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

[Fungus]

What it shouldn't be: a witch-hunt against Batteroo, with people leaping on every new statement, tearing it apart, looking for the tiniest apparent inconsistency (even when there isn't one) and crying victory.

"That UL logo looks photoshopped", ...  are opinions.

What it should be: taking testable claims, and testing them.

I agree.
There are plenty of testable claims,

UL certification is a testable claim.

[firewalker]

Is my "fear" for hazardous gases detectors valid? Could it go off without the ability to notify the user? If something like that occurs, what is the liability for Batteroo?

Alexander.

[TheAmmoniacal]

Is my "fear" for hazardous gases detectors valid? Could it go off without the ability to notify the user? If something like that occurs, what is the liability for Batteroo?

Alexander.

Well, yes. All devices use the battery voltage to estimate the remaining power, and thus display a battery indicator or low-battery warning. With the Batteriser no battery indicators will work, as the device will think it has lots of power up until the DC-DC converter shuts down and device stops working suddenly. 

[Fungus]

Data is beautiful!

So, I got it into excel and started massaging the data and got it into a nice graph.  VERY interesting results.  You can very clearly tell at what point the screen dimmed because it plateaus.  (again, I'm going to make a video today to put on youtube because people need to see this a real apples to apples test like this to show that batteriser is being very dishonest in their claims)

EDIT: this also means that if we didn't use the stupid servo to keep the screen on, this thing would probably easily do 20-25+ hours on fresh AA's

Nail is driven into coffin. Well done! If you make a video you can also show the different menu settings and the effect they have.

Can I make a suggestion? I know this is all your work, etc., but could you and Dave get together and collaborate on a video showing this result?

eg. Dave introduces the video saying, "Here's a guy who has the exact same GPS and has been testing it, the results are interesting...", then passes over to you with your test setup, your data, your screenshots, etc.

The reason is that Dave is under the spotlight of all things Batteriser. A lot more people will see it (and give it credibility?) if it's on EEVBLOG's Youtube channel.

(This is assuming You + Dave are interested, of course...)

[firewalker]

They should place a banner with huge letters stating that their product should not be used with any kind of safety device. They could find themselves in deep shit if something goes wrong...

Alexander.

[TheAmmoniacal]

eg. Dave introduces the video saying, "Here's a guy who has the exact same GPS and has been testing it, the results are interesting...", then passes over to you with your test setup, your data, your screenshots, etc.

The reason is that Dave is under the spotlight of all things Batteriser. A lot more people will see it (and give it credibility?) if it's on EEVBLOG's Youtube channel.

(This is assuming You + Dave are interested, of course...)

5ky did a video, here (read top comment):

[amyk]

While the footprint and circuit matches what they've shown, I don't think the ADP1607 is suitable - it only goes down to 1.8V output. If Batterisers are outputting 1.8V that could be too high for some devices. Two of them in series will be 3.6V instead of 3V, and four of them give 7.2V instead of 6V - 20% higher.

[Galenbo]

I didn't know about UL tests, but it sounds familiar to the TUV tests in Europe.
They only test security, not functionality.

I don't know about UL testing either, but from what I read in this thread they seem to offer more than just the certification.

Sorry I forgot, TUV also tests against compliance to the laws, security, radiation, EMC,... and in the medical/food field much more.
But doesn't (always, per definition) test the claims.

[samgab]

While the footprint and circuit matches what they've shown, I don't think the ADP1607 is suitable - it only goes down to 1.8V output. If Batterisers are outputting 1.8V that could be too high for some devices. Two of them in series will be 3.6V instead of 3V, and four of them give 7.2V instead of 6V - 20% higher.

Yeah, it's specced for 1.8-3.6V, but "The output voltage is set by a resistor voltage divider, R1, from the output voltage (VOUT) to the 1.259 V feedback input at FB and R2 from FB to GND". So you can get 1.6V out of it by tweaking the R1 and R2 values.

[Fungus]

5ky did a video, here (read top comment):

Two thumbs up for video. All the data, all presented loud and clear.

(A bit rambling towards the end but the important part at the beginning was spot on).

PS: I could do a Batteriser video if there's any left over (assuming they ever appear). If anybody else wants them I can post them onwards after I'm done...

[SundayProgrammer]

I don't see the batteriser doing more than about an amp, tops.

Based on my limited experience of GPS receiver modules I noticed (when I was designing GPS tracking system) that there can be some differences in power consumption whether the receiver is a) searching for satellites or b) locked to satellites. Ie the device is outside or inside for example. In that particular device can you see _any_ variations between those two operation stages?

Anyway... huge difference between test results! I am more and more interested to see how Batteriser works IRL. For example in action in this very same test (if you bought Garmin, you will probably buy couple of Batterisers too :-) )

Axel.

[Wytnucls]

incidentally the cheapest they are selling the batteriser for is $1.92 each for the AA and AAA size

I struggle to comprehend how they could possibly make that widget for such a low price and still turn a profit.
The IC alone costs 1.44$ in bulk (3,000 units) at Digi-key. If they rolled up their own (huge development cost), I doubt they could produce it for much cheaper than Analog Devices can sell theirs for.

The small package ADP1607 is rather impressive: More than 85% efficiency between 1mA and 100mA load current at Vin 0.8V.

[PeterL]

I just got this from igg, posting it here to have a useable timestamp:


It's not clear what John asked, this was a private message, but it looks like a reaction to 5ky's GPS test.

[dr.diesel]

I struggle to comprehend how they could possibly make that widget for such a low price and still turn a profit.

GIven what we've seen so far, I suspect their plan is to pay a few months salary to themselves and fold with no product ever being produced.  Be just another failed crowdfunding campaign.

[amyk]

While the footprint and circuit matches what they've shown, I don't think the ADP1607 is suitable - it only goes down to 1.8V output. If Batterisers are outputting 1.8V that could be too high for some devices. Two of them in series will be 3.6V instead of 3V, and four of them give 7.2V instead of 6V - 20% higher.

Yeah, it's specced for 1.8-3.6V, but "The output voltage is set by a resistor voltage divider, R1, from the output voltage (VOUT) to the 1.259 V feedback input at FB and R2 from FB to GND". So you can get 1.6V out of it by tweaking the R1 and R2 values.

The regulation will probably be horrible though - there's a reason AD did not specify lower than 1.8V. Boost converters with low input voltages also use the output to power themselves, so things like the internal voltage reference are going to stop working properly if the output is set too low.

[firewalker]

I believe 5sky is a backer. He should comment on their campaign... 

Alexander.

[Chipguy]

I believe 5sky is a backer. He should comment on their campaign... 

Alexander.

Yeah he said so in the video. But he also pointed out something like "only for testing them"

[Barny]

The funny thing is, that at least one sock pupet of Batteriser/Battero is in the comment section of the video from 5sky too.
(User:  Jones David)

But there he postet a really normal comment.

[Edit] Polished my crappy english.[/edit]

[dr_frost_dk]

I just got this from igg, posting it here to have a useable timestamp:


It's not clear what John asked, this was a private message, but it looks like a reaction to 5ky's GPS test.

They are REALLY grasping at straws, i do not believe AT ALL they have any UL certification or testing.
And what kind of "answer" is that, no data, no nothing, just a stupid (fabricated looking) UL report saying THE SAME as their "test".
Also one more thing, every body says it takes some time getting a UL done, so why are they using the RED sleved batteriser they recently posted on not the original stainless steel only batteriser ones from 1+ mouth ago. ?

ANYBODY can see this:
Batteriser GPS: 2 Hours normal - 10 Hours Batteriser...
5ky GPS: 17 Hours normal.....
AND Reviews for that GPS: ALL over 15 hours normal
 
Looking forward for another user here that bought the same GPS, i bet it will be within 10% of 5ky´s test

Good job 5ky, really good test and very good video.

[qoole]

You've been featured, indirectly, on slashdot now!

http://news.slashdot.org/story/15/09/06/1230220/ddos-style-youtube-dislikes-for-sale

Which links to the Dell Secureworks website:

http://www.secureworks.com/resources/blog/negative-feedback-attack-on-a-youtube-channel/

Watch out for incoming usage spikes!

[Fungus]

A couple of questions for 5ky now that he has his GPS and test setup:

(Both referring to Batteroo's graph from their video...)

a) Batteroo show a current draw of about 130mA at the start of their test. What input voltage would that correspond to? You can mess around with your power supply and find the voltage where it draws 130mA. Now look that voltage up on your chart, it might give us a clue about the state of their batteries at the start of their test. (Yeah, I know we don't know their burden voltage but we maybe we can get a ballpark figure).

b) Is the GPS a constant power device? Their graphs suggest it is. If so, why would Garmin even bother with that "NiMH" message? Your tests didn't seem to show a massive difference when the screen went 'dim' but could you measure power draw on either side of that bump? I'm interested in how much life would people lose if they select 'NiMH' in the menu to avoid that message then use alkalines?

[Fungus]

I tried to find the UL project number , UL 4787059213 ...

Lots of people have.

Has anybody found the report at the UL Labs ?   

Nope.

[TheAmmoniacal]

As has been described on the forum, the project number is only an internal number, you won't find anything on it. You can search for the company when UL testing is completely and a report has been published.

[5ky]

A couple of questions for 5ky now that he has his GPS and test setup:

(Both referring to Batteroo's graph from their video...)

a) Batteroo show a current draw of about 130mA at the start of their test. What input voltage would that correspond to? You can mess around with your power supply and find the voltage where it draws 130mA. Now look that voltage up on your chart, it might give us a clue about the state of their batteries at the start of their test. (Yeah, I know we don't know their burden voltage but we maybe we can get a ballpark figure).

b) Is the GPS a constant power device? Their graphs suggest it is. If so, why would Garmin even bother with that "NiMH" message? Your tests didn't seem to show a massive difference when the screen went 'dim' but could you measure power draw on either side of that bump? I'm interested in how much life would people lose if they select 'NiMH' in the menu to avoid that message then use alkalines?

right at 3v, but the current jumps around on that unit even when you aren't touching it.  it occasionally jumps up to 250mA, and then back down to 90mA, but the majority of the time it's right at 130mA right at 3v, and when it's at 130mA, when I crank voltage down 100mV at a time, I see the current go up in ~10mA increments so it's definitely constant power.  Plus, on my voltage graph you see downward dips and on batteriser's current graph, you see upward peaks which probably correlate with those downward dips

I'm guessing they started with fresh batteries, and their logger has 1-2 ohm shunt, and the burden voltage caused the brightness warning to come on a 2 hr mark, and that's where they called it and lied and said that the unit shut off.  (which we all know is BS)

EDIT: WORDS AND STUFF

[5ky]

Data is beautiful!

So, I got it into excel and started massaging the data and got it into a nice graph.  VERY interesting results.  You can very clearly tell at what point the screen dimmed because it plateaus.  (again, I'm going to make a video today to put on youtube because people need to see this a real apples to apples test like this to show that batteriser is being very dishonest in their claims)

EDIT: this also means that if we didn't use the stupid servo to keep the screen on, this thing would probably easily do 20-25+ hours on fresh AA's

Nail is driven into coffin. Well done! If you make a video you can also show the different menu settings and the effect they have.

Can I make a suggestion? I know this is all your work, etc., but could you and Dave get together and collaborate on a video showing this result?

eg. Dave introduces the video saying, "Here's a guy who has the exact same GPS and has been testing it, the results are interesting...", then passes over to you with your test setup, your data, your screenshots, etc.

The reason is that Dave is under the spotlight of all things Batteriser. A lot more people will see it (and give it credibility?) if it's on EEVBLOG's Youtube channel.

(This is assuming You + Dave are interested, of course...)

I would gladly pack up the entire setup and mail it to dave if he'd like to use it in any way, shape, or form.  Just let me know!

(A bit rambling towards the end but the important part at the beginning was spot on).

PS: I could do a Batteriser video if there's any left over (assuming they ever appear). If anybody else wants them I can post them onwards after I'm done...

Yeah, sorry about the rambling--I'm pretty bad about rambling.  I get excited and geek out about stuff like this. 

Anyway... huge difference between test results! I am more and more interested to see how Batteriser works IRL. For example in action in this very same test (if you bought Garmin, you will probably buy couple of Batterisers too :-) )

I have supported their campaign just to get some samples to test and am already planning to mail some out to vbloggers like dave who could put it through their paces