Of cases and batteries


Hey ED, are my numbers more or less correct? Going from a 20% failure rate originally to below 5% after making some changes for the CC and 1Ghz Pandoras? I can't remember how I know those numbers, they're just in my head and I'd like to be sure I'm not just making them up.
Not sure - I didn't follow the exact failure rate, but the original cables had a lot more than 20% (more like 70% within the first year).
 
Will the Pyra's shoulder buttons (both the switch on the board and the physical button on the case) be any more durable than the Pandora's?  Given that both of my Pandora's shoulder buttons are broken due to impacts from carrying it around daily, I'm a bit nervous seeing the buttons wrap around the edge again.

Minor concerns aside though, I'm glad to see the Pyra coming together, and I can't wait to buy one. Thanks for your continued work on this ED.
 
I remember noting on earlier shots of the peripheral board that the shoulder buttons looked to be better anchored to the PCB than they were on the Pandora. Also, I forget if I saw this or imagined it, but I can recollect a render of the plastics that seemed to show a hard stop to the shoulder built in, unlike the Pandora's design where the shoulder can easily run under the case and having done so is only resisted by the resilience of the mechanical switch.
 
I guess I have a problem owning a device "knowing" (or at least heaving the impression) it has a design flaw in it, and instead of doing something to remove the cause only the symptoms are treated...
I've got some unfortunate news for you: everything you own has a flaw. Every product ever produced in mass quantities has a greater than zero failure rate. The best anyone can do is design around the flaws and hope for the best. There is not a product in your house that isn't a ticking timebomb with some (sometimes extremely low) probability that it will fail: electronics wear out; batteries inflate; oven safe glass spontaneously shatter when heated; Pandora cables break.....

I still see the redundancy as strictly better.
As I wrote before, I know that everything will break at some point, and that is not what I am getting at: I still think that the whole design of having a display connection like that is a failure in itself. And I still think that having a redudancy may lead to bigger problems on a company (and indirectly customer level) due to that it may help in covering up that the faulty design principle is killing the cables. And the fact that the Pyra will be made in batches instead of having a constant, but low numbered steady production run (I guess that is the case with every electronic product out there where the factory line does not exist for the sole purpose of producing only the one product), may complicate the process even more. Think of the old cables, and how "long" it took them to fail. Depending on how well the Pyra sells, there could be several production runs already made and items shipped before the cable gets faulty and the source is identified.You may not agree with me - which is totally fine for me as I guess the only one with specialized experience in that field around here is ED himself. And as he explained in his previous post (thanks a bunch for that ED), he actually put a lot of thought into it, and did not approach that with "the last cable worked fine, so we are fine" attitude.

Having redundancy can be a fine thing, but as long as you haven't got a tried and tested product (which will happen with the Pyra as ED has no money to make case stress test that would create valuable results) it may not be so beneficial afterall.

Alternative bad option 2:  Flat exposed cable.

IBM & Lenovo have used this in many of their notebook computers.  The connector is a flat black ribbon cable ~40mm wide that comes up from a slot in the base and into a slot in the display frame.  The cable bends up to 180 degrees every time the unit is opened/closed.  The primary disadvantage is that this is an exposed cable - subject to wear and tear - and it's ugly on close inspection.
Show me one article/press release/whatever where a particular product of IBM or Lenovo had a failure rate that wasn't negligible in that particular case. About the "ugly" part: who cares ?
And lets not forget another option, doing it DELL style, by putting everything in one cylindrical cable, that bends with the lid - again no big fuzz (I know of) that there is a design flaw.

Both options seem a lot more reasonable to me, than having a cable that is constanly rubbing on itself and the case - at least as long as ED does not come around saying that the only way the hinge can be constructed is bascially the same as the Pandoras due to costs, size restrictions, whatever.

Well, an analyzation why Pandora cables fail would cost more than 10,000 EUR. It's not easy analyzing cables with such thin layers - and find out why exactly they break. That needs very high-tech equipment.
Having worked only in (small) companies till now where the owner wants to be proud of the product he produces, I guess I am a little naive in thinking that the manufacturer has its own motivation finding out why his product fails.
 
I'm perfectly fine if you let me know a better solution for the LCD cable.


The way laptops do it (a bunch of small copper cables) won't work simply because the hinge is too small.


Exposing the cable like the early G&W games or some laptops did would be a possibility - however, it doesn't look good and I'm not happy having the cable exposed. If the Pyra falls down on the ground and you're unlucky that there's a stone of small object, it could damage the cable right away.


Same with keys in a bag or pocket...


All other options (contact pads, wireless, etc.) won't really work with the resolution we got. That's the advantage of Nintendos 3DS: A lower resolution is a lot more sturdy.


About the shoulder buttons:


The Pandora also was planned to have the plastic stop the shoulder buttons.


This didn't work out for two reasons:


1. Weak plastic (that's why it gets stuck in the case)


2. Shoulder buttons not 100% accurate on the PCB, as they're surface soldered (which also means they're less robust to force)


The Pyra case should be a lot better regarding 1. (I'll test that) and the shoulder buttons are through-hole soldered, so they are exactly in the right spot and a lot more durable as their legs go through the PCB and aren't sitting on it.
 
Exposing the cable like the early G&W games or some laptops did would be a possibility - however, it doesn't look good
Do we have to deal with that as valid reason nowadays ? If this is really a show stopper for you (either for personal reasons, or you simply want some market appeal for your product) then there is no need to waste any other keypress on this topic.
and I'm not happy having the cable exposed. If the Pyra falls down on the ground and you're unlucky that there's a stone of small object, it could damage the cable right away.

Same with keys in a bag or pocket...
How much will the cable really be exposed to the outside ? Having the cable only go through the inner part of the lid cover should keep the cable covered enough (imho). Having the Pyra falling to the ground landing exactly on something very thin penetrating the rather small gap between displaylid and the case also happening exactly in the region that is occupied by the cable sounds extremely unlikely (especially with the Pyras center of mass probably turning the case towards the under or uppserside before hitting the ground).Keys in pockets or bags might be some sort of issue. While the likelyhood of the afformentioned sequence of events (but moving instead of falling)is surely higher, I'm having trouble to believe that people

- would do that for a long time as the keys would probably create a lot of damage to the case

- could not be held responsible for how they treat their Pyra.
 
Last edited by a moderator:
The way laptops do it (a bunch of small copper cables) won't work simply because the hinge is too small.
The point of mentioning IBM's old laptops is that they are using the same type of foil cable as the Pandora, they simply put it straight upwards to the screen. In the T60 it is being sandwiched by two black pieces of a more sturdy plastic foil that press it into bending up by about 90° (from the board towards the hinge), most of the dynamic bending applied by opening and closing the lid is happening at the point where the cable leaves these plastic foils to disappear into the screen - up to 90° in either direction. With the lower count of traces required this might actually be very feasible with the Pyra.
 
You may not agree with me - which is totally fine for me as I guess the only one with specialized experience in that field around here is ED himself
Yeah, that's fine, I was just making sure that we fully understood each other. Nothing bothers me more in a debate than not understanding what the other person is talking about. Ok, knowing what the other person is talking about and knowing it is complete BS with no way of convincing the person otherwise bothers me more, but that's not the case here, I mean like religious wingnuts, anti-vaccers, and windmill paranoia.
 
The way laptops do it (a bunch of small copper cables) won't work simply because the hinge is too small.
The point of mentioning IBM's old laptops is that they are using the same type of foil cable as the Pandora, they simply put it straight upwards to the screen. In the T60 it is being sandwiched by two black pieces of a more sturdy plastic foil that press it into bending up by about 90° (from the board towards the hinge), most of the dynamic bending applied by opening and closing the lid is happening at the point where the cable leaves these plastic foils to disappear into the screen - up to 90° in either direction. With the lower count of traces required this might actually be very feasible with the Pyra.
 Yes, that works if you've got enough space to fold the cable. Which we don't have :(  

Exposing the cable like the early G&W games or some laptops did would be a possibility - however, it doesn't look good
Do we have to deal with that as valid reason nowadays ? If this is really a show stopper for you (either for personal reasons, or you simply want some market appeal for your product) then there is no need to waste any other keypress on this topic.
No, the main reason is the one below. 

and I'm not happy having the cable exposed. If the Pyra falls down on the ground and you're unlucky that there's a stone of small object, it could damage the cable right away.

Same with keys in a bag or pocket...
How much will the cable really be exposed to the outside ? Having the cable only go through the inner part of the lid cover should keep the cable covered enough (imho).
It would be quite a bit, because the lid need to be opened and closed.

If the lid was fixed, it would just be a tiny bit - but in order to be able to open and close it, you need the full 180° range open, so that's quite a hole there.

Additionally, do you really think people would be more happy with a device that can break when they carry it around in their pocket than with a device that COULD potentially break after a few years of usage?
 
Additionally, do you really think people would be more happy with a device that can break when they carry it around in their pocket than with a device that COULD potentially break after a few years of usage?
 

As long as I can get my unit repaired, I would be happy with the second option.  I trust you to bring us all a fine product, ED. 

-Glyph Reader
 
By the way, does the Pyra LCD cable also needs shielding like the one in the Pandora (and caused additional trouble)?
 
Last edited by a moderator:
By the way, does the Pyra LCD cable also needs shielding like the one in the Pandora (and caused additional trouble)?
Oh, that's actually a really good question. The Pyra is going to have more radios in it than the Pandora did, it's going to need more testing. Hopefully the cable doesn't prove to be a problem. They figured out and fixed that problem fairly quickly as I recall, which is good because the metal tape they originally had to use was probably the leading cause of cable breakage.
 
The shielding was not because of the radios but because of the immensive signals you need for RGB.

With MIPI this is easier and we can use some traces for normal shielding, no extra shield layer needed.
 
The Pyra case should be a lot better regarding 1. (I'll test that) and the shoulder buttons are through-hole soldered, so they are exactly in the right spot and a lot more durable as their legs go through the PCB and aren't sitting on it.
I'm glad to hear this is being worked on and improved.  Thanks for the response ED.
 
This is coming along so much smoother than the OpenPandora did.  I'm looking forward to getting the pyra, and can't wait, but definitely want to wait to make sure things are done right.  Looks like we have at least a few months of work left though  :(
 
I myself am quite interested in buying a prototype. (see here : http://boards.openpandora.org/topic/17697-of-boards-and-displays/page-2#entry372118/URL])

But ED, what are the actual differences between the proto and the final units ?

The only thing I know of is that it will be 3D printed instead of molded. Does that mean it won't be as tough ?

Could there be a reduced lifespan or cosmetic changes ?

I'm totally ready for a prototype pre-order, but only if I know what I'm paying for, you see ? :lol:

If there's nothing more than paying for the on-demand case printing, then my wallet is all yours !

Edit : also, while we wait for official Pyra ports of emulators and stuff, could we use the complete Pandora software lineup ?
Any answer about these ?
 
Last edited by a moderator:
Since the 3D printed cases will just come as-is, while the final plastics will have been tested and selected for strength, I'd expect the 3d printed cases to be weaker. Whether that means limited lifetimes depends on how you treat it therefore, and you could always buy a retail case once they become available I guess.


In terms of running Pandora software, there's been plenty of discussion about that, but I've not seen anyone actively working on it at present. I tend to think it shouldn't be rocket science to set up a chrooted pandora env to get past the softfloat issue, although that still leaves scope for there being trouble finding a 3d driver that works in that env with the physical chip the pyra has, and ditto wifi and potentially any other chips in there. There's good odds it will arrive quickly therefore, but smaller odds there'll be lots of trouble and it'll take a while to come out IMHO.
 
Back
Top