Old engraver to Grbl conversions

Hi all, planning to covert a pair of old engravers to grbl control. The first is a vanguard 7000 (which has a built in controller that I plan to remove). The second is a dalgren 300 also with a controller (but so old it's useless). The vanguard is a 3 axis but the spindle motor may have issues, the dalgren is 2 axis and maybe pneumatic z.

The vanguard Steppers are 2.5amp and the dalgren are 2.9amp so I think I should be fine running them with grbl via cnc shield v3 and the more powerful pololu (sp?) driver sticks. Add a 8.5amp 24v power supply and then reuse limit switches etc. I may also keep the spindle controls if possible.

Anyone work on similar conversions?

I'd like to keep it on the cheap and cheerful at the moment and I don't really want Mach 3.

Sent from my LG-D852 using Tapatalk

Sorry, looked at the specs again, looks like gshield is the way to go or tinyg. Tiny bit more expensive but the 2.5 amp drivers should work better than the 2.2 available for the cnc shield v3.

Sent from my LG-D852 using Tapatalk

OK, so I'm going to risk a pibot kit.

The drivers are supposed to go to 4amp, which I doubt, but 2.9 should be OK I hope. The dalhgren will be the first project.

I've confirmed it is a pneumatic z so really only x and y and then figuring out how to trigger the air solenoid for cutting. It also seems to be missing limit switchs which the pibot kit had.

To go with it, I've ordered a 24v 10amp power supply (which I'll use for both projects).

Depending how the pibot kit works, I may repeat for the vanguard or try gshield.

Sent from my LG-D852 using Tapatalk

Well the pibot bits arrived via DHL, paid the taxes and import fees ($35) and I think I even dug up a 12v power supply I can use to test the steppers out until I get the 24v.

The drivers themselves are a bit underwhelming in look (fins bent during the drilling of mounting holes and heatsink is at slight angle under board) but should be fine for use. The rest of the board looks slick enough and is well labeled which might help as the documentation is otherwise lacking.

Anyhow - now to find some time...

OK, so I haven't gotten too far on this project. I have got the x and y axis jogging, and after swapping some of the stepper wire and giving it a good cleaning (the Dahlgren was quite abused in life I think) its jogging at the max 3000mm/min allowed in the default grbl set up. I've yet to properly set up limit switches but I think that should go OK. What I am trying to figure out now is how to operate the spindle. I've seen how to activate a relay via the spindle pin that should drop the spindle down as needed but I am not sure how to also use it to control the on off of the spindle and speed. So off I go to the grbl area here to see if I can find out. Oh and I have another pibot kit coming to start working on the vanguard 7000 I have,

Ok so I have gotten a bit further. Cleaned the heck out of the Dahlgren. I can now jog about 4000mm.min on the Y and 6000 on the x. Plenty fast for me. The binding is now on the Y not the x at that speed and it remains even after a disassemble and clean of the lead screw nut.

I've tested the pneumatics on the spindle and they work and also saw a NYCCNC video where they built a similar machine. The spindle power is also OK and I'll control speed manually with a fan controller (dimmer switch essential but starts high rather than low) and a relay.

The biggest issue I am having is with the Pibot and or GRBL. I cannot seem to get it to home or even try to home. I suspect this might be because there is no Z hooked up though I can't check the config.h file to see as the pibot comespre- flashed. I guess there is a way I can recompile but at this point I might abandon it and either move to gshield/uno combo or tinyg. I'm leaning to tinyg as that is what NYCCNC was using and I suspect it is a little cleaner on the user side to set up.

Attachment 352622

OK - so after a lot of time, a little progress last night. I downloaded the pibot grbl firmware and edited the config.h file to comment out the z axis and then (eventually - painfully) got it to sucessfully upload it. There are various videos on uploading grbl via the arduino IDE but I eventually found success following the pibot video (ie not adding any libraries to the arduino IDE that pibot has on their website).

Then I could get it to home (after having to invert the axis in a few settings). Yea!

The downside I found is that pibot does not like the 24v power at all.... Maybe a cooling fan (which they do say is needed for 24v) would help but I suspect its better to run it at 12v to the board. The silly part is all the stuff that is getting so hot seems to be for the 3d printer I think. So I'm going to hook up a wall wart that's 12v just for the board.

Also I should note that 1)I have bought a tinyg board so I will likely try that too and 2)before I ordered the Tinyg I had ordered another pibot and it did not ship in a reasonable amount of time (4-5 days) nor have they properly cancelled the order or refunded the money (they did send a credit receipt of some kind - which is useless except as an acknowledgement).

If I had to do it again, I'd just use an arduino directly with either gshield or external drivers for the steppers. Pibot really should be 2 boards, 1 for CNC and 1 for 3d printers. Too much extra stuff going on as it is.

Now I just need to work on the spindle activation/enabling via gcode (and later in the post processor) and setting up the proper values for the lead screws that I have so I have accurate movement. The tidying up the project, simplifying the layout and housing it in an enclosure that works.

Then I can start all over with the TinyG on the other engraver and see how that goes!

Parts so far: Pibot kit $130USD, Machine $600 or so (with shipping), power supply $30, Wall wart $5, fan motor control for spindle $10.

Time: Exploring 2 hours, cleaning/diassembly 8 hours, hooking up pibot 4-6 hours (with lots of fiddling), Firmware struggles 5-6 hours, research 10+ hours, and maybe another 20 hours to go to get it working. If this was not my first project it would have taken about 20-30 hours I expect. As it is, I'll be about double that. Still a very decent result IMHO.

Hopefully my next video shows it moving (I have no cutters yet so that's as exciting as it will get).

So another frustrating night. I did eventually get the spindle and z working with m codes via the relays, though for a while every time I initiated a command, a hard limit would trip (which I believe was due to electrical noise or so I read elsewhere). So I had to disable those using $21 to 0.

Anyhow they bigger issue is that on power up, the pibot board is somehow triggering the relays before setting the programmed state. In other words, the spindle turns on briefly and the z lowers until the program kicks in.

For me this will not work as it would spin and lower the spindle into a work piece or the table.

I've emailed pibot for support on it. It seems like a similar issues exists on these cheap relays for regular arduino sketches, where it happens if the pins are set as outputs before the default state of the pin is specified.

It seems like it can be remedied by setting the default pin value first, and then setting them as outputs. Anyhow, I guess I'll see if tinyg does it too or if its a grbl/pibot thing.

Sent from my LG-D852 using Tapatalk

Hey, any update on this? It's cool to read your progress.

Not really any progress - I seem to spend more time on this mid winter. The latest (last spring) was that I now have new proximity limit switches (the machine had none before) and a Tiny-g board. I wasn't in love with the pibot but have yet to wire it up to the tinyg. I'll be sure to update when I work on it again. Still have a Vanguard to work on after that too.

Gee.....having reads all of this my opinion would have to be.....go the Gecko route and don't look back......after a year and a half the system is still not reliable....what's the point in messing around with the GRBL approach when a Gecko would have worked better straight out of the box.

BTW....this is just my opinion as I'm new to the CNC game but am getting some ideas together for a mini CNC router build to play with, much on the same lines as the one on EBAY that uses GRBL to make it work and only costs $250 all up.

OK so the Gecko controller cost more than the whole EBAY mini router, but it's a DIY project in an all steel build to make one better......having watched the Videos on UTUBE by Myfordboy I got the yen to emulate his efforts too.
Ian.

Well gecko is just a driver, not a controller (see correction to follow). Pibot, tinyg and things like the uc100 are control boards or gecko gm215 and then they have the software controllers. So yes I could have - and still could - use a gecko driver set and I suppose I could use gecko motion control but I'd never heard of it until I just looked - nor does the information seem to be too deep.

But I know gecko products are well respected. My father has used them on some projects but I've not used them myself.

But its hardly been a year and a half. In project hours its been about 5hours or less since the last active post. Just been sitting there since. I have a 5x10' table at work and can engrave on it if I need to so this project is not critical.

The Gecko G540 is a 4 axis controller.

So have made some time to try and get tinyg working and so far it is not promising. The board powers fine, connects fine, drives the fan fine but it does not seem to drive the motors at all, even under no load and with the pots turned all the way up. They hum and the lights light up but they do not move. I will take home a spare stepper I have at work to see if it does any better as these issues may just be due to the fact the steppers are likely 30yrs old. However the specs seem similar to current steppers so not sure why it would be an issue and more to the point they worked OK on the pibot drivers.

The on board drivers are too small technically but I think they should move. I could still use tinyg with outside drivers but then I would need to solder on some pins and that might be beyond my skills - though I likely could have it done. If I have to replace the existing steppers, its not that big of an issue.

Most disappointing thing is that a few people seem to have had the issues on the Sythentos forum but no one has resolved it publicly (either by giving up or a solution).

Ok - so the main problem seems to be that I'm an idiot. I spent 2 nights with the motor's hooked up to motor3 and motor4. (default z and a) The hummed because they were in holding mode as it tried to move the non existent x and y motors.

Seems I can move on to the next step now....

So brought the machine into the shop proper (ie work - I'm the owner so I can get away with it and it might go into use at the end). I've essentially caught up to my grbl/pibot progress now.

Anyhow. I've tested the spindle via m3 and its works with the relay. The air cylinder z works with m8/m9.

The pibot relay set, which I was hoping to use seems like it might be low on rather than high on. What should be a N0 relay only works as a NC relay. A regular dollar variety arduino relay works correctly (but I broke a screw terminal).

Also, as it seems to be sort of known on the tinyg area, the 3v power provided by the board doesn't power the relay so I need to add in a 5v power supply. There was a diagram over on ox builds and I followed that. Kind of annoying to have 5v and 24v and 3.3v in play and still not 100% sure I couldn't have an accident and send 5v down the ground to the tinyg and toast something. But for now it works.

The issues that I had with pibot where the spindle and the air z would engage during boot up are not totally resolved on the tinyg I think. When booting, the spindle did briefly spin (much less than pibot) but the air z did not engage (go down). That I can live with. The spindle spinning is fine it it doesn't plunge into the table at the same time. So now I just need to isolate my limit switches for either 5v or 24v inputs (probably 5 now that I need it anyhow) and then I should be 'ready' to go.

Ready that is if you don't count the rather large task of cleaning up and finalizing the wiring, setting up the x/y travel, homing, limits and then actually getting to a testing phase.