[Mike4703]

I didn't see your post before I ordered. They didn't offer a one time deal, but maybe that's because the G251 drives are not repairable. The small dense multilayered SM components on this drive are not repairable per Gecko tech support. I was having trouble with two of the drives but when I tested them per the instructions in the FAQ section of the Gecko forum, only one failed. Now I'm wondering if there is a problem in the optoisolator board also, so while I'm waiting for the replacement drive everything is disassembled and back on the electronics work table. It would be nice to have a signal generator to test this stullf on the bench--but they are way too expensive. I'm trying to jury rig something right now.

[H500]

Why do you need a generator? The opto will prevent damage to your controller even with the worst case failure.

[Mike4703]

You have to understand you are dealing with a person who knows just enough to be dangerous, and knows it. Therefore I like to see everything work. I wanted to put a signal on each channel of the opto board and watch it show up on the output side with the scope. But a good versatile signal generator is way too expensive--tried a little vellman kit I had put together a while ago but the output isn't right, looks like the waves are alternating + and -. Still playing with it, seems like it should work some. Also trying the little blue box manual controller that came with the TB6560. Too much work to take everything out back and hook it up, or bring everything inside, but I guess that's what I'll end up doing.

[doorknob]

How about using this simple step and direction signal generator schematic from Gecko:

Support

[Mike4703]

Good idea doorknob, looks like it's just a power supply and schmitt-trigger IC. I actually recieved the new drive from Gecko today and installed it. Set everything up and it works! I traced the opto board problems to some bad point to point soldering and pulled up traces on the screwdown connecters. I'm going to epoxy them in place now that the board is working again. I still need to check out the C-axis for use on the rotary table. The DSP controller interface board has a 4th channel that is not supported in the software, so update software vs homemade controller vs buy one from Sherline ($375.)? I also recieved an e-stop switch today so will wire that up and install the home/limit switches, then start trying some code, if no more hardware problems.

[H500]

Do you dislike using a cheap PC running EMC or Mach as your 4 axis controller?

[Mike4703]

Why go with the DSP controller instead of a cheap PC? First of all my "shop" space is very limited and not well weather proofed. I'm only willing to do so much to fix this as I rent the space and don't own it. So my solution was to keep all PC stuff in the house and just take the program outside and run it on the DSP if needed. Also many times I would just want to be able to move the machine using the steppers but under manual control using the DSP as a "joystick" and DRO. Lastly I was a computer guy first and a machinist only later and the use of parallel ports and seventies looking computer programs didn't have any initial appeal. After my struggles with this, though I readily admit that it does have practical advantages. Oh well at least I've learned more about the electronics that if I'd just bought the regular stuff. What the world needs is a really good DSP at a reasonable cost. If I had the time I would start learning more about programming a microcontroller, but one that uses a BASIC type language, not the PIC that is used in rotary indexer project in the forum. I have no idea at this point how I'd get the microcontroller to run the G code but I'm sure it's possible.

[Mike4703]

Of course, I can't leave the build your own DSP/pendant controller idea alone. I now understand why the builder of the rotory table controller used the PIC controller. It's alot cheaper--but maybe harder to use than the ones made by Parallex. Right now just want to make a cheap source of step and direction pulses. The Gecko pulse and direction tester may be a starting point, but I would want some control over it--change direction and speed of the stepping. Anyone have any ideas?

[doorknob]

How much $$ is considered inexpensive?

Do you know how to program in C (or are you interested in learning)?

How about one of the LPCXPRESSO demo boards? I think that they come in versions that cost around $20 or $30 depending on which board you get. I have a couple of them that I'm going to be playing with for some sensor monitoring applications, but they may work just fine sending step and direction signals to some stepper drivers.

They have plenty of on-board processing power, there's free development software for them, and you can use built-in USB to download and debug programs.

See NXP OM11048 LPCXpresso Board and NXP Semiconductors - Microcontrollers [LPCXpresso] for info.

PICs can be fun to use, too, and there are (or used to be) a lot of inexpensive demo boards available that could be pressed into service.

I used to use some of the parallax chips in some projects, but I did not particularly like the built-in language interpreter, rather I preferred working "closer to the hardware" in order to better take advantage of processing speed.

I have considered doing some CNC experiments with an mbed board (I have a couple of them, too), which is quite a bit more powerful, but I think that they're close to $70. See Rapid Prototyping for Microcontrollers | mbed

(Oh, and there's always the arduino boards - there is already a G-code interpreter available that runs on arduino - look at GRBL and the grblshield http://grbl.tumblr.com/ and http://www.adafruit.com/blog/2011/03/02/the-grblshield/ )

[Mike4703]

I only have practice in Basic and Visual Basic and that is somewhat distant and a little with the basic stamp microcontoller. I don't really have the time to learn C although I would like to. My main experience is in clinical medical applications and databases, which doesn't directly help with this. My goal in getting into CNC is to create my own prototype designs, so getting to the making parts stage, and getting back onto the CAD learning curve is first. That said I have gotten interested in the CNC hardware and electronics but I do have to remember where I want to go with this. Thank you for the info. I will start checking out your leads but I may have to leave microprocessor developement to others because of time constraints and will end buying something. Haven't decided how much money will be too much or little. That often depends on how much is available at the time

[H500]

Here is a ready to run microcontroller board for less than $5. Digi-Key - 296-27570-ND (Manufacturer - MSP-EXP430G2)
The software can be downloaded for free. I just got one.

It's programmed in C, but what you're doing is so simple that I'm sure you can get a forum member to do it for you in a few minutes.

[Mike4703]

Wow, prices for these things have really come down. Might be fun to play with even if I end up buying a commercial controller. I have to think about this. For now I'm busy dialing in the DSP controller, and learning how to use it now that the hardware is working. I wonder if the C code for the PIC rotory table controller could be recompiled to run on this processor?

[H500]

The code cannot simply be recompiled. The syntax is slightly different and each chip have different accessories. It would be best to stay with the PIC if you don't want to mess with the code.

TI is probably selling the boards below cost in order to get more people try their chips. It's a good idea. The MSP430 is a very capable 16 bit processor.

[Mike4703]

Ok it's been a while but here's an update on this project. I went ahead and bought the Sherline rotary table controller ($375) because I realized how much time the DIY route took me. I had to keep with my original goal of learning CAD design and actually making something besides the machine. The Sherline controller can be interfaced to the other controller although it makes the G code more complex and CAM generated code would have to be edited. On the other hand it makes simple procedures on the rotary table very quick and simple, no G code required. I just need to design the interface between the 501 controller after deciding how to impliment it--should not be anywhere near as difficult as the first time. Here are some pics--I'm also building an enclosure. The last shot is the latest machining test, to see if my machine would handle a t-slot cutter. Did ok with real low feed rates.