Author Topic: EEVblog #751 - How To Debunk A Product (The Batteriser)  (Read 2369165 times)

0 Members and 1 Guest are viewing this topic.

Offline Howardlong

  • Super Contributor
  • ***
  • Posts: 4801
  • Country: gb
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7500 on: December 23, 2016, 03:37:41 pm »
I ordered a pack on their website a month ago just to see what they do, paying with PayPal and figuring that it will be easy to refund in the 90 day paypal buyers guarantee when they inevitably slip on their non-stop delays, but to my surprise they claim to have shipped my package yesterday. So kickstarter backers aren't even necessarily getting theirs before website purchasers!

And just like others have said, my tracking number does not work. It's a USPS number that draws a blank, and the email link goes to Fedex.
If your tracking number ends with "SE" and not "US", then it's not USPS, but a PostNord (Sweden Post) tracking number. Maybe try http://www.postnord.se/en/Personal/Pages/personal.aspx

Oooo, I just re-checked my email and an order that I placed on 22 Dec 2015 (yes, a year ago), was allegedly shipped on 21 Dec 2016. The tracking number's not in the system yet.
 

Offline FrankBuss

  • Supporter
  • ****
  • Posts: 2310
  • Country: de
    • Frank Buss
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7501 on: December 23, 2016, 04:21:59 pm »
A big thanks to Frank Buss for performing the very first tests on the Batteroo! 

Think of the headlines!!!

Batteroo guilty as charged with exaggerated battery performance claims

Batteriser campaign runs out of juice

Batteriser falls flat on 800% claims

Battery boosting campaign runs out on backers

Crowdfunding supporters discover their battery booster was free of charge all along

Lol, thanks, my pleasure. Looks like Batteroo didn't see my video so far, no dislikes and already one of my videos with the most likes, only topped by my Tetris video, with the awesome C64 music (not the annoying tune you might know for Tetris). Happy Holidays!
So Long, and Thanks for All the Fish
Electronics, hiking, retro-computing, electronic music etc.: https://www.youtube.com/c/FrankBussProgrammer
 

Offline snik

  • Contributor
  • Posts: 30
  • Country: de
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7502 on: December 23, 2016, 05:00:41 pm »
Quote

Lol, thanks, my pleasure. Looks like Batteroo didn't see my video so far, no dislikes and already one of my videos with the most likes, only topped by my Tetris video, with the awesome C64 music (not the annoying tune you might know for Tetris). Happy Holidays!

Here is the answer from Revive Batter(oo)ies with Video  ;) :

https://www.facebook.com/revivebatteries/posts/1303252479741119
« Last Edit: December 23, 2016, 05:05:10 pm by snik »
 

Offline quad

  • Regular Contributor
  • *
  • Posts: 113
  • Country: au
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7503 on: December 23, 2016, 05:09:33 pm »
Frank, I suspect a certain someone has found your video - it has one dislike now

One point, could I suggest you and anyone doing test videos to add both the word "Batteroo" and the word "Batteriser" somewhere in the searchable text, the title or description to get that extra search traffic

You are currently #1 video for "batteroo test" (private browser)
 

Offline CJay

  • Super Contributor
  • ***
  • Posts: 3494
  • Country: gb
  • M0UAW
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7504 on: December 23, 2016, 05:50:51 pm »
Even Oz's prospective distributor has verified our tests as accurate :)

https://www.facebook.com/revivebatteries/posts/1303252479741119
M0UAW
 

Offline quad

  • Regular Contributor
  • *
  • Posts: 113
  • Country: au
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7505 on: December 23, 2016, 06:11:37 pm »


Perhaps the batteries tested were AA and not AAA? The AA form factor could perform better?

Assuming it was a fair test setup -- which I think it was*
*Edit: It has been pointed out that the test should be repeated, with the Batteroo switched over to the other train to account for slight differences that may affect performance (such as friction).
It should also be conducted a few times to allow for differences in batteries (even when they're from the same packet). 

His results in condensed form:

Bare cell: Total Runtime - 7 hours 55 minutes
With Batteriser: Total Runtime 4 hours 55 minutes

With Batteriser it ran faster but for a shorter period of time
« Last Edit: December 23, 2016, 07:03:41 pm by quad »
 

Offline FrankBuss

  • Supporter
  • ****
  • Posts: 2310
  • Country: de
    • Frank Buss
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7506 on: December 23, 2016, 06:13:30 pm »
Here is the answer from Revive Batter(oo)ies with Video  ;) :

https://www.facebook.com/revivebatteries/posts/1303252479741119

Thanks, I commented it, and my photo he posted earlier. Amazing how fast Wayne could setup and execute an apparently carefully designed and documented test, with lap counter and speed measurement, if you compare this to his first flashlight tests.
So Long, and Thanks for All the Fish
Electronics, hiking, retro-computing, electronic music etc.: https://www.youtube.com/c/FrankBussProgrammer
 

Offline quad

  • Regular Contributor
  • *
  • Posts: 113
  • Country: au
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7507 on: December 23, 2016, 06:27:58 pm »
Thanks, I commented it, and my photo he posted earlier. Amazing how fast Wayne could setup and execute an apparently carefully designed and documented test, with lap counter and speed measurement, if you compare this to his first flashlight tests.

In his post, he says, "The test was set up by a reliable and honest person with experience in electronics."
« Last Edit: December 23, 2016, 06:31:05 pm by quad »
 

Offline Hensingler

  • Regular Contributor
  • *
  • Posts: 144
  • Country: gb
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7508 on: December 23, 2016, 06:29:06 pm »
Amazing how fast Wayne could setup and execute an apparently carefully designed and documented test

He doesn't say he did. It was done by an unnamed professional engineer. My guess would be employed by batteroo and they have been sitting on it because 7 to 4 hour battery life reduction is too embarrassing even with the batterooed train running faster.

I'm sure they can find other cheap junk products that ought but don't have regulators which show improved performance with batteroos, but, why spend money on batteroos and have the hassle of using them when you could spend it on something less junk in the first place?
 

Offline FrankBuss

  • Supporter
  • ****
  • Posts: 2310
  • Country: de
    • Frank Buss
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7509 on: December 23, 2016, 06:45:04 pm »
Thanks, I commented it, and my photo he posted earlier. Amazing how fast Wayne could setup and execute an apparently carefully designed and documented test, with lap counter and speed measurement, if you compare this to his first flashlight tests.

In his post, he says, "The test was set up by a reliable and honest person with experience in electronics."

Right, I missed this. Why don't they just say that Batteroo did the test? And probably Dave should implement the lap counter, they complaint this about my video. Still, I think the results are pretty obvious. Picking out one specific point in time in the video doesn't demonstrate much. And they should post their full test setup they were using, like the train model, so that it can be verified independently, the usual scientific method.
« Last Edit: December 23, 2016, 06:56:48 pm by FrankBuss »
So Long, and Thanks for All the Fish
Electronics, hiking, retro-computing, electronic music etc.: https://www.youtube.com/c/FrankBussProgrammer
 

Offline Towger

  • Super Contributor
  • ***
  • Posts: 1563
  • Country: ie
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7510 on: December 23, 2016, 06:47:50 pm »
Well DHL has let us down. they missed their own ETA and the current status is "The Estimated Delivery Date is currently unavailable. Please try again later."

Oh well, at least they have arrived down under.

Quote
Saturday, December 24, 2016   Location   Time   Piece
18   Customs status updated   SYDNEY - AUSTRALIA   00:22
 

Offline madires

  • Super Contributor
  • ***
  • Posts: 4888
  • Country: de
  • A qualified hobbyist ;)
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7511 on: December 23, 2016, 07:16:07 pm »
Thanks, I commented it, and my photo he posted earlier. Amazing how fast Wayne could setup and execute an apparently carefully designed and documented test, with lap counter and speed measurement, if you compare this to his first flashlight tests.

In his post, he says, "The test was set up by a reliable and honest person with experience in electronics."

And another "busted" for Batteroo's claims. The only benefit is having 1.5V until the battery dies, which is useful when performance of a simple load, e.g. a motor, matters. But with the drawback of missing an indication about the battery going flat. Actually, this renders the Batteroo sleeves also for the Aussie distributor's toys useless, because they might stop vibrating just at the right moment >:D
 

Offline razvanme

  • Contributor
  • Posts: 31
  • Country: ro
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7512 on: December 23, 2016, 07:27:31 pm »
We need to wait for the first Apple keyboard test. I think they can`t do / say anything about that. Keeping a key pressed and counting them with a small program. That should discharge the batteries a lot faster maybe 24hr`s? The number of key`s pressed + the battery indicator should clarify the situation with their own example.
 

Offline quad

  • Regular Contributor
  • *
  • Posts: 113
  • Country: au
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7513 on: December 23, 2016, 07:55:54 pm »
I'm working on small python script now that will count the number of keystrokes.

If I do one for Linux, Windows and Mac we can test all our wireless keyboards.

Any way to get a ball park estimate how long a wireless keyboard will last under these circumstances? < 12 hours? 
 

Online PA0PBZ

  • Super Contributor
  • ***
  • Posts: 4211
  • Country: nl
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7514 on: December 23, 2016, 08:06:00 pm »
In his post, he says, "The test was set up by a reliable and honest person with experience in electronics."

I think it's Dave  :box:
Keyboard error: Press F1 to continue.
 

Offline onlooker

  • Frequent Contributor
  • **
  • Posts: 385
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7515 on: December 23, 2016, 08:09:27 pm »
I guess that unnamed professional engineer must be from USA (or Liberia or Burma) since he is so used to the English system and logging speed using "feet per minute".

https://www.facebook.com/revivebatteries/posts/1303252479741119 chooses not to disclose the comparison of the total number of laps, so we may have some holiday math fun.

1). Their data:
         no_sleeve  with_sleeve
time  speed        speed           delta_laps
(h)    (ft/m)       (ft/m)
0       43.3         43.3               0
1/3    37.1-       37.1               0+
1       32.5         37.1              31
2       28.9         37.1              100
4       26.0         32.5              307
5       23.6         32.5              409
8       15.3         0

2). The track length:
Using the data for hour 5 and 4, duration=60m, delta_speed=32.5-0.5*(26+23.6)=7.7 ft/m, delta_laps=102.

track_length = 7.7*60/102 = 4.53 ft

3). extra laps during the last 3 hours for the test w/o the sleeve:
Using data for hour 8 and 5, duration=180m, 23.6 >= speed >=15.3, track_length=4.53 ft

    938 >= extra_laps >= 608

The actual number can be estimated to be the middle,
   extra_laps ~= 773 laps.

4). Overall, test without sleeve runs 363 more laps:
At the 5th hour when test2 stopped, it did 409 more laps, but after test2 stopped test1 did ~773 more laps. Therefore, overall, test without sleeve runs 363 more laps than with a sleeve.

Similarly, we can estimate the total number of laps:

test without sleeve  ~= 2700 laps.
test with     sleeve  ~= 2300  laps.

If only they are willing to release their full video, we can see the real numbers and compare.
« Last Edit: December 23, 2016, 08:38:19 pm by onlooker »
 

Offline FrankBuss

  • Supporter
  • ****
  • Posts: 2310
  • Country: de
    • Frank Buss
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7516 on: December 23, 2016, 08:37:49 pm »
The only benefit is having 1.5V until the battery dies, which is useful when performance of a simple load, e.g. a motor, matters. But with the drawback of missing an indication about the battery going flat.

And even this might be not the case. I measured the output voltage for different input voltages with different loads, see the second diagram here. But it is interesting that the output voltage is pretty stable, regardless of the current, but still much less thant 1.5V. As mentioned by amyk, the reason might be that battery indicators still work.

I still think they have a product that might be useful for some niche applications. They should just post all data so that the user can decide. Couldn't find this output voltage behaviour on their FAQ or production description page. They might be even be able to sell their chip without the sleeve with a good datasheet.

https://www.facebook.com/revivebatteries/posts/1303252479741119 chooses not to disclose the comparison of the total number of laps, so we may have some holiday math fun.

You can find this method in many postings, videos and articles from Batteroo. They disclose what looks good for their product and don't tell you what doesn't look good. Did you notice that they didn't tell anything about the MP3 player test results I posted? You can see all results in detail of the product tests so far here. You can ask Dave if you want to add your result, if you have some proof for it, like a video.
So Long, and Thanks for All the Fish
Electronics, hiking, retro-computing, electronic music etc.: https://www.youtube.com/c/FrankBussProgrammer
 

Offline richnormand

  • Supporter
  • ****
  • Posts: 371
  • Country: ca
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7517 on: December 23, 2016, 08:41:58 pm »

With several comments on clocks, keyboards and toys some tests could take a while or need datalogging in a compact form.
Has anyone though of using these units from Lascar. (Disclaimer: I am not associated with them but have used their products in the past to monitor car batteries.)
https://www.lascarelectronics.com/data-loggers/voltage-current/

The data probing can be set from seconds, minutes, hours up to months. I could see one to monitor battery voltage, sleeve output voltage and current delivered.
It would be then easy to compute energy, power, etc... over very long sampling periods. If the instrument being monitored is mobile the Lascar could just be taped on it.



« Last Edit: December 23, 2016, 08:45:10 pm by richnormand »
 

Offline quad

  • Regular Contributor
  • *
  • Posts: 113
  • Country: au
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7518 on: December 23, 2016, 09:33:57 pm »
Here is a quick keystroke counter script I wrote for Linux: http://pastebin.com/s8YxSpqz

Fire up the script and weigh down a key on your wireless keyboard, and wait. How long? I have no idea

Next up: Windows version
« Last Edit: December 23, 2016, 09:45:48 pm by quad »
 

Offline trophosphere

  • Regular Contributor
  • *
  • Posts: 217
  • Country: us
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7519 on: December 23, 2016, 09:55:48 pm »
@quad

Is there any difference between using the script you wrote and using something like Microsoft Word to show the letter count? Even Notpad++ has a letter counter which should be sufficient unless I am missing something.
 

Offline quad

  • Regular Contributor
  • *
  • Posts: 113
  • Country: au
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7520 on: December 23, 2016, 10:24:24 pm »
@quad

Is there any difference between using the script you wrote and using something like Microsoft Word to show the letter count? Even Notpad++ has a letter counter which should be sufficient unless I am missing something.

I'm not sure how long the test will go on for, and how many characters those programs will take without risk of lagging out and crashing... and the specs of the test computer i.e. RAM. I've dealt with some large text files, some text editors just don't handle them well even with 16GB of RAM and a fairly modern i7 processor. Sublime Text is very good with large files.

Also the number of characters will depend on your system's keyboard repeat delay/rate settings.

The python script doesn't store the characters, just the count.
« Last Edit: December 23, 2016, 10:28:42 pm by quad »
 

Offline trophosphere

  • Regular Contributor
  • *
  • Posts: 217
  • Country: us
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7521 on: December 23, 2016, 10:33:11 pm »
I'm not sure how long the test will go on for, and how many characters those programs will take without risk of lagging out and crashing... and the specs of the test computer i.e. RAM. I've dealt with some large text files, some text editors just don't handle them well even with 16GB of RAM and a fairly modern i7 processor. Sublime Text is very good with large files.

Also the number of characters will depend on your system's keyboard repeat delay/rate settings.

The python script doesn't store the characters, just the count.

Makes sense. Thanks for taking the time to make the scripts.
 

Offline samgab

  • Frequent Contributor
  • **
  • Posts: 423
  • Country: nz
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7522 on: December 23, 2016, 10:55:24 pm »
I dislike the lack of transparency with the "Revive" version of the train test.
I prefer to see the full data not "I won't bore you with the technical stuff which is probably too hard for you".
I prefer to see the full video time lapsed including evidence of the brand new battery blister pack being opened and the batteroo being put on one specific battery and being inserted into a specific train and another battery out of the same pack being put into the other, and then both trains being run in an unbroken cycle so I can see for myself how it was done (as was done by Frank in his video). Gives confidence in the methods being used, so I can see that they haven't made any glaring "mistakes".
I prefer to see the raw data, not a few cherry picked data points.
I also prefer to see the whole video, not just a little cherry picked snippet of the video.
I prefer to be given honest and open information about the person who is doing the tests, not all cloak and dagger stuff. It all reeks of "look away for a moment... aaaaand, look, the rabbit has disappeared! Just trust me, it really did."
If the "professional" in question posts a full video on youtube with more info, maybe it will be a viable test. Otherwise, all I see is a small video snippet with 2 trains going around 2 tracks for a short period of time, which establishes nothing useful whatsoever. The results are very different from what I saw in the other train test. Maybe as someone suggested it does come down to a difference between AA and AAA versions? But we can't know this because we know nothing about the batteries used or the toy trains used or the current drawn. What brand are the batteries? What expiration date is on them? Are they AA or AAA? Is it a single cell per train or a pair? Did the batteries in each train come out of the same blister pack? How many times was the test repeated? These are some of my questions.
 

Offline johndoe123

  • Contributor
  • Posts: 30
  • Country: au
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7523 on: December 23, 2016, 11:11:59 pm »
Frank, I suspect a certain someone has found your video - it has one dislike now

One point, could I suggest you and anyone doing test videos to add both the word "Batteroo" and the word "Batteriser" somewhere in the searchable text, the title or description to get that extra search traffic

You are currently #1 video for "batteroo test" (private browser)

Thats why revivebatteries test results should not be discussed on their facebook website. They will delete negative comments anyway.
 

Offline FrankBuss

  • Supporter
  • ****
  • Posts: 2310
  • Country: de
    • Frank Buss
Re: EEVblog #751 - How To Debunk A Product (The Batteriser)
« Reply #7524 on: December 23, 2016, 11:59:20 pm »
Any way to get a ball park estimate how long a wireless keyboard will last under these circumstances? < 12 hours?

I got an old Apple wireless keyboard, model A1016, which needs 4 AA batteries. It needs about 0.8 mA and with spikes of 2 mA when you type. So would last for about 2 months. When you search in Google, it says 1 month with 7 hours per day. I guess it needs more current when the voltage is lower, so expect at least a week. Newer keyboards might need less power.

Regarding an earlier post about this keyboard type: It showed 39% without Batteroo and 87% with Batteroo. I checked the voltages, some values for the Mighty Monitor (or the context menu in the bluetooth symbol; it has a really slow update, fastest is to turn off the power supply, wait until connection lost, change the voltage, then turn it on again) :
3.9 V: 5 % (it doesn't turn on reliable below this level)
4 V: 11%
4.1 V: 15 %
4.2 V: 18 %
4.5 V: 30 %
4.7 V: 36 %
4.8 V: 40 %
4.9 V: 44 %
5.7 V: 71 %
5.9 V: 78 %
6.1 V: 85 %
6.14 V: 89 %
6.2 V: 91 %
6.4 V: 97 %
6.5 V: 100 %

So the Batteroo sleeves boosted from 4.8 V to 6.14 V, which means from 1.2 V per cell to 1.54 V per cell. And the keyboard still works with 0.98 V per cell. This means there is only less than 10 % power left for Batteroo, not 80 %. In combination with the boost converter losses, which might be even higher for such low currents, I would expect no advantage, compared to running it without the sleeves. But you might need some months patience to do all the tests.

If you want to get the battery level from a script, you can do this:
Code: [Select]
ioreg -l | grep BatteryPercent

This searches the whole registry, could be optimized, but Apple changed the path between different MacOSX versions, so this should work always. You could paste the output of "ioreg -l" to a file and search for it, then use "-n" to get the attributes for the specified object, only, which is a lot faster. On Mac OSX Sierra this works:

Code: [Select]
ioreg -n "IOAppleBluetoothHIDDriver"|grep BatteryPercent

Output looks like this:

Code: [Select]
      | |   |   |   "ExtendedFeatures" = {"FactoryDefault"={"id"=69,"type"=2},"RecantConnection"={"id"=65,"type"=2},"DeviceNameChange"={"id"=80,"type"=2},"DeviceName1"={"id"=81,"size"=8,"type"=2},"DeviceName2"={"id"=82,"size"=8,"type"=2},"DeviceName3"={"id"=83,"size"=8,"type"=2},"DeviceName4"={"id"=84,"size"=8,"type"=2},"BatteryState"={"size"=1,"id"=48,"min"=0,"max"=2,"type"=0},"BatteryPercent"={"size"=1,"id"=71,"min"=0,"max"=100,"type"=2},"WillShutdown"={"id"=64,"type"=2},"FullFactoryDefault"={"id"=68,"type"=2},"UserMode"={"size"=1,"id"=67,"min"=1,"max"=3,"type"=2}}
      | |   |   |   "BatteryPercent" = 86

Should be easy to parse in Python, to log the battery percentages, to make it a bit more professional. Battery percentages can be directly mapped to voltages, no need for an external logger (except if you want to log the real battery output, too, before the sleeve, and the current, but this could be derived as well from the out output voltage, if there is no special behaviour in the regulator chip).

I don't have the Batteroo sleeves anymore, and not the time to do the month long test (I have a Mac Book and need it sometimes), but if someone wants the keyboard for a test, I can send it to you. Was cheap, used, and is an old model, I bought it just to measure the voltages :)
So Long, and Thanks for All the Fish
Electronics, hiking, retro-computing, electronic music etc.: https://www.youtube.com/c/FrankBussProgrammer
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo