New issue Rovio spins eternally

157 posts / 0 new
Last post
yellowboy63
yellowboy63's picture

Been having rovio a while now and finally ran into this problem, ive tried all firmware updates to date and never had this problem until now. After reading posts in this section i decided to check this probelem out my self. I opened rovio and slightly pushed on sensor on suspect wheel while making rovio spin in circles problem went away.That eliminated faulty sensor.Further investigation led me to 2 big issues with rovio. 1st is upon removing wheel from transmission i noticed a plastic bearing that is wearing out probably due to normal use and wear and tear. (Lipo battery can easliy help this.less weight)2nd issue is a design issue. If you remove any of rovio's wheels and look carefully at the upper wheel well you will see it is not attached to the rest of the shell which causes the wheels to flex up and down. In uneven terrain i think this can cause fatigue in low half of wheel well causing the counter wheel to become out of alignment with ir sensors. i bent wheel down slightly and epoxyed top middle of wheel well to the body, wheel doesnt move around anymore counter wheel stays in place problem solved. hope this helps some of you .

Mich
Mich's picture

With one sensor down and moved to the rear wheel things were fine for a while. Yesterday I noticed that Rovio is now slightly turning to the left during straight forward command. Is the right sensor going bad? With rovio openned, I confirmed with a voltmeter that two sensors are now bad. Voltages on the out put were zero on both defective sensors. They varied between 0 and 1 volt on the good one.
I decided to completely UNHOOK the three rotation sensors. Rovio had no problems. Now it goes forward and backwards like it should. HOWEVER the angular rotation buttons Definitely do not work. Until I can find some replacement rotation sensors, rovio will remain without any

cmilian
cmilian's picture

Your Rovio must have an incredible battery to spin eternally :)

ike
ike's picture

WowWee engineers are aware of the problem and are looking into it.

CanUSRovio
CanUSRovio's picture

My Rovio has developed the same problem. From reading this thread it seems as this is likely a hardware problem, not a software problem. Bummer if it is.

I submitted an online support problem ticket last week referencing the YouTube video that I think I found in this or another thread here.

Cheers,

Speculative

Leebo
Leebo's picture

ike said:
WowWee engineers are aware of the problem and are looking into it.

 

Smile

SCHiN101
SCHiN101's picture

Same issue for me with the constant turning

CanUSRovio
CanUSRovio's picture

Any updates Ike? Hardware or software problem? I'm hoping for a software fix.

Cheers

Usul
Usul's picture

It's definitely a hardware problem. The only firmware fix would be to ignore the sensors, which would be unwise. It might be neat if they gave us an option to disable the logic associated with the sensors in favor of a time delay, for such cases as ours. I'm holding out for replacement parts kits. It's very modular and easy to work on.

Mich
Mich's picture

Do not forget that you can just disconnect all three wheel sensors. The only issue with that is preset angular rotations. Driving forward in a straight line is no more a problem.
It is definitely a hardware problem confirmed with volatge measurements.

crony
crony's picture

Crap... I'm also experiencing this now.

Skibum
Skibum's picture

Same issue here....

milw
milw's picture

@Mich, did you find if it was the IR LED or the sensor?
Also I'm curious -crony and Skibum, and any others- what is your serial number? Can we identify a range of Rovios developing this problem?

Skibum
Skibum's picture

All I see is a MAC.
Where is the S/N located?

Thanks

BTW This ROVIO has about 30 minutes of total use. :(

milw
milw's picture

That definitely sucks. Look inside the battery compartment, mine is #26608WS, just bought in the past few weeks (I don't have the spinning issue - yet! knock on wood!).

Skibum
Skibum's picture

Spoke to WowWee about this spinning issue.
They assured me that they are aware of the issue and are working to resolve it.
The guy also said that at this point is does look like a hardware issue, but that was not confirmed.

He took my contact information and said that when a solution is available they will advise me and send the parts or software necessary for resolution.

That works for me. :)
(What more could they possibly do at this point?)

denodan
denodan's picture

The only problem with this, is if it is some other hardware causing the sensors to go faulty then the others will fail. Best disconnect them altogther.

MrScott
MrScott's picture

Unless you know something the rest of us do not, there is no confirmed failure mechanism for the scattered reports of malfunctioning wheel rotation sensors.

Without knowing why there are some failures, it is much too early for predictions that all will fail, or that leaving them plugged in will damage a Rovio that isn't exhibiting symptoms.

Making general statements like "they ALL will fail" is like Chicken Little claiming the sky is falling.

That statement is one potential outcome, but one of the least likely ones.

The symptom of a Rovio veering to the side over the length of a path is quite different than a Rovio that has a wheel that will not stop spinning. There are several other more likely failure modes that may be involved in creation of a curved path.

I base my statements on my 25 years of working in the semiconductor manufacturing industry. I know a fair amount of what's involved in functional specs, normal deviations, early life fails, and semiconductor reliability.

Lets talk facts, not fears.

MrScott
MrScott's picture

The two ways we'll get to the bottom of the failure mode is to either wait for WowWee tech support to tell us, or to make some observations on our own.

Does anybody here with a failing sensor have a bit of electronics debugging skill?
Can you tell us whether the transmitter is still emitting IR? If it's not, then the LED is not getting juice, or is blown. You should be able to tell if its emitting IR by viewing the transmitter through your digital camera or camcorder.

A multimeter would quickly tell whether the non-lit LED had a voltage across its leads. If it has a voltage, and it's not emitting, then it's a bad solder joint at the pin or a blown LED.

Replacement IR LEDs are common enough. Tack soldering a new one adjacent to the old one would tell whether a simple LED replacement is the fix.

If the LED is lit, then it's likely the sensor that's a problem. Testing the diode out of circuit would tell whether its failed as an open or short.

If there is not a voltage being applied to the emitter, then it's time to trace the wires looking for a bad solder joint or break. A cheap connector might be making intermittent contact.

crony
crony's picture

Skibum said:
Spoke to WowWee about this spinning issue.
They assured me that they are aware of the issue and are working to resolve it.
The guy also said that at this point is does look like a hardware issue, but that was not confirmed.
He took my contact information and said that when a solution is available they will advise me and send the parts or software necessary for resolution.
That works for me. :)
(What more could they possibly do at this point?)

It's good to know they're aware of the issue.  Makes me feel a bit better.

denodan
denodan's picture

I would advise disconnecting all Sensors, or at least, like I have done just unscrewed them. If it is something stuffing the sensors up then maybe they will all become bad.

There are a few here who have swapped sensors and find others going bad also, so something is causing sensor faults and possiabliy causing other good ones to go bad?

MrScott
MrScott's picture

Or, equally likely, a shipment of sub-par sensors made it to the assembly line, and a Rovio with one weak diode is more likely to have all of the diodes from the same suspect batch.

As a corollary, we had a flurry of reports of beacon LEDs burning out. There were a few cases reported all close together. Instructions were given for how to replace them at home. There was some concern that it would be a general problem for all Rovios.

I haven't heard of any other instances of dead beacons after that first two or three. That doesn't mean it's not happening. It may be that the folks aren't posting here. On the other hand, given how fast the bad news spreads, I'm more inclined to believe that the other beacons haven't failed.

Usul
Usul's picture

I'd say they have a significant quality control problem with the sensors. Probably not just one bad batch. My first Rovio was from Hammacher Schlemmer. My second, purchase two months later was from Wowee online. Now it has developed the same spinning problem. Not a big enough deal to warrant taking parts from one to fix the other, but definitely bad news for Wowee. I'd say that most, if not all, will eventually develop this issue. What are the odds that both of mine would?

MrScott
MrScott's picture

I don't know the odds, but I know that there are folks with older ones that haven't had the problem (yet).

denodan
denodan's picture

Just do what I have done and just unscrew the sensors, now mine runs stright, but the degree buttons don't work of course. Wowee should be either replacing the senors for good ones or recalling Rovios for replacements.

They do with other faulty products why not Rovio also? I think very few would not be having problems with sensors. Seems the problem does not develop stright away and the time it will happen to your Rovio maybe weeks or months, but many end up with Sensor problems.

Just disconnecting them at the moment is the best option till Wowee recalls Rovio back to be fixed properly or gives us the parts.

For us who bought outside the US or UK we don't have warrenties, so Wowee should be suppling parts to us.

I found disconnecting the Sensors Roivo still docks, but the very odd docking problem. The thing I notice without the sensors is when Rovio moves sideways to dock Rovio does an ark rather then stright sideways, so at times makes Rovio harder to workout it's home postion.

I could not stand Rovio moving left when going stright and unscrewing the sensors did the trick.

denodan
denodan's picture

I don't think many will go without problems, your lucky if yours is still Rovio and Rovio has not been out a year yet anyway, so plenty of time yet?

I think the failure rate of Sensors maybe very high just going by posts here and Wowee know this is a big problem, they yet have to know what the cause is. Why don't they just recall Rovio or offer free good sensor replacements and a step by step guide how to replace the senors boards?

MrScott
MrScott's picture

Glad to see you prefaced the statement with "I think", because that's all we have at this point are opinions.

Stating certainties on either side of the question is foolish without the facts to back it up.

Anecdotal evidence from a handful of people reporting problems is not a clear indication of a general problem. Other people have posted their counter examples of having no problems with the sensors.

At this point we can say it is failure mode seen on several samples. No more, or less, frequent than other issues that have cropped up such as dieing power supplies, burned out IR beacons, failure to maintain a router connection, short battery life, and inability to charge.

There have been a few reports of each of these problems. Not enough to say that all will be affected, but enough to begin to question the quality control during assembly. If several different types of product fails are turning up in the first few months of the product life, it may be no single thing that knocks a Rovio out of service, but a menu of problems that stack the odds against a long healthy life.

The only design problem confirmed at this point is the choice to integrate a camera with poor low light performance. I haven't heard a single report of a camera that performs well in low light, so it seems that is a general issue seen by all owners.

RovioFan
RovioFan's picture

I had two of the encoders fail. I first confirmed this by analyzing the output from the detectors using an oscope. An easier way of doing this is to use Roborealm and observe the left, right, and rear tick count variables (although this would not eliminate a number of other possibilities). The rear was not showing any counts and the right one was showing very few counts. The left one seemed to work fine. I couldn't determine the type of detector the encoders had but the emitters are pretty common so I replaced the emitters with some from other encoder assemblies that I had lying around. This repaired both of the non functioning encoders. Unfortunately the fix only lasted a few weeks and then the same two failed again.

Like Mr. Scott, I am doubtful that such a large batch of components is faulty. This leads me to the hypothesis that there are not sufficient current limiting resistors in series with the LEDs and or the voltage across the LED is too large. However, I am truly baffled as to why the left channel has always worked (would they have designed one channel correctly and not the other 2?).

I have seen some discussion about the curved path and whether or not the encoders have anything to do with this. In my opinion, they affect straight paths because the firmware expects to see the left and right encoders change at the same rate. When they do not the microcontroller determines that the motors are not spinning at the proper rate and then adjusts the current to them proportionally. I like the fact that the motor drive circuitry is very consistent for all of the motors when the feedback value zero. This is why the system is able to work so well even when the encoders are disconnected.

I am fairly content with the inability to make precise angle turns but I would still like for the encoders to work as Wowee might make critical use of them in future firmware releases. I wish that Wowee would provide the drive/transducer circuitry schematics and then maybe I could make a modification. I hate to send the whole thing back if itÂ’s something simple like adding a resistor in series with the emitter. I would at least like to get the datasheet for the emitter/detector they are using. Has anyone determined the part number/manufacturer?

denodan
denodan's picture

So sounds like some hardware design is causing Sensors to fail then? So may not be bad sensors afterall causing the problem.

MrScott
MrScott's picture

RovioFan, since you have a Rovio with failed sensors, and apparently some understanding of the electronics, can you take some measurements for us across the "good" and "bad" circuits for the sensors?

This is the kind of background information that's needed to determine what the root cause for the failures is.

It's still not pointing to a universal design problem, it's just shifting the area of investigation from the sensor, to the circuit driving them. Ultimately it will be a hardware problem. Software doesn't kill hardware that is designed to fail safe. Software can allow a weak design to stress itself and kill the weakest component in the link.

The question still on the table is what is the nature of the hardware problem causing the failures. Bad sensor, bad current limiting resistor, bad voltage regulator, etc..

If the design is not constructed with enough margin to allow for normal variation in component performance, then it will be a weak design. It might work for most, some, or no stress cases.

Until we know what the failure mode is, we cannot say how big a problem it is.

If you have a Rovio with three good sensors, one fails, and swapping a good sensor to the bad position causes that one to fail, too, then it points to the specific components in the specific circuit. Comparing the specs for good and failing circuits tells you whether they're built the same, and whether the components in the circuits are operating at spec.

Pages