[Big S]

I think it is only an issue of whether someone is willing to take the time to do the mod, and then we would be able to use Mach3 with the standard breakout board. I think there are a couple of other issues too, Like the lack of a charge pump to stop errant signals getting to the machine when Mach3 is booting. That is easy to solve by always making sure that Mach3 is loaded before powering up the CNC. The other thing that I am not sure about is the way the signals from the DSP are sent to the breakout, I don't think they should be any different from normal step and direction signals, but since there are 50 pins compared to the parallel ports 36 pins. There may just be a few more ground and + pins though.

I am all for the idea though and once we have a machine running on Mach3 it shouldn't be an issue to add another axis (and may even be possible with the stock breakout board)

Shannon.

[ksealey]

Shannon
Are the numbers in the circles on the dwg Pinouts?
Ken

[Big S]

Hi Ken.

I am not sure what the details on the dwg file are.

Shannon.

[Kent_Norway]

Shannon
Are the numbers in the circles on the dwg Pinouts?
Ken

Hi Ken

You're idea has been discussed between me and Shannon but the J1 pinout has of course not been shown. However you're find is excellent and I do believe you're observation is correct. I would try to get my hands on a oscilloscope and probe the breakoutboard on the machine while the DSP is connected.

It would be the best and easiest way of using the machine, where we could use the machine both with the DSP and Mach when that is needed. It could be that even a Scsi HW switch could be connected where we just switch between the setups

I have been reluctant to buy a separate breakout board since it seems like there is a lot of them and I feel not sure about any help if I get stuck.

Here is you're file by the way in PDF for view for them without Illustrator.

I'll update as soon as I manage to get my hands on a scope.

K

[Kent_Norway]

I did these shots last time I was home of the DSP "naked".
As we clearly see not all pins are in use so a big overhead in respect of pins is for sure....

K

[Kent_Norway]

Here is some more pictures

K

[ksealey]

Hi Kent

Unfortunately I have no practical electronic experience that I can contribute. I am happy to see that you appear able to add some more pieces to the puzzle and maybe even solve it.

I have asked Taney for the pinouts but have not received them yet. He believed that the Mach 3 couldn't be connected through the DSP board and I have asked for the reasons why but have not heard yet.

Are the numbers in the circles on the dwg pinouts?

Ken

[Kent_Norway]

Hi Kent

Unfortunately I have no practical electronic experience that I can contribute. I am happy to see that you appear able to add some more pieces to the puzzle and maybe even solve it.

I have asked Taney for the pinouts but have not received them yet. He believed that the Mach 3 couldn't be connected through the DSP board and I have asked for the reasons why but have not heard yet.

Are the numbers in the circles on the dwg pinouts?

Ken

Ken


I have asked Taney about the same many times and what I ended up with was a handrawn pice of paper with indicated some of the connections on the BB which is in use. No pinout of J1.

They did sell MACH3 setups earlier so I asked how they did this but no information has been obtained in this respect, so I think we are left to our self to figure it out.

To answer you if the numbers are pins I'll have to do a thorough check and as mentioned most likely need to probe some of the pins while the DSP is connected. This could be a "hairy" affair since the Oscilloscope could add noise and load the lines so it might make some errors in the operation.

I would try to look at it when I manga to get my hands on a some equipment. If I was at work I would simply grab a oscilloscope and try, but I'm at my home town so I guess I have to contact my teacher who teach me electronics when I was 17-18 years

K

[Matty Zee]

As we clearly see not all pins are in use so a big overhead in respect of pins is for sure....

while on the face of it i'd have to agree, bear in mind that is likely to be a multi layer pcb and connections to the mid planes won't be obvious...

[Big S]

I would say there should be about 11 or so really important ones to run the typical Excitech CNC.

They should be as follows
X dir, X step,X +5V, Y dir, y step, Y +5V, Z dir, Z step, Z +5V (gives 9 pins)
X home in, Y home in, Z home in, (for 12pins)
Then the other output is the spindle outputs (for 14 pins)
There will also be a + and - supply for the DSP (makes about 16 pins)

I have a hobbyCNC board with steppers running off a laptop beside me and that has very few pins active (6 pins for 3 steppers to be precise)

I would think that the DSP just uses a double sided board (costs lots more for multilayer boards and also harder to design. In the previous photos some of the traces that go to pins are not connected.

One of these days I will have to open my DSP and find out what all the pins are

Shannon.

[Kent_Norway]

I do agree in respect of multi layered board.

At work I have been looking at these types and the amount of circuitry is just so much bigger than the one on my handheld device. But that do however not exclude the possibility of multi layers.

Shannon;

The easiest would be to measure inside the CNC machine "engine compartment" where you could easily get in contact with the different pins.
Even easier would be to knock in a SCSI 1:1 connector which could be opened. I'm not sure if these exists for SCSI but we have those at work for RS232 and other serial connections.

I'll do a search later.

K

[Matty Zee]

Is it possible to see pics of the Connection board? From the doco posted earlier it seems like that is a far simpler board and afterall that is the one you'll need to connect a PC to for Mach3. I should be relatively easy to trace the pins on the board to their respective pins on J1.

[Big S]

Here's a couple.

Shannon.

[Matty Zee]

Great, thanks Shannon. It is too much trouble to get a close up of the under side of the connector end (need to see more detail of the J1 pinout and the vias around the two ICs (U1 & U2)

[Big S]

Ok I have the DSP apart in front of me, The circuit board used sure is a multilayer board. I am thinking that you can power it from the USB cable and then check the pin outs while jogging an axis etc.

Hope the pic is clear enough, its the best I have

Shannon.

[Matty Zee]

Its seems most of the important Step and Direction signals disappear into U1 & U2. I can't make out the part number from the photos (ands its likely the writing on the ICs will be mostly date codes anyway...) and without the info on the ICs i don't know which pins they come out of which then lead to J1.
This is what i can tell so far
+5VDC seems to go to pins 14 & 37
Gnd is pin 50 (there may be more though)
Pin 18 seems to be the Cutter Adjustment Device (CAD) input
Pin 32, X axis home input
Pin 33, Z axis home input
Pin 19, Y axis home input

[Big S]

I think U1 and U2 are MC3488 or MC3486, Its still very hard to tell from the full size photos I have though.

Shannon.

[Kent_Norway]

Ok I have the DSP apart in front of me, The circuit board used sure is a multilayer board. I am thinking that you can power it from the USB cable and then check the pin outs while jogging an axis etc.

Hope the pic is clear enough, its the best I have

Shannon.

Shannon.

The board has tracks on top and the bottom of the card which in one way is a multilayer card. However real "Multilayer" mean that there is stacked surfaces of tracks inbetween top and bottom which is a very expensive process to produce.

I have a sneaking feeling that there is just one track on top and one track layer on the bottom which makes us able to track each "track". But this do not help us to much, since we don't know the ICs functions.

I have requested a oscilloscope today and it would arrive next week.

But keep the good work going and I would update when I have some news in my end.

Kent

[Big S]

I was meaning that the DSP circuitry is most definitely a multilayer as some of the traces disappear inside the PCB and don't come out directly opposite, the breakout board may well be just a dual layer board though.

Shannon.

[Kent_Norway]

I was meaning that the DSP circuitry is most definitely a multilayer as some of the traces disappear inside the PCB and don't come out directly opposite, the breakout board may well be just a dual layer board though.

Shannon.

That's for sure a indication that there is a stacked multi layer board. If the via's are not going straight through, that's for sure a multi layer board.

Hmm, I guess there could be connections to the connector internally then...

I'll update when I have done some probing on the BB inside the machine or the DSP internals.

K