EEVblog #751 - How To Debunk A Product (The Batteriser)

[McBryce]

They are only handy if you play golf as the course layouts are down loaded into the unit and they can also tally up your shot count, I wanted to ask Sky if they save the shot count after the batteries die as that might be awkward in a serious match, anyway no good for general GPS usage just for ball whackers.

If your golf game is taking more than 15 hours, it's time to consider taking up a different sport

McBryce.

[FrankBuss]

That response from Batteriser just makes me want to go and get one of these GPS's and repeat the test myself...

waste of time

the 15 to 17 hour difference is ~ a 15% variance, not a 600% variance their test results indicated.

The 15% could easily be the difference of battery manufacturer used in their tests versus 5KY's test. Most of the difference is likely an "under promise and over deliver" factor in Garmin's claim for battery life.

No, the problem is that Batterise claims, it works for 1:52 hours, only, so an independent test might be useful. Not because we doubt the credibility of 5KY, but it is scientific practice to reproduce tests independently to be sure there is no error.

[Muttley Snickers]

If it's the same model Garmin G3 Approach Golfing GPS a quick search indicates that they are readily available from JB Hi Fi and Harvey Norman for around two hundred, I thought that they were a superseded model but obviously not.

And apparently they wont indicate where your lost ball is or help with obstacles such as kangaroos or wombats so pretty useless for my game really, 15 hours for me is one par 3, I wont be rushed by anybody.

[Jay_Diddy_B]

Hi group,

I have an experiment under way to measure the performance of a Duracell Procell AA battery.

The test setup

a) HP3478A DMM for voltage measurement
b) HP6060A electronic load for discharge current

The test and data logging is being performed by a PC using HPIB.

To compress the data, the time and voltage are recorded when the battery voltage changes by more than 5mV, rather than recording the data at fixed time interval.

In addition every time the voltage changes by more than 50mV, the load is turned-off for 4 seconds. The battery voltage is measured twice while the load is off. This additional data point can be used to calculate ESR.

The DMM leads and load leads are connected directly to the battery, (Kelvin connection).

Starting with a fresh AA battery, I am now 40,000 seconds (11 hours) into the test and the battery voltage is 1.1V, this is with a constant load current of 150mA.


Here is some preliminary data, from a partially discharged cell, gathered while I was testing the software:




I should have good data later today.

Regards,

Jay_Diddy_B

[Godzil]

If it's the same model Garmin G3 Approach Golfing GPS a quick search indicates that they are readily available from JB Hi Fi and Harvey Norman for around two hundred, I thought that they were a superseded model but obviously not.

And apparently they wont indicate where your lost ball is or help with obstacles such as kangaroos or wombats so pretty useless for my game really, 15 hours for me is one par 3, I wont be rushed by anybody.

There is one on eBay for (currently) less than £60:
http://www.ebay.fr/itm/Garmin-Approach-G3-GPS-/141779419943?hash=item2102b64327

[Delta]

Good stuff Jay.  How about a constant power discharge run, as I think that is the most realistic comparison to modern devices, like golf GPSs etc...

[Joule Thief]

That response from Batteriser just makes me want to go and get one of these GPS's and repeat the test myself...

waste of time

the 15 to 17 hour difference is ~ a 15% variance, not a 600% variance their test results indicated.

The 15% could easily be the difference of battery manufacturer used in their tests versus 5KY's test. Most of the difference is likely an "under promise and over deliver" factor in Garmin's claim for battery life.

No, the problem is that Batterise claims, it works for 1:52 hours, only, so an independent test might be useful. Not because we doubt the credibility of 5KY, but it is scientific practice to reproduce tests independently to be sure there is no error.

Two data points still would not be scientifically "significant".
I believe even if you redo 5KY's test to a statistically significant 20+ data points,  no Batteriser minds would be changed and no changes to claims would be seen on their web site.

[EEVblog]

Two data points still would not be scientifically "significant".

It doesn't work like that.
When you are dealing with batteries, they only have X amount of energy.
You can goof the test and get less time than you should, but you can't get goof the test and get more than you should. The product either works or it doesn't for X amount of time.
Even one test is proof enough, two would nail it home.

[EEVblog]

The Gizmodo article has over 165,000 views:
http://gizmodo.com/dont-buy-the-bullshit-this-indiegogo-campaign-is-sellin-1732402971
It will very shortly overtake my video in popularity.

[firewalker]

The Gizmodo article has over 165,000 views:
http://gizmodo.com/dont-buy-the-bullshit-this-indiegogo-campaign-is-sellin-1732402971
It will very shortly overtake my video in popularity.

The Gizmodo article is watched by "normal" people. I guess your audience is more technical people, that would understand the baloney by them selves. If Gizmodo had this article two months ago, it would be a significant hit for their campaign.

Alexander.

[Godzil]

The Gizmodo article has over 165,000 views:
http://gizmodo.com/dont-buy-the-bullshit-this-indiegogo-campaign-is-sellin-1732402971
It will very shortly overtake my video in popularity.

Dave I wonder, as this article make a link to you video does it have change anything on the audience of your video? I had one of my project on the Hack-A-Day front page at one time, and it made really big spike on my blog and related video visits for a few days

[hauptbr09]

Two data points still would not be scientifically "significant".

It doesn't work like that.
When you are dealing with batteries, they only have X amount of energy.
You can goof the test and get less time than you should, but you can't get goof the test and get more than you should. The product either works or it doesn't for X amount of time.
Even one test is proof enough, two would nail it home.

Agreed. It would be ideal in an academic setting, but this is more or less EE 101. A handful of tests are sufficient

Sent from my SM-N900V using Tapatalk

[EEVblog]

Dave I wonder, as this article make a link to you video does it have change anything on the audience of your video?

I don't have the stats for a few days, they are not instant.

I had one of my project on the Hack-A-Day front page at one time, and it made really big spike on my blog and related video visits for a few days

Hack-a-day used to give me video a big boost way back at the start, but now it's down in the noise, like a few hundred views.
Similar with Slashdot.
Hacker News gets a lot more views than both of those combined.

[Delta]

It is excellent that a more mainstream site like Gizmodo have picked up on this.

Ideally, Gizmodo should now get hold of a G3 gps themselves and repeat the test. No data logging needed, just a tapper, a clock and a timelapse camera.

No disrespect to 5ky, but it's one thing calling a youtuber a liar, but would they dare do that to a commercial setup like Gizmodo?

[amyk]

No disrespect to 5ky, but it's one thing calling a youtuber a liar, but would they dare do that to a commercial setup like Gizmodo?

If anything I think they'd be even more inclined to say Gizmodo is "working for big battery", but then again, I don't know how these people think...

[Delta]

I'm tempted to buy one myself (there's a couple of them on ebay uk), test it myself, then send it to Gizmodo...

[cezar]

There is an analogy between UL/Bateroo testing Garmin and Volkswagen testing engines :-)

[Fungus]

Anyway I was tenpted but walked out with an AEG 1300 watt sabre saw instead, handy for cutting sandwiches you know

That's a really neat idea!

Maybe you guys need to improve your bread making technique.

Use a high gluten flour and real yeast, not that 'self raising' stuff. Knead for at least 10 minutes (or five minutes in mixer).

[Fungus]

The Gizmodo article has over 165,000 views:
http://gizmodo.com/dont-buy-the-bullshit-this-indiegogo-campaign-is-sellin-1732402971
It will very shortly overtake my video in popularity.

Is that with or without Vietnamese viewers? I heard it was very popular over there.

[Godzil]

Dave: Thanks, good to know about hacker news

[Jay_Diddy_B]

Hi group,

Here are my test results from discharging a Duracell Procell AA Alkaline at 150mA.

The test setup

a) HP3478A DMM for voltage measurement
b) HP6060A electronic load for discharge current

The test and data logging is being performed by a PC using HPIB.

To compress the data, the time and voltage are recorded when the battery voltage changes by more than 5mV, rather than recording the data at fixed time interval.

In addition every time the voltage changes by more than 50mV, the load is turned-off for 4 seconds. The battery voltage is measured twice while the load is off. This additional data point can be used to calculate ESR.

The DMM leads and load leads are connected directly to the battery, (Kelvin connection).



Regards,

Jay_Diddy_B

[Galenbo]

There is an analogy between UL/Bateroo testing Garmin and Volkswagen testing engines :-)

I see an antipole.
One promises and talks a lot, has no clue and will not meet specs.
Two says nothing totally silent, knows everything, meets "specs".

:-)

[edy]

Here's an email from Batteriser announcing 14 hours left to go in the campaign.... Read the last sentence in the 2nd last paragraph (I've bolded)... Very troubling!

A message from Batteriser’s CEO
 
Dear Friends,

It is with great pleasure that we embark on the final hours of the Batteriser Indiegogo campaign. We have been blessed with tremendous support by over 7,000 contributors from 114 countries, raising over $387,000 to date. Our work is not done yet. In fact, it is just the beginning. Billions of batteries will be thrown away this year that would have lasted far longer with Batteriser.

Thanks to your help, we will make Batteriser a reality. I ask as we near our Indiegogo campaign’s conclusion, that you might take just a few moments to share a word about Batteriser with your friends and family. Every bit of support helps for development of 9V and rechargeable versions of Batteriser. We are also developing military-grade Batterisers so our troops will no longer need to haul 16 lbs of batteries on an average 72 hour mission.

It is only because of forward-thinking individuals like yourself that innovative products come to life. We look forward to introducing Batteriser to the world together.

Warmest regards,
Bob Roohparvar, Ph.D.
Chief Executive Officer
Batteroo, Inc.

Ok... So they are, and I quote "also developing military-grade Batterisers so our troops will no longer need to haul 16 lbs of batteries on an average 72 hour mission."       Way to pull at the heart-strings of the American public. Next they will be developing Batterisers for hospital medical use in Children's hospitals. Associate yourself with charitable and patriotic supporters... That always boosts sales.

[Jay_Diddy_B]

Hi,

A quick (in)sanity check:

16 lbs of batteries

AA battery typically weighs 0.8oz (Panasonic datasheet)

Therefore 20 batteries to the lb.

16 lbs is the equivalent of 320 AA batteries for a 72 Hour mission

I am sure that they probably use more exotic batteries, LTO etc.

And again I am sure much of the equipment has switching supplies.


Jay_Diddy_B

[rollatorwieltje]

Hi,

A quick (in)sanity check:

16 lbs of batteries

AA battery typically weighs 0.8oz (Panasonic datasheet)

Therefore 20 batteries to the lb.

16 lbs is the equivalent of 320 AA batteries for a 72 Hour mission

I am sure that they probably use more exotic batteries, LTO etc.

And again I am sure much of the equipment has switching supplies.


Jay_Diddy_B

To be fair, the expression "There's the good way, the bad way, and there's the army way" exists for a reason