Round 4. Fight! (2010-10-12)


I don't know if this has already been discussed, but would it be a good idea to make nubs that would go in a connector / socket instead of soldering them directly on the PCB? This way they would be easilly replaceable in case of failure and wouldn't require an RMA.
It has indeed come up - check page three of this very thread. ;)
 
Last edited by a moderator:
I don't know if this has already been discussed, but would it be a good idea to make nubs that would go in a connector / socket instead of soldering them directly on the PCB? This way they would be easilly replaceable in case of failure and wouldn't require an RMA.

Something similar has been suggested by other people, here. I am sure it is something they will look at in the future.
 
Hah, that's fine. :) It's one of the pitfalls of ending up at the top of a new page!
 
Thanks ED. It's nice to hear that you are still listening to your buyers and are still trying to improve the Pandora. This is why myself (and I'm sure quite a few others) still haven't lost interest in the Pandora after 2 or more years.
 
The one concern I have is that production will take so long that the Pandora will be running on long out-dated hardware. Sure, we'll still be able to do all the fun stuff that the hardware was designed for a couple of years ago, but as a gadget it has to be able to compete with other gadgets on the market. Paying £300 now and getting the device 1-2 years from now when the Nintendo 3DS / PSP2 or some other awesome and revolutionary device has been around for a while for half the Pandora price, would suck.


Hopefully, the Pandora 2 will be faster to go from design to finished product... to achieve this, 3 pointers to keep in mind.


1, Save some money from sales so you don't have to be dependent on pre-orders(or get to work while there still are some funds).


Several reasons, one being that unannounced long term pre-orders hurt the company image, in other words, consumer trust plummets with time and with that goes your future success. Another reason is financial, having the funding from the get go will create freedom, both in production and setting an appropriate price tag for later when the device is ready for sale(yes, the Pandora is a little "pricey" and I suspect that you set the price high because of unforeseen expenses).


2, Keep good contact with manufacturers, keep the one's that "behaves" in the current production, scout for new one's before hand so that the options are ready when production is ready to start.


3, Learn from experience, what works, what doesn't work, what can be done differently, what is more efficient, what reduces waiting time, what reduces costs, what boosts quality, in what parts are quality essential, anticipate problems and have back up plans ready... etc


Going from design to sale in short time while having a quality product is crucial for the company, right now, the problem seems to be marketing("2 Months" and "Before X-Mas Guarantees") and some minor design issues(nubs). Shipping seems to be top notch(all 900 finished units shipped). I'm not saying you are doing a crappy job, this project is a massive undertaking in organizing and your role in the production line(assembly and shipping) seems admirable, but there are room for improvements in organization.
 
The original 2008 estimate was massively overconfident, which was why I didn't order back then, and I was worried that the team would get despondent when the reality turned out to be harder - but they've kept at it, so I'm confident now that my order will get delivered. I'm cynical now about almost anything the team predicts, but I still wouldn't expect batch 2 to take another year now, let alone two.


Regarding your three step plan, from what ED and others have said on here I'm happy that they are learning lessons from this project and making contacts. Lets see how the iControlPad works out, since the manufacturing process for that is pretty similar to the Pandora - case, nubs, hardware, software.
 
Regarding your three step plan, from what ED and others have said on here I'm happy that they are learning lessons from this project and making contacts. Lets see how the iControlPad works out, since the manufacturing process for that is pretty similar to the Pandora - case, nubs, hardware, software.

iCP should be much easier assembled.


Cases are already finished for months (thanks to the UK), parts are already there (except for nubs), software is nearly finished (and we don't need a complete OS here. It basically shows up as BT controller / keyboard).


Also, the boards do not have 12 layers with complicated stuff like the OMAP, they are way simpler, you could basically even solder them at home.


So except for waiting for the nubs, I don't expect many problems with the iCP.
 
The one concern I have is that production will take so long that the Pandora will be running on long out-dated hardware.
It already did. The Pandora hardware already IS outdated. The current generation of Cortex A8 processors are clocked twice as high.


Furthermore, Cortex A9 devices have already been announced and will likely come on the market next year.
 
It already did. The Pandora hardware already IS outdated. The current generation of Cortex A8 processors are clocked twice as high.


Furthermore, Cortex A9 devices have already been announced and will likely come on the market next year.
actually, A9 devices are on the market already - doesn't make the pandora obsolete though, and for what I use it for not even too slow. Battery life on my Pandora beats the AC100... Form factor is the key thing.
 
actually, A9 devices are on the market already - doesn't make the pandora obsolete though, and for what I use it for not even too slow. Battery life on my Pandora beats the AC100... Form factor is the key thing.
It's not obsolete or too slow yet, definitely, but the lifetime I would have expected out of mine is shrinking. 2 years ago, I was thinking it would last for 4 or 5 years before I'd be looking for an upgrade. Now, I'm thinking it'll only be 2 or 3 years before I'll be needing something more powerful.
 
actually, A9 devices are on the market already - doesn't make the pandora obsolete though, and for what I use it for not even too slow. Battery life on my Pandora beats the AC100... Form factor is the key thing.
That's what I actually thought, but because I couldn't reproduce any link I decided to weaken my statement a bit.


It's not a big deal as we're going to want to run our Pandoras as slow as possible anyway (to save energy). However, a dual core Cortex A9 would enable playback of high-def video and emulation of systems which are impossible now.


So yes, if OPT would release an OMAP4 Pandora next year then I would actually order one. It never hurts to have a second unit. I kinda hope they will start the development of the Pandora 2 ASAP such that we can have an upgrade in 2012. Nevertheless, even if they don't, the Pandora is WAY more than just an OMAP3 processor and I will certainly enjoy it in its current configuration.
 
I kinda hope they will start the development of the Pandora 2 ASAP such that we can have an upgrade in 2012.
I kinda thought we are all getting an upgrade of sorts in 2012? Or the permanent downgrade?


Bring on the NUBS!


Peace & Pandora,


Link
 
It already did. The Pandora hardware already IS outdated. The current generation of Cortex A8 processors are clocked twice as high.


Furthermore, Cortex A9 devices have already been announced and will likely come on the market next year.

Can you please reference a currently SHIPPING product that is using a 1200Mhz A8 processor? I'm not talking about stuff that has been announced and is not yet shipping, but is actually shipping right now. I would be happy to be proven wrong, but I'm pretty certain that nothing is shipping with your claimed double clocking of the Pandora. And it seems the Pandora can be easily clocked to 800Mhz which brings it on par to most shipping A8 products out there right now. (And good luck trying to over-clock most of these locked down platforms so they won't be able to get a boost in performance like the Pandora can.)


The A9 right now is rare as can be right now and won't be common at least for a year or maybe even two years. The A8 still has a long life ahead of it as it is a well known commodity that will drive a lot of devices more than adequately.


It is so easy to say the Pandora is outdated when you look at stuff coming down the pipe. Heck the Blackberry Playbook specs (1.0 GHZ dual core CPU) look nice but that won't launch until the first quarter of next year. And lets not forget the upcoming phones the Pandora gets compared to are heavily subsidized by the cell phone companies. Buying any of these products without contract is very, very expensive expensive even compared to the Pandora. And the upcoming Samsung tablet looks wildly expensive too.


But compared to shipping products the Pandora still looks pretty good IMHO.
 
The 3DS will likely NOT be that much more powerful than the Pandora, unless Nintendo make it massively expensive - which is unlikely as Nintendo has been the 'cheaper' company in the gaming sector. And if it is more powerful, battery life will suck ;) . In my personal, honest opinion, nothing is ever obsolete. Ever. A Commodore 64 is still useful and can be used, even when a $0.50 PIC *can* likely replace it, it doesn't mean it has to :) (it's just much much much more efficient to do so - but you don't HAVE to, bleh, does anyone get what i'm saying?). The *challenge* of producing software for a device that is called by some to be obsolete is what good programming is all about (not my cup of tea, I need a 1337-skillz module installed in my brain first). Then your 'obsolete' hardware becomes 'obsoLEET'.


I think socketed nubs would be a bit risky, as without good, solid connections they can slip out of sockets - it'd have to be a socket like on some motherboards for heatsinks (or like a ZIF socket) that has a lever that holds them down. Or maybe those (annoying) little pin-like things that go through the board like a hooked arrow, that doesn't come out unless you squeeze the plastic spring-thing(offtopic: thing on a spring!) down. Both of which are currently difficult to fit in the Pandora (but the OPt has :ph34r: -skills). Naturally you couldn't use connectors like standard DIP sockets, yuck :) . Socketed nubs are very convenient tho, at the very least for the OPt, as opening the Pandora would void your warranty (unless you guys can think of an epic way to do it without opening up?)
 
Bear in mind the nubs are already attached to the PCBs by a number (8?) of holes through the PCB which are then soldered. A connector with the same number of points, positioned at places around the edge, should be almost as resistant to shear forces as just mounting it straight on the board. Biggest problem I see is finding a serious connector that's slim enough not to add much thickness to the Pandora shell.
 
Can you please reference a currently SHIPPING product that is using a 1200Mhz A8 processor?
The iPad run at 1Ghz. I thought that the Pandora was clocked at 500Mhz despite the fact that they list 600Mhz on the main page.

And it seems the Pandora can be easily clocked to 800Mhz which brings it on par to most shipping A8 products out there right now. (And good luck trying to over-clock most of these locked down platforms so they won't be able to get a boost in performance like the Pandora can.)
This is not a fair hardware comparison unless those other processors don't overclock at all. The fact that the Pandora is superior in its (open) software platform is obvious. Don't bring software into a hardware comparison.

But compared to shipping products the Pandora still looks pretty good IMHO.
I'm not arguing that it doesn't. I preordered a Pandora, not any of those other devices. Heck, I wouldn't want an iPad if someone gave it to me for free. The Pandora was awesome when I ordered it and it will still be awesome next year.


The poster I replied to argued that the Pandora hardware might get outdated soon. I merely made the point that the Pandoras processor has already been overtaken.
 
The iPad run at 1Ghz. I thought that the Pandora was clocked at 500Mhz despite the fact that they list 600Mhz on the main page.
It runs a default 500Mhz at stock voltage level OPP3. TI has rated the OMAP the Pandora has to clock at 600Mhz in "overdrive" OPP5, with the caveat that you'll get reduced longevity and battery life.


Of course, most people can clock to 800Mhz at stock voltage, and up to 1Ghz in overdrive, but that's besides the point.
 
In any case, the iPad CPU isn't an omap3 Cortex-A8, it's a different CPU. It makes little sense to compare frequencies.
 
Last edited by a moderator:
Back
Top