Marty-
I was about an inch away from totally giving up on this. I just wasn't getting anywhere, in fact after my initial somewhat successful attempt I was unable to duplicate it and wasn't getting anything at all. Today I went at it again and realized my error--once corrected it works perfectly. Not even an erratic signal, perfectly stable both forward and reverse.
Next comes working on getting the additional I/O from an auxiliary parallel port, then cut the new circuit board. It's coming together, slowly. Thanks.
Dan
Thanks Dan,
Probably have to do the same and replace everything. Did get the brake working so next is install the Bardac board and test the spindle.
Could you summarize the boards and software you are using. I believe Marty is using the centroid - Acorn setup.
Thanks
Don
Don--
I am using a Leadshine MX3660 for axis motor drivers; I've been pretty happy with it, seems to be reliable and has a small assortment of I/O's available. If your needs are modest it would probably suit you well for this machine. I've also replaced the original steppers and belt drive for a direct drive system. I tried the original steppers for a while but couldn't reliably get more than 72 IPM with that setup, so I made some adapter plates and replaced the motors. I can now get 120 IPM reliably; I've had recent days of using it all day long and have never experienced any lost steps. As for software, I use LinuxCNC. I'm not a Linux expert by any stretch. I built my own CNC router a few years back (my first foray into CNC building after programming and operating for several years) and went with LinuxCNC for the simple reason that it was cheap (open source, no cost) and my son was and is a Linux expert (some would say snob, but, he is my son :-)). In its favor it has a large user community and knowledge base, so if DIY CNC control is your thing I would recommend it. I don't have any experience with any others so can't recommend anything else.
Dan
- - - Updated - - -
Don-
Where did you find your machine?
Thanks Dan.
Machine was abandoned outside a local machine shop for years so I asked and got it. Fortunately as it was going to the metal recycling dumpster the next few days.
No luck yet with the Bardac spindle driver.
Looks like the MX4660 is better than the MX3660 with regard to the axis rapid travel. See:
Cheers
Don
Don, how about a few pictures of the prize?
Marty
Have the spindle working now with the Bardac 3200i board. Reversed the original mounting brackets for the servo drive, added an al backing plate to mount the board, a type D GE 6 amp circuit breaker, a switch and a potentiometer. The board needs up to 10v on termainal 66 and 68 which I supplied via the pot and a 9v battery. This controls the spindle speed for testing. However a switch is still required to activate, placed between terminals Run and Com. I wired the brake directly to 110v. The Bardac board has 240vac in and 180v dc out which goes through the original iso transformer bottom right of cabinet and also still drives the original amp meter on front of machine.
Last edited by Meri; 03-18-2018 at 01:15 PM. Reason: Added images
Looks like the important stuff is there. Good luck with it!
ATC is driven by a stepper motor, leave it and reuse it. Many get the machine going but few get the ATC going again.
If you can come up with an air cylinder for the power drawbar with sensors, and an air cylinder to swing the ATC in and out, again with sensors you would have it licked. There are two sensors on the spindle,one spindle orient and the other is an RPM sensor I believe.
This machine could do rigid tapping with the skip control.
Best of luck to you and Dan with your projects.
Marty
Dang, I haven't been getting any notifications from this thread since my last post a month ago, only logged on again because I had another question about the inverter circuit. I have a little catching up to do....
Don--I may be able to help with the presser bar circuit, I made some mods with a single button in place of the original two buttons on the head for the tool change. If I get some time soon I'll post some pix and diagrams. Mine's a bit simpler than Timothy's but it also involved making a new cam plate for the presser and two limit switches. When I first set up the machine I had removed the presser motor system completely and replaced it with an air cylinder for manual tool changing. After some time I realized that the Belleville washers for the drawbar were worn out and needed replacing (many of them were broken). After replacing them the force required to operate it was beyond the air cylinder's ability, so I went back to the original motor and made the mods to it. It wouldn't work if you plan to get the ATC working though.
Marty--you asked about the problem I had that I corrected--after a month I've almost forgotten, but it was just simply not connecting the -12V correctly. But I do have another issue I hope you may be able to help with (the reason I logged on today). I just now finally got the circuit board completed and temporarily in place when I ran into a funny issue with it that hadn't shown up before because I hadn't run the speed up fast enough for it to show. Last night I started testing by slowly increasing the spindle speed, when I got up to 3,000 RPM the spindle actually started slowing down with each bump upwards in speed. The more I increased the speed at the control the slower it went. I started checking the output voltage at the VFD on the MX3660, and it was increasing normally as expected, so I was getting a proper PWM signal out of LinuxCNC and from the MX unit. The only thing left in the train is the inverter circuit you gave me. I must have something wired wrong, but I've been over it several times, comparing mine with yours and can't find anything wrong. I've attached an image of the board I made, maybe you can spot something I've done wrong. There are two relays added to the board, one for enabling the spindle amp and another for passing the +5V signal to the inverter section. Any clues?
Thanks,
Dan
You need to monitor the INPUT voltage into the SERVO DYNAMICS Spindle Drive. See if it is closely linear with the signal coming from your MX3660.
The circuit is very simple. Outputs 0-10VDC until a +5VDC signal is inputted and then it reverses it. There is more needed to operate the tool changer however. You need the spindle orient signal, the brake, and a fixed voltage to the servo drive to do the spindle orient. That's only half the battle.....
You should just buy my DM4400 B-)
Marty
That's just it--the analog voltage at the spindle amp is actually reducing while the voltage at the vfd is increasing. This is the voltage coming out of pin 6 on the IC going to the input pin 2 of the spindle amp. It hits a threshold at about 3000 RPM then the signals start going opposite of each other. So increasing the speed at the control increases the voltage signal at the vfd (as it should) but starts decreasing at the spindle amp, reducing the spindle speed.
I'm not trying to set up the changer so not sure why you mention that.
Thanks,
Dan
Sorry Dan, not sure what's going on.
Marty
Okay Marty, thanks anyway. Did you look at my board and compare it with your design? I'd like to at least rule out something I may have overlooked with it. Appreciate it.
Dan
Problem solved. Electronics rookie mistake. I had the transistor backwards. I did a little research on transistors before completing this, but didn't spend enough time with it. I made a stupid assumption that the flat would be oriented the same way as your diagram, but the transistors I got were reversed. Once I turned it 180° it worked perfectly. I'm getting full speed on both forward and reverse now.
Dan