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

[Mr.B]

That chart above is pure justice.
Over a year ago, Boob originally claimed that battery powered devices cut out at 1.3v or so.
Here we see that the Batteriser/oo itself cuts out at approximately 1.3v.

Edit: My bad. As pointed out, it is the output voltage.

[twice11]

That chart above is pure irony justice.
Over a year ago, Boob originally claimed that battery powered devices cut out at 1.3v or so.
Here we see that the Batteriser/oo itself craps cuts out at approximately 1.3v.

 

Wait a second. That graph shows the output voltage of the batteriser. The input voltage is below 1.1V at that point. So the batteriser in fact helps to keep devices that cut out at 1.3v or so happy as long as possible.

[FrankBuss]

Wait a second. That graph shows the output voltage of the batteriser. The input voltage is below 1.1V at that point. So the batteriser in fact helps to keep devices that cut out at 1.3v or so happy as long as possible.

Right, this is the only useful application for the sleeve. When you carefully select the most crappiest device you can find, like the camera in the video from Batteroo, you might get similar results like they measured, 3.5x times longer for the camera. This is not 8x, but still pretty impressive, if it is true. But as we've seen with the Garmin GPS test, their testing methods might have some VW diesel car qualities.

[Delta]

 Who was the bloke on here who had that exact same Garmin GPS unit?  Have any Batterooiserer sleeves found their way to him yet...?   

[Mr.B]

5ky
He has not been here for just over a year unfortunately.

[samgab]

Yes, but he mentioned in his videos that he backed the project in order to get some sleeves for testing. So he's probably received the sleeves by now. I hope he's still around, and does an update. He has good equipment and test methods...

[Hugoneus]

Hi everyone,

I know these sleeves are in short supply, however if someone wants to send me one, I can perform multi-dimensional sweep on them because I have 4 SMUs. I can sweep the input/output voltage and current as well as pulse tests. I can also do tests as a function of temperature.

If you ever watched my review videos of the Keithley SMUs, I show experiments where I programm the SMUs to perform non-linear IV characteristics which may be interesting to try.

However I do feel that aside from satisfying curiosity, no test anymore will impact anything. This product has been busted for its insane claims already.

Cheers,

[Someone]

And would be cool to see your SMU in action to do the characteristics curves for efficiency, input/output voltage curves etc. I know you don't like to write software or setup some complicated test, but some viewers (like me ) might be interested how easy (or complicated) it is to control the outputs of it and read the measurements from a PC with a script, in comparison to EEVblog #957. Once you've done this, the experience you gained from it might be useful for other projects as well (uSupply?).

Its relatively easy to do when you have a multichannel SMU, there are some Keysight videos here:
https://youtu.be/Q4cX1FUe224
https://youtu.be/BKA00RxMCBM
Automating multiple operating points would be possible but its often quicker to put a human in the loop for the slower iterations. For other Agilent/Keysight SMUs similarly quick sweeps for these tasks can be done with the simpler Quick IV measurement software, but I'm unfamiliar with other offerings.

[FrankBuss]

Its relatively easy to do when you have a multichannel SMU, there are some Keysight videos here:

Thanks, I didn't expect that Keysight manufactures a dedicated instrument like the N6705B just for this type of applications. But looks like there is some demand for it. I guess if you develop or have to test lots of power supplies, it doesn't matter that it costs $7,765 (at Newark).

But for this price I would have expected better software. As you mentioned, looks like it doesn't measure multiple operating points, just one sweep for one current, or it might be complicated to automate it with the software. Even the output power is displayed inverted and the graph doesn't look very nice, missing labels, and ticks and values on the axis.

@Hugoneus: I just subscribed your Youtube channel, your TSP #36 video is really interesting and informative. Would like to see how you can script your two Keithley 2450 SMUs that you have. The graphical flow diagram interface looks a bit cumbersome, I would be faster writing some lines of code, but maybe it is useful. I hope someone sends you some Batteroo sleeves.

[Hugoneus]

@Hugoneus: I just subscribed your Youtube channel, your TSP #36 video is really interesting and informative. Would like to see how you can script your two Keithley 2450 SMUs that you have. The graphical flow diagram interface looks a bit cumbersome, I would be faster writing some lines of code, but maybe it is useful. I hope someone sends you some Batteroo sleeves.

Thanks!

I also have the review of the 2460 on my channel, there you can see some more multi dimensional sweeps.

[Someone]

Its relatively easy to do when you have a multichannel SMU, there are some Keysight videos here:

Thanks, I didn't expect that Keysight manufactures a dedicated instrument like the N6705B just for this type of applications. But looks like there is some demand for it. I guess if you develop or have to test lots of power supplies, it doesn't matter that it costs $7,765 (at Newark).

But for this price I would have expected better software. As you mentioned, looks like it doesn't measure multiple operating points, just one sweep for one current, or it might be complicated to automate it with the software. Even the output power is displayed inverted and the graph doesn't look very nice, missing labels, and ticks and values on the axis.

Thats just the base price for the chassis, you add some channels the price runs away quickly especially with the 2 and 4 quadrant SMU type modules. Its often cheaper to buy a "regular" SMU. The free software gets the job done, its good for investigation work but if you need flexibility then you quickly end up doing some sort of coding. Keysight offer BenchVue with the paid Test Flow option as their software for automation:
https://www.eevblog.com/forum/testgear/t57461/
But it has an ongoing cost to maintain a license, unlike the alternatives. I'm happy enough with the free tools and already offered to do testing of the batteriser batteroo sleeve if anyone was interested in what can be done with the "pro" tools.

[FrankBuss]

Keysight offer BenchVue with the paid Test Flow option as their software for automation:

I think I tested this software some time ago with my oscilloscope. IIRC it was a big bloated software package, and somehow slowed down the PC a lot.

I think open source software is perfect for automation. If you don't know it already, there is an interesting project where you can use Python scripts to control and read your instruments, see here:

https://www.eevblog.com/forum/metrology/raspberry-pi23-logging-platform-for-voltnuts/

All you need for Linux is an USB GPIB interface, and for more modern instruments just a USB connection. For example most Keysight instruments uses USBTMC, which is supported by a Python library, so it should work even on Windows, and with a Linux kernel module. There are already an impressive list of examples.

[HKJ]

My 10mA test is finished, the sleeve lasted for 10½ day. The first 10 day the voltage was above 1.3V, the last ½ day it dropped below 1.1V before the sleeve quit.

[dr_frost_dk]

My 10mA test is finished, the sleeve lasted for 10½ day. The first 10 day the voltage was above 1.3V, the last ½ day it dropped below 1.1V before the sleeve quit.

Well that's the last fail for the batteriser, it can't even go as long as without the batteriser and that is compared to a cutoff of 1.1V (without) and some devices even work down to 0.9V, and that just mean a bigger fail as you posted under the graph that it got 14 days down to 0.8V without the batteriser. 

[Blocco]

More proof that Batteroo significantly extends battery life....

https://youtu.be/x62rhxWfmgk

[Muttley Snickers]

More proof that Batteroo significantly extends battery life....

More proof that these idiots need to focus on the truth and stop spinning bullshit....

[samgab]

More proof that Batteroo significantly extends battery life....

I have to say, Batteroo, after learning and getting tips from this blog, have finally done a video demo test that covers off some of the very basic requirements. For instance they're comparing like for like, they show the brand new cells being taken out of the pack, they show the equipment and the timer, and it's in one take, and results are shown in a clear chart. Props to them for that. Pity they're still using that awful oleaginous-voiced woman to do the narration. And they stopped the test when the non-batteroo'd torch was still going, and still probably had hours of dim glow left in it.
And yes, as expected (and not denied by anyone), their product works as a boost converter (albeit one which can't maintain a steady fixed output voltage), and they've chosen the one-in-a-million device where using the Batteroo is maybe valid: A dumb, old-fashioned incandescent torch with no built-in boost conversion which runs on a pair of Alkaline D cells. Basically a switch and a bulb in a can.

Trouble is, the world has moved on. No one I know still uses crappy old incan torches anymore (even my parents who are in their 70's and aren't "techie" at all). Nowadays, everyone uses LED torches with fantasic buck/boost converters built in, multiple modes, far more efficiency and brightness, and much more robust than the old incan torches like they used to use in the WWI/WWII era.

Again; of course Batteroo works in this ultra-limited case scenario, but it's solving a problem which simply doesn't exist in 2017.

But again, props to them for doing a much better video test demo than they previously seemed capable of.

[EEVblog]

More proof that Batteroo significantly extends battery life....

https://youtu.be/x62rhxWfmgk

Starting at 3:10 it's bullshit that you'd throw away the batteries when you reach 50% brightness. You likely wouldn't even be aware that it's dropped to 50% brightness.
They are absolutely desperate to get anything to back up their "times X" claim.
They also won't tell you how long the non-Batteriser gives out usable light after the Batteriser dies.
They also won't mention how crap a voltage regulator the Batteriser is, why isn't it a constant brightness the whole time?
They also won't tell you what effect the Batteriser has on bulb life.

[ez24]

But again, props to them for doing a much better video test demo than they previously seemed capable of.

I agree they have found two products that work. 

[EEVblog]

I have to say, Batteroo, after learning and getting tips from this blog, have finally done a video demo test that covers off some of the very basic requirements. For instance they're comparing like for like, they show the brand new cells being taken out of the pack, they show the equipment and the timer, and it's in one take, and results are shown in a clear chart. Props to them for that. Pity they're still using that awful oleaginous-voiced woman to do the narration. And they stopped the test when the non-batteroo'd torch was still going, and still probably had hours of dim glow left in it.
And yes, as expected (and not denied by anyone), their product works as a boost converter (albeit one which can't maintain a steady fixed output voltage), and they've chosen the one-in-a-million device where using the Batteroo is maybe valid: A dumb, old-fashioned incandescent torch with no built-in boost conversion which runs on a pair of Alkaline D cells. Basically a switch and a bulb in a can.

Trouble is, the world has moved on. No one I know still uses crappy old incan torches anymore (even my parents who are in their 70's and aren't "techie" at all). Nowadays, everyone uses LED torches with fantasic buck/boost converters built in, multiple modes, far more efficiency and brightness, and much more robust than the old incan torches like they used to use in the WWI/WWII era.

Again; of course Batteroo works in this ultra-limited case scenario, but it's solving a problem which simply doesn't exist in 2017.

But again, props to them for doing a much better video test demo than they previously seemed capable of.

Yep, they went to a LOT of trouble to do this test and it's the best they can come up with.
Maybe I should get myself that Garmin GPS now that I have the AA Batteriser?

[EEVblog]

My 10mA test is finished, the sleeve lasted for 10½ day. The first 10 day the voltage was above 1.3V, the last ½ day it dropped below 1.1V before the sleeve quit.

Can you post the data/graph please?

[samgab]

My 10mA test is finished, the sleeve lasted for 10½ day. The first 10 day the voltage was above 1.3V, the last ½ day it dropped below 1.1V before the sleeve quit.

Can you post the data/graph please?

In case HKJ isn't online at the moment:
http://lygte-info.dk/review/batteries2012/Batteroo%20AA%20UK.html

"A fairly low current test, the Duracell without sleeve can run for 11 days down to 1.1V and for 14 days down to 0.7V.
With the sleeve the runtime is 10½ day down to 1.2V and a few hours more to 1.1V and 0.7V.
From my logfile I can see that the sleeve deliver a total of 3.67Wh, without the sleeve I got 3.02Wh to 1.2V, 3.46Wh to 1.1V and 4.15Wh to 0.7V (These two batteries are not from the same pack, i.e. the comparison is not completely valid).
In this case it looks like the sleeve delivers slightly more energy to device that only work down to 1.1V, but not anywhere 8 times or even 2 times."

and:

http://lygte-info.dk/review/batteries2012/Duracell%20Ultra%20Power%20AA%20UK.html

[Mr.B]

More proof that Batteroo significantly extends battery life....

Is it me, or does it look like the torch with the batteroo is focused in more of a spot on the lux sensor, versus more of a flood on the other torch.
Visible from about 1:18 to 1:27 where he is fiddling with the torch trying to get the spot centered on the lux sensor.

[samgab]

More proof that Batteroo significantly extends battery life....

Is it me, or does it look like the torch with the batteroo is focused in more of a spot on the lux sensor, versus more of a flood on the other torch.
Visible from about 1:18 to 1:27 where he is fiddling with the torch trying to get the spot centered on the lux sensor.

It seemed that way to me, but to be fair, he zeroed both the light meters before the start of the test and they were both reading 189 or whatever it was, the same reading at the start of the test, so I don't think it would be fair to cry foul over that detail. They both started from an equal measurement and went from there.

These are crappy budget torches with crappy reflectors, so they won't have a perfect beam pattern.

[Mr.B]

You are right on the lux reading at the start.
However, given the companies history, I was suspicious of the two apparently different beam patterns.