Thanks, Dave.

Today I got the end of the strands stripped and got the winds checked for shorts. That MidroDan wire is *not* easy to bare, I suppose that is good and it probably accounts for the fact that I got this far without shorting a winding.

I checked each strand and got good continuity, also checked each for shorts to the other four ends and to a bare spot on the end of a hammerhead) and didn't find any shorts.

I get the same resistance across each of the three phases, 000.3 Ohms (the test leads are 000.2 Ohms shorted).

I've been reading "Dry Testing" thread over on WattFlyer and some of the other troubleshooting and testing stuff here and think I have a handle on the rest of the testing I want to do.

But before I went ahead and joined the phases I wanted to ask if there is any other testing I should do with the phases still not joined. It looked all the rest of the testing is with the phases joined.

I think I followed your and Kevin's discussion on the lengths or the motor leads but I'm going to ask, if I bond the wire with a solder joint near the windings and at equal distances like you describe, the length of the motor leads from there back to the ESC does not matter, right? I don't need to try to keep the lengths of the motor leads the same the way back to the bullets?

Jack

Originally Posted by jackerbes

... if there is any other testing I should do with the phases still not joined.

The only thing I test before terminating is for short to stator and phase equalty ... so that's that. You're ready to rock'n'roll. But! Since you don't have a milliohm meter, you could do a spin test (drill press, drill, whatever) and get a better reading of the phase matches. Depends on how much trouble you want to go to.

...if I bond the wire with a solder joint near the windings and at equal distances like you describe, the length of the motor leads from there back to the ESC does not matter, right? I don't need to try to keep the lengths of the motor leads the same the way back to the bullets?

That's partly why I don't bother. Fact is, if you want the wire lengths to match across all phases, the wire has to be the same length from the ESC all the way to where the phases join. And the benefit, if there is one, will approach zero so closely that it's almost imponderable.

Fuhgeddit. Wire not wrapped around a tooth is almost irrelevant. (This is _not_ true on the other side of the ESC, where you're running DC current. There you have to be sure not to make the wires too long. But if you don't add any extra wire, don't worry about it. The mfger usually has the sense to make those wires within reasonable limits and allow for the battery wire).

But I digress.


Dave

Jack, it sounds like you're headed in the right direction. One trick o' the trade I'll add is the "feel how it spins" test (<- note the correctly elided f ). It's quick & dirty, but IMHO a very powerful diagnostic tool...

After winding (also after termination), put the rotor on and spin it with your fingers. If it spins like it's soaked in molasses (won't "freewheel to a stop"), then you know you've done something wrong... time to break out the multimeter and start investigating for cross phase shorts, stator shorts, incorrect winding direction, incorrect termination. If it spins freely (with cogging, but it "freewheels" like it did without wire), then you're cleared of all the above problems; the only winding errors possible would be internal shorts and added/missing turns (you need a milliohm meter to sniff out those problems).

Cheers,
Kev

Thanks guys, those are the things I intend to do next, to connect the phases, put it back together, and start feeling it and measuring some voltages.

Jack

Originally Posted by Truglodite

Rodrigo, the 36turn I'm talking about does use a single strand of 26awg. The parallel part is just the way it's terminated. Make sure you understand well what a parallel wind is in this context before you start winding. It's not a multistrand wind, but it is a bit different than the standard wind. Here's a parallel delta schematic... sorta:

http://www.rcgroups.com/forums/showp...7&postcount=88

It's not a schematic, but you might get the idea by just looking at the pictures. It's wound like a standard DLRK delta, but the 4 poles per phase are split in to 2 pairs, and terminated parallel to each other.

Jack,
The DT750 spreadsheet is located on my blog. I think the version on my blog is a little old; my newest spreadsheet is much better... uses the latest DCalc constants and removes a lot of the fudging we often see with efficiency and thrust numbers. I'll upload the latest version after this post.

Cheers,
Kev

Truglodite Thanks!

I rewinding the motor with 36 turns of wire 26 awg single and using this scheme has no error. I will rewind it like this, anything you say.

Truglodite wrote:

"..The DT750 spreadsheet is located on my blog. I think the version on my blog is a little old; my newest spreadsheet is much better..."

The rewind_DT750.zip file on your blog page right now:

http://www.rcgroups.com/forums/showa...5&d=1273886610

contains one file, "rewind DT750.xlsx" and I cannot open that file with OpenOffice.org Calc as I can the DT750.xls file I obtained somewhere previously. I get an error from OpenOffice saying that it is an unsupported excel format, OpenOffice has dealt with your other DT750 spreadsheet fine. I tried renaming it to *.xls and that did not help.

I'm not in a rush or pressing you to get the new one up, I just thought maybe you wanted to know this. I still have DT750.xls and it is working fine for me.

Jack

OK, I cobbled up a testing rig and used my drill press to test the AC voltage on the phases.

I cut the leads to even and equal lengths, and soldered the ends together, and marked them 1-2, 2-3, and 3-1.

My drill press is running at 600 RPM and I got voltages as follows:

1-2 = 1.39V
2-3 = 1.43V
3-1 = 1.40V

I wasn't sure what the variation could be or might be so I took a new DAT-750 and ran it through the same test at the same drill press RPM. The results on that were:

RED-WHITE === 1.661V
WHITE-BLACK == 1.626V
BLACK-RED ==== 1.639V

On the new motor it took 4-5 seconds for the voltage readings to climb from about 1.30V until they stabilized at the readings above. On the rewind the voltages stabilized much quicker.

I feel a little better about the variation in mine after seeing the new motor's readings. So who had the worst day, me or the six year old kid in China?

And anything you want to share with me about the readings will be interesting of course.

Jack

An approximately 3% error is consistent with the idea that one phase has one turn too many or too few. It is, however, not infallible. All the test equipment has variances also. Before I got all messed around about it, I'd retest a few times and average the results. If it's consistent, you probably missed a count somewhere (there are other explanations, some not even unlikely). It would also be instructive to desolder and test the phases independently.

Dave

Truglodite!
Now I understand what you told me, I reboinar DT750 motor with 36 turns of wire 26 awg each motor stator. and I do like desenhologo below. am I correct?

http://www.powercroco.de/schemamodus12N_YY2.html

timocharis wrote:

"..It would also be instructive to desolder and test the phases independently..."

I had pressed on with the project before I saw that. I wish now it had occurred to me.

At any rate I have the motor all finished up and ready for a test with an ESC. I figured that if the stock winds on the DAT-750 run OK with higher variations, mine should run OK too with what I saw there.

I shorted the leads and the cogging is very consistent, no chugging or roughness at all. The windings all Ohm out about the same on my Fluke 79 and I'm ready to take a chance on it with my oldest and cheapest ESC.

If it does not go well, I'll do it again and do it right!

The photo is of the finished motor, I hope it runs as well as it looks...

Jack

"..If it does not go well, I'll do it again and do it right!..."

It fired up and seem to be doing fine. Got the following numbers:

no load = 1.86A @ 11.91V = 11,000 RPM

no load = 1.6A @ 9.4V = 870 RPM

I Let it run for a minute at half throttle and nothing is more than barely warm.

Looks like a keeper. I had put a new set of bearings in it and I'm hearing a noisy bearing already. The quality of the two for a dime bearings I got on close out from BP Hobbies is not awe inspiring. I noticed that the seal had fallen off of a bearing in one of the sets and they show no evidence of lubrication. I'll lube the new set I am using and see what that does for things. I might be hearing a dry bearing.

Jack

I just replaced my bearings on an older motor. I found that the shaft clearance may be 1 or 2 thousandths clearances. Not a lot but I can see it moving at close inspection. I believe that the new BP bearings I put in are ok but the clearance on a loose fit shaft alows some floating and creates some of the noises.
**Neons** Bob

I'm waiting for someone to make ball-less ceramic bearings at this scale... much more precise, doesn't get hot, and lasts forever.

No load looks about right, Kv looks a tad high. In any event, it's worth taking a run at it!


Dave

Originally Posted by jackerbes

I get an error from OpenOffice saying that it is an unsupported excel format, OpenOffice has dealt with your other DT750 spreadsheet fine. I tried renaming it to *.xls and that did not help.

Jack

Ah, thanks for letting me know. I upgraded to Excel 2007 recently, so looks like I need to redo those files using compatibility mode. I'll do that soon.

Thanks,
Kev