Speedict Ebike - Anybody seen/tried this?

[Ken Taylor]

This little orange Mercury....

I'm having issues so I'd like to hear from you guys...

OK,I'll try.

It's very difficult to analyze what is going on, is there a HW problem, SW problem in the client, FW issues, Bluetooth issues on my phone or what???

More than one problem with mine, made more difficult because I didn't know if I was doing things right. Whenever the Speedict detects a loss of Bluetooth it goes into reconnect mode and a green led flashes. This happened a lot with mine even when the phone was close and I suspect some issues follow from this.

First one was bricked during FW upgrade, ok that can of course happen...
Danny sent me two new ones, a lite version to be used for FW upgrade test and one with the correct FW as a replacement for the bricked one.

I'm wondering if there is enough memory to store the new and old firmware simultaneously in the device?

What happens on phones is the new firmware file is loaded over the air onto the phone, the file is checked for completeness and only then will it upgrade. I suspect the Speedict writes over the old firmware as it transfers the new one across the Bluetooth connection which would mean bricking if the Bluetooth connection fails part way through.

Installed in the bike I had issues synchronizing data from the start, I couldn't figure out what made it work, because sometimes it worked.

I suspect this is a side effect of an unreliable Bluetooth connection. The sychronisation can not recover from loss of connectivity when partially complete.

All of a sudden it just died... I took the bike apart and had a look... Only red LED would come on.

You are the 3rd or 4th person to report that. It looks like a firmware issue that is only recoverable by returning the device for re-flashing.

Live View looses signal about 4 m from the Speedict.

Is that consistent or does it drop out and reconnect and just drop out more often further away? I seemed to lose the connection periodically even when very close. The easiest way to spot the drop outs is the green led coming on. I'm guessing the logic for detecting drop outs is not very forgiving and it goes into reconnect mode too easily.

Today, the data looks very strange indeed and the trip record downloaded is about 360 entries long, even though it should be 35 minutes...

Could that be an incomplete synchronisation?

I'm speculating a lot here and I'm sure Danny could correct my errors. My Mercury has gone back for repairs and I'm keen to get it working.

 

[Ahoog]

Thanks for your input guys

I don't know if this is related to anything else but I got a new version that would solve some national settings bug - In Sweden we don't have decimal point but comma instead.
On recommendation from Danny I have turned the GPS off and now, all of a sudden most things seems to work.
I had the GPS off the last time I rode on it, with the older version of the client, and the data was all strange then.

The range where I'm now able to successfully sync the device to the phone is now greatly increased.

 

[Ken Taylor]

Thanks for your input guys

I don't know if this is related to anything else but I got a new version that would solve some national settings bug - In Sweden we don't have decimal point but comma instead.

Before you said:-

All of a sudden it just died... I took the bike apart and had a look... Only red LED would come on.

Does this mean you were able to upgrade the firmware when it was in the state "Only red LED would come on"?

On recommendation from Danny I have turned the GPS off and now, all of a sudden most things seems to work.
I had the GPS off the last time I rode on it, with the older version of the client, and the data was all strange then.

Seems an odd solution. The GPS is not a transmitter.

The range where I'm now able to successfully sync the device to the phone is now greatly increased.

Maybe the firmware upgrade addressed syncing issues as well as the "national settings bug".

 

[Ahoog]

Well, I see that I need to clarify things a little...

The speedict in my bike now is the third one... Which has always worked to some extent at least.

First Speedict, bricked by FW upgrade on my desk.
Second Speedict, died for no apparent reason while on the bike.
Third Speedict (lite version) Had all sorts of issues related to transfer of files and bluetooth range. At least it looked that way.

Now it seems to me that it might not have been bluetooth range, but rather something with how the phone writes GPS data to the speedict.
How that would affect the range while syncing I don't know but it works now. Maybe there were some other bug fixes in the new client version as well that I don't know about.
I haven't been able to do anything to break the sync on it yet.
So, the new version of the software is on the phone, and not the FW on the speedict. Speedict FW is still 1.0.3 on the one currently on my bike.

There will unfortunately not be any tests today as I had a small accident yesterday where I tried to turn into a 90 degree curve a with a little too much speed
Ice in the entry started the slide, and also on the apex which made the front wheel disappear from underneath me.
Fun to see the log file from the speedict where speed suddenly drops from 26 km/h to zero.
My right knee didn't like the contact with the asphalt much though

 

[Ahoog]

I did some tests yesterday anyway, it seems that it is actually the GPS that is somehow interfering with the transfer of files. If I turn it off I can easily sync the data from 2 m away.
If I turn it back on again it's back to no sync unless closer than 10 cm.

Very strange behavior indeed...

I made the same test with another phone and it was exactly the same.

 

[jackhandy]

I have the same problem - no synch with gps on. fine without, but frequently get garbled trip data (crazy speeds & no gps data at all).

samsung galaxy note 4.0.4

 

[KCvale]

Just a little levity to this phone geek fest...

Did you know that some people actually survive just dandy in this world without any cell phone let alone a 'smart' phone'?
It's true!
Sure I bought a disposal phone to carry with me on trips for emergencies but the prepaid minutes expired long before I used any becasue well, I didn't have an emergency and simply don't want let alone need to 'stay connected' to everything all the time.

I have a nice PC with 2 big screens and fast Internet with e-mail I shut off when not using, and a land line with an answering machine and get along just fine running my own business and going about my daily life without some fragile little device constantly annoying me.
I used to carry a digital pager back in the day and found it annoying but I could ignore it and it would take a massive beating and still keep working.

I'm not dissing any of you smart phone app crazy people, this is just some levity remember? It is just that what I see is society in a sort of mass hysteria over a fragile little expensive device they pour there whole lives into to the point of it taking over your life and pocket book is all, and of course that is whole business model behind them.

How far did I ride my bike today? How fast am I going? What are my total miles ridden?
Have you seen the revolutionary new dedicated device with a huge hands free display that will tell you all of that info with a simple quick glance down?

$27 bucks, virtually indestructible, never needs batteries, works with no cell or Wi-Fi service and in any weather conditions, untraceable, and not worth steeling.

I'll leave it at that and smile as I click the Submit button ;-}

 

[Ahoog]

I have the same problem - no synch with gps on. fine without, but frequently get garbled trip data (crazy speeds & no gps data at all).

samsung galaxy note 4.0.4

Well, this was rather interesting... I thought mine were the only ones doing this.

So, now... Does anyone have this feature working?
Maybe some kind of bug otherwise

 

[Ahoog]

I'm not dissing any of you smart phone app crazy people, this is just some levity remember? It is just that what I see is society in a sort of mass hysteria over a fragile little expensive device they pour there whole lives into to the point of it taking over your life and pocket book is all, and of course that is whole business model behind them.

Well, first I'm totally hooked on my smart phone, I admit...
However... That is not what goes onto my bike, it's a cheap Internet connected android device which I got for almost no money at all, or as in the case with the htc desire... Free

It's a very nice speedo you got there though

 

[evb]

Found a bug:
Mercury 1.0.8, Samsung Galaxy Note, 4.0.4
When view a chart of saved record, zoom "+" several times, and move to end of chart, software crashes.

 

[Ken Taylor]

It is just that what I see is society in a sort of mass hysteria over a fragile little expensive device they pour there whole lives into to the point of it taking over your life and pocket book is all, and of course that is whole business model behind them.

That's what these guys said.

Frame Breaking-1812

How far did I ride my bike today? How fast am I going? What are my total miles ridden?

How much power did I produce?
How much effort did I exert as measured from heart rate?
Am I getting fitter?
Does electric assist motivate me to exercise or the opposite?
Is their a control strategy that maximises fitness but minimises pain?
How much power did the motor produce?
Where did I go?
How steep were the hills?
How fast did I go compared to last time over this section?
How fast did I go compared to everyone else over this section?
How fast am I pedalling and which gear gives the best results?
How much difference does drafting make?
How do I compare with the pros? https://theconversation.edu.au/how-did-amateurs-compare-with-pros-at-the-tour-down-under-11946
etc.
And all this magical data, becoming practical to acquire.

Boring to some, fascinating to others.

 

[Ahoog]

Well, I thought I had the answer to getting this device to report accurate data and syncing ok.
Today I'm proven wrong. I still get data that is not correct, even though GPS is turned off.
While syncing it doesn't complain at all and then reports sync successful.

I could have lived with the fact that GPS-data wasn't available, I could easily have stored it with another app and merged it when I needed to analyze data based on location.
What I can't live with is not being able to trust the data being collected by the Mercury.

Something that would have been nice to have is a list of working device configurations, phone/android version. At least I could have just bought one of those and be fine with it, but no list that I'm aware of.
Now I have two android phones acquired just for this and none of them works as they should. I'm refusing to buy another just to try if it works.

As it is now, this looks more and more like a beta program where we need to work out the bugs in both sw and hw. Or maybe the ones I'm reading about here and elsewhere are just unlucky bastards like myself?

I'm almost ready to give this thing up and buy the CA instead. The Speedict/phone solution is a lot cooler I think, but it has to work.

/Anders

 

[Ken Taylor]

Well, I thought I had the answer to getting this device to report accurate data and syncing ok.
Today I'm proven wrong. I still get data that is not correct, even though GPS is turned off.
While syncing it doesn't complain at all and then reports sync successful.

That is consistent with my theory there is a problem with the Bluetooth communications logic. If it needs accurate timing for polling, the less other things the phone is doing the better it will work. That would explain the odd situation of the GPS affecting communication reliability. If you can make sure your phone is not running any other background tasks, it might improve. I can't test anything as mine is back for repairs.

Danny said "most of the issues or cases we have received so fas were software related such as android custom rom caused firmware upgrade failed !" which is more evidence of a problem with the Bluetooth communications logic. The phone should only be responsible for transferring a complete firmware file. Then, once the file is on the device and checked, it should upgrade. In this case a fault in the phone could cause upgrades to be impossible but no phone fault could cause an upgrade to brick the device.

Something that would have been nice to have is a list of working device configurations, phone/android version. At least I could have just bought one of those and be fine with it, but no list that I'm aware of.

I don't believe it is the phone that is the problem.

As it is now, this looks more and more like a beta program where we need to work out the bugs in both sw and hw. Or maybe the ones I'm reading about here and elsewhere are just unlucky bastards like myself?

Looking back through this thread, historically there is little talk of problems, however we have a new model. The old one had less functionality and firmware was upgraded through the USB connection. In the the new Mercury the USB is disabled. I suspect the issues discussed here are as unexpected for Danny as they are for us.

Where else is the Speedict discussed, there is not that many people reporting problems here? I don't mind being a beta tester but I would like a lot more feedback which would facilitate less speculation and confusion. I wonder if they've got the resources to sort out the issues and whether they've sold so many that the cost of supporting them will be too high.

I'm almost ready to give this thing up and buy the CA instead. The Speedict/phone solution is a lot cooler I think, but it has to work.

Conceptually it is the way to go and there is lots of ways I'd like to extend the Speedict functionality. You get a lot of functionality in a handset at very low cost compared to custom displays in the CA and the CA offers a clunky serial connection for extracting data. The CA implementation seems very good though.

 

[megacycle]

The CA seems very last century in comparison to Speedict.
The wireless arrangement even though it's having issues, the device is so novel and new, got to be expected, especially when yeh we're asking for so many functions, suprising no one's asked if it can make a cup of coffee, yet :wink:

 

[t3sla]

I've been trying to find a good handlebar mount for my phone so that I can ride hard and still have the phone with speedict in view without worrying if it will fly off.
Finally came across slipgrip and bought one, took just over two weeks to get here from the US.
It allows me to keep my otter-box (which is a very tough case) on my phone whilst its on the mount, perfect!

[youtube]048m_omDb00[/youtube]

 

[speedict]

I'm speculating a lot here and I'm sure Danny could correct my errors. My Mercury has gone back for repairs and I'm keen to get it working.

ken, your new speedict mercury with customized wire

 

[romelec]

Danny said "most of the issues or cases we have received so fas were software related such as android custom rom caused firmware upgrade failed !" which is more evidence of a problem with the Bluetooth communications logic. The phone should only be responsible for transferring a complete firmware file. Then, once the file is on the device and checked, it should upgrade. In this case a fault in the phone could cause upgrades to be impossible but no phone fault could cause an upgrade to brick the device.

That is acceptable to have some software/firmware issues, In my company we deliver some products with limited functionalities and update them after. But the main thing is to be sure that the update system is robust and is always possible by testing all the error cases.
It could be an issue with the available flash size to not do what you said, but in this case there should be a bootloader or rescue updater application able to always perform the update on Bluetooth or USB.

 

[speedict]

Danny said "most of the issues or cases we have received so fas were software related such as android custom rom caused firmware upgrade failed !" which is more evidence of a problem with the Bluetooth communications logic. The phone should only be responsible for transferring a complete firmware file. Then, once the file is on the device and checked, it should upgrade. In this case a fault in the phone could cause upgrades to be impossible but no phone fault could cause an upgrade to brick the device.

we have identified the problem was caused by some custom rom and there were a loop hole allow android to inject some incorrect GPS data into speedict rom during firmware upgrade, and now been resolved by speedict android client 1.0.8.

 

[Erikjan]

Nice you find the problem, hope i can try it soon.
Did you already now if my speedicts are repairalble ?

 

[speedict]

Nice you find the problem, hope i can try it soon.
Did you already now if my speedicts are repairalble ?

repaired and sent today

 

[Ken Taylor]

ken, your new speedict mercury with customized wire

Thanks Danny. Kind of you to add the extra cable. On the Aprilia it will be long enough to to reach from the battery compartment to the controller comfortably without additional cable. I'm looking forward to trying it out.

... But the main thing is to be sure that the update system is robust and is always possible by testing all the error cases...

Yes.

we have identified the problem was caused by some custom rom and there were a loop hole allow android to inject some incorrect GPS data into speedict rom during firmware upgrade, and now been resolved by speedict android client 1.0.8.

Thanks for the additional information, its an odd problem. If you don't check the firmware file integrity on the Speedict prior to upgrade there is always a risk of bricking. This will occur if the bluetooth connection fails during file transfer which will inevitably happen occasionally. The spec at http://speedict.com/user-guide/mercury/speedict-mercury-user-guide-rev-1-0-0.pdf says there is 1MB of internal memory which ought to be enough to store the complete new firmware so that file integrity can be checked prior to upgrading.

 

[Ken Taylor]

The CA seems very last century in comparison to Speedict.

Perhaps, but their engineering process is as good as it gets. Methodical, open and thorough. There is an amazing amount of detail at http://www.endless-sphere.com/forums/viewtopic.php?p=571345#p571345 . I'm an admirer, even if not a customer.

The wireless arrangement even though it's having issues, the device is so novel and new, got to be expected, especially when yeh we're asking for so many functions, suprising no one's asked if it can make a cup of coffee, yet :wink:

Yes, the possibilities are large.

 

[speedict]

Thanks for the additional information, its an odd problem. If you don't check the firmware file integrity on the Speedict prior to upgrade there is always a risk of bricking. This will occur if the bluetooth connection fails during file transfer which will inevitably happen occasionally. The spec at http://speedict.com/user-guide/mercury/speedict-mercury-user-guide-rev-1-0-0.pdf says there is 1MB of internal memory which ought to be enough to store the complete new firmware so that file integrity can be checked prior to upgrading.

yes it did file integrity check even on every batch of bluetooth file transmit, the reason was the area (for temporary store data not the entire 1m data area) for temporary store completed firmware file was contaminated by incorrect gps data which is also temporary stored there.
when on beta test phase, we have no such error reported by our testers or even our development team, now we came up with conclusion :

1. when perform firmware upgrade, most of them were did it indoor where gps was unlock without sending data to speedict mercury
2. yes we have one tester did upgrade outdoor however she didn't turn on the gps
3. not all android has gps issues, even samsung s3 only happened to some rom ...

we have internal meeting last month, argue about the up coming product with firmware upgrade or not, usb or bluetooth, talking about usb upgrade none can guarantee 100% without glitches (remember on speedict ebike model, when major upgrade - from simple logger to add throttle control features, some customers didn't follows correct step that caused brick of their eBike too).

one of the proposed methods during our meeting was on our speed controller project we "might" take out the firmware upgrade features on basic model and let live view data logging features perform on speedict mercury !

 

[speedict]

Thanks for the additional information, its an odd problem. If you don't check the firmware file integrity on the Speedict prior to upgrade there is always a risk of bricking. This will occur if the bluetooth connection fails during file transfer which will inevitably happen occasionally. The spec at http://speedict.com/user-guide/mercury/speedict-mercury-user-guide-rev-1-0-0.pdf says there is 1MB of internal memory which ought to be enough to store the complete new firmware so that file integrity can be checked prior to upgrading.

by the way ken, your speedict mercury was not brick, it was damaged by over voltage

 

[rosich]

Speedict on iphone ?




[youtube]LXddbiizUaI[/youtube]