Back to top

Goldline Motor Configuration with the AKD

Documentation to wire a Kollmorgen Goldline servo motor for use with the AKD servo drive.  Basic drive parameter settings are also provided.

Comments

Mike_Kilroy's picture
Mike_Kilroy
It is easy to get idea of torque angle control backwards;  I believe the author here has done so also...

Author here states:
image

I believe this comment is wrong...  TAA REDUCES Kb & so too Kt as speed increases...  This means a Goldline motor with its buried magnets on a servostar drive with TAA produces LESS torque as speed increases.  So a drive WITHOUT TAA does NOT reduce torque at higher speeds!

Second, the REAL issue is not torque IT IS SPEED!!!  As example, our B604B motor rated 3150rpm can only achieve 2700rpm top speed with 240Vac line input on AKD!  

So your NOTE should not mention torque, but should mention SPEED reduction instead!

Let this be yet ANOTHER request by me to ADD TAA into the AKD!!!  We do many many retrofits reusing the present goldline motors, and this reduction in performance is very dramatic sometimes!  PLEASE take a few moments and add the algorithm in!

Mike Kilroy

 
Mike_Kilroy's picture
Mike_Kilroy
I would be interested in how the MOTOR.PHASE values for different pole counts were determined...

I gave up a long time ago trying to make a set value for phase on these retrofits;  auto wake and shake just works too well, and it seems the phase is sometimes 30 or 60 degrees different on identical motors and wiring with the AKD - I know, that does not make sense.
 
Mike_Kilroy's picture
Mike_Kilroy
Since I am being picky, I wonder if tranformation ratio for XT motors is a typo:
image
My understanding is the ratio variable in AKD really works over about 0.2 to 3.5 range.

I believe we have various AKD drives running various MT motors with resolvers, and IIRC all has default .50 ratio.
Michael Brown_455's picture
Michael Brown_455
Several different types of Goldline motors were tested and we found that the AKD drive has a 120 degree phase shift compared to S600, S300 and S700 commutation.
kenny.hampton@kollmorgen.com's picture
kenny.hampton@k...
The nominal transformation ratio is correct, the information supplied in this article was determined from actual Goldline and XT motors tested in our lab.  
This information was not calculated.  
Mike_Kilroy's picture
Mike_Kilroy
That is weird Kenny as I have:

2pc MT506C running on AKD with FB1.RESKTR=0.50 for 4 years
6pc MT504A running on AKD with FB1.RESKTR=0.50 for (oldest) 3 years

And those are only the ones I could find quickly. We have more XT resolvers running in the field.  NONE ever got this default FB1.RESKTR changed.  We DO run some Moog motors on printing presses that have required this to be tweaked down a tad to around .45 to operate correctly.  No other brand we have ever run required adjusting the transformation ratio.  As you know, we sorta specialize in retrofitting old obsolete drives of all manufactures with ours, specifically now of course the AKD. 

It is my strong opinion Kenny that a motor requiring 0.01 resolver transformation ratio will absolutely not run if set to 0.50.  Consider it:  say 10Vrms reference, that means the output of the resolver will be 0.1Vrms!  We both know the AKD will not run with that.  Said another way, in 40+ years of working with resolvers, I have never seen ANY resolver with a transformation ratio of .01  (1%).  That means to get the required 3-4Vrms output from it you would need to put in 400Vrms into the reference.

We DO run old 64 pole Reliance resolvers on spindles that have a transformation ratio of approx 4% (.04).  For these to work with drives, I use a "20v line" audio transformer between the drive and this resolver reference:  I boost the 10Vrms reference to 130Vrms in order to get the required 5Vrms output from sine and cosine.  Kollmorgen does not, and as far as I know, has never, sell ANY product with a 1%  (0.01 value here) transformation ratio.  

It is for this reason I flagged this as a typo error.   I respectfully submit that it is a typo and, if one checks back to their notes from the actual tests, they will find the mistake.

I will be back in my office in a couple days, and if you wish, I will add a comment here with the Harowe resolver part number in the XT motors and a copy of its spec sheet showing the 50% transformation ratio.  I would likewise be happy to send you our .akd parameter files running the above mentioned standard MT50x series motors if you wish.
 
Mike_Kilroy's picture
Mike_Kilroy
Hey Kenny,  I asked one of my guys back at the office to look up the resolver part number in a standard MT50x resolver motor. 

It is 21brcx-600-j33 on the MT50

spec sheet shows:
image
So you see that the default .50 is indeed close enough to use.  If you are a purest, you could make it .47 as Michael  did for the B series:  we do  not do that in the field - we have enough OTHER problems on a typical retrofit to contend with!  If we have ANY concern, we will scope the sine/cosine and sum of the squares to be sure all are close to middle of the range. 

If you do not want to correct the typo in this white paper, that is fine with me.