Are you interested in a new Pandora LCD cable?

[comradekingu]

I never said the case files were libre software, or open hardware.

I never questioned your right to do what you want.
Your ability to do what you want is what you could use to your benefit. It isn't about how good the format is, but what sort of
opportunity it presents to others coming after.

From my point of view, the involvement in the first type of device to be made like this, is the greater point to make, in achieving
yet more groundbreaking firsts. As your company seems to only be involved in the libre ecosystem, the advertisement of being
paid to do things in a more open manner seemed like a good venue of opportunity to me. The libre software world is
fortunate that Kicad exists, and it is the perfect match for the pandora now, taking into account the Pandora predates that easily.

I didnt ask you to use Kicad _on_ the pandora, seems I caught you on a bad day.

 

[levi]

I'm not sure what your asking for. The schematics as released were in the forms of gerber data (and a courtesy pdf for the curious), and it's that data that's apparently going to the fabricators to make. Asking for more work just seems likely to persuade hns to give up on doing this for an ungrateful community.

 

[hns]

fortunate that Kicad exists, and it is the perfect match for the pandora now, taking into account the Pandora predates that easily.

I just do not understand what KiCad has to do with retrofit LCD cable production for the Pandora. And I am not sure what it is a perfect match for. It is one tool within several alternatives (BTW: there are other fully libre alternatives) and each one must match a given task and not a device.

 

[Wally]

Don't worry about Kingu, he's often lost in his own little fantasy world on IRC

 

[comradekingu]

Often at times I am, you will see me use the nickname "Wally" when it gets really bad.
Just making cables isnt the whole picture. Being able to utilize files for a discontinued device for crowdsourced production of spare parts
is something that adds itself to a long line of novel and open feats of the pandora. Using Kicad to do it would pull an additional rabbit out of the hat.
It also has to do with what sorts of companies the effort can align itself with.
ED found some displaycables when cleaning his shop, but I don't know how many.

 

[hns]

Good news. The new cable samples are shipped and are announced to arrive next Tuesday. Then we will see how they look like...

 

[levi]

Just in case you didn't notice, ED discovered some new old stock LCD cables hanging around his labs during a recent clear out. That might affect the number you can sell if everyone's satisfied by that stock, but I don't know to be fair how many he's found, nor how many actually want one. But just a heads up in any case.

 

[crazyhorse]

Is ED selling them yet?

 

[levi]

I don't think so; he said he needs to do some tests on them before he can stock them. I assume they'll be stocked here when they become available.

 

[hns]

Just in case you didn't notice, ED discovered some new old stock LCD cables hanging around his labs during a recent clear out. That might affect the number you can sell if everyone's satisfied by that stock, but I don't know to be fair how many he's found, nor how many actually want one. But just a heads up in any case.

Yes, he told me the quantities.

 

[Farox]

I'm still interested in this new cables. Thanks @hns

 

[hns]

Ok, LCD cable samples have arrived!
They look quite good, except that the stiffner at the display end seems to be a little thinner than ordered, but that is not a deal breaker for evaluating the prototypes.
I am still waiting for some components (capacitors, diode, connector for the panel) and then I can try to solder cables.
Then we can start to test them. Well, it will not be possible to test durability, i.e. how long the cable survives until it breaks. Important is that it works basically.
On the fotos there is one sample of the older cable design for comparison.

 

[gunrock]

No photos for me, in your post @hns

 

[Linux-SWAT]

No photos for me, in your post @hns

It's because it's a fake news :^)

 

[levi]

No photos for me, in your post @hns

Yeah, not sure what's up with that. We're seeing the alt text of the images only, presumably because if I dig out the image urls and visit them myself the server reports 'oops'. I'm just not sure how hns managed to post like that. Unfortunately 'oops' doesn't really tell me whether it's a permissions issue of simply a systematic typo or something.

 

[hns]

Well, I have no problem seeing them in Safari. But Firefox fails.
So here they are uploaded again:

Post automatically merged: Apr 7, 2020

I did go to "Edit" both versions and switched to BB mode.
The first one:

[IMG alt="DSC01081.jpg"]https://pyra-handheld.com/boards/attachments/dsc01081-jpg.35949/[/IMG]

the second one which works in firefox:

[ATTACH type="full" alt="DSC01081.jpg"]35957[/ATTACH]

 

[hns]

I finally found time to setup my pick&place machine (which needed some tweaks because the PCB data and placement plan is for a bottom side coordinate system and needs to be mirrored).

But then I was able to add the capacitors to 3 of the 5 cables we have had produced. This worked fine and doesn't need much time.
I also manually added the 60 pin LCD connector before doing reflow soldering, but that makes trouble as you can see on the second foto:

The pins are not well connected at one side and are up in the air.

The reason seems to be that the PCB is too flexible and warps during the reflow soldering process. If it warps in a convex shape the connector is only soldered in the middle (or one side). If it is concave it is only soldered at the ends but not in the middle.

So I have to find a way to keep the thin PCB really plane and flat during soldering. I think about using double-side adhesive (Kapton) or forming a paper clip to press down the flex PCB in the middle and at the ends.
I still have 2 untouched sample cables so there is room for experimentation before we need to have built another set of samples...

A minor issue is that the solder paste printing stencil is a little too big which seems to be the fault of the Gerber data having too long solder pads for the connector. This sometimes ends up in solder bridges between adjacent pins. It can be solved manually with a soldering iron but takes too much time for mass production and should be avoided right from the beginning. So I have created my own stencil data and a new stencil has been ordered.

Generally the PCBs look fine and have some electrical contact (I have tested 3 or 4 wires). So the choice of the manufacturer was perfect and we are getting close to a good end result...

 

[Linux-SWAT]

Impressive !

 

[Farox]

Great.
Thanks a lot for your work.

 

[Phosphate]

Thanks hns!