[santiniuk]

Hi All,

Like a lot of people on here, this forum has captured my interest.

I now no longer pass heaps of scrap metal without wondering what use they can be used for....

I'm from the U.K and would like to pass my thanks to Alan at www.fromorbit.com as his Picstep controller design will be the 1st elements of my project.

It's early days but hopefully writing my log here will be an incentive to keep going...

Enough rambling.... thanks for having me on here....

[santiniuk]

Stage 1

I need to make 3 controller boards and 1 interface board.

Had some spare photo-resist board so after printing the artwork on a HP laserjet printer to clear film It was time to hunt out the UV exposure box.

Some of the pads are not as thick as I would like, especially as they need to be hand drilled.

Photoresist was not too bad so I will continue.

[santiniuk]

After exposing the 4 boards to UV light, the exposed resist was removed.
Not one of my favourite jobs !

The boards were then placed in the etching tank and overall I was quite happy with the track detail.

After washing the boards I hunted for some tin plating solution I knew I had somewhere.....

This certainly helps when soldering to the boards and it stops the copper contamination.

So with 4 boards etched I call it a day... I will drill them tomorrow. Oh how I will enjoy that....

(As a side note can anyone recommend a pcb prototyping company in the UK ?)

End of day 1

Image of boards after using the tin plating solution

[arvidb]

(As a side note can anyone recommend a pcb prototyping company in the UK ?)

http://www.pcb-pool.com/ppuk/info.html

Arvid

[BobLWeiss]

I must say they came out really good. I am also in the process of making these boards although I wanted some more bells and whistles so I redesigned the layout but still using the same firmware and basic design he layed out. I am interested to see how they perform when you get that far. What size motors are you using for this? What size power supply?

[MrBean]

I'm also building these boards for my router. I'm trying to, "isolation mill" the tracks on the boards using my CNC. Still some work to do on them to get them usable. You can check out the progress at:

http://www.terry-is.f2s.com/


Regards Terry.....

[BobLWeiss]

Well I started etching the driver boards and they came out really well. I used the "toner transfer" method to make them which has to be the best way to make a board I have ever come across. (that is until I get my CNC working :boxing: )



And then I tinned them. I made all 3 drivers boards and the breakout board.



Thanks Alan for the design!

[santiniuk]

I should know better !

Things never go to plan. Somewhere in my workshop is a large collection of pcb drills. The problem is I searched high and low and didn't find them. I ordered a few mail order after a frustrating hour or two.

Unfortunately these have still not turned up. No progress at all.

I have found a few useful looking items at work. Linear rails and what looks like a sealed ball-screw assembly. I'll grab some pics and post shortly.

To answer a few questions posted.

Thanks for the link arvidb. Will check this out for future pcb's. I mailed one company who are quite local to me. They have quoted £12 approx per pcb. If anyone is interested I can pass on the details.


@ BobLWeiss, I'm afraid I have not really planned this project very well. At this stage I knew that the Picstep controller would be suitable, I'm more biased to electronics than mechanical construction so I thought, Get the controllers built as an incentive and then research the mechanical bits. I'm still at the research stage

The toner transfer method certainly looks like it produces good quality boards. I have never used this method but would like to understand it a bit more. Time to ask Google how it's done.

(If I make any more of the Picstep boards I will change a few pads as some are on the small side.)

I will be interested how your construction goes. Maybe you can keep me on the right track !

@MrBean, Really enjoyed visiting your site. I have often wanted to do pcb's in the method you are using. Good to see someone else in the UK. (North East here).

Thanks all,

(Alan if your reading any of this. Your artwork still shows v2.0 for the V2.1 boards. - I thought I had etched the wrong ones !)

[BobLWeiss]

You can find more about the toner transfer method by going to the yahoo-group for homemade pcb's and also here:

Toner Transfer Method

I am also waiting on bits to arrive so I can drill out those boards and begin contrustion of the drivers. I have most of the machine finsihed and can post pictures later on when I get the camera back..(wife...hehe)

[Garfield2]

(Alan if your reading any of this. Your artwork still shows v2.0 for the V2.1 boards. - I thought I had etched the wrong ones !)

Doh! I'll fix that now. I've also got to upload my latest firmware that fixes a small bug. It's nothing major and almost completely cosmetic but it should be fixed (I brain faded and used a call instead of a goto in the main loop, so it just overflowing the stack each iteration. No harm done).

I'm glad you guys have built these boards with such furvor. It's impressive! Let me know if you have any problems or if you need help. PICStep has been throughly tested, but you never know what happens in different situations.

Thanks,
Alan.

[santiniuk]

I have just spent a good 10 minutes on the site you posted BobLWeiss, quite impressive !

This is certainly a method I will investigate in future. Can you comment on how easy / successful it is ? From the images you posted it looks great.

It sounds like your well ahead of me if your machine is built. I look forward to seeing the pics. (And grab some ideas...)

@Alan,

Good to see the master is watching ! Thanks again. Will post any queries here.

Cheers

[BobLWeiss]

Alen, what kind of speed (rpm's) are you getting with your design? I built one as a test and I cannot get it to move fast at all. I changed the source to make it run continuoulsy so I can test my hardware and I had to put a delay loop in between each step (40ms). With this delay its painfully slow. It could be something on my end that is wrong but I just wanted to see what speed you were able to get by using the computer to generate steps (TurboCNC, etc). Thanks!

[BobLWeiss]

Can you comment on how easy / successful it is ? From the images you posted it looks great.

To me it can't be any easier, after printing it out (I use Staples Picture Paper and haven't tried anything else), I press a hot iron on it for about 7 minutes and move it around to make sure I get every inch. After that I let it cool for about 5 minutes then place it in warm water with some dish soap in it. I let it sit for about 15 minutes and peel the first layer of paper off. Then soak some more and do this until you get 90% of the paper off (total of 30-45 minutes). I then use an eraser on a number 2 lead pencil and rub off the remaining bits (make sure you do this in the water and not dry).

Sometimes it rubs off a little piece of toner so you will need to look at it with a magnifying glass and touch up the spot(s) with a fine pointed Sharpie. Now your ready to etch away. I can get as small as 10mil traces with 10mil spacing between them. I haven't tried any smaller yet but I'm sure if your careful it can be done.

Bob

[Garfield2]

Andy, what kind of speed (rpm's) are you getting with your design? I built one as a test and I cannot get it to move fast at all. I changed the source to make it run continuoulsy so I can test my hardware and I had to put a delay loop in between each step (40ms). With this delay its painfully slow. It could be something on my end that is wrong but I just wanted to see what speed you were able to get by using the computer to generate steps (TurboCNC, etc). Thanks!

Andy or Alan?

Umm I have a video on my website of the PICStep running a motor via TurboCNC on a crappy Celeron 300. I guess the motor was turning at about 1000 - 2000 rpm.

That's if you're talking about PICStep that is..

I'm confused.

[BobLWeiss]

I ment Alen, sorry. Yeah I saw that video..so it must be my mistake somewhere. I am anxious to get one of them completely built and try it out using TurboCNC.

[Garfield2]

I ment Alen, sorry.

Closer! It's Alan.

Yeah I saw that video..so it must be my mistake somewhere. I am anxious to get one of them completely built and try it out using TurboCNC.

Silly question, you using a 20MHz resonator/crystal setup? I ask as you mentioned you've modified the design a bit and I'm wondering if your clock is one hundred percent.

Cheers,
Alan.

[BobLWeiss]

Closer! It's Alan.

Ignore my stupidity...sorry again. I am bad with names, both in person and in type.

Silly question, you using a 20MHz resonator/crystal setup? I ask as you mentioned you've modified the design a bit and I'm wondering if your clock is one hundred percent.

Yes, I am using a 20Mhz crystal and its running 100%. I checked it with a frequency meter. I tried to speed it up by making the loop 30ms and it won't run at that interval. It starts to work at 40ms and slower. I am not going to worry too much about it unless it does the same thing on the new boards (which I didn't modify at all as you can see from the pictures I posted.) I am waiting on the resonators and drill bits so I can put those together then I can fully test it using the computer to drive it.

[MrBean]

Here's some piccys of my freshly CNC milled board. No holes yet, as it was just a test to see how the modified track layout would work. Looks like it should be usable.

Regards Terry.....

[Garfield2]

Yes, I am using a 20Mhz crystal and its running 100%. I checked it with a frequency meter. I tried to speed it up by making the loop 30ms and it won't run at that interval. It starts to work at 40ms and slower. I am not going to worry too much about it unless it does the same thing on the new boards (which I didn't modify at all as you can see from the pictures I posted.) I am waiting on the resonators and drill bits so I can put those together then I can fully test it using the computer to drive it.

Hmm weird, by my ruff calculations with a 16KHz pulse you'd get around 600rpm with a 1.8degree motor at 1/8th step.

16000 steps per second
1600 steps per 360 (360 / (1.8 / 8))

16000 / 1600 = 10 revs per second
10 * 60 seconds = 600rpm

I've had my motors up to at least that high.

Also by my calculations using the 200ns instruction time would mean one step per 0.0135 milliseconds with the current firmware. This could theoretically be achievable by the 20MHz PIC.

1000 ms / 0.0135 = 74074 steps per sec
1600 steps per 360 (360 / (1.8 / 8))

74074 / 1600 = 46 rev per second
46 * 60 seconds = 2760 rpm

So what does it do with a 40ms delay that's different from the 30ms? Are the motors singing like the current limiting is working? Also what drive current do you have set? Can you email me your source code for your driver so I can see if there isn't something else going on.

Cheers,
Alan.

[BobLWeiss]

I just emailed you the code. As far as the motors singing, all it did was step, step, step, step...sort of like 4 steps per second if that. I know something is not right as it should got alot faster than that. I used a 10K resistor for 2amps per phase and I used a 24vdc powersupply. If I changed it to 30ms it does absolutly nothing. No noise, nothing.