Need information on a Datametrics encoder.

[mjm]

Hello everyone. This is my first post here. Lots of good information on this site. I took that plunge and bought an Excello 605 cnc knee mill for my home shop. It has a Dynapath system 10 for driving the axis.

After some examination of the control I think it needs some steamlining (read gut it). I will keep the power supply and some of the periphials needed to make a system.

I got the model number off the Datametrics encoders (S-9705) and went searching on the net for info as to what ppr's they might be. Was'nt to successfull. Datametrics web site doesnt even mention anything about encoders in there product line. Do any of you know if perhaps Datametrics sold off there encoder producing portion?

If so to who? I would like to retain them if they are compatible for the retofit.

Thanks for any help.......Mark

[torinwalker]

I took the plunge too and now have a brand-spanking new old and used Ex-cell-o 605 CNC, and just like you, I am stuck with ElectroCraft E727 motors and Datametrics S9705 encoders for which I have no specifications.

Another gentleman here had BEI determine a replacement BEI encoder compatible with the S9705s, but he failed to mention which BEI encoders he ended up with!

I think I'll find that thread and see if he still follows it, or send him a direct email.

Where did you end up with all of this? I got nice copies of the original blueprints for my 605. Too bad there's no information about the encoders or motors (*sigh*).


Torin...

[Al_The_Man]

I found the info on the 0727 ElectroCraft, which I assume is a similar motor?
These are 28 lb-in continuous torque,
Peak torque 130 lb-in
Max terminal volt 120
Max rpm 2700
Max pulse amps 73a
Voltage constant 14/krpm
The encoders should not be to hard to decifer or reverse engineer, I would not think?
Al.

[torinwalker]

Dude, you're awesome.

I just got this beast home tonight. We took it off the flatbed trailer and literally slid it under the garage door with 0.5" to spare (after removing the motor)!

So I took off the covers and low and behold, three giant Nema 42 E727 motors from ElectroCraft and Datametrics S9705 encoders (whew - not resolvers!)

The encoders are nicely terminated at the back of the encoder with screw terminals and nicely printed with the terminal numbers and their function. The only thing I think I need to do is determine what the encoding rate is per revolution. I can probably do this with a counter. I probably have a pulse counter around here somewhere :-)

If you have an educated guess as to what it might be, that'd be just as helpful. Ultimately, it'll be determined when I plug in the ppr into emc3 and watch it rotate. If I get one nice revolution, I've guessed the right number.

I was SOO disappointed to learn that the guy selling me the machine (who is a wonderful guy, btw) took off the controller and electrical housings and the manual oiler and misters along with it. Damn!

I think I'm rambling at this point. Chalk it up to excitement over my new pre-christmas toy.

Thanks, Al. You've been a big help, as usual.


Torin...

[Al_The_Man]

For that age of machine I would hazard a guess at 1000 or 1024p/rev encoder, this would be a typical resolution.
If feeding in to a controller that multiplies the quadrature count X4 then you may see a result of 4k or 4096.
Al.

[mike_Kilroy]

I will look in my old Datametrics catalog when I get back to the office next week..... we used to be there sales rep in OH back in around 1978-ish...yes, they were bought out if I recall by Sumtak, who was bought out later by Daido, then Advantech. they were another company name b4 datametrics too.....u r making we think too hard....... aha! yes, they were Trump Ross encoder company b4 datametrics. they were all in same building up in little town in MA. I have fellow in from Israel all next week so may take me a while to get to my ol Datametric catalog. Maybe this history might help find the part no b4 I do?

Al is prob right, you should assume 1000PPR until you know otherwise.....

E727 was a common EC motor; sounds like Al's 0727 might be it; IIRC E727 was about 2.5: dia x 6" long..... it likewise is speced in my old EC 6x8" book on my darn shelf at office.....

[torinwalker]

All this information flooding in is absolutely wonderful. By the way, these motors are like 4" in diameter, not 2.5.

I'm now spec'ing out the 5i20 FPGA Anything card and a pair of 7i29 Dual-2KW H-bridge driver cards. The max current through each of the 2KW halves of the 7i29 are listed as 22A, with a max voltage of 165V (well over the motor's 120 rating.)

I know I could easily get away with a Gecko, but they only go up to 8A@80V and there's so much extra power here for fast jogging. I'm not so much for taking heavy cuts (ok, perhaps I'll evolve to that later ;-) but I would like to squeeze out the extra performance for, say, some really fast (read: cool) 3d surfacing. Besides, these Mesa cards really aren't any more expensive than the Geckos, so I'm going for it.

Incidentally, the guy who sold me the machine also provided some other details including these figures:

1.77in-lbs/amp
29lb-in stall torque.

So I'm going to go out on a limb here and figure that a nominal torque of 28lb-in would mean 15.8A. Plugging in a peak of 130lb-in and hey! -> 73Amps just like you said. So I think I'm more good with the 22A per channel on this 7i29 card.

How should I use this voltage constant to determine the best max voltage to drive them at? At 14/krpm, I get 37.8V for 2700 RPM. Why does this seem so low?

At 24V, these motors rotate (move the table) pretty slowly. Perhaps the video I saw was actually closer to the max speed than I anticipated.

Does this (jogging speed) look fast or slow to you guys?

X Axis: YouTube - x axis ex-cell-o cnc 605 for Sale mach 3 retrofitY Axis: YouTube - y axis ex-cell-o cnc 605 for Sale mach 3 retrofitZ Axis: YouTube - z axis ex-cell-o cnc 605 for Sale mach 3 retrofit
In 10 seconds, it traversed (one direction) slightly more than 1/4 of the table's travel or about 14 ~ 15 inches. That puts it very roughly at 84 ~ 90 ipm.

Is it fair to reason that at 38V, 132 ~ 142ipm wouldn't be that bad.


Torin...

[Al_The_Man]

At 29 lb-in continuous it looks close for what I have for these motors, they are just over 4" in dia, which I would expect for a motor of that size.
The 14/krpm is for no load condition, check the present power supply voltage, I would suspect you will find at least 80VDC, if not more.
Al.

[torinwalker]

Al,

Yeah, that's the problem. The who sold it also removed all the electronics and power. All I have is the iron with a few motors and bare wires. That's why I'm struggling, because I have no references to original equipment.

My current thinking is this: With a max voltage of 120V, the nominal (bus) voltage for these motors should be pretty high - at least way higher than the 38V I calculated earlier. And yes, now that you mention it, 38V for a no-load speed makes perfect sense. I guess I should now determine the voltage drop across the power supply for these motors when loaded to, what, 16A? (That's potentially three motors at 16A full load). Assuming a perfectly suited power supply, with each motor loaded such that they're drawing a nominal current value (which I don't know yet), the final bus voltage should be 38V @ full load.

I'm good with electronics, but not familiar enough yet with motor control dynamics yet, so the above is a guess. So, so long as the unloaded voltage of the system is less than 120V, and the power supply is stiff enough to maintain the voltage under load conditions, I could potentially settle for any voltage below 120V, right?

My real question is this: Will I need to be closer to 120V to be able to acheive the top speed and power availed to these motors, or should I worry more about finding a power supply that is rated for the expected loads (for example, a DC power supply of 48V @ 22A, or something like that.)

I'm not even sure if my assumption of 1.77/A and 29lb-in stall yielding 16A is a correct. This would be SO much easier if these electrocraft motors had documentation!

Say, Mike? This office you spoke of where the ElectroCraft book sits, it wouldn't happen to be the same office you'd likely be at on Monday, would it?

What I think I'll do now is go and look at any other motor that most closely matches the specs of this motor and extrapolate to determine the best bus setup for my motors. I mean - how different can it be if all the specs are nearly the same, right?

[Al_The_Man]

Here is a usefull link for power supply calc.
Power-Supply Considerations For Servo Amplifiers
Also see section 2.7.2. here on P.S. sizing.
http://www.a-m-c.com/download/manual...tallManual.pdf
I believe the previous value is wrong, I found an updated Ke value of 21v/krpm.
You can determine this by removing one of the motors and back feeding it at a known rpm and reading the generated armature voltage, and extrapolate using rpm & voltage.
As the article points out, there is no direct link from P.S. to motor, there is a drive in between.
Al.

[mike_Kilroy]

1.77in-lbs/amp 29lb-in stall torque. So I'm going to go out on a limb here and figure that a nominal torque of 28lb-in would mean 15.8A. Plugging in a peak of 130lb-in and hey! -> 73Amps just like you said. So I think I'm more good with the 22A per channel on this 7i29 card. How should I use this voltage constant to determine the best max voltage to drive them at? At 14/krpm, I get 37.8V for 2700 RPM. Why does this seem so low?

You've asked a couple time for verification of motor rated current of 16amps - yes, you are correct. it is simply 29/1.77 - assuming the numbers are correct.

Al is right that the 14/krpm is wrong. Kt & Kb are actually the same number, just different units. so you can double check by looking at any other motor's and ratio them to yours (as long as same units). so I picked a motor with Kt=1.79#-ft/amp or x12"/ft for your units = 21.48#-in/amp, its Kb=147.1v/krpm. so Kb is bigger than Kt so take your 1.77*147.1/21.48=12.1v/krpm

so if your 1.77 is right, then Kb=12.1 not 14

so max no load speed at 120v u said was max terminal voltage is 9900rpm - they would have limited it to mechanical limit of probably 6000rpm - this does not jive with specs Al found and posted previously, sorry. Also, we adjusted windings to get volts and amps that made sense in motors back then as today, so why would a winding be made this way? It does not make sense to waste such speed like this if indeed the motor is rated 120v. Also, most of our similar size motors have Kt that is almost 10x higher than this 1.77 - this is very very low for a general purpose motor, and again IIRC, the E727 was pretty popular in its day (sorry for missing its size dia b4 - I aint perfect!)

(Yes, my electrocraft book is at office I normally would go to monday morning but I have engineer in from israel all week I need to take to customer, starting 7:30a tomorrow (: I will try to break away and get there tues or see if one of my cohorts there can look it up and scan it to pdf for me along with datametrics if I didnt throw my old sales binder catalog away - I like to keep old data but I cannot picture my trump ross 3 ring binder on the shelf...)

[mike_Kilroy]

I see on ebay one of the encoders;

Dynapath Datametrics Optcoder Encoder #S-9705A-500 - eBay (item 140407427877 end time Dec-08-10 11:30:34 PST)

it reminded me that the PPR was after a dash after the S-9705-500 here is 500ppr encoder... urs has no more part no than s-9705?

I found another post in another forum on this too? same thing? I posted "I just found e727 for sale on ebay - seller says it was x axis on HURCO machine..... maybe if there is a Hurco group on cnczone u shud post over there for motor spec too??"