Reply to topic  [ 73 posts ]  Go to page Previous  1 ... 4, 5, 6, 7, 8  Next
My projects 
Author Message
User avatar

Joined: Mon Feb 06, 2012 3:04 pm
Posts: 258
Real Name: Alex Crane
Bob, if you can live without the V carving that is attached to V-Carve pro, Cut2d will do nearly everything that you would want to do on a vertical mill. Boise State was looking for some software to use with their full CNC bridgeport, and for $150 you cant beat it. There are not really any operations you would normally do on a mill that you can't do with Cut3d, and it is the same interface that you are used to. It is what I would use, if I didn't spring the extra $450 for the V carving features for sign making.


Sun Feb 10, 2013 11:35 pm
Profile WWW

Joined: Wed Mar 07, 2012 10:40 am
Posts: 123
Real Name: Robert
Hi Alex--yes, I looked at Cut2D--since most of what I want to do is cutting front panels for my electronics projects, this is probably the smarter solution. But, it would be more problematic for my mill conversion, and since I got lucky and ended up with some spare money recently, I went and got the V-carve program.

My PC board layout is done, I'm doing some last minute checking and I think it will go out tomorrow. I ended up making the 1Ghz analyzer/200Mhz scope as an adapter to an existing FPGA board, trying to minimize the extra work, even though this will look kind of kludgy with a PC board mounted on a PC board. By the same logic, I made the Raspberry Pi board have a generic interface that will talk to an existing FPGA board and to the RPi rather than plug directly into the RPi--that way I don't have to redo all the interfaces I did on the FPGA board. This will also hold the PC board cost down since the adapters are quite a bit smaller.

Unfortunately that is going to look rather kludgy too--but I'm really feeling like i need to consolidate my time more efficiently and re-use existing layouts--too many projects are taking too long to get done. The PC board layout sucked up all my spare time for the last month. I'm discovering I am far faster at laying out boards than turning them on.

On the OLI Arduino/motor interface board--with no takers, I decided to skip that effort but will reconsider if some of you are interested. Maybe that can wait for the next run, I seem to be doing around 2 times a year or so.

Next up, once the PC boards go out, is to see if my 1Ghz sampling idea works on my FPGA board. If that works, I'll use it to debug a DDR DRAM timing problem that my current analyzer is too slow to capture. The new analyzer will be more portable--kind of like a really fat logic probe but with an LCD panel display interface. Once I get *that* working, I'll try writing a simple RPi OS that will talk to the RPi interface board.

Bob


Mon Feb 18, 2013 10:15 pm
Profile

Joined: Wed Mar 07, 2012 10:40 am
Posts: 123
Real Name: Robert
Kewl! I got my FPGA 1Ghz analyzer working very well! You wonder, why do I want an analyzer? Well, if something electronic like your arduino is not talking correctly to your stepper motor driver, for example, noise on the step and dir signals, an analyzer can pinpoint the problem quickly. For example, I am getting errors from my DDR memory circuit, but I don't know why. I can go easter-egg hunting and hope to adjust the DQS and data delays blind--hoping I hit on something that works, or I can collect analysis samples and decide exactly how much to adjust the PLL phase delay.

My old analyzer had a 10ns sample rate (200Mhz sampling, but the circuit could only detect at about 100Mhz). That DDR picture below is running clock at 80Mhz, so the old analyzer could sometimes collect a single sample per clock cycle--way too crude to tell me how to adjust the PLL. The new one is squeaky clean at 1Ghz with 13 samples per clock cycle and 1ns resolution in the display. You can see the read cycle DQS and data phase alignment is perfect, so now I will look at the write cycle. This demonstrates the enormous power and flexibility of an FPGA design.

Note that the analyzer interface was written in a few days using the Cycling74 MAXMSP graphical programming environment--kind of like the Lego Mindstorm environment but much faster and substantially more powerful.

Bob


You do not have the required permissions to view the files attached to this post.


Sat Mar 23, 2013 1:54 pm
Profile

Joined: Wed Nov 28, 2012 2:48 pm
Posts: 58
Real Name: Dan
Looks great! I still need to get my fpga board out again to play with it some more. After hooking up a little adder with in within the first few days I haven't had much time to play with it. Here's hoping I'll find some time soon :-)


Tue Mar 26, 2013 8:08 pm
Profile
User avatar

Joined: Wed Jun 06, 2012 9:30 pm
Posts: 334
Real Name: Randy
Looking forward to seeing the FPGA project here in person. This looks really nice Bob!


Sun Apr 07, 2013 12:11 am
Profile

Joined: Wed Mar 07, 2012 10:40 am
Posts: 123
Real Name: Robert
Yes, I'll bring it in--I need a case for it, it's just a raw board assembly right now. Wish I had a 3d printer! I am building up another board--it's so great to do PC board runs, because my original analyzer is a big one-of-a-kind hand assembly. Now I can build several and leave them attached to each of the projects I'm working on. I show the second analyzer board assembly in the pic below--I don't have all the parts installed yet. And the big panel is something I made this morning on the CNC router--it's the new control panel for both the router and the CNC mill conversion. Any ideas for creating labels for it?

I powered up the mill CNC powersupply for the first time--no smoke, everything is looking good. Next is to get Mach3 talking to a free spinning stepper motor. To do this, I'm building the parallel port interface for Mach3--it will be an FPGA board with direct control of the stepper motor driver signals (STEP and DIR) and will take feedback from the mill DRO to compensate for backlash. The way I'm proposing that that will work is that the FPGA board will monitor the steps issued, and if there is no motion, extra pulses will be issued automatically by the FPGA board. Of course, that wont work if two axes are supposed to move (one axis might have backlash, the other no backlash)--still thinking this through--I'll have to be careful and plan on some messes. I thought if there is two axis motion, I could detect quickly if there was backlash in one axis and stop movement on the other axis until the first axis moves--minimizes the impact of asymmetrical motion. I briefly researched backlash compensation methods, but didn't find any that could handle the 2 or 3 dimensional case. In any event, by putting this control through the FPGA board, I can incorporate feedback from the DRO and maybe come up with something better than nothing (my mill has about 4 thousandths backlash, more on the Z axis).

Bob


You do not have the required permissions to view the files attached to this post.


Mon Apr 08, 2013 8:38 pm
Profile
User avatar

Joined: Wed Jun 06, 2012 9:30 pm
Posts: 334
Real Name: Randy
Are you taking the direct DRO interface from the scales or just tapping/splitting it?


Mon Apr 08, 2013 9:28 pm
Profile

Joined: Wed Mar 07, 2012 10:40 am
Posts: 123
Real Name: Robert
I'm planning to tap the quadrature signals coming from the DRO scales going to the display panel. It should be pretty simple to make a digital quadrature counter in the FPGA that will have enough accuracy--but if I need more, that FPGA board has an eight channel ADC on it (only four channels wired to a connector, so I'll have to do a kludge...)

Bob


Tue Apr 09, 2013 8:44 am
Profile
User avatar

Joined: Wed Jun 06, 2012 9:30 pm
Posts: 334
Real Name: Randy
What is the step rate/distance per pulse on the scale? .000 or .0000? So digital rather than analogue steps from the scale?


Tue Apr 09, 2013 9:13 pm
Profile

Joined: Wed Mar 07, 2012 10:40 am
Posts: 123
Real Name: Robert
Quote:
What is the step rate/distance per pulse on the scale? .000 or .0000? So digital rather than analogue steps from the scale?


If I remember right, it's 0.0000 (I'm at work right now--not working, obviously :oops: ). I suspect that there is significant interpolation of the quadrature sine waves done to get tenths of a mill--I will be figuring that out shortly. I'm not going to hack the readout panel, I'll just tap the quadrature signals for my own nefarious purposes. But in the meantime, I decided to do a shortcut and yesterday just made a straight buffer board for the parallel port to the stepper motor drivers--I want to see this thing work, and doing an FPGA board would take at least another couple of weekends. I also got the big control panel interface in place, now all I need to do is wire up both and hook up a stepper motor. Almost there!

I also (with some major help from that new analyzer) got the new home control PC board working, I had been having trouble with the DRAM timing. It's now running a randomized DRAM data test for hours with no fails. Time to starting writing code--I really need a kitchen panel with multiple timers. I tend to cook something, then go downstairs to my office and look at webpages/video game/project work/whatever and forget (I cant hear the oven timer down there) and something burns. Since the home control system is hooked up to our intercom, I will set timers that announce completion on the intercom--I *really* need this!!

Bob


Tue Apr 16, 2013 8:58 am
Profile
Display posts from previous:  Sort by  
Reply to topic   [ 73 posts ]  Go to page Previous  1 ... 4, 5, 6, 7, 8  Next

Who is online

Users browsing this forum: No registered users and 2 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group
Designed by ST Software