[LeeWay]

At this price, I think it would be great to have that extra third drive as a spare. That way you would never have any trouble at all with the other two.

[Mariss Freimanis]

I think at this point it will have to be 4 axis only. The mechanical integrity of the package depends on all 4 drives being present. Otherwise the mobo will bend or sag at the missing drive location when a DB connector is mated with the unit. FR4 fiberglass is not particularly rigid.

The idea for this assembly method is to minimize the labor component. I am looking at some ideas for eliminating the eight 2-56 screws from the assembly as well. These ideas cannot be tested until I have a prototype unit built up.

On another note, we have had very gratifying experimental results trying a new manufacturing technique called through-hole reflow. This is where the surface mount solder paste stencil has openings for through-hole component pads. Bottom mounted through-hole parts are inserted and fixed in place with UV-set epoxy and the panel is screened. The pick-and-place machine places the SMT components as usual, then the PCB panel is populated with top mounted through-hole components afterwards. The whole mess is run through the reflow oven which makes solder joints for both the surface mount and through-hole components.

The resulting through-hole solder joints are superior to hand-soldered and solder-potted joints. Bad hand-soldering technique can result in cold joints and is labor intensive. Solder-potting requires good technique but still has a tendency for solder-bridging and excessive thermal stressing of the assembly. Through-hole reflow soldering neatly addresses these issues. I will post some photos of the results using this new method next week.

Mariss

[keebler303]

Mariss

Great work so far. Maybe you could add some dummy spacers made of delrin or similar. It would be a shame to have 1 or 2 drives not being used just for structural support. Sell the spacers with the unit for a few bucks and people will be able to use 2 or 3 axes without spending the extra money on unneeded drives, after all, this is a cheap drive!

Matt

[wolfdagon]

Wow. Great work Marriss. I can't wait to see how much the kit will cost and when it will be available. I am still in the planning phase of building my router, and trying to get all of my parts together. I had originally planned on trying to use Pminmo's L297-8 drives, but I was not looking forward to putting them together since I am not that good at soldering. These new boards look like just what I need. I will keep an eye on this thread for a release date.

[Mariss Freimanis]

keebler33,

Good idea but.. "Barbie Dolling" out a product is a killer. Some would want it with 2 drives, some with 3 a few with only one. It becomes a logistical and stocking nightmare. Many, many thousands of these will be built in the first year; a smooth running production line means stamping out identical copies, putting the finished product in stock and shipping. Think of production stock as a reservoir or flywheel, its function is to match pulsing demand with pulsing supply. It evens out or smooths supply to demand. Once in stock, modification requires the finished product to be disassembled, customized, then reassembled again. This undoes two production steps of labor and makes it more expensive.

Things get very different when building 10,000 of anything compared to building 100 copies.:-)

Mariss

[keebler303]

Mariss

How about an assemble it yourself model with the case/mobo and drives separate. Then you could plug in however many drives you wanted and there would still be no wiring to do. I know what you are saying about logistics but I think you would be shooting yourself in the foot by "demodularizing" a product which is intended to be modular. I think the "cheap" will be less apparent when $60 worth of "spare" drives have to be purchased to retrofit your lathe. If you don't want to mess with it, I understand your reasoning. Maybe a distributor could sell these units customized for the needed number of axes while the bare mobo and drives would come from gecko.

Matt

[phomann]

The pictures:

Mariss

Hi Mariss,

Just a couple of comments from the pictures of the layout of the motherboard . With it as shown, it may be a little difficult to mount the module efficiently into a CNC enclosure.

Ideally the DB-9 and DB-25 connectors could be accessible externally from the CNC enclosure. You could almost do it with the current layout by mounting it on a right angle plate so the connectors protrude through the rear of a cnc controller enclosure. The problem with the current layout is the screw connector block.

It you changed the connector block to a 'green' 3.5/3.81mm pluggable connector block then they could also exit the rear panel also.

That said, it would be desirable for the connector block to be accessible from the inside of the cnc enclosure though.

Is it possible to have R/A DB connectors mounted along rear edge? That way they could protrude through the rear panel of a CNC enclosure, white the screw terminals are accessible from the inside. The db-25 could also be substituted with an idc-26 connector, that way it could be left where it is now.

Cheers,

Peter.

[tauntdesigns]

On the subject of 2 and 3 drive set-ups:
If they sell the cheap drive by them self?...... Then
I'm sure there will be bob makers who will design a bob set-up for those who only want 2 or 3 drives.

On the connectors:
I'm thinking of building an enclosure that houses the computer and drives/mobo inside. Power supplies.... inside or outside?.... not sure yet.
Such a enclosure may fix any problems with the connectors

Thinking out loud,
Jack

[phomann]

Hi Mariss,

Just a couple of comments from the pictures of the layout of the motherboard . With it as shown, it may be a little difficult to mount the module efficiently into a CNC enclosure.

Ideally the DB-9 and DB-25 connectors could be accessible externally from the CNC enclosure. You could almost do it with the current layout by mounting it on a right angle plate so the connectors protrude through the rear of a cnc controller enclosure. The problem with the current layout is the screw connector block.

It you changed the connector block to a 'green' 3.5/3.81mm pluggable connector block then they could also exit the rear panel also.

That said, it would be desirable for the connector block to be accessible from the inside of the cnc enclosure though.

Is it possible to have R/A DB connectors mounted along rear edge? That way they could protrude through the rear panel of a CNC enclosure, white the screw terminals are accessible from the inside. The db-25 could also be substituted with an idc-26 connector, that way it could be left where it is now.

Cheers,

Peter.

Looking at pictures 3 and 4, you could replace the connector block with the 'green' 3.5/3.81mm pluggable connector block, but mounted on the underside of the motherboard.

That way it would allow the assembly to be flush mounted onto the rear of the cnc enclosure panel with all the db connectors exposed, and the terminal block would be accessible from inside the CNC enclosure.

The down side is that it may be more difficult to solder in the connector block.

Cheers,

Peter.

[acondit]

Hi Mariss,

Just a couple of comments from the pictures of the layout of the motherboard . With it as shown, it may be a little difficult to mount the module efficiently into a CNC enclosure.

Ideally the DB-9 and DB-25 connectors could be accessible externally from the CNC enclosure. You could almost do it with the current layout by mounting it on a right angle plate so the connectors protrude through the rear of a cnc controller enclosure. The problem with the current layout is the screw connector block.

It you changed the connector block to a 'green' 3.5/3.81mm pluggable connector block then they could also exit the rear panel also.

That said, it would be desirable for the connector block to be accessible from the inside of the cnc enclosure though.

Is it possible to have R/A DB connectors mounted along rear edge? That way they could protrude through the rear panel of a CNC enclosure, while the screw terminals are accessible from the inside. The db-25 could also be substituted with an idc-26 connector, that way it could be left where it is now.

Cheers,

Peter.

Mariss,

I am partially in agreement with Peter. I would like to see the DB9's (and DB25) on a different face from additional screw connections, but I don't really see a problem with the DB25 on the same face as the DB9's.

Alan

[Mariss Freimanis]

Peter,

You are a mind reader. that connector and its location is something I haven't been happy with either. Panel mounting the thing occurred to me as soon as the the first rendering appeared on my monitor. I will be trying different treatments to see how they come out.

Mariss

[Mr.Chips]

Peter,

You are a mind reader. that connector and its location is something I haven't been happy with either. Panel mounting the thing occurred to me as soon as the the first rendering appeared on my monitor. I will be trying different treatments to see how they come out.

Mariss

How about just replacing the DB 9 and 25 with IDC 10 and 26. And the customer would add jumpers of the necessary to the cabnet side. This would give the customer mounting flexability for convience and fan cooling.
Being internal the connection would not necessarly need a mech locking device like screws and it might even be less expensive for you.
Hager

[acondit]

How about just replacing the DB 9 and 25 with IDC 10 and 26. And the customer would add jumpers of the necessary to the cabnet side. This would give the customer mounting flexability for convience and fan cooling.
Being internal the connection would not necessarly need a mech locking device like screws and it might even be less expensive for you.
Hager

I could be wrong, but the ribbon cable for the power connections doesn't seem to me to be a good idea. One, it is parallel runs, two it is not shielded and three the ribbon cable tends to be very small wires.

Alan

[ger21]

How about just replacing the DB 9 and 25 with IDC 10 and 26. And the customer would add jumpers of the necessary to the cabnet side. This would give the customer mounting flexability for convience and fan cooling.
Being internal the connection would not necessarly need a mech locking device like screws and it might even be less expensive for you.
Hager

The DB 25 will let you plug it right into a parrallel port. Putting IDC connectors would make it more difficult for most people.

And I believe Mariss has said that no fan is necessary.

[Mr.Chips]

The DB 25 will let you plug it right into a parrallel port. Putting IDC connectors would make it more difficult for most people.

And I believe Mariss has said that no fan is necessary.

My thought was to get is away from the side of the cabnet and the little jumpers would have enabled that, but I can see the possible electrical prob with a parallel jumper.
But having the DB 9 and 25 facing the same direction they could all come through the side of the cabnet and have very direct connections. That would just leave the screw connectors to address in this type configuration.

[CoAMarcus]

Mariss

How about an assemble it yourself model with the case/mobo and drives separate. Then you could plug in however many drives you wanted and there would still be no wiring to do. I know what you are saying about logistics but I think you would be shooting yourself in the foot by "demodularizing" a product which is intended to be modular. I think the "cheap" will be less apparent when $60 worth of "spare" drives have to be purchased to retrofit your lathe. If you don't want to mess with it, I understand your reasoning. Maybe a distributor could sell these units customized for the needed number of axes while the bare mobo and drives would come from gecko.

Matt

We are trying to keep the drive as simple as possible, and that means assembling it all here. If we do it all here, we all have control over the quality of the build and the absence of foreign particles. If we start shipping them out as "kits" as you suggest, then the quality control becomes a major issue.

We are looking for simplicity on the supply side. Unfortunately the kit idea does not help us out in the long run, because we would have a higher rate of failed drives do to people putting them together incorrectly.

-Marcus Freimanis

[LeeWay]

Thats what I was thinking too. I make blade guards and they are all built to order. Your choice of color, model, port size and acessories. I am not fast enough making them to actually be able to build up any stock, but I make so many different models that it really wouldn't be possible anyway without selling a bunch of kits. I would rather put them all together here and know everything will fit and is in the box for that particular model and order. A little different, but handling a bunch of kit parts is more costly and time consuming than having everything in one build to ship.

When I look back over all the threads on just this forum where someone blew a drive and had to wait on a replacement or someone wanted to or needed to add another motor or axis, I think the 4 banger Gecko setup will be a hit. (Dang! What a run on sentence.)

[CoAMarcus]

Exactly LeeWay, people should look at it as having a spare drive built in. If you blow a drive and have an open one, all you have to do is move the DB9 connector over to the unused drive (after you have solved the problem that blew up the drive to begin with of course). I am afraid if we do the kit idea that it would turn out like a do-it-yourself RC airplane: Someone will put the wings on upside down, then blame the manufacturer for why it won't fly.

Aside from that, the main reason we don't want to do the kit idea is because of the inability of testing the whole package. If we did the kit, we would have to assemble it, test it, disassemble it, and ship it out to be assembled again. That is a lot of wasted work on our part and on the customer's part.

-Marcus Freimanis

[Mariss Freimanis]

OK, I redid the packaging entirely. I now have a 12 position Phoenix-style 3.81 mm connector plug sliding in from the side. The new packaging allows for panel mounting in a control box with the DB connectors protruding through the control box wall while the Phoenix style connector stays entirely within the control box.

The people on CNCzone is what makes it great. It was their thoughtful observations and suggestions that caused these changes to be made. I'm really pleased with the way it looks now and I will post the model renderings this evening.

Marcus has been made the project manager for this control and it will be his responsibility to shepherd it forward from this point on through to production.

Mariss

[phomann]

Hi Mariss,

I'm keen to hear how you are getting on with this technique. I've been playing around with it for a couple of months now. It has reduced my hand soldering from 15-20 minutes a panel to about 1-2 minutes (Just cleaning up a few suspect joints)

I'm using a pneumatic paste dispenser for applying paste to the through holes (Using a stencil first for the smt parts). The hardest part is getting the amount of paste right.

I'm not confident yet to get a stencil made with the oversize apertures for the through hole components yet.

There is a quite a bit of info on the net. I've been working off these papers ;
http://www.speedlinetechnologies.com/docs/1371.pdf
http://www.speedlinetechnologies.com...P_SMTA2004.pdf



Cheers,

Peter.

On another note, we have had very gratifying experimental results trying a new manufacturing technique called through-hole reflow. This is where the surface mount solder paste stencil has openings for through-hole component pads. Bottom mounted through-hole parts are inserted and fixed in place with

Mariss