New Rovio Beta Firmware Available - Feedback Needed!

254 posts / 0 new
Last post
flowersrj
flowersrj's picture

cowboycctx,

I assume 420000 is the delay. How long does 420000
represent?

Thanks,
Rich

UPDATE: Disregard question as I just found the comment in the html code

Jimcking
Jimcking's picture

 

cowboycctx,

Although I am very impressed with your work, am not so sure it is a good idea to repeatedly send back up commands.

There are only two connections between the charging base and the robot.  Therefore Rovio is probably only measuring battery voltage and charge current.  Rovio displays "Docked" when it is Charging and "Roaming" when it is on the dock and "Fully Charged".  A back up command causes the motors to load the battery thus lowering the battery voltage and starting a charging cycle which displays as "Docked".  These repeated backup commands will initiate a lot of charging cycles which is hard on the battery.  Batteries are normally rated on the number of charge cycles they will endure.

 I have left Rovio unattended on the charging dock for several days and the battery shows full charge when Rovio is driven off the dock.  Rovio sometimes displays as few as two bars when it is on the charging dock and fully charged.  After it is driven off the dock the charge indicator will go up to full.  The charge indicator is set to read lower when it is on the dock and charging.  The charge indicator is just a simple voltage indicator.  Voltage is not a very accurate way to indicate state of charge in a NiMH battery.

 Rovio's battery will run dead in about two hours when it is really roaming and not sitting on the dock.  When it is sitting on the dock it will go from roaming to docked on it's own when the battery needs charging.  When Rovio is sitting on the dock think of "Docked" as meaning "Charging" and "Roaming" as meaning "Fully Charged". When is it not on the dock and reads "Roaming" it means it is discharging.

I think Rovio's charging circuit is good.  What it needs is a firmware fix to properly indicate what is happening.

 Jim K.

 

RoboBrain
RoboBrain's picture

Jimcking said:
cowboycctx,
Although I am very impressed with your work, am not so sure it is a good idea to repeatedly send back up commands.
There are only two connections between the charging base and the robot.  Therefore Rovio is probably only measuring battery voltage and charge current.  Rovio displays "Docked" when it is Charging and "Roaming" when it is on the dock and "Fully Charged".  A back up command causes the motors to load the battery thus lowering the battery voltage and starting a charging cycle which displays as "Docked".  ....  Rovio's battery will run dead in about two hours when it is really roaming and not sitting on the dock.  When it is sitting on the dock it will go from roaming to docked on it's own when the battery needs charging.  When Rovio is sitting on the dock think of "Docked" as meaning "Charging" and "Roaming" as meaning "Fully Charged". When is it not on the dock and reads "Roaming" it means it is discharging.
I think Rovio's charging circuit is good.  What it needs is a firmware fix to properly indicate what is happening.
 Jim K.

 

Jim, You speak of this issue as if you have intimate knowledge of how the charging indicicator works -- is this fact or conjecture?  I have been wondering the same about these issues as I see similar behavior from my Rovio. 

While keeping an active web interface runnig while it is docked, I have not seen it change to Roaming on it's own, but if I reload the web interface after it has been docked for a length of time, it frequently says roaming when it is fully docked.  The fact that it is streaming video however while "docked" does consume battery life and most times if I drive it straight off the dock without "topping off" in a fashion similar to cowboycctx's work around, it shows down 1-2 bars on the battery.

I would like to know the truth on this issue though and wish Wowee could provide more insight here into the expected operation, when it stops charging, etc.

The one that puzzles me though, when mixed with the information you provide is that when I do noticed it physically docked but indicating roaming, removing the A/C adapter from the wall and reconnecting (without disturbing rovio's position), it will go back to charging.  So,  if the charging cicuit thinks the battery is full and doesn't need a charge yet, why does it go back to charging simply by toggling the charging input?  fwiw, I still have the 8V charger it came with. 

I have been tempted to monitor the charging indicator from a script and send an on/off to an X10/Insteon outlet to record how often this state transition happens, etc.

 Thanks...

...RB 

Nipahc
Nipahc's picture

Thanks for the info Jim K.

I was thinking that this might not be the best idea, but I wasn't really sure why. I assumed it was firmware error since my Rovio doesn't really move. I just make sure to check occasionally, and it seems to not be that bad.

I was hopeing if we had any firmware issues that were outstanding and damaging, Wowee would keep us up to date. Least alert so we could follow up with such a situation.

I also assumed it may just be a firmware error and still charging??? It hasn't been going down so far I'd prefer to see charge, thats why I move it.

Hickory
Hickory's picture

Had a little time to mess with my Rovio today and was recording a path when it ocurred to me, there's no icon to replay the path on the dashboard. How does one play back a path that is recorded without a playback button? I also noticed and fault Wowwee for kind of telling a lie, you can only record where there is a beacon signal. It's not like my roomba that records it's way around the room and replays back to it's charging doc is it? The more I play with it particularly now that I can access it remotely and it will find it's dock 99% of the time the more I like it, but am still convinced that Wowwee overpromised and underproduced to entice many of us to make a purchase. The lighting issue alone would have been enough to convince me not to buy, I've strapped a flashlight on it and that works good enough, for now.

cowboycctx
cowboycctx's picture

RoboBrain said:

Jimcking said: cowboycctx, Although I am very impressed with your work, am not so sure it is a good idea to repeatedly send back up commands. There are only two connections between the charging base and the robot.  Therefore Rovio is probably only measuring battery voltage and charge current.  Rovio displays "Docked" when it is Charging and "Roaming" when it is on the dock and "Fully Charged".  A back up command causes the motors to load the battery thus lowering the battery voltage and starting a charging cycle which displays as "Docked".  ....  Rovio's battery will run dead in about two hours when it is really roaming and not sitting on the dock.  When it is sitting on the dock it will go from roaming to docked on it's own when the battery needs charging.  When Rovio is sitting on the dock think of "Docked" as meaning "Charging" and "Roaming" as meaning "Fully Charged". When is it not on the dock and reads "Roaming" it means it is discharging. I think Rovio's charging circuit is good.  What it needs is a firmware fix to properly indicate what is happening.  Jim K.

Jim, You speak of this issue as if you have intimate knowledge of how the charging indicicator works -- is this fact or conjecture?  I have been wondering the same about these issues as I see similar behavior from my Rovio.  While keeping an active web interface runnig while it is docked, I have not seen it change to Roaming on it's own, but if I reload the web interface after it has been docked for a length of time, it frequently says roaming when it is fully docked.  The fact that it is streaming video however while "docked" does consume battery life and most times if I drive it straight off the dock without "topping off" in a fashion similar to cowboycctx's work around, it shows down 1-2 bars on the battery. I would like to know the truth on this issue though and wish Wowee could provide more insight here into the expected operation, when it stops charging, etc. The one that puzzles me though, when mixed with the information you provide is that when I do noticed it physically docked but indicating roaming, removing the A/C adapter from the wall and reconnecting (without disturbing rovio's position), it will go back to charging.  So,  if the charging cicuit thinks the battery is full and doesn't need a charge yet, why does it go back to charging simply by toggling the charging input?  fwiw, I still have the 8V charger it came with.  I have been tempted to monitor the charging indicator from a script and send an on/off to an X10/Insteon outlet to record how often this state transition happens, etc.  Thanks... ...RB 

You could be right.  If you or anyone has a concern of my method PLEASE do not use the HTM file.  If you do have a concern and still want to use the file you could consider increasing the 420000 (in the setInterval("Charge();" , 420000);)  number to a much hinger number.  Again use at your own risk.

Jimcking
Jimcking's picture

RoboBrain said: Jim, You speak of this issue as if you have intimate knowledge of how the charging indicicator works -- is this fact or conjecture?

I have no intimate knowledge other than observation, measurements, and familiarity with NiMH charging circuits.  I took Rovio apart to do the headlight mod:

http://www.robocommunity.com/forum/thread/14773/Yet-Another-Headlight-Mod/

It is normal for a NiMH charger to initiate a charging cycle when it is first plugged in as it has no history of the battery.

Visit the Rovio Web Page and wait a full minute for it to stabilize.  Observe the number of bars on the battery charge indicator when it is fully charged "Roaming" (Mine reads 2 bars).  Drive straight off the dock and let it sit (again wait a full minute for it to stabilize).  They normally go up (Mine reads 5 bars full charge).

Let us know what you find out on the X10 experiment?  It will probably never go to "Roaming/Full Charge" when it is streaming video as the addtional current required to stream video will not allow the battery to reach full voltage.

Thanks,

 Jim K.

 

 

johnny d
johnny d's picture

Still cant get email pic to work. Also disconnects internet claiming UnPnP error

however Im using airport extreme which doesnt use that. When I refresh the browser it reconnects. what is the IR checkbox for?

Thanks
John

Giobatta
Giobatta's picture

@Johnny d:
with the extreme you need to use manual mode (mine too refuses UPnP mode), just copy the web port into the upnp section, and disable upnp automap

the IR radar detect and avoid obstacles during the execution of a path (eg: the usual curious pet :)
its useless in manual mode

(speaking of emails I'm using the mobile.me service, ... its weird Mail client show me an empty mail, if I read the email straight into webmail -at me.com- i see also the picture O_o)

Kreblon
Kreblon's picture

Two things that need to be addressed:

Change the charging routine so that even if the battery is very low, it automatically goes right into charge mode if you park it on the dock.

And

Add a fourth head position selector button for a position in-between full up and
slightly above at rest. (positions 2 and 3)

This will give us a full vertical range of coverage using the focal length of the camera's lens now.

Bob

Robot007
Robot007's picture

I also get the following error even THOUGH I can access from the outside world. Strange bug?

Under Internet Access on the Rovio network setup, Rovio was able to obtain my Internet IP without problems, but puts a question mark on the screen for system status and says

RoboBrain
RoboBrain's picture

Charging Mystery Update - Don't be fooled by the UI, there's a bug!

Ok, So we all got these fangled new Rovio's and all we have to go by is the external apearance, the UI, to tell us what's going on. Well, darnit if it can't be trusted.

I started monitoring my Rovio with a little script that queries the rev.cgi for MCU and Status reports and GetStatus.cgi. In doing this, I determined that after charging fully, it does go into a battery maintenance state.

Here's an example of Docking after driving around just a little:

Mon Jan 5 16:40:57 2009: Changed Status: state=0 [was 2], battery=112 [was 114], sm=15 [was 27], charging=80 [was 0], head_position=203 [was 65]
Mon Jan 5 16:41:21 2009: Changed Status: battery=116 [was 112]
Mon Jan 5 16:41:27 2009: Changed Status: battery=113 [was 116]
Mon Jan 5 16:42:04 2009: Changed Status: battery=114 [was 113]

[So above, went from charging=0 to charging=80 - charging at full throttle]

after charging for about 30 mins:

Mon Jan 5 17:11:05 2009: Changed Status: battery=119 [was 118]
Mon Jan 5 17:11:28 2009: Changed Status: battery=117 [was 119], charging=72 [was 80]
Mon Jan 5 17:11:36 2009: Changed Status: battery=118 [was 117]
Mon Jan 5 17:11:54 2009: Changed Status: battery=117 [was 118]

[Battery topped off, charging=72 now]

At this point, the charging value remains at 72 while it is docked. According to the API docks, 0-79 is "not charging," but as the battery continuously hovers between 116-119 while docked and being queired by my script, I expect it is really a tirckle-charing state. My script is querying Rovio every 6 seconds. I have not seen charging value change away from 72 at all while it remains docked.

Oh, and while I'm at it, driving off the dock, the battery indicator does go up:

Mon Jan  5 22:01:49 2009: Changed Status: battery=126 [was 118], charging=0 [was 72]
Mon Jan  5 22:01:55 2009: Changed Status: battery=125 [was 126]


Now to the UI bug
: rovio.js has absolute values being used to detect the charging state. When you stay on the Web GUI, it never reaches a different state as far as that code is concerned. It appears to be looking for charging=64 for it to turn off the battery charging indicator but still indicate being Docked. If you reload the Web UI, because it is not at one of the 3 absolutes the code is looking, it shows Rovio as Roaming and Not Charging. This is misleading and I think has led a lot of us to question the charging of our little robots. I am not sure if everyone's Rovio says charging=72 when it's in this state or not, if it does, then perhaps that is the only change required to the JS code. Why there is such a range of 1-79 for the charging value is unclear. I have only seen 0, 72 and 80 in my logging for the last ~30 hours. fwiw, Rudolph's text_mode_rovio client has a bit more flexible logic in this regard and does display the right information. It varies a bit on the charge % indications though, haven't dug into that too deep yet.

So don't despair, Rovio is charging and not mysteriously jumping off the dock!  Cowboycctx -- while your backup script was a novel approach, I think with this information it may be uneccessary. 

Mystery Solved?

Ok.. Now has anyone figured out how to enable FTP so we can fix these trivial coding issues and add other on-board widgets of our own?

...RB

 

Kreblon
Kreblon's picture

OK...I just checked battery charge current when docked.

There is NO current flowing into the Rovio unless the Rovio goes into
full-on charging mode. It does not trickle-charge the Rovio when docked.

This is why the Rovio goes into charging mode periodically when just sitting there
docked- as it is running on it's internal battery only. That's sort of silly, as
a low trickle-charge could allow the Rovio to stay topped-off while docked.

Actually,it would be possible to add a small resistor and diode directly from the contacts on bottom - to the battery. I could determine the quiescent draw of the Rovio while it sits there- and add a little more current for battery topping, and a transistor to switch the trickle power only when the Rovio is switched on. Ahh, wait...why not switch between a tiny trickle charge when off - to enough current to run the Rovio and top-off charge when on...? Better!

Charge current when in Charge mode is 600ma. It slowly goes down as the battery comes up.

It is extremely likely that the battery NEVER gets fully charged because of the lack of a trickle charge at the end of the charge cycle. Normally you would do a "bulk charge" at 20% of full AH capacity (600ma)- then switch to a trickle to top it off (like 50ma). Our batteries never get topped-off. That means they are only charged to about 85%..maximum...maybe less.

This is a hardware work-around for the problem of the code being faulty.

Let's keep investigating.

Bob

Kreblon
Kreblon's picture

Dam- That won't work because then the Rovio will never "know" that the batteries
are down..as the trickle charge will lift the apparent voltage of the batteries
up to where they look OK...as soon as it docks.

OK...Come on you Firmware Gurus at WowWee...help!

Bob

johnny d
johnny d's picture

Firmware seems to have helped for smoother operation. still cannot get email snapshot to work. It would be nice if the camera was infrared also.I suppose this could be modified. A motion detector for the cam might also be beneficial

cmilian
cmilian's picture

Guys, so much work into looking this charging issue. Is there someone from Wowwee that maybe reads these threads that can clarify?

PeterM
PeterM's picture

I was more than happy to hear about the beta. Since running internet via tv-cable I do not have the option running WEP, so have to stick to WAP. Beta should run WAP, but I can not run setup since it tells me that I must run WEP.....any ideas?
BTW: Advanced Setup does not work either, because the screebns I see are simply too different from what I read in the manuals.
...would like my rovio up and running soon...have it since 3 weeks...

Thanks in advance for any help!

Michelle C
Michelle C's picture

cmilian said:
Guys, so much work into looking this charging issue. Is there someone from Wowwee that maybe reads these threads that can clarify?

We're well-aware of your concerns regarding this issue and are looking at this on a case-by-case basis, which is why I urge anyone with this concern to contact our Customer service department... They'll be able to tell you how to proceed.

RoboBrain
RoboBrain's picture

FYI, there's a little bug in rovio_mobile.js - updateVideo is calling changeResolution.cgi?ResType= when I think the calls I see to it are expecting ChangeCompressRatio.cgi?Ratio=

Also, mobile/facebook.htm and mobile/iph.htm both incorrectly try to set value=3 (instead of 2) for High quality video.

When this actually changes I found I needed to cause the image stream to reload as well.

Partial diff below.

...RB

----
function updateVideo(value){
- sendCmd("../ChangeResolution.cgi","ResType=" + value, doNothing);
+ sendCmd("/ChangeCompressRatio.cgi","Ratio=" + value, doNothing);
+ setTimeout("updateImage();", 2000);

+function updateImage() {
+ $('oCamCtl').src = 'GetData.cgi?' + Math.floor (10000 * Math.random ())
+}

Paul T
Paul T's picture

The problem I'm having is that I finally was able to connect to Rovio externally using DynDNS, but it only lasts for a little while then I lose connectivity to Rovio. The only way I've found to resolve it temporarily is to power down my Linksys wireless B router and power it back up again. Then I can connect to Rovio from an outside network again for a little while then it starts all over again, but I can still access the Internet even though I can't access Rovio. From the internal network, I always have connection, it's only externally where I have the connectivity lose. Has anyone else experienced this?

cmilian
cmilian's picture

Check your router for updates. A weird thing happened to me. I had no problems with my router. One day out of the blue started loosing connection. I needed to reboot my router every few hours. I bought a new router of another brand and still had the problem. I started to hate Comcast and all they stand for. Change modems and the same problem. I decided that Comcast had a conspiracy against me. I have 3 computer at home and it only happened when one of them was connected. I started to troubleshoot that computer for weeks. Formated the computer and reinstalled. Same problem. I was going crazy. I Learned to live rebooting my router every few hours. One day out of the blue decided to upgrade the ROM on my new router. Guess what? THE PROBLEM WAS GONE !!!!!

MrScott
MrScott's picture

I feel your pain with regards to updates by service providers suddenly disturbing the stability of a long running router. In my case it was Adelphia, which has since been bought up by Comcast.

Just like you experienced, my router suddenly started locking up or dropping connections several times a day. Calls to Adelphia resulted in "Nothing we did.".
Nothing on my end had changed, so I went looking for explanations online.

Based on discussion group feedback, I was advised to get the latest flash for the router. After it was updated, stability was restored.

cmilian
cmilian's picture

Yeah, comcast told me that my problem will be gone if I rented their wireless router. Yeah right !!!

Paul T
Paul T's picture

I'm already at the latest firmware version for my router (1.52.02), but I found an article on Linksys' website that has a possible solution that I'll try. I have a cordless phone a few feet away from the router. I'll try a different channel.

http://linksys.custhelp.com/cgi-bin/linksys.cfg/php/enduser/std_adp.php?...

Paul T
Paul T's picture

I've found that by just turning off Rovio and back on again I get connection back. Then drops again after a few minutes. Seems like the problem is with Rovio not my router, I can connect to the internet just fine, it's only Rovio I can't connect to externally.

MrScott
MrScott's picture

All I can offer is that other people do not seem to have the same connectivity problem, so it's not a problem in the firmware that affects all users. It is most likely some variation on the settings you've chosen, and the specific configuration of the network you're tying to.

I seem to recall some reports in threads here of dropped connections early on, but I do not recall the specifics of what people did to resolve them. You might want to review the earlier threads looking for keywords like "dropped connection", "connection lost", etcetera.

Some people have discovered that updating their router firmware provided the subtle support changes that Rovio was sensitive to.

cmilian
cmilian's picture

Wasnt there going to be a newer firmware beta last week?

RoviNo
RoviNo's picture

I too am testing the 5.00b7 firmware and wanted to share my experiences with this firmware version. I am very interested whether the "Forward Motion Pulling To One Side" issue has been experienced by anyone else.

My Rovio is 2.5 weeks old.

CONNECTIVITY ISSUES
In my particular situation, I have found that after connecting externally most of the time I loose connectivity to the Rovio. If I ping the Rovio from an inside network PC, I find it will not respond. The light on the Rovio is usually blinking when this occurs. I always have to power off and power on the Rovio to get it to connect.
I never experience this if I am connecting internally.

LIGHT INDICATOR INACCURACIES
I also find with this firmware that the Rovio light indicator usually is not accurate and/or erratic.

FORWARD MOTION PULLING TO ONE SIDE
My Rovio often pulls to one side(to the right). This is quite pronounced at times. Any one else experiencing this?

SAVED PATH EXECUTION PROBLEMS
Sometimes my saved paths don't work and I have to reset the Rovio. Are executing a path the Rovio just runs into a wall or goes in the wrong direction.

HOME PATH EXECUTION PROBLEMS
Sometimes when sending the Rovio home it gets totally confused and goes in a direction that is 180 degrees off. A reset cures the problem.

PULSATING MOTION
When driving the Rovio, the motion is sometimes pulsed (as described by someone else in the forum). I believe a reset usually solves this.

PICTURE DARKNESS
And of course, the picture is way to dark to be useful in most lighting situations 'in my house'. This is with natural window lighting.

CONFIGURATION - FIREFOX SUPPORT
When setting a manual IP using the web interface and Firefox V3 I find that the IP is garbled and doesn't properly set the IP. I have to use IE or the setup utility instead. As a matter of fact - after setting the IP using Fireox and then looking at it with the setup utility the IP is totally corrupted.
Example: The setup utility shows IP = 192.168.100.15 / 255.255.255.0
Firefox shows IP = 92.168.100.150 / 0.255.255.

BATTERY INDICATOR
In most cases, while on the charger, the battery indicator seems to be totally inaccurate.

AUDIO NOT WORKING IN FF, CHROME
No audio in Firefox V3, Chrome. Not recording either.

I would love to test a new firmware... ;) Please hurry.

Any comments are appreciated,
-Floyd

Paul T
Paul T's picture

That is my connectivity issue to a tee!!!

savino
savino's picture

I'm not getting the ability to change the security to WPA with the beta. The settings say Version: v5.00b7, but only offer WEP.

Shouldn't the setup software (attached to the usb) allow you change the security to WPA also? Or does it get the available security modes from the robot?

Pages