[GrahamY]

Hi kwackers,

In the beginning you posted the source code (c-code) for your programmer, however for the new updates you have only posted the .hex file.

If you are willing to re-post the latest source code I would be interested (I may contemplate adding functionality to use the indexer as a power feed on my mill). You said before that you are using the Hi-Tech C compiler with the Mplab IDE, is this the "lite" version or a full paid up version?

While I haven't built an indexer yet I do intend to, it looks like a very useful and well thought out tool.

Graham

[kwackers]

Hi kwackers,

In the beginning you posted the source code (c-code) for your programmer, however for the new updates you have only posted the .hex file.

If you are willing to re-post the latest source code I would be interested (I may contemplate adding functionality to use the indexer as a power feed on my mill). You said before that you are using the Hi-Tech C compiler with the Mplab IDE, is this the "lite" version or a full paid up version?

While I haven't built an indexer yet I do intend to, it looks like a very useful and well thought out tool.

Graham

Hi Graham,
yeah sadly I took the decision not to release later versions of the source after I came across a couple of sites with modified versions of the code (in which they'd removed my copyright and replaced it with their own). One of these sites was very obviously attempting commercial gain.
I'll probably review this decision at some point.

However with regards using it to create a mill power feed then the core motor driving code hasn't changed an awful lot and the original code would be well up to doing that. It's certainly an idea I've been playing with on and off, although I've allowed mine to get a little over complex and become more 'simple cnc' than feed.

The compiler used is the full version of hi-soft C (around 8.x I think), although I would think it'd be fairly easy to get the code compiling on another compiler.

[bmurray]

Hi Steve

I have just come across your elegant design and its very nice documentation. It should do what I want but I would like your comment. I use a Bridgeport with a DRO that gives me one pulse per 0.0005" table Y travel. I would like to use this pulse train to drive the rotary table at a fixed preselected increment so that I can make helical gears by coupling the rotary table to the Y axis traverse.

I guess the Division mode would be the correct one to use. However having looked at your latest schematic V2.0 I don't see where the input pulse should be connected.

Secondly, while I am an experienced mechanical and analog electrical engineer, I only know enough digital stuff to be truly dangerous. Therefore I wonder if you are willing to supply a programmed 18F452 ready to go. I read your post a couple of years ago saying that you were not supplying circuit boards. Any change of heart on this?

Thanks

Bruce

[elforum]

Hi,

I put 72,000 step per 360, not work

I use to move a linear and correspond to 1 degree = 1 mm

merci

[kwackers]

Hi,

I put 72,000 step per 360, not work

I use to move a linear and correspond to 1 degree = 1 mm

merci

Can't do that - max value is 65535 (16 bits), sorry...

[elforum]

problem solved, merci
I replaced the gearbox of 2 / 1 with 1 / 1 and we set 36000 steps per 360
i have 0,01mm per step, screw is 2mm per revolution

morning trying to see what came out

photo old gear

[MGTp93]

I ordered a pickit3 but way later than the parts, coz i was trying to figure what the best option to go (havent programmed anything else yet with microcip so i am a noob about it).

When parts first arrived i had an itching hand + a dead brain (bad combo). I started the work and couldnt figure where to stop in time. Wouldnt be a problem if i had any skills in the first place, because i had a drip of solder leaking into inside socket without noticing, and when pushed the ic in to see how it is, i suceeded to damage 2 pins (19&20). Pins meters ok with pcb and between each other so it should be feasable except its blank. I would rather try to program in circuit instead of trying to take out the pic. Havent done anything so far, but i am in a dilema. I have read just a litte from the pic datasheet and pickit3 manual (damn should of went with just a simpler the programmer), but cant figure out clearly what to do.



So, a noob's question:

Is "in system/circuit programming" - possible for this circuit?

1) I noticed that vpp and vdd seems to be in short, whereas the documentation i got with pckit3 "reccomends" 4.7k-10k rezistor between. So far, havent figured if this resistance is mandatory or not.
2) i think PGC and PGD (39&40) seem to be clean which is what encouraged me to this direction.
3) I notice C4, C5 condensers on vpp~vdd line.

If i remove the C4, C5 condensers, isolate the pin 1 island with a cutter and link it to vdd with a resistor, will this be programable environment for the 18F452?
I would rather do anything like that, to keep the chip as it is, because it feels like those are fragile and will break upon removal. Maybe, with some jumpers, something like this could be added as a feature to the pcb layout for the future noobs without any skills - like me


Also, a secondary note: there is a rezistor for the backlight (as it is noted on silkscreen with 82). That is that a 82 ohm value rezistor yes? I'm asking because i cant put my finger on it on schematic or partlist. Just asking to be sure, as i said im a noob .

Sorry for the trouble.

[hhogendoorn]

You can program the pic "In_Circuit", if you search for example Microchip AN851 sheet 20 you see a MAX232 connected to pins 25 and 26 of the 18F452.
together with the free MPLAB software and a serial port on your pc it's possible to program the hex file.
Please do not remove C5 an C4 , it's necessary for decoupling.
The Backlight serial resistor is +/- 82 Ohm.

Hans

[MGTp93]

Sorted it. Only had to isolate pin 1 (cutting the vdd trace near it) and connect it back to vdd through a 10k resistor, then made all connections as shown in pic and programmer documentation(s).

Just had some errors where MPLAB IDE couldn't id the chip, the reason was i had quite a few small interruptions on the pcb, at those fine traces, one even on the pgd line. I patched the pcb and thickened all of those fine traces, then programming worked like a charm. I will continue when i get the other parts, just for now chip seems to be ok.

[joegib]

Hi,

Can I raise the vexed question of suitable drives/stepper motors again, please?

I imagine this excellent project appeals to a lot of people like me who've no intention of getting into full-blown CNC but who would like to exploit modern electronics to automate some tiresome manual engineering processes like dividing. Personally, I don't expect constructing Steve's control unit to be a problem and, likewise, I can cope with the mechanics of hooking up the stepper to the the dividing head I plan to use. Where I'm all at sea is in specifying/obtaining a suitable drive and stepper motor.

There was some discussion of these matters on pages 15/16 of the thread (post 174 onwards). That gave me some idea of what to avoid (the cheapo Chinese drives offered by many traders on Ebay). I'm still fuzzy, though, about what kind of torque spec I need for the motor. I appreciate this depends up to a point on the duty required.

So, can I ask people who've built Steve's project to drive 4 - 6 inch rotary tables of the kind used by home workshop guys (Vertex and suchlike), what torque spec does your motor have? It would also be helpful if people with some expertise in these matters could say what the minimum torque spec for this kind of duty would be.

Thanks

Joe

[kwackers]

Hi Joe,
I've used a number of steppers on small (4/6" tables) some are just unknown motors I've found in 'bargain buckets', I've also used the following 180Ncm from Arc Euro ( Arc Euro). On a small 3 or 4" table I'd expect (but haven't tried) the smaller 36Ncm motor they sell to be suitable.

Prices of stepper drivers seem to be on the rise, but you may be lucky if you look around.
I use something like this routout driver, Arc Euro also have a driver although it's getting a bit pricey.
Basically you need a driver capable of producing the necessary current for your stepper (probably 1 to 3A), you can run the motor on a lower current setting with some loss of torque - I run my 2.5A motor on 1.25A, the loss in torque not really being an issue for this purpose.

Once you've found a motor and driver, post on here for opinions.

Steve.

[joegib]

Thanks Steve. That's given me a starting point.

Joe

[Kekar]

Is it possible for someone to collect all the parts and make a "kit"? I can solder, but no chip programming or board etching.

Kenneth

[GrahamY]

Hi,

I am thinking of using a DC-DC converter in the circuit instead of the 7805 regulator. I was wondering what the current draw of the whole circuit is?

By my calculations,
the PIC draws 2mA (ish)
The LCD draws 5mA or 80mA with Back Light
The Stepper Driver would be around 75mA (the inputs are opto-isolated for the ArcEuro drivers)
The Sounder, Sense, Ack circuits look like they are very low current.

Therefore a current of around 250mA would be approximately OK (so a 2W DC-DC for safety)?

One last question is... Does the PIC store the saved machine profiles even if the power is removed or do they have to be re-entered each time the unit is power cycled?

Thanks, Graham

[kwackers]

Hi Graham,

250mA sounds about right, main current draw is the LCD backlight. Not sure about the opto isolated inputs on your driver, mine are quite a lot less than that.

Machine profiles are saved in EEPROM on power off - along with the settings for the current mode.

Regards
Steve.

[GrahamY]

Hi Graham,

250mA sounds about right, main current draw is the LCD backlight. Not sure about the opto isolated inputs on your driver, mine are quite a lot less than that.

Machine profiles are saved in EEPROM on power off - along with the settings for the current mode.

Regards
Steve.

Steve,

Thanks very much for your immediate reply (you are sure good ). I am drawing the circuit up in ExpressSCH and ExpressPCB (free software and linked to a PCB fabrication company, 3 small boards for $51 + postage) and I hope to post the files on here when complete for others to use if they want.

Graham

[gandalf1]

OK second attempt at posting, first has vanished into Cyberspace.
Firstly may I add my thanks to Steve for posting all the details for building this piece of equipment. Next I would like to ask if anyone has had experience of using the latest version of software in which it uses motor steps per chuck or table revolution as I am using a 36:1 ration worm and wheel gear set and 200 step motor? I spent a most pleasureable evening going through all the posts on the project but could only find a single reply as to how well it worked. Also for my own convenience is there a DFX drawing or set of GCode I can use so that I can mill the circuit on my home brew CNCRouter?
I ordered the display from China and for those in UK the PIC from Saltburn on Sea so now am awaiting anxiously the delivery from the Far East although most likely the local delivery will take as long. In the meantime I can make a Nema 23 motor mounting bracket to attach to the little dividing head I am, converting and continue collecting the rest of the controller components.
Thanks again Steve for the inspiration to convert my dividing head.
John.
PS Perhaps in recognition I should print out a Duck logo and stick it to the control box.

[kwackers]

Hi John,

It's probably a bit 'off' for me to answer your post but perhaps I can allay your fears. ;-)

Firstly 36:1 and 200 steps will work fine with either version. The older version of the software makes an assumption of 400 steps so you'd need to fib and tell it that the ratio is 18:1 instead of 36 (since 36*200 is equal to 18*400).
However I'd suggest you run the motor in half step mode (you should be able to set your driver up for this) in which case your motor will be 400 steps.

If you use the later version of the software the only thing that has changed is that previously when the ratio was entered it was multiplied by 400 and stored. Now you enter that number directly which is why we can have strange ratios that weren't previously possible.

Hope this helps.
Regards Steve.

[gandalf1]

Steve,
I can think of nobody bettwer qualified to answer my question so quickly. Now I have to decide which way to jump, at the moment I cannot think of any odd divisions needing other than those divisible by 2 but there may come the time when I need an odd number of teeth on say a toothed belt pulley so p[erhaps the latest version is best for me. Pity the pension does not run to more than one PIC or else I could swap things around a bit. Of course not having used these things before that might not be possible. Sadly my motor controllers only support single stepping which is why I thought to ask first before I considered adding a 2:1 belt reduction into the system.
Thanks for your quick reply.
John.
PS For those interested the PIC 18F452 from mechanique.co.uk in Saltburn the cost was £10.90 including postage

[kwackers]

Hi John,

You may as well use the later software, it's identical in every way with the exception that you need to multiply the worm ratio by the motor steps before entering the number, since this is done once only on setup and gives you the versatility to easily move the device to another piece of division hardware with an oddball ratio sometime in the future then there doesn't seem any reason why not.

I'm slightly confused by your post though and think it's possible you've misunderstood how the device works?

Just to clarify. The number you enter in the setup is simply the ratio used by your dividing hardware. Once this is entered you never change it (unless you change the hardware).
When using the software you can tell it how many divisions you need directly, e.g 7, 16, 537 etc etc.
This is identical on both versions of the software and doesn't require the setup to be changed.

So for example, if you want to divide by 19, you turn it on, select division and enter the number 19, that's pretty much it.

Steve.