Re: driver performance

In a message dated 08/24/1999 7:04:31 AM Hawaiian Standard Time,
cnk@... writes:

<< I've been looking at Microkinetics' DM4050 drivers (40V, 5A) but only using
them for half-stepping. Matched with their 34M315 motors (size 34, 315
oz-in) and PWR36 supplies (36V, 10A unreg.) they've got curves that say they
can do 170 oz-in minimum out to 500 rpm and then ramp down to 50 oz-in at
1900 rpm. (half-step mode, parallel wiring, at 5A). With 2:1 gear
reduction and .200" pitch ballscrew, I've been hoping to use them out to
1100 rpm at 100 oz-in on my custom machine. On paper, that's supposed to
offer 354 lbs thrust before losses. >>

Carlos,
I had been using the DM 4050 drivers on a heavily modified Emco Maier
FB-2. I bought mine a few months before this group started. Had I known of
Dan's drivers earlier...
A couple of comments regarding the 4050. The driver card has a H bridge
mosfet output as expected. The mosfets are small DIP packages soldered on
with surface mount style leads. There is no provisions for a heat sink. At 4
amps, the output chips are almost too hot to handle. At 5 amps, my drivers
got erratic. The driver closest to the fan was OK but the axis farthest away
was skipping. I never got the fault light to lite up. I machined large
aluminum heatsinks and clamped them on, added another fan, the erratic
behavior went away. I called the factory and they agreed that 4+ amps was not
recommended:( I thought that I needed a higher voltage power supply so I
went to some higher voltage/current drivers (AMP 7080). In terms of motor
torque, I can attest that the curve appears to drop much faster than you'd
first surmise. I'm convinced that the torque falls faster than the 2:1 ratio
can compensate for. What I mean is, though you have doubled the speed of the
motor, running twice as fast seems to lose more than half the torque.
The power required to drive a dovetail will be your major loss in these
small machines. You should use very light way oil. For a stepper based
system, linear ball bearings, whether Thompson round or THK style rails would
be a better choice.
To get 60+ inches a minute, one must use large power supplies and robust
motors and drivers. The axis travel distance is so short, to get up to speed
requires a lot of muscle. I would consider a 10 amp 80 volt driver if I was
building another high speed machine. This machine would have a direct drive
ballscrew with the motor running on quarter or eighth steps. (One thing they
don't tell you is at the finer microstep rates, motor torque falls off
somewhat. I've never been able to quantify the loss. I suspect it is quite
large, like ten or fifteen + percent.) The 4050 would be OK for geared down
steel munching work. Don't expect more than 10 or 15 inches a minute maximum
G0 rates.

I am ordering a Rong Fu-45 today. It will be for cutting metal only, no
high speed wood or plastic cutting. I will most likely end up putting the
4050's back into service running this mill. I will use .750 X .200 precision
rolled ball screws from Nook. For motors, I think I will run 34m470's in
series connection with a 2 : 1 XL cog belt reduction. I saw it written that
power supply current should be at least three times motor current so I will
need to build three 40 volt 20 amps supplies. (I scoped the output of my 10
amp supplies and could see the voltage dip during heavy acceleration) If the
4050s let me down, I may opt for Dan' s or AMP's drivers.

Peter
PT Engineering Co.