Another bridgeport retrofit, need advise

[sohcpwr]

what do you think of this contactor? It has a built in amplifier on the signal side so low power (down to 9mA) 24v signals can activate the coil. So I'm gathering I will need 2 in total, one for the quill drive so it can be disabled separately and one for all the others. Will 50a be enough for x,y,knee, and A axis? The DMM manual says its recommended to use a 50A contactor for each drive!? Yet it only drawls 8 amps on the AC input... I guess when you do a full power direction change they can really pull some power for a fraction of a second.

https://www.ebay.com/itm/Telemecaniq...19.m1438.l2649

I was planning on tapping off the receptacle for my welder thats right behind the machine. Its the biggest branch circuit in my garage at 50A with 8ga copper wire going to it. Beginning to think this isn't going to be enough. Hell you'd need my entire 200A service cable to follow the recommended breaker size for 5 drives

[sohcpwr]

A planetary system would work, but how do you disengage it for manual use?

Originally I planned on a gearbox type sliding gear with a shift fork type of deal to engage and disengage the drive coupling. But for a gear to slide on the splines of a shaft, I will likely end up with at least a small amount of lost motion. Newest idea hit me from thinking back to some of the magnetic drive pumps I used to work on at the refinery. If I could take a servo brake of considerable power and basically modify the way it works a bit, it would still free wheel when energized but with it would drive the coupling mounted on the other end when disabled. Still not sure if this is even possible cause I don't think both halves are meant to turn, but it sounds good.

[Jim Dawson]

Can you elaborate on the circuit of contactors for the drives? And other than the aforementioned contactors, is there anything else you see I'm missing? I'd like to get everything mounted this weekend but I'm hesitant to do so until I know I have all the devices in there

Drawing attached This one is showing 5 volt control power where yours would be 24 volt. But the wiring is basically the same. The E-stop shuts power down to all output devices. The control power button enables the system. Your wiring will be a bit different, but functionally the same. Rather than the control power button, yours will be wired to a Galil input and an output will power up the contactor coils and a relay for other devices. In either case, the E-stop button will kill control power outside of the Galil. You never rely on a computer to shut down power.

what do you think of this contactor? It has a built in amplifier on the signal side so low power (down to 9mA) 24v signals can activate the coil. So I'm gathering I will need 2 in total, one for the quill drive so it can be disabled separately and one for all the others. Will 50a be enough for x,y,knee, and A axis? The DMM manual says its recommended to use a 50A contactor for each drive!? Yet it only drawls 8 amps on the AC input... I guess when you do a full power direction change they can really pull some power for a fraction of a second.

https://www.ebay.com/itm/Telemecaniq...19.m1438.l2649

Never seen one of those before, nice to know they exist. Probably overkill for this application.

You are never going to do a full power direction change, you would break something.

I was planning on tapping off the receptacle for my welder thats right behind the machine. Its the biggest branch circuit in my garage at 50A with 8ga copper wire going to it. Beginning to think this isn't going to be enough. Hell you'd need my entire 200A service cable to follow the recommended breaker size for 5 drives

The largest power draw is the inrush from just turning the drives on, but that only lasts a fraction of a second. The current draw is huge, but not a problem in low energy systems. Just the resistance of the wiring limits the inrush current. In high energy industrial power systems capable of delivering hundreds or thousands of amps to a system you have to think about inrush current limiting.

In normal operation the drives will draw a fraction of their rated input power. I'm running 2 servo drives and 3 steppers, plus the rest of my controls, all plugged into one 115V, 20 Amp plug on the wall. The spindle VFD is separately wired to a 230V feed.

Originally I planned on a gearbox type sliding gear with a shift fork type of deal to engage and disengage the drive coupling. But for a gear to slide on the splines of a shaft, I will likely end up with at least a small amount of lost motion. Newest idea hit me from thinking back to some of the magnetic drive pumps I used to work on at the refinery. If I could take a servo brake of considerable power and basically modify the way it works a bit, it would still free wheel when energized but with it would drive the coupling mounted on the other end when disabled. Still not sure if this is even possible cause I don't think both halves are meant to turn, but it sounds good.

You could use an electromagnetic clutch, maybe like an automotive air conditioner clutch. I think those are good for about 10 HP or so. Industrial electromagnetic clutches are crazy expensive, but you should be able to pick up an automotive one pretty cheap. Magnetic pump drives are pretty mushy, would not work well when you need positive coupling in both directions.

Backlash compensation is accomplished with an air spring to bias the quill upwards. That way the whole system is preloaded and the servo is working against the air spring, 100% backlash free. Without the air spring, my quill system has about 0.020'' backlash, but I am able to consistently hold 0.0001'' with the system.

You mentioned above about powering the knee to use as the Z axis for making the quill axis parts. Not really needed, we have been making acceptable parts with a manual knee for over 100 years. You just need to plan your stopping position so the tool doesn't chew through the side of the part or something at the end of the cut Makes the G-code simpler too.

[sohcpwr]

Started mounting and wiring stuff today, have a few more questions for you experienced folk:

I am going to use 2 push on push off switches on the door to enable the servo drives and a seperate one to enable the quill for manual use. When I went to order them today, they come with N.O. or N.C. contacts. I don't understand how it would matter what the contacts are for a push button switch?? Is it not in the same position either way so I would have thought they would be interchangeable. Have a feeling since these will be tied into the e-stop circuit that I should use N.C. though.

Does the AC input and the DC output of the 24V power supplies need to each have a fuse? I just ordered 2 that seem to have an overload feature built into them. Not sure if thats on the input or output but would again assume the DC out.

Finally, I initially thought I would just use 220v for everything but still curious about the proper way to have 120v in the cabinet as well. Obviously I believe the truly correct way is to use a transformer correct? But if the load is really small, could you just use one leg and ground for 120v? Was looking at that last night and saw that if you do that I guess you end up with an unbalanced load on the 2 power legs. I'm really only trying to do this for the controller power supply. The controller says it uses 12-16W which at 24v equates to drawing .6amps? Is that enough to upset the main power supply? Thanks!

[Jim Dawson]

You want momentary contact switches, not push on/push off. The momentary switches come with one NO (green) contact block. The servos are shut off with the E-stop switch, comes with one NC (red) contact block. The E-stop switch should be twist to release. The quill on/off is done on the screen. No switch required.

Servo Enable Switch
https://www.automationdirect.com/adc...tended/GCX1103

E-stop Switch
https://www.automationdirect.com/adc...inated/GCX1131

To really do it right, both the input and output of the 24V supplies should have fuses. The outputs are current limited on some power supplies. What is the input on the power supplies? Most modern ones will accept 120V or 240V

The proper way to have 120V in the cabinet would be to have a transformer. Never use the ground as one leg of a 120V, but you could bring in a neutral. The ground and neutral are electrically the same, but should only be connected together at the breaker panel or at the utility pole depending on local codes. But I think you already have a small transformer. All of the switches will be running on 24V, I don't think there is any need for 120V in the cabinet. The power supplies won't affect the balance at all.

[sohcpwr]

The power supplies do all accept up to 240v. However if I choose to fuse the ac input to each one it will require 2x the amount of fuseholders like that.

I do not have a transformer, was plannning on running everything off 230v.

I was planning on having a push on push off 120v switch (can use 2 contact blocks for 230v?) wired off the input of the main power on off switch. This would go to the 24v power supply for the Galil and allow me to keep it on for days on end with the rest of the box shut off. Since the galil powers the encoders, I figured this would be a good way to guarantee not loosing position of any of the axis

[Jim Dawson]

The push on push off switch would be a great way of powering up the Galil, and would maintain the position data. I pretty much never shut my controls/computer down. You could switch the 24V side to the Galil, and power up the power supply off of the mains. Since you don't have a transformer, you could bring in a neutral to get the 120V.

[sohcpwr]

That's how I thought I would do it at first, only reason I suggested the above is what I read in the manual the other day when planning out the circuit:

"The DMC-40x0 power should never be plugged in HOT. Always power down the power supply before installing or
removing power connector(s) to/from controller.

NOTE: Any emergency stop or disconnect switches should be installed on the AC input to the DC power supply.
Relays and/or other switches should not be installed on the DC line between the Galil and the Power supply. An
example system is shown in Figure 2.7 with a DMC-4080-C012-I000-I000-D3040-D3040:
The green power light indicator should go on when power is applied."

All of my power supplies accept 230V input so to save the hassle of pulling in neutral terminal as well, I think I will run everything off this voltage.

[Jim Dawson]

OK, fair enough. We can connect the 24V side directly to the Galil, and switch the AC side of the PS. The E stop won't be connected into that circuit anyway, we want the Galil to stay hot all the time, but we will break the 24V power to the outputs on E stop. I have been doing that for years with no issues.

Correct me if I'm wrong here, the 240V will power the servo drives, the spindle, and the two 24V power supplies. What else do you have in mind that needs 240V power?

[sohcpwr]

So the majority of the box is wired up except for the control to the knee servo drive and 1 leg of 10ga power wire to the drives that I ran out of. I think after that all I am missing is the linear encoders, a couple fans for the enclosure, and some random odds and ends. Found a place to mount the box behind the machine on the wall instead of on the side, the height and width of it was going to make it too difficult due to the slope of the column at the ends.

Back to the encoders, I found some very well priced Balluff magnetic read heads on Ebay. They are 5um resolution but reading through their documentation it says they are for 5mm pole spaced tape, which i cannot find. Will the 2mm pole spacing of the Renishaw tape still work, and actually work better, or willl it not work at all because the sensor's electronics are meant for 5mm? I would like to think its the same thing as having a higher density optical disk with more pulses per rev/inch and actually work better but not really sure.

Progress so far:












Gonna start on the motor mounts this weekend and hopefully make the final push twords the finish pretty soon. I'm super excited, I was able to score autodesk's HSM cam program for Inventor that is on my work computer. Little bit of a learning curve from Mastercam X that I used in college but I'm starting to get the hang of it. Here's a screenshot of the timing cover for my 1/4 scale engine I'm working on:



Found out after I programmed this toolpath that adaptive clearing is really meant to be used in a different manner, cutting at close to full depth but with a light step over. Gotta re-do it still but on the back burning until the machine is running.

[Jim Dawson]

You can't possibly be done with the panel, there is still a little bit of room between the drives. Looks good, and good use of panel realestate.

Using 2mm tape with 5mm read heads???? If it works at all it should be higher resolution. But I think the tape required would depend on the spacing of the sensors in the read head. It might screw up the phasing of the sine waves. Is there a link to those, the ones I saw were a bit pricy.

[sohcpwr]

You can't possibly be done with the panel, there is still a little bit of room between the drives. Looks good, and good use of panel realestate.

Thanks for the kind words! Once I had a better idea of the wiring I had to find a way to keep the controller off the door, I was dreading running that many up and around like you did. It came out a little more cramped than I would have liked, but it all fits. I really wanted another row of cable track in between the fuse blocks and the controller but had to sacrifice that to fit everything. The space between the servo drives is for.... another servo drive. Haven't decided if I am going to use the other electrocraft one I have for the quill or if I'll end up going with something else by then. Cross that bridge when I get there. Planning on using one of the controllers internal drives for the "A" axis.


As far as the encoders go, that's what I was thinking at first as well. But then I saw the ebay auction for the tape and it says the 2mm pole stuff works with read heads up to 25um resolution? This is with the same Renishaw brand ones though, not sure if that makes a difference or not.

Here is the literature for the Balluff heads I found on Ebay, they are S1B's. There is a different manual for a few different kinds of them, havent looked to see what the difference is yet.

http://usa.balluff.com/manuals/BML%2...0MANUAL%20.pdf


List price is quite expensive, and would def get the Renishaw ones before paying that... but ebay seller has more than 10 of these S1B sensors available for $30 each. I'll do a bunch of digging if I can end up getting a quality encoder for that cheap.

[Jim Dawson]

Did I give you the Ditron read head information.?

I'm pretty sure the tape pole spacing needs to match the read head. You should be able to get the 5mm tape from Balluff, I think they have some US distributors.

[sohcpwr]

you did, I gotta go back and find it though. Dont remember if it was here, through PM, or email.

I think your right about the tape, when I looked through the other manuals it lists the different pole spacing for different read heads. There are 2 different specs for the "accuracy" though for those Balluff sensors: The resolution is 5um but it says the overall system accuracy is +/- 50um. Thats almost .002" each way, dont think Renishaw documentation lists this spec but I think you said your 5um heads were accurate to .0001" as a system? If that system accuracy statement is true it sounds like that's not really what I'm looking for. Getting ready to give in and go with the ditron ones

[Jim Dawson]

I have 1um read heads and yes, they are accurate to < 0.0001 inch. I can normally hit +/- 1um or better. That is just going by encoder counts, I have absolutely no way of actually measuring anything that close in my shop. The one test we did that was checked by the metrology lab at an aerospace manufacturer indicated my readings are accurate to 1um. A sample size of one item is hardly conclusive, but the result was promising. I have found that the parts I make are on size to < +/- 0.0005, I can't reliably measure much closer than that, nor for most work do I need to.

I have no idea what they mean by +/- 50um, over what distance? That spec doesn't make any sense.

[sohcpwr]

Yea it really does make no sense, why have the ability to read to 5um when the obtainable accuracy is literally over 10x that!!

copied and pasted from the manual:

9 Technical Data
Electrical data Type BML-S1B0-Q
Output digital RS422 or supply voltage level (HTL)
Output signal A-Signal, B-Signal, reference signal
Reference signal no, one, periodic signal
Resolution 5 µm, 10 µm, 25 µm, 50 µm, 100 µm, 200 µm, 500 µm, 1000 µm, 2000 µm
Output voltage Differential signal per RS422 or same as operating voltage (HTL)
Limit switch Umax = 28 V, Imax = 20 mA, N.C., GND switching (cable break monitor)
System accuracy ±50 µm up to 1 mm distance, above that ±60 µm
Hysteresis depends on air gap 3 to 7 µm
max. non-linearity of the ±20 µm
processing electronics unidirectional
max. non-linearity of the ±50 µm within any one meter at a distance range up to 1 mm
overall system (sensing head + tape) ±60 µm within any one meter at a distance range 1 ... 2 mm
Temperature coefficient 10.5 x 10-6 K-1
of overall system like steel
Max. traverse speed depends on model, see table p. 10
Reverse polarity protected no
Overvoltage protected no
Supply voltage 5 V ±5% or 10...30 V
Current draw at 5 V operating voltage <50 mA + current draw of controller (depending on internal resistance)
Current draw <40 mA + current draw of controller (depending on internal resistance)

Looking at a couple RSF brand encoders that come with the mag strip already installed in steel plates that I would need to make anyway for Y and Knee axis. They also seem to have limit and home switches built in to the strip with adjustable magnets or something. Pretty cool, save some wiring and make for a clean install.

[Jim Dawson]

I am having a little trouble translating exactly that that says. Does that mean that you could have +/-50um in 1 mm? That's a lot of error. +/- 50um in 1 meter is a lot but might be acceptable. I'm a bit confused here. I suggest that there is a misprint in the specs and where it says 1mm it should read 1M

Here is the Renishaw spec. Precision class for MS scales ±20 μm/m and ±40 μm/m

[sohcpwr]

I was confused to but after reading through the rest of the manual and pondering a bit, I think its referring to the ride height of the sensor over the tape. If I can find the tape cheap enough I may pick a couple up and give it a try. I can always turn it into a DRO for my lathe if it doesn't work out

[sohcpwr]

SCORE!! I knew my persistence would pay off eventually... for some reason after I received the quote from Ditron for all 4 encoders and tape at $600, I decided to hold off because it was quite a bit more than I expected for china parts. Low and behold on my weekly Ebay search, a set of 4 renishaw LM10's came up in 1um resolution and with the reference option. I was hoping to get a better deal, but ended up paying $170 each for them. But in the end, an xtra $150 or so for Renishaw encoders over the Ditron's seemed hella worth it to me. Next question is: can the stick on reference marks, which use the index pulse of the encoder, be used in place of a home switch?

[Jim Dawson]

[QUOTE=sohcpwr;2135954]SCORE!! I knew my persistence would pay off eventually... for some reason after I received the quote from Ditron for all 4 encoders and tape at $600, I decided to hold off because it was quite a bit more than I expected for china parts. Low and behold on my weekly Ebay search, a set of 4 renishaw LM10's came up in 1um resolution and with the reference option. I was hoping to get a better deal, but ended up paying $170 each for them. But in the end, an xtra $150 or so for Renishaw encoders over the Ditron's seemed hella worth it to me. Next question is: can the stick on reference marks, which use the index pulse of the encoder, be used in place of a home switch?

/QUOTE]

Yes, that might work. Connect to the HM input, I think. I need to think about that a bit. The normal way of using this is to hit an actual home switch, then move to the index mark to set the actual home position.

Having said that, I see a home switch on the Z as being useful if you are using a tool changer, but a home on X & Y is useless IMHO. Even the Z home is not that useful, because you can set zero anywhere you want it and use that for the reference position. I don't have any home switches on my mill.