_Man]

This was the mod to the original treadmotor controller sold by surplus outlets it ran a Universal motor in the DC mode, it had a fairly crude digital tach on the rear of the motor for speed control.
The speed control mod is pretty simple. You just remove transistor T3 from the board.This allowed the motor to run at pot. speed when turned on, rather than turn the pot down and back up.
. What it entails is to cut the trace to pin 6 of the chip and then solder a jumper from pin 6 to ground (pin 8). This mod was the most important since it allowed the motor to have the power required out of it. What it does is removes the slow soft start feature (it's still fairly soft just not extremely). This allows the controller to respond to sudden heavy loads such as the cutter engaging the work in power feed without lagging so much as to let the motor bog down.
I still have to confirm the Cap values and see if I can reverse engineer the tach.
I have this on my Band saw which is converted from wood to metal.
The advantage of this controller it does not require power transformers and low voltage supplies or large caps as when DC FET switching is used.
Due to the fact it is full wave triac switching into a bridge, the output is 120hz pulsed DC. Refer to the TDA1085C spec sheet for details.
Al

[Mcgyver]

thanks Al,much appreciated