Did I Bork My Nubs?

[mali]

^ From my observations it seems just the mouse mode is buggy, joystick mode works fine(see Mupen64+). Looks pretty much like a software fault, else the nubs would lose calibration in Mupen64+, too.

 

[EvilDragon]

It's a mixture of hardware and software, I think.
If N64 works, but Mouse mode acts a bit, it's a thing to be optimized in software.

Additionally, some nubs are borked. The manufacturer therefore tests EVERY SINGLE NUB currently to make sure we won't get anymore broken ones (or, at least, not that many).

Yep, those are the early problems we have - but we're working on getting that better.

 

[sebt3]

Like i said before ,this issue the op has does not reflect the general condition most nubs have (BUT) most nubs do not seem to calibrate properly at all

I don't agree.
I have my pandora for 3 weeks now. I use my pandora daily and its nubs too (mostly mouse mode). I've yet to encounter any problem with nubs.
I doubt to be alone. In this forum, I've seen about 10-20 faulty nubs sent back to OPT and about the same amount of "one nub-er".
There have been about 700 pandoras shipped.
So by my count, we have around 1350 working nubs and about 50 brokens. I cant see how this is most....

 

[Bosbeetle]

Like i said before ,this issue the op has does not reflect the general condition most nubs have (BUT) most nubs do not seem to calibrate properly at all

I don't agree.
I have my pandora for 3 weeks now. I use my pandora daily and its nubs too (mostly mouse mode). I've yet to encounter any problem with nubs.
I doubt to be alone. In this forum, I've seen about 10-20 faulty nubs sent back to OPT and about the same amount of "one nub-er".
There have been about 700 pandoras shipped.
So by my count, we have around 1350 working nubs and about 50 brokens. I cant see how this is most....

My nub worked fine before it borked so I agree with sebt3

 

[paddy]

well sorry to say this but can any you guys show your anologs working fully on desktop mode via a video ? i would like to see you do a few full circles with both nubs and then acurately move the cursor around the desktop.

Fact is i have had nothing but trouble getting anolog controls to work properly and if my third unit is the exact same as the first two then i must be extremely unlucky ,and if i am right about the software as i think it most prolly is a software issue then it shuld be looked into further becase these nubs could very well be working fine but the software is bad and users could be damaging those anolog controls in the process of trying all manor of calibration methods only to weaken the life of the nubs.

 

[WizardStan]

Like i said before ,this issue the op has does not reflect the general condition most nubs have (BUT) most nubs do not seem to calibrate properly at all

I don't agree.
I have my pandora for 3 weeks now. I use my pandora daily and its nubs too (mostly mouse mode). I've yet to encounter any problem with nubs.
I doubt to be alone. In this forum, I've seen about 10-20 faulty nubs sent back to OPT and about the same amount of "one nub-er".
There have been about 700 pandoras shipped.
So by my count, we have around 1350 working nubs and about 50 brokens. I cant see how this is most....

There's probably 2 or 3 people not part of the community that have a problem, and would simply return it without saying anything. For argument sake, let's say 5.
So that's (a hopefully pessimistic) estimate of 250 broken to 1150 good nubs; a disturbingly large number, but still not nearly "most".

 

[sebt3]

well sorry to say this but can any you guys show your anologs working fully on desktop mode via a video ? i would like to see you do a few full circles with both nubs and then acurately move the cursor around the desktop.

I've no video equipment, you'll have to believe my word. For the sake of it, I've just tryed what you said after a fresh boot *without* calibration (I've never done this anyway...), played a bit and then retried. I still have no problem moving my cursor acurately.
But I've lower the mouse speed with the OPT nub configurator since the first 5 minuts I've my P. (was too fast for me)

Fact is i have had nothing but trouble getting anolog controls to work properly and if my third unit is the exact same as the first two then i must be extremely unlucky ,and if i am right about the software as i think it most prolly is a software issue then it shuld be looked into further becase these nubs could very well be working fine but the software is bad and users could be damaging those anolog controls in the process of trying all manor of calibration methods only to weaken the life of the nubs.

You're not the only one unlucky with nubs (Prometheus is also waiting her 3rd P because of nub problems). But I realy think this come down to luck.
Sorry for you


EDIT :

So by my count, we have around 1350 working nubs and about 50 brokens. I cant see how this is most....

There's probably 2 or 3 people not part of the community that have a problem, and would simply return it without saying anything. For argument sake, let's say 5.
So that's (a hopefully pessimistic) estimate of 250 broken to 1150 good nubs; a disturbingly large number, but still not nearly "most".

How does adding 5 silently returned P. can grow my 50 broken count to 250 ?
I think your counting is realy pessimistic (but may be mine is too optimistic too)

 

[Exophase]

I gotta be honest here, I don't have a 100% confidence that the nubs are gonna make it in the long run. I think way more than 10-20 people have expressed problems and the number keeps increasing, I see no reason why new cases won't keep developing as people have their Pandoras longer.

Yeah there are probably some outright defects the manufacturer can test for but we're seeing a lot of early mortality with use and descriptions that sound more characteristic of design flaws than defects. I don't think the manufacturer is going to reproduce the conditions necessary to weed out problems like "eventually stops returning to center." This is with nubs that don't seem to be very rigidly autocentering to begin with (my experience and that of most people I've seen). Frankly, with the frequency of problems we're seeing, I don't think that the problems are at all a one off thing, although maybe some will be luckier than others.

A lot have praised the nubs but I'm concerned.. even if they're all released with the highest level of screening quality possible I think we're going to be suffering with some big flaws that won't be fixed. Namely the large deadzone, loose return to center, no resistance to spinning, and questionable calibration procedure locked in firmware. Maybe I'm just complaining excessively, but a large part of me hopes that some other nubs become an option at some point and the ones being used get replaced altogether. This is going to become an especially big deal if the 3DS nub is as good as it looks; yeah you can't really tell these things by watching people slide it alone but I dunno, I just have a much better feeling about it.

 

[Bosbeetle]

Yeah it is a bit worrying the nub is essentially a little piece of rubber I dont know if it can handle a lot of 'mousing'

 

[Prometheus]

I don't think the manufacturer is going to reproduce the conditions necessary to weed out problems like "eventually stops returning to center."

I must admit that I am a bit concerned about this myself. It took about a week to manifest in my first Pandora, and about two in my second one.

(And I'm not sure that it would be reasonable to expect a manufacturer of anything to spend one to two weeks, or more, thoroughly testing each individual item that comes off of their production lines.)

 

[paddy]

Just now got my 3rd pandora so i will check the anolog nubs out once the battery charges up ,but i am already pretty sure those nubs will be the exact same as my last two units in calibration loss ,im not happy at this not just for myself but for OPT and everyone else as i don't want to see the pandora fail in anyway but if these nubs are a total flaw by design then i may as well sell my unit and move on as i wanted a gaming handheld device not a small linux umpc style device with limited software, useless nubs are a deal breaker for me and this is why i would love to hear that it is in fact a software issue but i don't feel 100% about that ,i will also try waverace and mario64 as there are tell tale signs of wonky calibration tho on mario64 it's not as apparent.

Also shouldn't quake 3 be ok with the nubs ? i tried that on my first unit and it was terrible and i read others had simular problems with quake 3.

If those nubs are a failing then i would be looking to wait off untill OPT put new ones in its place possibly from another company then after reading good reviews on them i would want to buy a pandora
but not before ,i know this is what you get for buying into such a custom new system from a small company but main games controls are one of the biggest selling points of the pandora and it's the reason i preordered so long ago and waited for an eternity.


just telling it like it is for me.

ps ,i agree with Fusion_Power yep Ed and craig should kick their ass for such low quality ,those nubs simply mean a full unit needs to be replaced and that messes up the entire unit ,not good enough ,opt don't need this on top of trying to mass produce these units only to get them returned within a month.

 

[fusion_power]

It's a mixture of hardware and software, I think.
If N64 works, but Mouse mode acts a bit, it's a thing to be optimized in software.

Additionally, some nubs are borked. The manufacturer therefore tests EVERY SINGLE NUB currently to make sure we won't get anymore broken ones (or, at least, not that many).

Yep, those are the early problems we have - but we're working on getting that better.

thats are around 8000 Nubs to test? I hope that more than one person does this job.
And I guess this don't solve the Quality problem of the nubs. After all that readings here about self-broken nubs I guess they are such a poor quality, that they eventualy break after some usage. So even if the shipped Nubs are OK, they seems to break after some time at the user.
Time to check the general design of these Nubs, OPT ordered High Quality Stuff so they deserve High Quality stuff and not the last crap of Hardware. Re-Desin should be no longer a taboo, of course the Manufacturer has to pay for these extra costs.

I want Hardware that works even in 10 Years as good as at Day one (Except the Battery of course). And because Craig always call the Pandora a quality Product I guess I can expect this.

 

[WizardStan]

How does adding 5 silently returned P. can grow my 50 broken count to 250 ?
I think your counting is realy pessimistic (but may be mine is too optimistic too)

I'm pretty sure I used the word "pessimistic" more than once

Yeah there are probably some outright defects the manufacturer can test for but we're seeing a lot of early mortality with use and descriptions that sound more characteristic of design flaws than defects.

At the same time, we're seeing prototypes that have been working for months without these eventual failures. I played with Jeffs Pandora prototype months after he'd already been using it and they were fluid and perfect. EvilDragon has also said his prototype is still working wonderfully.
At the same time, we've really only heard from a subset of users. I wonder how many people have had their Pandora for a month and are not speaking up about how well it works.
You're right, it seems like something that can't really be tested until after the fact, but the fact that it happens to some and not others sticks in my mind to say that there must be some difference between the ones that die and the ones that work fine. Maybe the manufacturer knows something they can test for: a certain "squishiness" to the rubber, variables in the conductivity, who knows.

 

[Prometheus]

I quite agree about the prototypes. Evidently the company's gotten it right before.

Personally, I'm just putting it down to mass-manufacturing teething troubles, and the pitfalls of early adoption.

 

[Exophase]

At the same time, we're seeing prototypes that have been working for months without these eventual failures. I played with Jeffs Pandora prototype months after he'd already been using it and they were fluid and perfect. EvilDragon has also said his prototype is still working wonderfully.

Are these from a different batch of nubs? Not that it doesn't discount the "design flaw" concern. I guess it's also a question of how much they were used, how they were used, and just what criteria exist for "smooth and perfect."

At the same time, we've really only heard from a subset of users. I wonder how many people have had their Pandora for a month and are not speaking up about how well it works.

I wonder how many people have had their Pandora for a month and are not speaking up about how disappointed they are with the nubs. I wonder how many people don't want to rain negativity in on a product they've waited a seriously long time for. I wonder how many people are in outright denial over this. I especially wonder how many people are unwilling to send back a unit that could take a long time to get back to them (mine took two weeks to get to me in the first place) and might even be held up due to replacement part shortages or god knows what.

Basic fact is that we can only really go by what is actually being said and not give credit to a pleased silent majority that we have no evidence really exists. This talk about "luck" confounds me; multiple people getting multiple bad units in a row says something about basic statistics and talking about bad luck is outright ignoring a real problem. Something that at the very least is probably characteristic of the current batch of nubs out.

You're right, it seems like something that can't really be tested until after the fact, but the fact that it happens to some and not others sticks in my mind to say that there must be some difference between the ones that die and the ones that work fine. Maybe the manufacturer knows something they can test for: a certain "squishiness" to the rubber, variables in the conductivity, who knows.

I don't think the manufacturer has such indepth knowledge over such subtle qualities, because that suggests the kind of experience that isn't suggested by a company that up until now hasn't been looking for it. What I think is that they're tightening the tolerances on their acceptance plan for static things and that it might not actually improve these other sorts of things, just the DOA nubs.

Either way, a lot of people are reporting being unhappy with nubs that OP themselves tested, so either they're going bad later or OP's acceptance criteria are not consistent with user acceptance criteria. Either one is a problem.

 

[paddy]

Didn't you think of the nubs from the protypes as company test parts ,those parts would obviously work perfectly otherwise OPT would have gone elsewhere,the mass produced parts are the main example not company test parts that must get checked fully before they are sent out to possible customers ordering them in the thousands.

 

[mali]

This talk about "luck" confounds me; multiple people getting multiple bad units in a row says something about basic statistics and talking about bad luck is outright ignoring a real problem. Something that at the very least is probably characteristic of the current batch of nubs out.

This might be related to "in house" testing on the "assembly line", which hasn't been that clean as I would expect. Ok it's rather kind to send some tobacco flakes, but what if the user is non-smoker? Personally, I expect the externally tested nubs to be of better quality. I also hope that scratched/broken cases go into the trash bin when the new cases arrive.

 

[EvilDragon]

At the same time, we're seeing prototypes that have been working for months without these eventual failures. I played with Jeffs Pandora prototype months after he'd already been using it and they were fluid and perfect. EvilDragon has also said his prototype is still working wonderfully.

Are these from a different batch of nubs? Not that it doesn't discount the "design flaw" concern. I guess it's also a question of how much they were used, how they were used, and just what criteria exist for "smooth and perfect."

Yes, they were from a different batch and are still working (after 7 months) like they were working from the very beginning.

As said, the manufacturer checked them and told us something has gone wrong with the manufacturing (a tolerance problem).

My guess is that even some nubs that are working at the beginning have a tolerance problem and will eventually fail. Some others fail from the very beginning while the good ones are reliable even after a long time.
As we cannot test the tolerance thing and therefore just test if they are working, there are some of the tested and working ones that still fail after a while.
That's why the manufacturer is now doing this, fixing what they can fix and reproduce the ones that are really bad with more accuracy.

As we have 13000 nubs (8000 for the Pandora, 5000 for the iControlpad) but the Pandora has a higher priority, those tests should go pretty fast. It will delay the shipping a while, but as we prepare everything else (lids, LCDs, bottom parts, etc.), all we need to do is stick the boards in and ship, once we get the next batch of boards, things will go way faster.

i.e. it takes about 17 - 18 minutes to build a single Pandora. About 10 - 12 minutes is the assembly of the lids, LCD cable, bottom part, etc. That's everything that can be done without having boards
As soon as the boards arrive, it's about 5 - 6 minutes left per unit... and way fewer units should faile because of broken nubs, so assembly will then be a lot faster and smoother.

 

[WizardStan]

As we have 13000 nubs (8000 for the Pandora, 5000 for the iControlpad) but the Pandora has a higher priority, those tests should go pretty fast.

Wait, you have 13000 already? As in, 13000 nubs produced with this poor manufacturing? As in approximately 13000 (or lets say only 25% failure, 3250) nubs effectively need to be thrown out and remade?
Or do you mean "have" as in "have ordered and just waiting for them to be made and sent"?
On the one hand, I had hoped that all nubs would have been produced by now. On the other hand, it'd be good if only a couple thousand had been produced so they would be easier to replace.

 

[EvilDragon]

As we have 13000 nubs (8000 for the Pandora, 5000 for the iControlpad) but the Pandora has a higher priority, those tests should go pretty fast.

Wait, you have 13000 already? As in, 13000 nubs produced with this poor manufacturing? As in approximately 13000 (or lets say only 25% failure, 3250) nubs effectively need to be thrown out and remade?
Or do you mean "have" as in "have ordered and just waiting for them to be made and sent"?
On the one hand, I had hoped that all nubs would have been produced by now. On the other hand, it'd be good if only a couple thousand had been produced so they would be easier to replace.

Yep, we had all 13.000 since a while
Well, yeah, it's good and bad
It's good because even if there are a lot of broken ones while testing, we'll still have enough for the first Pandora batch

Bad because they need to be thrown out and remade... it's not our costs, but still a bummer