I'm working on a new driver board based on a Atmel's AVR's (single chip micro) for the open source forum here. Long term intent is to have a low cost DIY three axis board, that will also have the pport connection and interface for limit switch and spindle. Attached is a grapic of a one axis board layout and schematic for software development. I doubt if I can get to doing an this actual board for several month's. If somebody would be willing to etch or mill one for a minimal fee, I could get to work on firmware.
Or if somebody want to do some code, or.........

The Attiny2313 is fast enough to handle a simple edge filter to make the input filter circuit on my other boards unnecessary. It is also inexpensive. It should be able to handle half step as well as full steps simply. The 3 axis board would have three of them.

Cost wise the 2313 is about $2.25 and the fets show are about $1.50 each. That adds up to a 3 axis board that could probably be built for under $45 worth of parts.

Phil

If all you want a couple to evaluate the design with, I can pop off some quick TT style boards..
just give me a mailing address..

got some going in the etch tank tomorrow am.. can do easily,,all I need is Gerber, or Eagle / Pulsonix..

Vladdy,
Look at your email. Thanks,
Phil

cound you tell me what the fed numbers are.

Irlz44

I'm playing with a similar designed based on a at90s2313 and a irl530. The result was very unsatisfactory. To get speed, it is necessary to use a high (35-45v) power supply and a chopper to control the current. A chopper can be implemented by adding a lm324 op amp to use as a pair of comparators to detect the coil current. With the o/p going to interrupts, the mcu can turn off the drive for a fixed time whenver the setpoint is exceeded.

I will try it this week as soon as I have time to write the program.

One of the issues here may be the drive capability of the micros. The irlz44 etc
have fairly large gate charge requirements (66nC). You may need to use some sort
of buffer circuit to drive the fet gates to get them to switch fast.

jazz

Irlz44
If the FETs aren't the logic gate types...then that processor most likely won't be able to turn on the FET. Also I didn't check if you could do bit operations on the Micro's ports.

H500,
To get the best speed out of a stepper, you have to run a higher PS voltage and current limit.
Jazz,
That is a concern, one of the reasons I wanted to protype.
Viper,
The AVR is bit programmable, and yes the FET is logic compatable.

Phil

Got the PCB's out by snail the other day, bit of a delay, toner cart ran low in the HP, and topped up with generic toner, not as clean as before, so photo etched...
last photo is TT, piss poor..but good enough for a 'trial' anyways

you should have them in a few days..

enjoy..

Got the PCB's out by snail the other day, bit of a delay, toner cart ran low in the HP, and topped up with generic toner, not as clean as before, so photo etched...
last photo is TT, piss poor..but good enough for a 'trial' anyways

you should have them in a few days..

enjoy..
did you etch you boards or have them done?

vladdy,
Thanks.

Phil

did you etch you boards or have them done?

etched in the basement.. the copper appearance single PCB was done with Toner Transfer, but my toner on the old HP was getting poor, and the 'generic' toner I used to top up wasn't quite fully compatible..pretty messy..
factory toner in that HP [300dpi] gives half decent transfers at 16mi, maybe 95% acceptable at 12mi, to get consistent 12mi I should actually have a 600 dpi laser.. For most of my hobby stuff I run 20mi signal and 24mi supply widths..no screwups, quick results..
The bluish set of three PCB's was photo etched, a little faster, cleaner, better quality, but more expensive.. probably $4 CDn for actual photo sensitized board cost for that size..
Silkscreen is better for [me anyways] for quantities of 5 plus.. a little slower because of drying time, and probably the cheapest of all for larger quantities.
The top placement overlay on the bluish ones was Toner Transfer again, useable but not 'professional' quality..
I did a quick silkscreen with 160 on these, but 12mi isn't very 'neat', useable, but not really nice looking, need 255 or better for 12mi, I use 160 for 20/24mi and the quality is reasonably good..
After all, this is just a hobby??

enjoy..

hows this going?

Vladdy's board hasn't arrived yet. WIll update when it does.

Phil

and you still have some work to do before that..
code to check..
board to populate...
oh yeah, and you said you would drill the board ...:)

that's the part I'm not too fond of..

Just might have to make a drilling table.. just a hand one, probably faster than doing a cnc setup and registering, especially for small [read 4 or less] quantities..

should still have some small linear slides around here someplace..probably drill from bottom up, need some kind of alignment pin..dremel mount, another dremel [again], maybe a small magnifying glass, just in case...
maybe not, that way there will be a built in excuse if they are 'just a tad' off center..:)

great, another project...

vladdy,
Very true, I thought I was the only one out there that juggled a dozen projects. :-) Have to say a pcb router is dancing in my head...............

Thanks for your help,
Phil

vladdy,
Very true, I thought I was the only one out there that juggled a dozen projects. :-) Have to say a pcb router is dancing in my head...............

Thanks for your help,
Phil


Ya... I think my mill will be the second thing I finish.. pcb router has been moved to first.

Tek

The coating on the set of three is the photo resist, I didn't bother removing it after etching, makes a good protective layer to help against corrosion, especially if it's going to sit for a while..It's not too bad to solder on top of, I use a Haako set around 550 degrees, gives a bit of a poor mans solder resist otherwise..can be scrubbed off with acetone, or polished off with a scotchbrite pad, tin if you want.. their your's now..:)

The single one was Toner Transfer.. not very good..traces should be widened up a bit, and a little more isolation, especially on the one end with three small traces to make it a little more 'idiot proof'..
I did a quick silk screen for these, but all I had was 160 line silk, results on a few trials came out 'fair'.. either a wider [024, 020 ??] trace would work better, or 255 line silk to get proper definition.

You should have mentioned you needed some 2313's.. I've got a few dozen left over from a old satellite project...let me know if your's are going to take a while to come in..mail got there faster than I expected it would..

how's the code coming??

enjoy.

You should have mentioned you needed some 2313's.. I've got a few dozen left over from a old satellite project...

An old satellite project ehh? Sounds pretty cool. :cheers:

vladdy,
Just sketching it out, listing attached, haven't checked it against pin numbers, and half stepping. Really forming ideas for a basic start.

Phil

Free smal Basic compilier for AVR's:
http://www.mcselec.com/bascom-avr.htm

Phil,

Are you still working on this project?
I'm interesting in learning microcontroller too.
I had coded something similar to yours but on an AT90S8515 (coded in assembler). I was able to simulated it on the STK500.

Please let me know how this project goes.

Kind of,
I sent a blank board to Ito Brazil (cnc zone handle) a couple of weeks ago. I actually in the last couple of days have started looking at it. I had stolen the FET's for the L297uniboard prototype and ordered a different Mosfet for this avr board which I now have, I'm looking at the half step code also. Will keep you posted.

hows this Going Phill? Im looking for some drivers that can run @ 55V dc. I like the DYI world.

thanks

Look at the IR2010 H bridge drivers.
They can drive most FETs
Voltages to 200 volts
Directly interface to microcomputers
Can be bootstrapped so no high voltage supply source is needed.

http://pdf.alldatasheet.co.kr/datasheet-pdf/view/91738/IRF/IR2010.html

There are a number of other H bridge drivers with similar performance

Add current chopping and you have a high current - high voltage bi-polar driver.

Hi guys,
Teckno,
Actually I'm in work on a expanded version of the AVR board on the first post. 100V, unipolar, heavy current (can't say yet, the FETS are more capable than the board will be) microstepping. The unipolar will come first as the hardware is almost ready to order protoboards on. Microstepping has been the hurdle to implement in a low cost, KISS design, but that too I believe is not an issue, at least to 1/8th step.
Also in parallel is a bipolar design.

BCwandererer, I'm familiar with the 2010, 2104, etc. Right now I'm playing with P channels on the top side though. Their simpler, their cost too implement is less, 120V 14A Pchannels don't cost much these days and their pretty robust.

Both designs are limited to 100V.

I would like to contribute some ideas for a unipolar, pic18252 based design.
I have been working on the board design.
I have not started coding as I have been working on the layout.
The microproccesser is a pic18252 because I have some on hand.
Resonator is 20 mhz
I am driving the FETs with a high current driver chip.
Microstepping will be by using a resistor ladder with 4 micro outputs to input into the comparator that controls the current in each leg.

The board layout is still under review as there may be some bypass caps to add. Values of resistors on the layout are not right. They will be fine tuned when calulated.

When the unipolar unit is working, a bipolar unit will use similar coding.

I am throwing it out for ideas.
Maybe someone has done this before and has ideas that will shorten the learning curve.

Layout and schematic attached (forgive the non-technical layouts - I am still new at this)

Schematic reloaded in larger size

BCwanderer,
I can't read your schematic can you enlarge it?

Sweet. are you planiing on a chopper board or resisters? and or are you going to expand this to Bipolar.

Sorry I havent replyed just been a lerker of late.

Tek

Hi guys,
Teckno,
Actually I'm in work on a expanded version of the AVR board on the first post. 100V, unipolar, heavy current (can't say yet, the FETS are more capable than the board will be) microstepping. The unipolar will come first as the hardware is almost ready to order protoboards on. Microstepping has been the hurdle to implement in a low cost, KISS design, but that too I believe is not an issue, at least to 1/8th step.
Also in parallel is a bipolar design.

BCwandererer, I'm familiar with the 2010, 2104, etc. Right now I'm playing with P channels on the top side though. Their simpler, their cost too implement is less, 120V 14A Pchannels don't cost much these days and their pretty robust.

Both designs are limited to 100V.

tekno,
if the question is to me, they are chopped.

THanks Phil it was to you.. sounds like a sweet idea. I have been thinking of how to make a Power Supply since I read this.

tekno,
if the question is to me, they are chopped.

HI, very new to this cnc idea but would like to know if anyone has built a controler ussing the Alegro a3983 driver chip, if so do you have a circuit diagram and board layout and how did you cope with the heat disipation. I was thinking of mounting it upside down and gluing or soldering a heat sink to it.

I just bought the Butterfly proto board about a month ago with the c programming manual kit. Thought about doing a cnc system with it . I joined the forum and being a linux user went right to the open source after introducing myself... Was wondering where this section of the forum was heading and if AVR Butterfly could be of help... They are cheap and I am thinking about buying em for christmas gifts.