Convert from Anilam Crusader M to EMC 2

only picking up this thread now, and please ignore if this is of no use or overly obvious:
my bridgport was an anilam crusader and fitted with a montec card.

I have had the green and red cards over the years.

Not sure wha tit is but these older card have a tendancey to blow the power tansistors. one can replace them and tbf that is what I generally do(rs sell the parts).
norwin say they changed the boards because the boards tended to blow, and the designated replacment transistors have a lessor spec then the originally design spec.
that being said, my mill has been stable for two years now.

fit independant servo drives enable lines on each servor drive card. When the transistors go U/S, they can go short cct and those axis flippin MOVE!!!!

I have the norwin setup pfd if you need a copy, pm me. it is important to get that servo drive card stable before PID tunning in software.
looking at your P and D values suggest things are ok, they are simular to mine.

Graydog, thanks.

I agree that getting the drives stable is an important first step.

Here is what I did:

1. open the following error way up to allow large error without aborting.
2. Test PID at zero gains for one axis at a time
3. Adjust the balance pots on my drives to minimize drift.

In my case, if the drift was in the "positive" direction, I would turn X-axis CCW, Y-axis CW, Z-axis CCW

At this point, using only P-gain, performance seemed decent, with some significant error during accel and decel

After that, I followed the advice here: wiki.linuxcnc.org/cgi-bin/wiki.pl?Gantry...Pulse_Generator_Mesa

and set my FF1 parameter to 1, and kept only the P-gain. following error was less than .0005" on each axis individually which seemed to me that was probably as good as I could hope for given the encoder resolution.

After decreasing the allowable following error to .002", and not running successfully, I increased it back to .005" and ran a test engraving successfully. My guess is that increased error is due to small backlash when changing direction.

In any case, I believe that observed performance is likely "better than new" compared to the original control, and I am a happy customer!

Thanks.

That is slightly different to how I turn my axis, well the start bit. The Nowin guy suggest tieing the -+10V to zero iirc then running though thier pfd setup guide.
one done run the above type method. This ensures you servo control card is not fighting you software settings.
what is interesting on older machines like ours Nowins suggest inducing a very small drift. you have to get this right else the system snaches, and ripples edges.


at the end for the day, if it is working right magic

respect
Graham

I just picked up a Bridgeport mill with Anilam Crusader II control. Looking to convert it to Linuxcnc. I'm leaning towards reusing the exiting SEM motors. They have the tach feedback. I'd like to keep the existing scales for DRO manual purposes and maybe add encoders to the motors. The mill has backlash, not sure how much of it I can take out, hence reading the encoder off the motor instead of the scales would be better?
I'm not sure how robust the existing drives are. Should I re-use them/? AMC drives has been suggested, what model should I look at? I'm hearing high frequency whine coming from the power/drive cabinet. I assume it's coming from the drives. I this normal? I don't like hearing it, so I'm also looking for drives that are quiet.
I most likely will run these with MESA hardware. So any advice is welcome...

Hi,

I converted an Anilam Crusader M w/ the SEM motors and West Amp servo Amps. I think there are some similarities to the systems and a bunch others have done it as well. There's another thread on this forum of guy doing just what you did, I posted a few things for him in his conversion, so look for that when you get a chance.

The whine you hear is the 5kHz switching frequency of the old West Amp servo's. 5Khz is w/ in the audible range for our ears. New servo amps are higher, around 20kHz or more and quieter:-)

I have looked into the replacements for the servo amps. First off, a manual on the sem servo motors:: www.hhrobertsmachinery.com/Accessories/SEM/SEM-MT.pdf

The SEM servo's are 30amps peak, so a 30 amp peak drive would work. I like the copley 423. However, I have 2 brand new copley 422's that I want to use that are only 20Amps peak, but I bet they'll work great. I'm still using hte original West Amps and that whining is annoying...

West Amp went out of business and the 'story' is that many of the engineers went to Servo Dynamics and the Servo Dynamics SD1525-10 is a 15amp cont/25amp peak servo amp that looks a lot like the west amps and is a replacement.

One thing to watch out for is that the DC bus voltage is higher than 100Vdc, I believe, and many of the AMC or Copley amps only goto 80Vdc. I would measure the input power on your mill to verify, I think mine is near 120vdc, but I don't recall off hand. The copley 422 and 423 go to 180Vdc. www.copleycontrols.com/Motion/pdf/412.pdf

I am currently using tach feedback to the original west amp servo's and closing the loop around the linear scales. It 'mostly' works and I get some 'movement' that can be called hunting/pecking/dither where the control system is hunting between two encoder spots. I was bent out of shape at first on this, but then when you think about the linear scale resolution is 0.0004", the hunting really isn't going to affect anything I'm doing. There are fixes for it, put a 2nd encoder (like you mentioned) on the servo and dual loop, or use single loop w/ an encoder on the motor only. Either would work.

I am using a 5i25/7i77. I just bought a 7i92 to try on a small lathe I'd like to retro-fit. You might consider going to a 7i92/7i77 setup with the PREEMPT kernal. I think it will take a smidge more work to setup, but that seems to be the future of linuxCNC as the RTAI kernal is not supported as well (I guess, not an expert in this). Either will work and will take time/learning to get going. I like the ethernet because you can string it a decent distance and should be 'ok'...

Lastly, John Thornton has converted a mill like yours and has info here: gnipsel.com/linuxcnc/
I like these guys at Sector 67 for info: code.google.com/archive/p/sector67-sandb...rojectSheetCake.wiki
Here's one where the guy used PICO systems and he has a schematic I found useful: sites.google.com/site/bmklawt/home/emc2 (also, west amp servo manuals)

Good luck,
Mark

Mark, thanks for the detailed response. I will look into the copley drives as I can't deal with the 5Khz switching frequency.
I think I will go with the 5i25 7i77 combo but will probably need an additional encoder card? What options does MESA offer for additional encoder boards? I'm leaning towards keeping the linear scales for manual positioning and not using them in the pid loop. I would have to read in possibly 8 encoders with MPG and 4th axis. I don't want to keep the old display for that.
Also, does anyone have specs on the rotary encoder anilam used on the Z?
Last for now, I could not not find any documentation on the Anilam Crusader mechanical Bridgeport CNC conversion components. Does anyone have an original parts/assembly manual?

OT-CNC,

One thing I forgot to mention is that when shopping for servo drive replacements (copley, AMC, etc), patience is your friend. I see that some Copley 423's are for sale for > $400, however, if you wait you might get them less than $100 each, just depends. Also, if htey have a "make an offer," offer about $100... Also, look at the AMC drives, one issue is the 30A8V drives have too low of input voltage (only 80Vdc), so look at the AMC website/datasheet to see what servo amp has higher input voltage.

I second Andy's suggestions on the mesa expansion, you have some options. Don't forget a spindle encoder if you want to do rigid tapping.. And, if you're doing that, plan for that early, I'm learning that lesson now.

OT-CNC WROTE:
"I could not not find any documentation on the Anilam Crusader mechanical Bridgeport CNC conversion components. Does anyone have an original parts/assembly manual?"

I have a supermax machine w/ the anilam, not a bridgeport, so I wouldn't know the specifics. I did have a linear scale for the Z-axis, though. Not sure how your's is setup. Also, I have a bunch of manuals or/and data sheets for the supermax that I found by searching cnczone, linuxcnc forum, practical machinist, etc...


This was from the emc mailing list. You can search that, and probably should:

Re: Anilam A10 Encoder Pinouts
RogerN <regor@...>
2012-08-02 04:02:56 GMT
I'm guessing it's the same as my Crusader II:

Anilam DRO Scale Pinout of Bendix 10-6 connector

PIN OUTS of 1 axis
....A - Channel 1
....B - Channel 2
....C - +5 Volts
....D - Ground
....E - Shield Ground
....F - Marker pulse

Roger Neal

This was in the linuxcnc forum:
Ok, I have the encoders hooked up. I used the previously attached A10 pin-out...traced through the connectors and determined the original machine (plugged into PCB 504) used the following colors. I simply cut off the stock connectors, wired into screw terminals on TB3 of 7i77...and it works!

pin 1: A chan (white)
pin 2: NC
pin3: COM (black)
pin4: B chan (green)
pin5: NC
pin6: +5V (red)
pin7: marker pulse (brown)
pin8: NC

Good luck. If you have any specific questions, fire back, might be able to help out.
Mark

FloppyDisk wrote:

OT-CNC,

One thing I forgot to mention is that when shopping for servo drive replacements (copley, AMC, etc), patience is your friend. I see that some Copley 423's are for sale for > $400, however, if you wait you might get them less than $100 each, just depends.

Another option is to use the Pico Systems PWM servo amps and PWM controller. (They can also be used with Mesa controller boards that can output PWM, if you already have the Mesa board.)

Jon

Jon,

Thanks for the info.
I assume your boards don't read the tachometer?