[Kipper]

Those are small indeed! Small but perfectly formed I hasten to add!

[erniebro]

The G540 looks great!

I know it's probably too late in the design process, but it sure would be nice to have a version of this with a USB interface similar to what Warp 9 are doing- http://www.warp9td.com/

New pc's don't include a parallel port anymore...

[CoAMarcus]

erniebro,

That would require a microcomputer inside, raising the price drastically. Step and direction loses all timing information when sent over USB; it comes down to either spending an extra $150 on each board, or buying a $10 parallel card.

-Marcus Freimanis

[the4thseal]

what is the purpose of this ?Smooth Stepper? is it a controller or a driver?

[skmetal7]

its a 4 axis stepper driver, like the xylotex, but a lot better

looks like it would work well with the SmoothStepper

[ger21]

what is the purpose of this ?Smooth Stepper? is it a controller or a driver?

.
The smooth stepper emulates two parallel ports with a USB connection, and also generates step and direction signals to 4 Mhz. It comes with a Mach3 plugin and is said to work identically to a parallel port, and allow much faster speeds. The G540 can be plugged directly into a SmoothStepper and run using USB, but they are two very different products.

[the4thseal]

so we could use this with the drive in this thread?

[ger21]

Any step/direction drive will run on the SmoothStepper. It acts just like two parallel ports.

[phomann]

Hi,

Is there a mechanical drawing for the G540 that provides dimensions and locations of mounting holes, connectors, etc?

I seem to recall that it was done but I cannot find it.

BTW Good to see its using 3.5mm pluggable connectors.


Cheers,


Peter.

[Mr.Chips]

Any step/direction drive will run on the SmoothStepper. It acts just like two parallel ports.

Two parallel ports? Do you mean it has 50 pin output????

Need some translation here.

[ger21]

It has two parallel port connectors on it.

[phomann]

Two parallel ports? Do you mean it has 50 pin output????

Need some translation here.

No, it connects to 2 db-25 connectors, that are used in place of 2 parallel port cards in a pc. Have a look at the accessories picture at http://www.warp9td.com/

BTW, this is probably the wrong forum to discuss this.

Cheers,


Peter.

[CoAMarcus]

Hi,

Is there a mechanical drawing for the G540 that provides dimensions and locations of mounting holes, connectors, etc?

I seem to recall that it was done but I cannot find it.

BTW Good to see its using 3.5mm pluggable connectors.


Cheers,


Peter.

Peter,

The exact dimensions are pretty much refined, but not quite finalized. We are still working on prices of parts, and as such may need to redesign the layout just a little bit to compensate for different size parts. We should have the footprint some time fairly soon.

I am meeting with some suppliers throughout this week, and that should finish with what parts we will use. From there, we will know exactly where we should put the screw holes.


Also, the 2x15 connectors will be socketable, but there will not be 3.5mm connectors on the G540. The G251 will have the little 3.5mm connectors, but they will not be socketed.

-Marcus Freimanis

[phomann]

Peter,

The exact dimensions are pretty much refined, but not quite finalized. We are still working on prices of parts, and as such may need to redesign the layout just a little bit to compensate for different size parts. We should have the footprint some time fairly soon.

I am meeting with some suppliers throughout this week, and that should finish with what parts we will use. From there, we will know exactly where we should put the screw holes.


Also, the 2x15 connectors will be socketable, but there will not be 3.5mm connectors on the G540. The G251 will have the little 3.5mm connectors, but they will not be socketed.

-Marcus Freimanis

Hi Marcus,

Thanks for the reply, I can wait. I thought the 12 pin black terminal block was 3.5.. spacing. From what you've said and re-looking at the photos, I'm guessing they are 5.00mm pitch?

Cheers,

Peter.

[Mariss Freimanis]

Peter,

I'm finalizing the sheetmetal drawings this evening for fabrication of the G540 enclosure. By this time tomorrow I will have complete dimensioned drawings of it and I will post them here.

On a note of interest for technical geek types: While we are awaiting the G250 panels to arrive Thursday, we ran some final thru-hole reflow production tests using TO-220 MOSFETs. These of course are much bigger than the TO-251 MOSFETs that will be used in the G250 and are more of a challenge thermally.

The G250 price absolutely depends on having no post-reflow labor content. This means the power MOSFETs located on the bottom of the G250 board must be soldered in place using the same process that solders the SMT components on the top side of the board. The process is as follows:

1) Dispense thermoset glue dots on the bottom of the panel for the MOSFETs.
2) Place the thru-hole MOSFETs onto the glue dots (96 per panel).
3) Press the dotted MOSFETs against a flat-plate to insure co-planarity.
4) Run the dotted panels thru a conveyor oven at 150C to cure the epoxy.
5) Stencil the top side with solder paste for SMT and thru-hole components.
6) Run the panels thru the pick-and-place machines as usual.
7) Reflow the panels to complete the assembly process.
8) Wash, dry and inspect the panels.
9) Gang JTAG the CPLDs.
10) De-score the panels into individual drives.
11) Test the drives and pass on to shipping.

The tests run today and the results (see thumbnails) apply to steps (1) thru (5) except no stencil was used during solder paste squeegee. The lack of a stencil left the top side of the board solder-starved and accounts for the concave solder meniscus evident in the top side picture. The bottom side picture shows excellent solder meniscus on the MOSFET pins.

The test board was introduced into today's production reflow of G203s. The oven profile was set to our standard RoHS top only SMT component temperature profile. No accommodations were made for the bottom MOSFET thermal mass of the test panel. The oven is our brand new (2 weeks old), $23K, RoHS capable, 38kW splurging new capital equipment addition. I'm very impressed it did such a good job without any profile adjustment. It bodes well once it's profiled for G250/G251s.

Mariss

[phomann]

Hi Mariss,

Thanks for sharing. I was wondering how you were going to apply the paste via a stencil when the FET leads were poking through the holes. From your picture, the legs are short enough to not protude through the hole. Smart.


I've played around a bit with using paste for through hole connectors. I've been using a pneumatic paste syringe to apply the paste. Quite time consuming but still a lot faster than hand soldering.

What I do is;
1. use a stencil for the SMT components,
2. use the pick and place machine for the SMT assembly.
3. Reflow the board in the oven.
4. Once cooled, apply paste for the THC,
5. then hand place the THC stuff
6. then, back through the oven.

The process works quite well. Initially I was skipping step 3, but it was too slow placing the THC without bumping the SMT stuff.

I haven't got around to using a stencil for the THC as I'm not sure as to what size pad apertures I need for the various component pads. At $240 per stencil, I don't want to do too many iterations.

There were only 2 pics, not 5.

Cheers,


Peter.

[keebler303]

Peter

It was steps 1-5, not pictures.

Matt

[Mariss Freimanis]

Update: The G250 and G251 panels arrived this morning. The stencils arrive tomorrow. Today will be spent programming the Juki pick-and-place machines with G250/G251 component data (tedious hand-entered data keyboard work) and a test placement will be run on a board taped with clear double-sided adhesive tape to verify correct parts placement. A first panel proof will be run-off, reflowed, washed, JTAGed and tested tomorrow or Friday (April 17 or 18).

Mariss

[CoAMarcus]

Hi Mariss,


I haven't got around to using a stencil for the THC as I'm not sure as to what size pad apertures I need for the various component pads. At $240 per stencil, I don't want to do too many iterations.

What we are using is JB-Weld. This is an easy mixing epoxy, strong as anything you can find, cures at 150F, and is good up to 600F. Just put a spot on the board, stick the MOSFET down on it, run it through the oven at 150F, and you are good to go. Make sure the leads aren't sticking above the panel and you can just stencil it like normal.

As I type this the machines are being set up to run the G250s. We get the stencil tomorrow, and if everything looks good we will run limited production some time very soon. As promised, there will be lots of pictures. ;-)

-Marcus Freimanis

[jalessi]

Marcus,

I am curious, will every 250 or 251 you produce require using J-B WELD epoxy?

If so how do you mix and dispense it?

Jeff Alessi