Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: BR549 on September 21, 2012, 10:09:30 PM
-
Has anyone did this hookup yet. Been a long time here since setting up any ac servos. The sanyos use a straight up commutation 120 deg hookup abc/abc on the old servo dynamics drives but the ABs use a different sequence. Can't wrap my head around the change in sequence to make it work.
I keep getting commutation angle errors (;-)
(;-) TP
-
What AB drives?
Hood
-
1398-019 's
(;-) TP
-
I see part of my problem the UltraMaster software ver 1.60 I have will NOT allow one to modify the motor files. So there is no way for me to setup this drive to a custom motor.
(;-( TP
-
Hi Terry,
Open Ultra Master Help file and look for "Setting Ultra Master Advanced Mode". This will tell you how to access the advanced features of Ultra Master. Custom motor creating is one of them.
Dan
-
Change the shortcut target to C:\ULTRAMST\ULTRAMST.EXE /A then open UltraMaster from that shortcut and in File menu you will see Edit Parameter Files.
If however you do not have any commutation signals on your motor you will not get very far.
Hood
-
HIYA Hood yes I found your software trick with a web search. The /A switch did the trick.
Worked out the commutation problem . Now I am trying to test with vel mode testing but I get a overspeed fault. The input field for setting the rpm will not take any values keeps sayin to enter a value between 0.0 and 0.0 . I ener 100 but it rejects it.
Auto setup does not do anything yet(;-) Clip on start but nothing happens yet(;-)
Working on it. The AB software is strange compared to the old servo dynamics setup program (;-)
(;-) TP
-
What have you got set for the RPM, see pic.
Hood
-
HUM I don't remember that screen anywhere. Let me check again.
(;-) TP
-
Double click Drive Parameters.
Hood
-
OK found it it is at 3000 rpm which is the upper limit on the servo motor by spec.
I will study this some more. I am not positive the commutation error is total fixed.
I worked out the sequence from an example of both sequences and do not get the error on start up. But still have not gotten it to move yet.
Thanks for the help, (;- )TP
-
Use the Encoder alignment option to see.
Hood
-
HUM still missing something here, No luck. Th epahsing seems simple enought BUT the halls do not exactly line up so there is an offset but how much? That is the $100 question.
OH well maybe AB will have some insight, Yea right
(;-) TP
-
Soes the encoder alignment tool not tell you the offset you are needing? I think in the older drives (ones youhave) you can then tell the drive to offset it automatically.
If you want to align the encoder properly you can do it either via that tool or you can do it with a dual trace scope. I prefer the scope method personally. If you want to now the scope method give me a shout as I should have the info here somewhere, sure AlTheMan on the zone told me it years ago.
Hood
-
Fiddle some more with it and seem to have the commutation correct and the encoder show zero offset. So maybe that part is right now.
Still no movement when using the control panel to test . something in the motor file is preventing it from accepting a vel number . I can load an odd file (not for that motor and the vel number show up in the test panel.
I hack at it some more tommorrow.
I am suely missing wsome just have not figured out what yet )(;-)
(;-) TP
-
OK no luck so far. ANY AB drive experts out there?
I have the spec sheets for both components BUT not enough brains to work it out.
(;-)
-
You've got Hood here, Terry ;) Can't think of someone with more experience with AB drives around here.
Dan
-
I am more used to the newer drives but have used the DDM's and actually currently have one on the Conects spindle.
Post your details of the motor and maybe screenshots of your custom motor setup.
Hood
-
Here are the specs on the Sanyo Denki P5 motors.
What I can't get straight is the proper commutation AND maybe an offset of the Hall setup.
The Ultra software is on the shop puter so I do not have a screenshot. I have tried various combinations all failed to work (;-).
-
I presume you have U connected to pin 13 on the drives J1 connection , V to 14 and W to 15 and have left the U- V- and W- unconnected?
Red motor wire to R on drive, White to S , Black to T and have the Yellow/Green to Gnd on the drive?
Hood
-
Yes that is correct.
(;-) TP
-
Ok so then it may just be the custom profile you have made up. Cant remember if you can save the drive parameters to a file with the older drives or not. If you can you could zip and attach along with your custom motor profile and I could see if I see an error.
Oh actually not sure if you can attach the custom profile, with the DSD drives you can save the motor data base but cant remember if you can with the DDM's, will have a look. Also the DDM's are restricted to one custom motor profile where you can have as many as you like with the newer drives, not that that matters to you, just thinking out loud really. Anyway will start up mater and see.
Hood
-
Ok just looked and you can save the drives parameters, file menu then Save or Save As.
I assume if you have the custom motor profile loaded then that will be saved as well.
Hood
-
I will create a new file and try again. The best I can do with testing is make the motor jump about 60 degs with auto tune and toggling back an forth with forward and reverse.
That seems to indicate a commutation error but there are ONLY 6 possible combinations to try and I have tried all of them as well.
I am thinking MAYBE it is a HALL offset problem . To my eye it seems the commutation is off by 120 degs but that is just my looking at the sequincing and I could be wrong. BUT I have tried 0/60/120/240/300 deg hall offsets ,all failed as well.
The drive ONLY stores 1 custom motor config, but in the software you can create as many as you like and reload .
(;-) TP
-
On the custom motors, yes that is true but you lose that custom motor profile when you write a new one so if you want it back you need to recreate it. Obviously if its in the drive then thats fine but anyway its nothing to do with the issue at hand so is just a distraction, I was just typoing as I was thinking to remind myself of the DDM's ;D
Is there a chance the encoder could have been moved?
Do you have a dual trace scope and some means to rotate the motor such as a lathe?
Hood
-
The motor was new out of the box. AND I have 3 more new and 4 older ones off the mill.
I will go out and open a nother new box(;-) and try that.
(;-) TP
-
The best I can do with testing is make the motor jump about 60 degs with auto tune and toggling back an forth with forward and reverse.
(;-) TP
Please excuse the interrupt men,
I had a very similar reaction once when I had the wrong encoder count entered in the motor data. Corrected that and all was well.
I assumed 2000 and it was actually 500.
Carry on.
Regards,
Russ :)
-
OK broke out a new one and same results. I have tried all 4 drives and same results so it has to be a setting somewhere.
(;-) TP
-
On the old servo drives (SDs) 2000 was the count and it worked fine as well in the MACH3 setup.
But I agree a wrong encoder count will cause the motor to jump(clog) on startup.
(;-) TP
-
Attach your parameter file and I will see if I can find the issue.
Hood
-
I dont see the amount of poles listed in that pdf, maybe I am missing them?
Hood
-
SHould be 8 poles if I am not mistaken BUT I have tried all the pole settings as well. I am waiting on confirmation from Sanyo Denki of that value.
(;-)
-
The motors are
Sanyo Denki P50B08100VCKS7 1kw Serial 06119001
There is VERY little info out there that I can find.
(;-) TP
-
Put the motor wires together and rotate the shaft by hand and you should be able to count the poles.
Hood
-
I have a pdf with info on the P series motors but no poles listed, weird as usually that is one of the first things motor makers state.
I have looked at their drives/motors before but info is poor and seems they drop it from there downloads if they stop producing them so I grab all info when I get a chance.
Hood
-
OK all leads together you get smooth drag no cogging. Any 2 leads together and you get 4 clogs.
(;-) TP
-
4 pole then.
Hood
-
Tired that setting again no luck. I am going to try all the offsets next with those settings.
One thing if I test the encoder offsets it will move slightly to find the encoder then falut with a commutation error. BUT I have also tried all 6 combos of the hall seqence and it stays the same.
(;-) TP
-
OK i verified that the motor has 8000 counts per rev so that is a 2000 count encoder.
As to the rest I don't have a clue as to the magic settings required or IF the drive can even RUN the motor AC servo/drives can be very tricky to match to a foriegn motor.
Always had good luck up until now (:-)
(;-) TP
-
Might be a moot point, but are you sure that ol' antique AB drive is in good shape ? ::) :)
Have you run any AB motors with it ?
Russ :)
-
I also have 3 more I have tested SURELY one of them would be still good (;-). They go through all the motions.
I don't have an AB servo to test with.
I think I am just missing the secret handshake. OR it may be time to pack it all up and call the scrap man. (;-)
(;-) TP
-
HIYA HOOD question for you. What line drivers are you using for the step and direction on the AB drives?
(;-) TP
-
On the big lathe I use the acustep BOB that CNC Building blocks used to make, it has diff out. On the Beaver Mill and Bridgeport mill I made up my own boards with SN75172 quad drivers and on the wee lathe I use the CSMIO/IP-S controller which has differential outs.
Hood
-
Hi Terry,
Here is what Hood helped me to set up for mine. :)
Russ
-
WOuld you happen to have a schematic of that setup. I can solder (;-)
Note on the drive/motor does not look good. IF I could figure out the crossover between the two systems it would work BUT there seems to be about a 1000 combinations to try to make it work.
So far after 3days of combinations it still does not match up.
I was hoping to find someone that could take the Phase/Hall sequence that AB wants and look at the Phase/Hall sequence that the SDenki does and figure out the conversion.
I have drawn them out in cad and tried what I thought to be the good combo and all the variants of that without luck.
Several Servo techs have offered their good lucks (;-) IF AB were cooperative they could probably look at the specs and get it right BUT IF a frog had wings he would not bump his rear when he jumped.
BUT to be fair the AB drive does have the tools to do it. BUT it is like having an atomic bomb and NO plutonium.
(;-) TP
-
Using a scope to verify the phasing, like Hood suggested is the best way.
Dan
-
I have the sequence graphs of the motor AND the drive BUT the combos I come up with do not work.
(;-) TP
-
TP you dont seem to want help from me so I will bow out :)
Hood
-
Hood where was it you read that I don' t want your help?
(;-) TP
-
WOuld you happen to have a schematic of that setup. I can solder (;-)
(;-) TP
I just used the data sheet.
Russ
-
Well you dont seem to want to upload your parameter file for me to look at to see if there is any issues in it.
There may not be but I dont think its commutation as the drives can compensate if you tell them to and you said you have.
I have set up quite a few motors that were not intended to run off these drives and not had a problem. As long as you have an encoder with the correct commutation signals/ pole count then you only need the data asked for in the motor file.
I currently have a motor on an Ultra 200 drive that originally had a resolver on it, it runs the spindle on my manual lathe and it was no problem t set up.
I have also set up 2 Yaskawas one which I had to fit a new encoder on and also an indramat that was resolver and I put an encoder on it. All were straightforward to do although I chose to line up the encoder rather than let the drive offset. However it does work as I have offset it before actually rotating it physically and it worked fine.
Hood
-
Sorry I missed that request.
Here it is, (;-) TP
-
Motor file looks fine with the exception of the hall offset but think you said you had changed that, would set to zero and see anyway.
What I was also wanting was the drives parameter file.
If there is no problem with it you will really need a scope to see if the wiring is correct with respect to the halls, I dont think it will be wrong as I think you said the drive said commutation was not needing offset but a scope would confirm.
Hood
-
I think the problem is that the AB drive does comutation in a CW direction and the motor does it in CCW. The only way I can see to even get close will be to scope it see IF it can be done.
It may be that the encoder will have to be shifted as well AND I believe the Halls are part of the encoder on this motor. May not be able to do what needs to be done.
I think this is the Drives FIle.
Thanks For the help, (;-) TP
-
I am sure you can invert the commutation in the motor profile.
Hood
-
Here it is TP. might be worth a go.
Hood
-
THe manual recommended using Hall/hall but I will give it a try.
(;-) TP
-
No you will need to use HALL/HALL, it was the invert bit I was meaning. The only reason the ABS is in that pic is because its that as default in the custom profile and I only opened it to look to see where the invert was.
Hood
-
It attempts to start/spin but then errors out on a commutation angle error.
(;-) TP
-
Will have a look at your parameter file when I get home tonight, bit busy at the moment.
Hood
-
TP thast just the drive list fle, what I am looking for is for you to save your parameters from the file menu and make a par file.
Hood
-
Hopefully this is what you wanted.
(;-) TP
-
BTW I just looked at the commutation of the Sanyo and it is what AB wants so doesnt look like it is the problem.
Will have a look at the parameters and see if I can find an issue but maybe scoping to see if the signals are correct with wire colours will be the only way.
Hood
-
All looks well in the parameters but I would say to put the offest back to zero and try.
Have you tried to auto tune, the tuning is what I would expect to make a motor run so unlikely its the problem.
Hood
-
The best I can make it do is start to spin and then error out with a commutation error.
In one combo I can make it spin very slow about 120 deg and then error out.
THere are about 360 combinations to try.
AUto tune does about the same thing as run does. IF the motor can't spin up then it can't be auto tuned.
I dug out the scope but can't find the leads so I will try to get it scoped and see what is what. IF I can remember how to setup the scope(;-) I used to have a young eng in house but he got married and moved out .
(;-) TP
-
Its a strange one indeed, should be a piece of cake to set that up. The pdf shows the commutation is exactly what AB wants or at least as far as I can see it does, same phases same halls for those phases.
Only thing I can think of is the wires are not corresponding to the pdf but find that hard to believe, if it was a cheap Chinese motor then yes I could believe it but not with a Sany.
BTW I found the code I used to make my first diff board, similar to pics, not sure if its any use to you as its metric ;D but also made to fit directly into a PMDX 122, suppose you could just fit connectors instead of the heavy copper wire.
Hood
-
Can yall review the schematic and PLEASE explain what the ABCD's are . Not sure of some of th eempty pins the picture was not that clear.
I am not an electronics wiz..
(;-) TP
-
See if you understand this.
Hood
-
BTW that is a D (Direction) next to the S (Step) at the top and there should be another one beside the second S but I forgot it ;D
If you look in the other pic you will see the wire jumper that connects the Gnd to the Gnd pin on the driver and you will also see the red wire connecting the 4 pins on the two drivers and then the tail of that goes to your BOBs 5v supply.
Hood
-
ok which side connects to the AB DRIVE
-
The Top, Step + Step - Dir + Dir -
Hood
-
Please double check me one more time.
Is the chip the correct side up by the indent at the top?
(;-) TP
-
Looks fine.
Hood
-
I may be wrong, but I think your view is not from the top. It is the belly view, the bottom of Hoods PCB.
Just a mirror image of the Data sheet TOP view .... I think.
Russ
-
This is the one I did for the Scorpion, I put Gnd connections to go to the drive, they are what the resistors are. Its supposed to be done that way but never had issues without them on the Beaver Mill which has the first style but you do want to connect the Bob Gnd to the Drives Encoder Gnd.
Hood
-
Russ, thats correct, the drawing TP has is the copper side of the PCB.
Hood
-
BTW when I said encoder Gnd on the drive I am meaning the aux encoder power common. (Pin 2 on J1)
Hood
-
Heard back from AB today it was a sweet sorry about your luck email and we really would like to sell you some new drives and matching motors (;-)
(;-) TP
-
Heard back from AB today it was a sweet sorry about your luck email and we really would like to sell you some new drives and matching motors (;-)
(;-) TP
They are a little cheaper if you buy 100 of each. :)
-
Ah for the good old days when engineers at a company took pride in knowing there products inside and out. AND they took it as a challenge to PROVE they knew how to make things work.
Now it is so sorry cannot make ching ching work you need new ching.
Those days are long gone, (;-) TP
-
Well I won't win BEST OF SHOW but the line driver is together. All the boards parts, board,socket etc can from radioshack. The chips came from mouser.
(;-) TP
-
Looks neat :) I would suggest you twist the diff step and dir wires into pairs (step + and Step-, Dir+ and Dir- ) though.
Hood
-
OK so far I have not been able to make it work. Everything looks right(;-)
SO is there a way to test the output of the line driver ?
Is there a way to test the inputs of the DRIVE to make sure that part works?
(;-) TP
-
Scope to test the outputs. As for testing the drives inputs suppose the only way is to apply a pulse stream and see.
Hood
-
SO if I apply the steps output from the BOB to the steps input on the drive pin 14/15 it should at least move?
(;-) TP
-
Yes.
Do you have the Gnd reference of the Driver to the encoder Gnd pin?
Hood
-
YES it goes to pin 2. I have a good set of signals going into the LD on the scope but nothing coming out I check from S+ to com and Dir+ tto com.
Back to the drawing board and get out the magnifing glasses again. I need to double double check the chip for the key location to make sure it is in correctly.
I also checked the drive by running the step lines to the drive S+ and S- and got nothing from the drive.
I also double checked the drive set and it is set to FOLLOW step and direction.
(;-)TP
-
Do you have the enable on the chip connected, you have to have it either connected to Vcc or Gnd depending on whether you want active Hi or Lo.Cant remember the pins will look in a bit.
Hood
-
Ok its pin 4 and 12, one needs to be connected. Pin 4 to Vcc or pin 12 to Gnd.
Hood
-
AHHHH NOPE I did not know about enabling the chip(;-)
OK I take that back I see on the board that pin 12 IS going to VCC
I found a WISKER hiding on the board hopefully that may have been the problem.
(;-) TP
-
Pin 4 should be to Vcc or Pin 12 to Gnd if I recall.
Hood
-
ITS ALIVE, Well at least the A axis is moving under MACH3 control again. Easy part now is to rewire the other 3 axis to match the AB drive.
(;-) Thanks GUYS , (;-) TP