[pminmo]

fyffe555
Thanks,
I was hopin the tab would be the source as the are all common to ground. I suspect when you see them share a heatsink the fet is sourcing the voltage via the drain connection, in that instance the drains of the souce drivers wouldn't have to be isolated from each other. I'n not hta familiar with HEXFET's, is there an internal diode that protects against the spike's from the switching states, eliminating the need for clamping diodes?
Eagle keeps me learning, there is always something it can do I haven't seen. Has a few shortcommings, but the price is sure right for the hobbiest.

Phil

[pminmo]

Ok, more info:
http://webpages.charter.net/pminmo/cncelectronics.htm

There is a single channel board using discrete logic compatable with Benny's request and compatable with my modular format. You can build three boards and directly interface to a pc with some wiring. Also there is a start at a document in pdf and the zipped .doc file. Eagle files are there also. I'd like to have another set of eyes look at the schematic and eagle files. I did a single sided layout with 5 jumpers. Once I get a nod as somebody has reviewed I'll post a 600dpi bitmap of the copper that a person can do laser transfer with. Will also combine grahically so 3 or 4 axis, can be on the same board, as I can't reasonably fit 3 or more with the eagle free version. Down the road maybe somebody can do a mill file. Documentation or design help wuld be appreciated.

Phil

p.s. I found a miswire in the 5804 schematic so that will have all have to change. Simple mistake reversed pins 3 and 5 reversing step and direction so it was the odd one out. Messed up the html file so the pointers to the L29x files is missing.

[fyffe555]

Phil,

If I didn't make it clear, I was a lazy SOB when I picked the IRF512 when I couldn't find a 'real' Fet, IRF512 is pin pin/tab compatible but is *not* any good to drive a motor with. You need an IRFz44 or similar Hexfet. You'll need to change your doc and make the part in eagle. Sorry should have made that clear before. mea culpa. Don't want anyone building it with 512's, it might work but.......

IRFZ44 is 60V, 50A and fast switching. IRF512 is a mosfet 100V and 4.9A, Datasheets at;

http://www.irf.com/product-info/data...ata/irfz44.pdf
&
http://cecsfp.cecs.evansville.edu/pdf/irf510.pdf

Clamping diodes in Fet's, some have them, some don't. Both the above do but as you'd expect the IRFZ44 is tougher... Up to about 35volts they probably wouldn't need a heatsink.

Nice effort on the new board, Do like the inbuilt logic probe...

Andrew

[pminmo]

Andrew,
Got it on the FETS, thanks. Changes incorporated in the website.

Phil

p.s. Anybody good a drawing/art? I'd sure like to have a decent drawing of wiring that is in the pdf file to incorporate in the document.

[ynneb]

You Guys make me laugh. You are a small team on here, that seem to enjoy your electronics chat. I think most dont know what you are talking about, and just let you do your thing.

So is the "Benny's request" ready then? If so, can we post it in the zone? I will make a huge noise about it here so that everyone will be aware of our first open source board. You guys will get full credit for your work. Personally I would prefere Benny's request to be changed, I am a bit red faced with the name since I did zip to help you guys. Any ideas for a better name?

Just let me know when I should start making a huge noise about the board on the Zone.

I will do a support page for it that has a step by step tutorial explaining how to make a chemical etched board. In the mean time if we get your file posted we can add to it as we go.

You guys are terrific.

[pminmo]

Benny,
I'm comfortable with the single axis board: http://webpages.charter.net/pminmo/ostop.png
minimal documentation at http://webpages.charter.net/pminmo/osunidata.pdf
But I've been in the business to long to know that the first board built usually has bugs. 20 years ago it was a lot of bugs, today via software such as eagle it's much reduced. For example my Allegro A3977 board I did for myself, I ruined the first board because of one simple error. For some reason a simgle trace was terninated abruptly. I don't know wether eagle was to blame, or I goofed somehow, but the error was in the design file and eagles rule checker did'nt find it. I blew the 44 pin surface mount part troubleshooting the error, and didn't have the proper tools to remove it. Plus there wound up to be some parts that were tight. I mean the board (I built 4) wound up with one patch wire and is a pretty good looking board, but it wasn't perfect, the second batch I expect to be. So what I'm trying to say is it would be nice for one to be built to see if there are any issues. I'm very willing to support a first builder if it doesn't work right out of the gate. One of the reasons I put the simple logic probe on it was to help long distance troubleshoot. And if all else fails if one is sent to me with a return prepaid mailer I'll go over it to verify it or troubleshoot it. On a second note I would really like to see the ZONE here have a dedicated webpage to each element of open source work. I've done something crude on my webhost, but it is a constant link that can be referred to. For example (http://www.cnczone.com/opensource/dr...teunipolar.htm) and in conjunction with that page a single forum thread for that module that people can read to learn each others mistake. Do you want me to place artwork up?

Phil

[chuckknigh]

One of my "local" suppliers has a large number of Allegro A3951 chips -- does anyone have a driver design based on one of these?

The price is definitely right...

-- Chuck Knight

[ynneb]

Yes please Phil, place the artwork up and I will build the first board and test it. This way we will definitely know if its OK to promote. You lot have done a fine job. Thanks guys.

[pminmo]

Benny it's up there :http://webpages.charter.net/pminmo/cncelectronics.htm

600 dpi bmp file. If you need it mirrored or reversed let me know.

Phil

[Greg Fill]

Chuck

A design for the A3951 can be drawn up but I think the A3951 has been discontinued. What is your suppliers price and how many does he have? You may not want to go down the path of obsolescence.

[pminmo]

Chuck,
The A3951 is really a DC motor driver, not a stepper driver.

Phil

[Greg Fill]

Chuck

Even though Allegro calls the A3951 a DC pwm motor driver it can be used as a bipolar stepper motor controller. A bipolar stepper is actually just a two winding DC motor. Two A3951s will be required for each bipolar stepper motor. Unfortunately like most of the older stepper motor (DC motor) controllers you need to use a micro with these chips as it does not have a built in Step/Dir translator like the A3977 has. But fortunately micros capable of this task are cheap < $2.00 US and connect to the A3951 easily. I can help or code the micro if wanted. For those who are interested they can download the spec. sheet for the A3951 from www.allegromicro.com.

So even though the part is going obsolete from Allegro if your supplier has a lot and if the price is right, ~ $1.00 ? each, a 2 Amp 50 volt bipolar stepper controller could be put together easily. The only problem I would have in using the A3951 is I would find the 2 Amp maximum a little on the low side.

If you are interested look into what your supplier would charge for the chips. I am sure it is another alternative that we could use for a DIY controller.

[pminmo]

Benny,

http://webpages.charter.net/pminmo/cncelectronics.htm There is now an optoisolated breakout board there as well. But it is 2 sided.

Phil

[tachus42]

Good work pminmo,
I had a look at the design and tidied up a few things.
1) Relocated the power conection to the output side of the board to avoid having heavy motor currents flowing through the small signal area of the board.
2) Terminated the unused gate inputs - its not good practice to leave them floating. As they can start oscillating and doing funny stuff.
3) The transistor in the logic probe section isn't really needed so i removed it.
4) Moved a few parts a little and rerouted the board keeping the traces thick where possible - particuly where a component
leg joins. I've found the pad lifts when you drill the board otherwise.

Eagle files in the attached zip.

[pminmo]

Great work tachus42, looks much better than mine, Good catch on the unused gates.
Benny wait till I upload the new files.

Phil

[davesaudio]

nice board, fill in the space around the ground leads
(even with some rectangles) - less to etch
looks easy to do at home

[pminmo]

Tackus42,
There are some clearance issues a couple of places. I'm in the process of fixing. Davesaudio, will post the board both ways.

Phil

[tachus42]

I didn't actually check the clearance on the board just used my standard rules.
I find when you actually print the layout there is more clearance than there appears to be on the screen. I use the "toner transfer" method to make boards and find it easier to correct a clearance problem with sharp knife than is to repair a broken track.

The board could be made a little more adaptable if there was a position with a cutable track for a resistor in series with the mosfet gates, then darlington transistors could be used instead. With a transistor level shifter or your opto BOB and a 12 volt regulator you could use standard mosfets with 12v of gate drive.

[davesaudio]

Mr T has 24 thou between dip pins
I worked up thru the sizes and ended up there as well for TT.
Are the printer setting really inconsistant with the gerbers?
- would expect the resolution is sufficient for them to be comparable

drill aid ulp with .4 setting is nice too for TT.
( with the above I have pretty much exhausted my eagle knowledge)
still workin' on the real "fill" thing

Dave

[pminmo]

I uploaded new files based on Tachus42 work. MUCHO THANKS! (eagle brd and sch, 600dpi copper bitmaps, schematic and board layout in graphics format). I did make a few changes. Added more copper area per Daves request. (Eagle Polygon command). I also added on more jumper and eliminated one of the traces that was causing the problem. Moved the pads on the discrete signals slightly to fix the other clearance problem. Redid some of the pads on the resistors and the 10pin IDC for more copper. Tachus42, while your method worked for more copper I didn't like having airwires dangling in case a trace had been missed. I work as much as possible to get the Design Rules Check to be valid. (but not necessarily warning free). I think the resistor idea for using regular transistors is a good idea, just didn't get to it. Haven't looked to see if there is a TO220 pwr transistor that would just drop in the mosfet spot. I have added wire pads on previous projects to allow device substitution.

Am working on your suggested improvements for the BOB. Using the 6N138 actually would allow the removal of the 2003's, by going to a much lower led current that you can rely on a PP. Still thinking that through. Will upload those eagle files also if somebody wants them. I will eventually load all the files here to the zone. Anybody want to take a crack at improving the instuction document? It's uploaded also.
http://webpages.charter.net/pminmo/cncelectronics.htm

Phil