Machsupport Forum
Mach Discussion => Mach4 General Discussion => Topic started by: kf2qd on June 03, 2023, 07:04:28 PM
-
Just getting started on the planning stages of a retrofit to a 1985 Bridgeport Boss. This version has servos instead of steppers and the spindle is 2HP with 5 inches of Z travel. The computer boots, but as soon as it is done with all its system checks and goes into the actual run time it gets a little confused. Not a real problem. I bought the machine at a very good price and the mecahical portion of the machine is almost pristine.
I have been using a Sieg KX3 CNC with Mach3 for several years now, have not wanted to spend the money on Mach4 as I was also trying to get some tooling for the lathe and the Hardinge UM horizontal mill. Mach 4 will require getting used to a little different appearance and some changes in the logic of operating the machine. It is not Mach3, so I know there is going to be some learning curve.
My plans are to go with a Ethernet Smooth Stepper as the machine interface. I have been using one, and it has behaved very well for me. How difficult would it be to configure Mach4 so I could use some 20mm industrial push buttons for my Start, Feed Hold, and Stop buttons? I will have a keyboard mounted up where the current keypad for the Boss is mounted, and I would like to get it so I don't need the mouse quite as much as I do with my current setup.
I am also looking at the DG4S-16035 High voltage servo drives and probably the BRKC-180 braking units. from CNC4PC. The current drives are rated 140VDC and I am not sure what the performance would be with a drive limited to 80VDC. And besides, I already have the higher voltage DC supply in my machine...
Any thoughts?
-
Hi,
How difficult would it be to configure Mach4 so I could use some 20mm industrial push buttons for my Start, Feed Hold, and Stop buttons?
That is dead easy. You say you are using an ESS but do not say what breakout board or boards you are using. You'll need one spare input for each button you wish to use.
Ten buttons, ten inputs.
I am also looking at the DG4S-16035 High voltage servo drives and probably the BRKC-180 braking units. from CNC4PC. The current drives are rated 140VDC and I am not sure what the performance would be with a drive limited to 80VDC. And besides, I already have the higher voltage DC supply in my machine...
You don't say what the servos are? The servos will usually have a recommend max voltage, and that's what you should aim for. If you supply a lesser voltage the servo will still
work but never get to it maximum speed.
As a rough guide, lets say your servos are rated to 2500rpm at 180V, and you have only a 80V supply then the servo will get to about 80/180 x 2500=1111rpm. The torque should remain the same
its just that the servo will not go as fast until or unless you provide a higher voltage.
Craig
-
Hi,
I use 750W Delta B2 series servos on my machine, and a number of retrofits that I've done.
https://www.fasttobuy.com/flange-80mm-239nm-ac-motor-driver-kits-with-3m-cable-220v-075kw-cnc-servo-motor-for-cnc-router_p28084.html (https://www.fasttobuy.com/flange-80mm-239nm-ac-motor-driver-kits-with-3m-cable-220v-075kw-cnc-servo-motor-for-cnc-router_p28084.html)
They have a 160,000 count/rev encoder, 2.4Nm(rated), 7.1Nm(overload),3000rpm(rated), 5000rpm(max). They are AC servos, so all the bells and whistles. Delta
provide free setup and tuning software, which is a must have if you've had little or no exposure to modern servos. They are $438USD per kit (servo, drive and cables).
They do not require a power supply, they hook direct to the 230VAC line, easy.
You might like to consider that replacing the entire servo and drive might be an easier way to go. Many people have tried to get old school DC servos to run using all manner of different
drives and often encounter either compatibility issues or tuning issues. Buying a new and matching servo and drive obviates those sorts of issues.
Before you spend big dollars on the drives and/or braking units you really need to answer the question ' what is the likelihood that this purchase will adequately drive these servos?'
If the answer is not 100% confident then there is a risk that you'll end up with a sub-standard result....and after spending good money would be extremely disappointing.
I do recognise that its likely more expensive to replace the servo and drive, but the certainty of the outcome is appealing. You may also find some other 'twat' who wants to buy
the old servos and defray the extra spent on AC servos, there is a 'twat' born every minute!
Craig
-
Hi,
those drivers you mentioned require the servo to have an encoder.
Have you checked the encoders on the servos? Are they in fact encoders at all? From 1985 they could be resolvers or even tacho generators. Resolvers
or tacho generators will not work with the drive you've mentioned. You need to check....what you don't want to do is get new drivers at $190 each
only to find that what you thought was an encoder is in fact a tacho generator.
Craig
-
The servo motors have an encoder on end of them. Don't know the number of counts at the moment. They should still be okay, but might want a few more counts per revolution.
-
Hi,
are you 100% sure that they are encoders? Resolvers and tacho generators look very similar but will not work as you might imagine.
Also around 1985 optical encoders were in their infancy, they were expensive and fragile. I think it wise to check on their operation before
you spend money on drivers. Remember these things are 38 years old, you need to check.
Craig
-
Yes they are encoders. BEI 250 line encoders. Might want to go for something with a slightly higher count. In 1985 they were pretty much the norm.
-
Hi,
kool. 250 line is 1000 count/rev, so at 3000 rpm the signal rate is 50kHz. The drives you listed are max input of 400kHz, so you could put
better encoders but you can't afford to go too far otherwise you'll not be able to signal fast enough. 2500 line or 10,000 count would be too much,
you could not signal the drives fast enough to get the servo to do 3000rpm, although you'd get away with, just, 2000 line.
I'd suggest a max of 200kHz single ended and 500kHz differential. The max signal rate to my servos is 416kHz, and so I had to use differential signalling, I tried
single ended but it started to crap out at 240kHz.
Still makes sense to te3st the encoders, if one or more are shot you really should know about it.
Craig
-
I am continuing to check things out and examining the wiring. Seems that the Drive Enable for this machine is to pull in the individual drive relays which makes the connection between the motor and the drive amp. Old transistor devices will allow that, the newer stuff with MOSFET and IGBT ouput stages would not tolerate that. I think my drives are not enabling because of the problems inside the computer section of the machine. been working on pulling the motors off so I can set them on the table for testing and not having to worry about what might run into an end stop.
Checking on the encoders - they are 250 counts per revolution, which because of the 2:1 ratio between the motor an the ball screw means I have, in a practical sense, 500 counts per revolution. As the ball screws on this machine are all 5mm pitch I am coming in at just about a thou per count.will work to start, but a 1000 count encoder seems like a better minimum count for this machine.
I have ordered some replacement capacitors for the DPS (Drive Power Supply) board. All the large electrolitic filter caps had blown at some point in the past. The power supply still seems to function ok, but the servos might behave better with cleaner DC power.
-
Hi,
Yes they are encoders. BEI 250 line encoders.
Checking on the encoders - they are 250 counts per revolution,
Two different quotes from two different posts. Which is correct?. There is a difference between lines and counts. One line results in four counts.
In the first quote you say they are 250 line or 1000 count/rev and in the next quote you say they are 250 count/rev. I can't help but think the last quote is wrong.
All optical encoders have four counts per line, so 250 counts is 62.5lines????? A Half a line????
There are encoders of 256 counts/rev or 64 lines.....but I'm dubious that's is what you've got......but I don't know.
Craig
-
Hi,
the part number of the BEI encoder should tell you the 'number of cycles' the optical disk it has been fitted with. If that number is 250, which seems highly likely,
then it is a 250 line/rev encoder and therefore a 1000 count/rev encoder. That, with the reduction you mention on 5mm screws would result in 2um
resolution, or 0.05 thousandths of an inch.
Just as a matter of clarification, have you measured the screws? Being Bridgeport I would have guessed 0.2 inch pitch screws or 5.08mm. They are so close to 5mm
that its easy to mistake one for the other.
My Delta servos have a 160,000 count per rev encoder (40,000 lines), but that would result in ridiculously fine resolution and be totally impractical to signal
at the rate required to get max speed. Using 'electronic gearing', and almost universal feature of AC servos, I have an effective encoder of 5000 count/rev. The servos
are direct coupled to 5mm pitch screws for a resolution of 1um or 0.025 thousandths of an inch, if you prefer. The point is that this resolution is as fine as I could
conceivably use, and in most cases finer than I can measure.
Your set up, if indeed you have 250 line encoders, would still be very fine resolution indeed. I do not think there is any need to swap them, the resolution as is is probably finer than
you can use anyway. That is certainly my experience. Additionally.....do you think it likely that when the machine was built they would use encoders that had less resolution than
is required? I rather think not. Bridgeport do not make those sorts of mistakes, then or now.
Craig
-
The first batch of Bridegport Boss machines came with Inch screws. Shortly thereafter, they switched to metric screws. So they went from .2 inch pitch screws to 5mm pitch screws. That along with 2:1 belt drives between the motors and the screws. So the system sees 500 encoder lines for one rotation of the screw. 500 lines for 5mm. Y axis rotates the screw, X & Z rotate the nut, with the screw on Z being about 2" + in diameter at the top of the spindle.
Got the caps for the servo low voltage power supply and found a blown trace, the one cap really blew out hot, and will be trying it out over the next few days.
-
Hi,
kool, that all sounds promising.
Craig
-
Well, Replaced the caps in the low voltage DC section of the drive amps and the next set of caps decided to blow. Some smallish tantalum caps made a couple of very nice bags and stunk the shop up. Looks like i will be needing to order parts to get this working. So I will have to get my KX3 sold to generate the cash for the next phase.
At least the High Voltage DC supply for the output of the drives is working. It is a couple half wave rectifiers and a string of large capacitors with nice copper bus bars down the top. And the necessary relays and contactors for the high voltage DC section all seem to be behaving.
So now to pull the computer section, 2 boards with Motorola 68000 processors and the related memories. Will have to put them away and see what I might be able to do someday getting a 68000 computer to play with... Doesn't seem t be much demand for those boards these days. Probably others are doing the same thing I am.
Now begins the hunt for some tool holders. Not a lot of CAT30 stuff out there. Might need to look at BT30 and ISO30 but they will require changing the drawbar from 1/2-13 thread to 12mm. Should not be overly difficult, or modifying NMTB30 holders by getting rid of the straight nub on the end. Have done that with a NMTB50 for the horizontal mill at work.