[TOTALLYRC]

Thanks for the kind words Todd.

I am not that technically inclined to answer your udp vs tcp vs serial questions.

I will say that since the "step generation" is done onboard the dsp and not the pc you can have small delays and still not miss a beat, as opposed to the pport.

All I know is that it works and works well.

My pc is a amd athlon xp 1ghz if I remeber correctly with 512 ram. Just a buch of stuff I was given and put together.

I am using a discrete video card because I had one lying around. I am running win2k on the machine control computer. I do recommend not having the computer hooked up to anything else thru the network port. I figure it this way, do you want to take the risk of something getting thru and ruining that 12hr machining job?????

I did all my development and testing on my laptop with win Vista and it works as well or better than my dedicated machine. Since the current pc won't power down by itself, I will be installing a different computer and see what else I can do.

I have spoken to several current and future customers for Rufi, but I don't have any of the deep info you are looking for.

One other thing that I forgot to add to the list of things to come is that in the future you will be able to have step and direction output from the dspmc/ip. The coding for this will be done once the curent version has the MPG inputs, Lathe threading and any other minor things worked out.


THESE ARE MY OPINIONS AND I AM NOT FLAMING ANYONE.

Not wanting to take anything away from the cnc brain as they look like the are working real hard to put out a nice product, but you asked for my opinion.

CNC Brain My thoughts

I took a good look at the brain before I purchased the DSPMC/IP.
The major setback was, that at the time I was looking, there was no provision for analog outputs to control the servo drives. I have followed the cnc brain thread and I would still not buy it today based on the following reasons.

1. Analog servo control is an addon and I would like to see more machines running them before I took the leap.

2. The major reason to run the brain is double closed loop and that is still very costly to implement, needs linear glass scales or something to measure the tables actual movement in addition to a motors encoder. Although I understand that they are working on a lower cost solution.

3. Costs are about the same when running in single closed loop.and if you go double closed loop, the costs go up even farther. Look real hard and they heavily suggest that you need optio isolation to protect the device. That $499 looks good at first but it soon climbs when you want to do it properly.
In my case I would need 5 DAC boards at ~$90 each, plus optio protection, plus the linear scales. And I still can't use mach3.
I couldn't find what the number of I/O is standard but they offer I/O boards at $89 each for 13 I/O.

4. Mach3 plugin. Apparently it is being woked on but you have to use their software to run the brain, I didn't need to learn another control program.

5. The dspmc/ip has 8 axis of control, six motion with encoders and 2 more for things like 0-10v spindle and anthing else that you need. I am not sure how the brain handles analog spindle. People will say why do you need six axis any way, but I using 3 right now, x y z and will be adding a(rotary), w(knee) and I just might use b or c for the tool changer, plus I am using 1 of the non encoder outputs to run the spindle.

6. I/O 36 inputs and 16 outputs. I may have to get creative when I hook in the tool carosel, but there are more than enough to run the mill with power drawbar and coolant and stuff.
Output from the breakout boards are 2 amps each!!! as long as you provide a big enough power supply. Since most things that you attach will draw much less current, you will size your power supply based on total current drawn.
Additional ways to add more I/O are in the works.

7. Do you really need double closed loop on a 3klb piece of cast iron that does what you want with good surface finish? On the size machine we are talking about, Todd, if the mechanicals aren't good enough for the job, we will improve them or repair them until they are. Even something as simple as backlash compensation, if the backlash was .010" or more, we would fix it before using it.


Enough rambiling, time to go outside and cut the new control panel face.

Mike

[T63-A-700]

Thanks for the post. This helps me and hopefully others understand some of the newer products that are being released.

This product is getting more and more impressive as I learn more about it. I like it because adhears to the KISS methodolgy.

Double closed loop sounds impressive and only help, but for my purposes its overkill. I like the fact they it has some analog outputs as well, I was almost convenced that I needed to trash my drives for a step/directional control device.

What will you control with your BOB? With my limited amount of inputs and outputs I doubt that I would opt for the additional I/O channels on a BOB?

Sorry I missed your call this morning, I was helping with the nursery at the church this morning. We had eight one year olds keeping the floor clean for us.

I will be available for the rest of the day.

[T63-A-700]

Can the Vital Systems DSPMC/IP operate in mixed mode, i.e. a combination of servo and stepper outputs.

My thinking would be a 4th axis rotary table driven by a stepper. This would allow for a cheaper drive to be used on the rotary table and the servos can be used where the real muscle is needed.

Just some thoughts.

Todd

[TOTALLYRC]

If you had 8 1 year olds cleaning the floor, I like that. It is never too early to put them to work.

Mixed mode will be available when the stepper coding is done. I told Rufi the same thing, that many people will want a stepper driven 4 axis. I might even go with a stepper for my tool carousel.


You will need one of the I/O boards to isolate the dsp (#7535 16 inputs and 8 outputs) and at least 1 on the encoder connector boards.

I was called in to do an emergency repair and then had to get some sleep.
I will pm you my email and we can arrange a good time for a call.


Mike

[tjs88yj]

Mike,

Any idea on when step/dir might be available?

Tom

[TOTALLYRC]

Mike,

Any idea on when step/dir might be available?

Tom

Last I spoke to Rufi about it, He said when the analog version was more fully inplemented. Which means threading, probing, and MPG integration.

I am hoping for June of 09, but I am sure it would be sooner if he had more requests. I can see these selling to quite a few people with them using all steppers and just 1 analog for spindle. Ethernet interface and 36 inputs/16 outputs, none of which will have to be used for stepper outputs as there is another port on the dsp setup for the future stepper outputs.

I will be installing an analog version on my lathe conversion when I decide on which lathe to convert.

[NZ_Jase]

Hi,

I've been reading these forums for a bit, but this is my first post.

A few months ago I emailed Rufi and made enquiries as to the suitability of using one of these MCs to replace the ageing and temperamental Fanuc 5T brains on my Mazak 2L lathe.

I had done a lot of searching and trawling forums and product websites for info to make sure that whatever I ended up with was worth the time spent, and was pleasantly surprised at the price vs features of the DSPMC/IP unit.

I have considerable IT experience, mostly implementing Microsoft server-side product. Wiring, electronics and simple controls are also no problem for me, as I have been exposed to high end motorsport for a number of years also.

I decided pretty early on that closing the control loop using cheap breakout boards and forever crossing fingers hoping that Windows doesn't hiccup at the wrong moment was no way to move forward from the issues I already have. I also want to keep the hardware that is in perfect working condition and has not presented an issue - inch ballscrews (even though I will be programming in mm), grunty DC servos, power supplies and analogue servo amps/controllers. The lathe also has a 3 speed gearbox that has an electronic over hydraulic gearchange and 8 turret toolchanger that I need to keep operational, plus the +-10V spingle controller that hasn't missed a beat.

Reading the posts below, it sounds like the analogue side of things is all up and going. Can you confirm that I can use a 2000ppr encoder on X, 2500ppr encoder on Z (already there and working), and a 6v/1000rpm tach on the spindle?

It looks like I'll have to use a fair few IO channels and code some logic for the changer and gearbox but I'm sure I can figure that out.

Basically, I want to replace the Fanuc control and only the Fanuc control. As far as I can find, the only thing that will do this in one box is the DSPMC/IP. Yes/No??

Only problem: no threading control...yet.

At this stage I think I'll have to shell out for a new Fanuc board to get the beast profitable (been making no parts for a bit now...), and will be very eager to order one of these units once threading is working.

Regards, Jase.

[TOTALLYRC]

Hi,

I've been reading these forums for a bit, but this is my first post.

A few months ago I emailed Rufi and made enquiries as to the suitability of using one of these MCs to replace the ageing and temperamental Fanuc 5T brains on my Mazak 2L lathe.

I had done a lot of searching and trawling forums and product websites for info to make sure that whatever I ended up with was worth the time spent, and was pleasantly surprised at the price vs features of the DSPMC/IP unit.

I have considerable IT experience, mostly implementing Microsoft server-side product. Wiring, electronics and simple controls are also no problem for me, as I have been exposed to high end motorsport for a number of years also.

I decided pretty early on that closing the control loop using cheap breakout boards and forever crossing fingers hoping that Windows doesn't hiccup at the wrong moment was no way to move forward from the issues I already have. I also want to keep the hardware that is in perfect working condition and has not presented an issue - inch ballscrews (even though I will be programming in mm), grunty DC servos, power supplies and analogue servo amps/controllers. The lathe also has a 3 speed gearbox that has an electronic over hydraulic gearchange and 8 turret toolchanger that I need to keep operational, plus the +-10V spingle controller that hasn't missed a beat.

Reading the posts below, it sounds like the analogue side of things is all up and going. Can you confirm that I can use a 2000ppr encoder on X, 2500ppr encoder on Z (already there and working), and a 6v/1000rpm tach on the spindle?

It looks like I'll have to use a fair few IO channels and code some logic for the changer and gearbox but I'm sure I can figure that out.

Basically, I want to replace the Fanuc control and only the Fanuc control. As far as I can find, the only thing that will do this in one box is the DSPMC/IP. Yes/No??

Only problem: no threading control...yet.

At this stage I think I'll have to shell out for a new Fanuc board to get the beast profitable (been making no parts for a bit now...), and will be very eager to order one of these units once threading is working.

Regards, Jase.

Hi Jase,
I too eagerly await the threading capabilities to be done.
It looks like I am goining to convert my Tsugami lathe to cnc as it has an 8 tool turret and a 5c spindle with collet closer. It will be able to hold much better tolerances than anything else that I have.

The 0-10v spindle and the +-10v servo drive both work very well. I am going to update my progress on the mill this weekend but I need to take more pictures. All I can say now is that I am very happy with the dspmc.

Since you have programing experience, writing the macro for tool change and gear box speed control should not be a big deal. There are plenty of I/O to work with and I have made a few more suggestions to Rufi on what things I would like to see in the future.

I am using 500 count encoders in quad (2000cpr) on all axis and I get 300ipm rapids (20,000steps/inch) without any problem. With the 2mhz encoder limit, It should go 600ipm but my motors won't go that fast. As long as your encoders are diferential output you should be OK. I am using single to differential converters and they work fine.

I am unsure if the spindle tach will work. I think that for threading you will need some sort of encoder or pulse from the spindle which would probably be used for rpm input as well.
Of course we could check with Rufi to be sure.

I am trying to get him to start his own thread for people to make suggestions but we will see.

I was test cutting today and it is nice to hit the e-stop and not loose position on the encoders. The encoders are read into mach3 from the dspmc box and you can move an axis manually without loosing position. Makes for an expensive dro., but you get what I am saying. If you needed to move an axis manually on a regular basis, you would just need to disable the drive and then re enable when you went back to cnc.

Mike

[TOTALLYRC]

I am about a week late in posting some updates.
As many of you know all too well, life gets in the way of what you want to do. Several emergency call to work and three inside of a day and a half have really wiped me out. This is also not to mention all the beautiful snow that is falling in CT this time of year. As I write this we are expecting 4-8" on top of the ten we got friday.

Enough of the weather, back to Mike with the CNC update.

I apologize if this is getting out of sequence but here is where I am today.
I was doing some test moves and some basic test cuts and I was not thrilled with the Z or Y motor response. The machine from the factory was spec'ed at 200 ipm on the z and 300 ipm on the x and y. I can only go 175ipm on the X, 150 ipm on the Z, and only 75ipm on the Y. Keep in mind that these are very possibly 25 year old servos that have rebuild stickers on them.

Since I shop ebay when I am bored to look for CNC bargains, I found some BRAND NEW, IN THE FACTORY SEALED BOX,SEM SERVO MOTORS. With encoders?? for $30 each. When they arrived, I opened one up and there wasn't any encoders inside. I contacted the gentleman and he opened up his last one and sure enough his didn't have any either. He graciously refunded some money to pay for my troubles. I have 4 of them.

I also found some Renco encoders on Ebay Item #200084103247 for $20 each.

I grabbed 4.

Now 75 ipm isn't that slow, but I didn't like the way that the Y axis ran and the servo amp tuning was very different from the Z. I am running the Y and Z in torque mode, which means no tach feedback. The X is currently in velocity mode using tach feedback.

With the new motors, they will all be using Torque mode. as it works very well and the new motors don't have tachs.

The SEM motors have 1/2" shafts on both ends, and the encoders are for 1/4" shafts. I chucked up the motor, full assembled into the Atlas lathe, used the tailstock and started turning as seen here.

This is the early stages. I am taking shallow passes and I was leaving plenty of stock for the live center to hold on to.

http://i401.photobucket.com/albums/p...c/0420568a.jpg

More turning progress

http://i401.photobucket.com/albums/p...c/96f1cc0a.jpg

http://i401.photobucket.com/albums/p...c/7ce145d9.jpg

After sanding the shaft multiple times, here is the final result.


http://i401.photobucket.com/albums/p...c/97edabf2.jpg



Next I machined up a plate to hold the new Renco encoders based on the drawings from their website and the measurements that I took from the motors. I also added the 4 holes to hold the differential converters that I had made previously.

http://i401.photobucket.com/albums/p...c/ab3cd4e8.jpg

Notice that the plate is held down by the three perimeter screws. The screws allow the plate to be rotated to any position, along with the slotted mounting holes of the encoder itself, which is very useful when using index pulse encoders.

Since Rufi has added index pulse homing to the dspmc plugin, I will be using this feature in the future as I believe that the Renco's have index pulses and it will allow for very precise homing of an axis. This will be much better than the current switches for homing the Z as it needs to be in a very specific spot, to allow the power drawbar to punch out the tool.
If I am correct the index pulse homeing will be accurate to +- 1 encoder count, which in my case would be +- 0.00005".


Testing the new motor and encoder combination yielded the following results.


:banana: 300+ ipm on the Y and Z, :banana:
I then retuned the X servo amp and have them all set at 250 for the moment.
More than enough Rapid speed while I am learning how to run the machine.

Mike

P.S. If you can't find your camera, ask your children to check their coat pockets. LOL.

[turbostang]

Great Project!

Most of what you've said electronic-wise has gone over my head, but I can relate to the pictures! LOL!

Anyways - I am soon to be redoing my machine as well. It's a J head BP with Centroid controls. The mobo has recently checked out, so it's time to use the servos and what not to convert it over to MACH3.

Let's see some chips!

[TOTALLYRC]

Great Project!

Most of what you've said electronic-wise has gone over my head, but I can relate to the pictures! LOL!

Anyways - I am soon to be redoing my machine as well. It's a J head BP with Centroid controls. The mobo has recently checked out, so it's time to use the servos and what not to convert it over to MACH3.

Let's see some chips!

Due to the passing of my Mother, I have not been able to get much work done. She went into the hospital on New Years eve and passed away Wed night (Jan 14th)without regaining conciousness.

I have made a few test cuts and I took some more pics. I will get started again next week if all goes smoothly.

Thanks for the encouragement.

Mike

[turbostang]

I am sorry to hear about your mother, RIP ma.

[abdulrafiq]

Basically, I want to replace the Fanuc control and only the Fanuc control. As far as I can find, the only thing that will do this in one box is the DSPMC/IP. Yes/No??

Only problem: no threading control...yet.
...

Regards, Jase.

Yes, you can replace the fanuc controls with DSPMC. If your system is using resolvers, you will have to replace them with differential quadrature encoders.

we have just completed the Threading feature programming with Mach3/DSPMC. Mike is going to do some testing with his Lathe as time permits. I have made sure the Z axis feedrate is synchronized with spindle RPM and started with the index marker using G32 and other threading G-codes. so as far as I am concerned, threading is ready to go. pls feel free to contact Mike or myself at rufi _at_ vitalsystem.com for more information.

thank you
Rufi

[TOTALLYRC]

After taking care of Mom's final needs and getting life back on a normal track, I have had a chance to do more work on the mill.

I was using a 2?? gal cooling tank and pump from Harbour freight, but just doing some test cuts, I was able to put enough coolant on the table that the sump ran low enough to interupt the coolant flow.

I didn't want to use it because of the floor space it takes up, but it works reall well now that I have installed it. It is the cooling tank/catch pan that was given to me with the machine when I picked it up. It even has a red bridgeport decal on it, I don't know if it is a bridgeport product or what, but it does fit the base of the machine nicely.

As you can see in the attached photo's, I still have to neaten up the wiring, but the project is coming along nicely.

The pump is the pump from the harbour freight unit, it just drops in where the original pump would have gone. It is enough flow for starters, but in the future I will upgrade to a pump with more flow/pressure, as this one is a little weak.

Next up the new control panel.

Mike

[TOTALLYRC]

I went into the scrap pile and found a large piece of aluminum that was big enough to make a control panel including housing the monitor.
I also had a nice bunch of Allen Bradely push button switches.

I wanted to make the panel much larger than the final size I settled on, but the machine will only go 18"x12" and I didn't want to do a multiple setup for my first real machine work on this machine.

As you can see in the pictures, I was able to put in 10 pushbuttons, and the computer monitor. I will be adding a lexan face to keep coolant out of the screen. Since I have a really large pile of 4-40 screws, I used those to bolt the whole thing together. I plan on using a thin aluminum cover to go around the perimeter to seal the whole thing up, plus having holes for ventilation, and possibly a small cooling fan to keep the monitor cool.

Ignore the fact that it looks like it isn't flat, It is an optical illusion caused by how close I was with the camera.

I also want to put an e-stop button on the right side, just in case I am over there and need to stop it in a hurry, but I will wait and see if I need to add anything else and put in in an auxillary contol box.

While the perfectionists out there will notice that the monitor isn't perfectly centered in the opening, and there are a few notches along the top of the monitor opening, I am very pleased with the result, and will probably dream up a really overdone one in the future as time allows, but at least the machine now has all the wires inside the control panel and not rubbing against the z axis belt or exposed to the coolant spray.
Of course if I am happy with it in its current form, I will just weld up the notches and refinish the opening. If that becomes the case, I will inset the lexan cover from the inside as the panel is thick enough to allow for it.
I will also round over or bevel the monitor opening. So many improvements come to mind, once you are done.

As I am sure that many of you can tell, I spent hours with my custom sharpie marker labeling the switches. LOL.

:wave:I am still open to suggestions :wave:as to what functions should be assigned to the other switches. I am considering coolant, spindle, and tool index, when the tool changer comes online, but I am always open to ideas.

In the coming weeks, since Rufi was busy and got the MPG code written, I will be working on installing them. While the gamepad that I am currently using works well, I like the option of using an MPG.

Mike

[TOTALLYRC]

I was going crazy trying to figure out where to put the keyboard and mouse.
I had tried several places and alway kept dropping one or the other, or getting them wet with coolant, so I came up with this arangement to see if it works.

http://i401.photobucket.com/albums/p...c/SDC10996.jpg

And this.

http://i401.photobucket.com/albums/p...c/SDC10997.jpg

The wonders of double sided tape never cease to amaze me. There is a strip of it on the lower edge of the keyboard amd on the back where there is a bracket on the machine. The back of the mouse is covered in it and they are both staying well.

I am quite pleased, and since I don't do a lot of input at the machine, it is working out quite well.

I was playing around with task lighting and I think I saw this on the web somewhere, and since I had several cheap LED flashlights, not to mention a bunch of lineloc stuff, I thought I would give it a try.

http://i401.photobucket.com/albums/p...c/SDC10998.jpg

I just drilled a hole in the end of the flashlight and the found a washer that wouldn't go all the way thru the line loc and voila, a flashlight mount.

It gives a nice bright spot of light right where you need it. I will be adding a power supply with the wires run thru the line lock tubing, but for now the batteries last quite a while.

[TOTALLYRC]

I needed to make a bracket for my cnc router to hold the new motor I wanted to install, so I took the time to video part of it, I actually remembered part way thru the second time I drilled the holes, because I used the wrong drill the first time.

The video is approx. 14 MB so it may take ahile to load.

http://s401.photobucket.com/albums/p...t=SDC10995.flv

I will be putting up some more video once I get a tripod to put the camera on, as it is tough to hold the camera and keep track of the machine.

Here are a couple of pics of the mount with the NEMA 23, 425oz motor test fitted. Sorry for the blur.

http://i401.photobucket.com/albums/p...c/SDC11000.jpg

http://i401.photobucket.com/albums/p...c/SDC11001.jpg

http://i401.photobucket.com/albums/p...c/SDC10999.jpg

The hole to allow for the centering ring on the motor is exactly 1.5" in diameter, and on the first pass it was about .003-.004" too small. I went into the code and tweaked the final pass numbers and it came out at 1.501. The motor now has too be wiggled ever so slightly to go into place and the bolts go right in without any problems. I need to tweak the steps per inch ever so slightly as distance traveled is off just a tad, and I found my magnetic mount dial indicator to test the steps per inch.
The hole is slightly flat spotted at the 4 points due to the machine backlash. The backlash for the plugin is being rewritten as we speak, so I am not using it, but it is cutting fine.

All in all, I am very pleased with the operation of the machine and the DSPMC control box. I will be implementing the following feature in the upcoming weeks and months.

1. MPGs.
2. Index pulse homing, at least on the Z, as it is crucial for the tool changing that the Z be in exactly the right place for the drawbar to grip the tools properly.
3. Rigid tapping. Now that Rufi has the lathe threading code written, rigid tapping on the mill is on the way and I do a lot of parts that need to be tapped.
4. Integration of the auto tool changer, which is the major reason I went with this particular mill in the first place.
5.4th axis rotary and 5th axis for the Knee. As one of the other guys mentioned, it is a great way to implement tool ofsets for the Z axis.
6.Auto tool length detection.

[TOTALLYRC]

I was cutting some more small parts on the machine today, trying to put the router back together, when all of a sudden, the machine faulted.
On the display it says that the X axis has exceeded the following error (Thanks Rufi for having it tell us "which" axis has faulted).

I was about to check the encoder and everything else, when I decided to open the control cabinet and I saw a little red led that is not normally on. The X axis drive had faulted, and the back plate of the drive was almost to hot to touch.

Those of you who have read this whole thing might remember this picture

http://i401.photobucket.com/albums/p...c/24cc92d0.jpg

In my rush to keep working with the mill, I forgot to turn on the high performance, specially made, drive cooler!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!:withstupi

Well, since running the machine with the drive cabinet open is not the smartest idea anyway, I decided to fix the problem more permanently.

In my box of fans, I had a 220v fan that was just what the doctor ordered.

The fan is located so that it blows directly on the servo drives.

I first cut the hole and drilled the mounting holes

http://i401.photobucket.com/albums/p...c/SDC11002.jpg
and
http://i401.photobucket.com/albums/p...c/SDC11003.jpg

And the final product.


http://i401.photobucket.com/albums/p...c/SDC11004.jpg
and
http://i401.photobucket.com/albums/p...c/SDC11005.jpg

all that was left was to run some power over to the fan, I used a nice peice of extension cord so it would flex with the door opening and closing.

When I turned the fan on for the first time, the airflow out of the small openings in the cabinet was impressive. I will decide on where to cut an exhaust hole later, as it is working well enough for me to finish the rest of the router parts.

I will be adding a smalll filter to keep swarf and other undesireable stuff out as well.

Mike

[TOTALLYRC]

I was doing some small cuts yesterday and all of a sudden the z came crashing down and snapped a 1/8" carbide cutter. Well was I pissed because I thought I was done with this crap.
Turns out that I never upgraded the z axis to the new renco encoder and the heds went bad. Swapped out the encoders and it is up and running again.

More pics and videos to come later this week.

Mike

[TOTALLYRC]

As I type the machine is making parts, a small lot of 12 parts in three different setups.

I need to make a trunion setup if I keep doing parts like this.

I would post a video, but the guy I am doing this for said no, not at this time.
I am going to do some brackets for the lathe, and i will post those video's when I do the cutting.

I lucked into a bunch of tooling that fits the spindle and the tool carousel.
Included are some floating tap heads and asorted others so I will try to do some "semi rigid tapping" untill Rufi gets the rigid tapping code done.

All of the other stuff is still on the list of things to do.

As always, if you have questions, feel free to ask.

Mike