fearedbliss Posted August 24, 2016 Share Posted August 24, 2016 On 8/21/2016 at 6:44 PM, esaj said: The worst people to test the software are those who wrote it... automated unit- and integration-tests work great for catching small programming and logic mishaps, and to check that refactored/changed code still behaves as it should from interface-standpoint, but you should never, ever use the same people who write the software to do the actual "functional" testing There's a psychologic factor, once you know how the software is written (ie. how it works under the hood), you're inclined to use it exactly the way it was written for... well, that was probably badly worded, but I've seen so many cases where a tester (who doesn't know anything about the internal workings of the software) makes it crash / misbehave by doing something "unexpected" when using the program There are definitely benefits to automated and unit testing, and you are completely right about having the non-writers test out the code. I'm very happy with the trend of new languages (My favorite being Rust) that are giving a lot more responsibility to the compiler, and making the language more safe (while not reducing power and accessibility of low level components). The safer the language, the safer the code. Link to comment Share on other sites More sharing options...
EricGhost Posted August 24, 2016 Share Posted August 24, 2016 On 23 agosto 2016 at 0:59 PM, KingSong69 said: Since yesterday i am also on V1.22...driving KS18.... My Experience so far: Yes, the driving experience is a bit smoother...hard to describe, the Movements between pushing and braking are kind of better and i also have the Feeling the acceleration is faster....but that can also just come from the sunny, bright day :-) But: I also have a new Sound Spektrum, when pushing hard on hills or hefty braking i got a new dizzling, crispling Sound in the Background....Not like the Standard Tinnitus Sound, more like a Sound like electricity flowing or sparkling! Just in the Background and only on hard manouvers.... Somebody else Hearing something like this? Please...no more questions if i'm mad ;-) Today I noticed a sound in the background a periodic "clak clak" that raise with speed, I switch off the wheel and rotate it manually but nothing, I tried by raising off ground the switched on wheel changing the angle to have less or more speed and at high speed I got the sparkling/dizzling but not what I hear while riding even at low speed ?!? Link to comment Share on other sites More sharing options...
Jason McNeil Posted August 24, 2016 Share Posted August 24, 2016 5 hours ago, Chriull said: data file the hop is 0,64 km within 30 ms while driving Sure thing. Is this the odometer reading? Link to comment Share on other sites More sharing options...
Chriull Posted August 24, 2016 Share Posted August 24, 2016 27 minutes ago, Jason McNeil said: Sure thing. Is this the odometer reading? I assume its the current kilometers driven (the values started at zero...) I have to ask @Blunznwhich app he has installed - i only got the datafile after he logged my wheel. the ks16 app(s?) should be on github. And if of interest/help i can provide the datafile, too. Link to comment Share on other sites More sharing options...
Blunzn Posted August 24, 2016 Share Posted August 24, 2016 23 minutes ago, Chriull said: I assume its the current kilometers driven (the values started at zero...) I have to ask @Blunznwhich app he has installed - i only got the datafile after he logged my wheel. it was WheelLog, latest version i think. the values in the csv log should be like date, time, speed, V, A, Watt, Battery, km, temp, fan 2016-08-23,13:48:30.761,12.81,60.47,4.75,287.23,65,0.64,40,0 2016-08-23,13:48:30.789,12.03,59.72,9.47,565.55,60,0.64,40,0 2016-08-23,13:48:30.800,12.03,59.72,9.47,565.55,60,0.64,40,0 2016-08-23,13:48:30.830,12.03,59.72,9.47,565.55,60,1.28,40,0 2016-08-23,13:48:30.865,13.42,60.56,4.89,296.14,66,1.28,40,0 2016-08-23,13:48:30.876,13.42,60.56,4.89,296.14,66,1.28,40,0 2016-08-23,13:48:31.078,12.50,60.35,5.25,316.84,64,1.29,40,0 Link to comment Share on other sites More sharing options...
JumpMaster Posted August 24, 2016 Share Posted August 24, 2016 At the end of a ride I'd recommend checking the distance with the official app. WheelLog checks if the new distance reading is greater then the last reading. Which should always be the case unless the wheel has been switched off. In the event the wheel is switched off/on it stores the previous distance and adds it to the new reading. If for some reason KingSong are incorrectly reporting values lower then previously reported it could show values incorrectly doubling as you've seen. Just checked a few of my logs and I don't see a doubling of distance at anytime. Link to comment Share on other sites More sharing options...
HEC Posted August 25, 2016 Share Posted August 25, 2016 15 hours ago, KingSong69 said: For the light mode remembering: The board has limited Memory: If they do the "remember light" Thing...we will loose AUX button functionality completly! (Thats perhaps why AUX works after powering on wheel, but not after using light Setting in the app) I'm sorry to say so but that's a total bullshit and most stupid excuse I've heard ever! To save the Light mode settings value all it takes it's measly 2 bits (well technically "one and half" if combined with some other values to save) and to retrieve that value and apply it to light after power up about 10 - 15 bytes of code max. @tinawong please get some real programmers instead and don't believe all that nonsense they're telling you ... And IF (and that a big if) that's really due to the "memory limitation" drop some useless feature like multiply blinking LED modes (just have 2 instead of 3, for example 1 and 3, the middle one looks like crap anyway) instead of dropping the most important safety feature - having the lights on is more important and even required by law in some countries unlike the "flying circus". Link to comment Share on other sites More sharing options...
HEC Posted August 25, 2016 Share Posted August 25, 2016 3 hours ago, JumpMaster said: At the end of a ride I'd recommend checking the distance with the official app. WheelLog checks if the new distance reading is greater then the last reading. Which should always be the case unless the wheel has been switched off. In the event the wheel is switched off/on it stores the previous distance and adds it to the new reading. If for some reason KingSong are incorrectly reporting values lower then previously reported it could show values incorrectly doubling as you've seen. Interestingly enough I was just about to ask you how "smart" is this feature as in some scenarios it could incorrectly can detect the loss of connection as "turned off" wheel and add the driven distances. What about verifying if the wheel has been really turned off or just lost the connection by reading and comparing also the Total Mileage and / or Working Duration values? This way you'd know for sure. Link to comment Share on other sites More sharing options...
JumpMaster Posted August 25, 2016 Share Posted August 25, 2016 5 hours ago, HEC said: Interestingly enough I was just about to ask you how "smart" is this feature as in some scenarios it could incorrectly can detect the loss of connection as "turned off" wheel and add the driven distances. What about verifying if the wheel has been really turned off or just lost the connection by reading and comparing also the Total Mileage and / or Working Duration values? This way you'd know for sure. It doesn't check if the wheel has turned off or disconnected. It compare the new distance to the previous which should always be the same or higher. If the distance is lower then the wheel must have been turned off (or some sloppy programming from KingSong). I'm going to add a 10 master buffer just in case. Link to comment Share on other sites More sharing options...
HEC Posted August 25, 2016 Share Posted August 25, 2016 2 hours ago, JumpMaster said: It doesn't check if the wheel has turned off or disconnected. It compare the new distance to the previous which should always be the same or higher. If the distance is lower then the wheel must have been turned off (or some sloppy programming from KingSong). I'm going to add a 10 master buffer just in case. I see. Technically though you can have the distance higher after reconnecting and still have the wheel switched off. Let's say you drive just a few kilometres, loose the connection, restart the wheel (without app / phone re-connecting for whatsoever reason) drive a little bit more and only then reconnect. Hence the recommendation to check either the total mileage value from the wheel as that's the only one "reliable" (leaving aside the FW distance calculation bug indeed). I do understand that that's quite unlikely to happen as in my example on other hand this way the app will be bulletproof for any period of disconnection / loss of signal. Link to comment Share on other sites More sharing options...
DS Posted August 27, 2016 Share Posted August 27, 2016 On 8/22/2016 at 10:23 PM, KingSong69 said: wouldnt it be better to check a "max Watt" on a full batterie? Here we go. Fully charged, after 10 hours on a 2A charger - 66V. I do not feel a difference when the wheel "folds", but the numbers seem to increase to 2400 Watts when fully charged. Now, maybe a good idea to test the wheel when only 10% charged? Link to comment Share on other sites More sharing options...
Chris Posted August 27, 2016 Share Posted August 27, 2016 So did they just change when the "Be caution, overpower" message is heard or did they acutally lower the power output? Does the euc fold earlier with this firmware? Link to comment Share on other sites More sharing options...
JumpMaster Posted August 27, 2016 Share Posted August 27, 2016 I believe I had an over power event this week. I have the voice disabled so there was no audible warning. I stopped at a traffic light in front of a speed bump so pulling away was also up hill. The wheel folded instantly. According to the graph in WheelLog the wheel peaked just above 40A. I would say the power is irrelevant as it's going to be minding that 40A fuse regardless of the voltage. Link to comment Share on other sites More sharing options...
HEC Posted August 27, 2016 Share Posted August 27, 2016 3 hours ago, DS said: Fully charged, after 10 hours on a 2A charger - 66V. I do not feel a difference when the wheel "folds", but the numbers seem to increase to 2400 Watts when fully charged. Now, maybe a good idea to test the wheel when only 10% charged? Maybe use the WeheelLog app for more detailed results? Link to comment Share on other sites More sharing options...
DS Posted August 27, 2016 Share Posted August 27, 2016 3 hours ago, Chris said: So did they just change when the "Be caution, overpower" message is heard or did they acutally lower the power output? Does the euc fold earlier with this firmware? Well, I would say that they actually lowered the power output since 1.20 FW. Now it is easier to over lean the wheel and harder to burn fuses/mosfets...IMHO. Link to comment Share on other sites More sharing options...
DS Posted August 27, 2016 Share Posted August 27, 2016 3 hours ago, JumpMaster said: I stopped at a traffic light in front of a speed bump so pulling away was also up hill. That's how I've noticed first time the new "feature" of my wheel in a similar situation. Then decided to do tests to see if the situation is repeatable....well, it is. Link to comment Share on other sites More sharing options...
DS Posted August 27, 2016 Share Posted August 27, 2016 1 hour ago, HEC said: Maybe use the WeheelLog app for more detailed results? Good idea. Here is it. For 10 seconds between 0:50 - 1:00 I jiggled without pushing the wheel hard. Link to comment Share on other sites More sharing options...
David S Posted August 27, 2016 Share Posted August 27, 2016 Anyone else think the tiltback is too aggressive on 1.22? If from a slow speed i max accelerate to top speed it tilts back really hard, the first time i did this it felt like it was about to throw me off the wheel. Link to comment Share on other sites More sharing options...
EricGhost Posted August 27, 2016 Share Posted August 27, 2016 5 hours ago, DS said: Good idea. Here is it. For 10 seconds between 0:50 - 1:00 I jiggled without pushing the wheel hard. If you activate the log it will register the power every 0.1 sec Link to comment Share on other sites More sharing options...
Chriull Posted August 28, 2016 Share Posted August 28, 2016 6 hours ago, David S said: Anyone else think the tiltback is too aggressive on 1.22? If from a slow speed i max accelerate to top speed it tilts back really hard, the first time i did this it felt like it was about to throw me off the wheel. Imho thats the normal behaviour with all wheels. The stronger you accelerate, the harder the tilt back kicks in. But i never tried this with 1.22 or any older firmware... This could be refined easily by a more intelligent firmware - the firmware knows your speed and acceleration and so could start the tiltback earlier to make a softer warning... Link to comment Share on other sites More sharing options...
vido Posted August 28, 2016 Author Share Posted August 28, 2016 @David S tilt back feels the same to me. Link to comment Share on other sites More sharing options...
HEC Posted August 28, 2016 Share Posted August 28, 2016 18 hours ago, DS said: Good idea. Here is it. For 10 seconds between 0:50 - 1:00 I jiggled without pushing the wheel hard I've meant using the actual Log function / Log file instead of video. That could be then also plotted into graph ... Link to comment Share on other sites More sharing options...
JumpMaster Posted August 28, 2016 Share Posted August 28, 2016 On 25/08/2016 at 9:51 AM, HEC said: I see. Technically though you can have the distance higher after reconnecting and still have the wheel switched off. Let's say you drive just a few kilometres, loose the connection, restart the wheel (without app / phone re-connecting for whatsoever reason) drive a little bit more and only then reconnect. Hence the recommendation to check either the total mileage value from the wheel as that's the only one "reliable" (leaving aside the FW distance calculation bug indeed). I do understand that that's quite unlikely to happen as in my example on other hand this way the app will be bulletproof for any period of disconnection / loss of signal. I've changed this on the next release as it appears to be working well. start odometer = odometer - current distance. Then going forward. Calculated distance = start odometer - odometer. So no matter how the wheel is connected and disconnected, the distance will be correct. Link to comment Share on other sites More sharing options...
DS Posted August 28, 2016 Share Posted August 28, 2016 5 hours ago, HEC said: I've meant using the actual Log function / Log file instead of video. That could be then also plotted into graph ... I have not used that log function. Maybe someone else more familiar could test it this way. Link to comment Share on other sites More sharing options...
JumpMaster Posted August 28, 2016 Share Posted August 28, 2016 Here's the log for the event I mentioned. Forgot I had auto log switched on. Great feature idea @HEC https://drive.google.com/file/d/0B_qIN0Xd80hCMnZQOUMxeWhRLVE/view?usp=sharing Peak amps of 40.98 @ 09:00:25. I've also improved writing to logs as the Wheel logs at 5hz and not 10hz. Hence the duplication of data. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.