Author Topic: Batteroo testing  (Read 265134 times)

0 Members and 1 Guest are viewing this topic.

Offline dexters_lab

  • Supporter
  • ****
  • Posts: 1889
  • Country: gb
Re: Batteroo testing
« Reply #475 on: January 10, 2017, 05:41:46 pm »
The slight drop at high current and low input voltage is probably the voltage drop at the input wires, because you didn't connect the sense wires directly at the Batteriser pcb.

i fixed that and re-measured
https://www.eevblog.com/forum/projects/batteroo-testing/msg1110445/#msg1110445

« Last Edit: January 10, 2017, 05:45:41 pm by dexters_lab »
 
The following users thanked this post: Kean

Offline FrankBussTopic starter

  • Supporter
  • ****
  • Posts: 2365
  • Country: de
    • Frank Buss
Re: Batteroo testing
« Reply #476 on: January 10, 2017, 07:56:48 pm »
is there anyone else doing similar testing to confirm/reproduce?

Looks close to my measurements. My efficiency is about 4 percent points lower, but I used the AAA sleeves and I guess the AA sleeves have bigger inductors.

Over 90% for the AA sleeve over a large range of current and input voltages is pretty good. They could make good money selling the chip, only, at Digikey.
So Long, and Thanks for All the Fish
Electronics, hiking, retro-computing, electronic music etc.: https://www.youtube.com/c/FrankBussProgrammer
 

Offline Hensingler

  • Regular Contributor
  • *
  • Posts: 144
  • Country: gb
Re: Batteroo testing
« Reply #477 on: January 10, 2017, 08:23:42 pm »
Over 90% for the AA sleeve over a large range of current and input voltages is pretty good. They could make good money selling the chip, only, at Digikey.

But regulation is poor, ripple is terrible and who wants 1.5v out?
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16531
  • Country: 00
Re: Batteroo testing
« Reply #478 on: January 10, 2017, 08:28:48 pm »
But regulation is poor, ripple is terrible

A bigger inductor and more capacitance on the output should fix that. Batteriser is really constrained by size.

 

Offline samgab

  • Frequent Contributor
  • **
  • Posts: 423
  • Country: nz
Re: Batteroo testing
« Reply #479 on: January 10, 2017, 08:42:32 pm »
Now why couldn't batteroo have just provided technical datasheets containing this sort of data in the first place? Would have placated a lot of the detractors somewhat...
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16531
  • Country: 00
Re: Batteroo testing
« Reply #480 on: January 10, 2017, 09:40:34 pm »
Now why couldn't batteroo have just provided technical datasheets containing this sort of data in the first place? Would have placated a lot of the detractors somewhat...

No it wouldn't.

90% efficiency still leads to less than 1x battery lifetime, they were claiming 8x.

 

Offline Brumby

  • Supporter
  • ****
  • Posts: 12281
  • Country: au
Re: Batteroo testing
« Reply #481 on: January 10, 2017, 10:23:30 pm »
Now why couldn't batteroo have just provided technical datasheets containing this sort of data in the first place? Would have placated a lot of the detractors somewhat...

No it wouldn't.

90% efficiency still leads to less than 1x battery lifetime, they were claiming 8x.

That's a slightly misleading statement.

The instantaneous power efficiency is not the only factor to be used in determining battery life.  There is the rather well discussed parameter of "run time".  These two must be considered together.

Having said that, though, the "run time" certainly has not been the celebrated success that Batteroo were claiming.
 

Offline dexters_lab

  • Supporter
  • ****
  • Posts: 1889
  • Country: gb
Re: Batteroo testing
« Reply #482 on: January 10, 2017, 10:36:49 pm »
Now why couldn't batteroo have just provided technical datasheets containing this sort of data in the first place? Would have placated a lot of the detractors somewhat...

i am not sure it would, we knew from the get-go that when a battery is down to around 1v there is bugger all energy left in the battery to make use of and claims from batteroo about device battery cut off voltages was nonsense. Their dc-dc converter could be 99.99% efficient and we'd still be telling them they are talking bollocks.

there is also another reason they couldn't give us data... back then they didn't have the silicon to test!

Offline JiggyNinja

  • Regular Contributor
  • *
  • Posts: 52
  • Country: us
Re: Batteroo testing
« Reply #483 on: January 10, 2017, 10:59:44 pm »
Charts!

I hope you don't mind, I had a little play and did a couple of overlaid charts which show all current rates overlaid:
How dare you take what I took from someone else and do something different to the thing that I did!

Quote
I think the results, the batteroo performance, looks pretty good from this data.
It's important to remember that Dexter is testing with a power supply feeding the input. The real performance, factoring in the internal resistance of a battery, will be worse.
 

Offline dexters_lab

  • Supporter
  • ****
  • Posts: 1889
  • Country: gb
Re: Batteroo testing
« Reply #484 on: January 10, 2017, 11:26:13 pm »
Charts!

I hope you don't mind, I had a little play and did a couple of overlaid charts which show all current rates overlaid:
How dare you take what I took from someone else and do something different to the thing that I did!

Quote
I think the results, the batteroo performance, looks pretty good from this data.
It's important to remember that Dexter is testing with a power supply feeding the input. The real performance, factoring in the internal resistance of a battery, will be worse.

indeed, my tests are simply the overall efficiency of their converter with a steady-state load and supply


Offline Hensingler

  • Regular Contributor
  • *
  • Posts: 144
  • Country: gb
Re: Batteroo testing
« Reply #485 on: January 11, 2017, 01:21:55 am »
Quote
I think the results, the batteroo performance, looks pretty good from this data.
It's important to remember that Dexter is testing with a power supply feeding the input. The real performance, factoring in the internal resistance of a battery, will be worse.

The internal resistance of the battery will be there without batteriseroos. The peaky nature of the batteriseroo current draw where it runs at its maximum current capability to charge the output capacitor +200mV then sleeps till it discharges will make internal resistance losses much worse for low current applications.

Anyone think they know what that does with battery chemistry? Is 0.5A pulses which average 10mA worse than 10mA dc?
 

Offline Brumby

  • Supporter
  • ****
  • Posts: 12281
  • Country: au
Re: Batteroo testing
« Reply #486 on: January 11, 2017, 02:42:21 am »
... what that does with battery chemistry? Is 0.5A pulses which average 10mA worse than 10mA dc?

A very good question.

While we might have a really good understanding of the principle of operation of the Batteroo sleeve - the claims made for it come down to a fundamental issue of battery chemistry.  That is the limiting factor here - not the electronics.  (Using the 80/20 rule, I'm looking at the 80% of common cases.)
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37626
  • Country: au
    • EEVblog
Re: Batteroo testing
« Reply #487 on: January 11, 2017, 02:48:08 am »
Huh  :-//
I thought the Batteriser gave a constant 1.5V out regardless of input voltage?
 

Offline samgab

  • Frequent Contributor
  • **
  • Posts: 423
  • Country: nz
Re: Batteroo testing
« Reply #488 on: January 11, 2017, 03:37:18 am »
If I had a sleeve or two, I'd run some discharge curves on my graphing smart charger/analyzer alongside bare Alkaline cells of the same type.
Perhaps, 50mA, 100mA, 250mA, 500mA, 1A, and 2A discharge curves.
Then I'd do a chart with those discharge curves alongside the plain Alkaline cell curve, and an eneloop curve, so we can see which is the best method at different constant current loads, side by side.
But my analyzer can't do a constant power load or a constant resistance load, or a programmed load with dips and spikes in load.
The BK Precision 8500 programmable DC electronic load can do all of that, except maybe not the pulsed load (edit: Okay, the transient setting is for doing pulsed loads. Cool. Daves BK 8601 would be ideal).
Charts of the outputs of some of that testing with that would be interesting.
« Last Edit: January 11, 2017, 05:00:52 am by samgab »
 

Offline HKJ

  • Super Contributor
  • ***
  • Posts: 2892
  • Country: dk
    • Tests
Re: Batteroo testing
« Reply #489 on: January 11, 2017, 07:51:21 am »
Huh  :-//
I thought the Batteriser gave a constant 1.5V out regardless of input voltage?

Definitely not, it is not just a boost converter, it has some more in it.
 

Offline HKJ

  • Super Contributor
  • ***
  • Posts: 2892
  • Country: dk
    • Tests
Re: Batteroo testing
« Reply #490 on: January 11, 2017, 07:54:42 am »
If I had a sleeve or two, I'd run some discharge curves on my graphing smart charger/analyzer alongside bare Alkaline cells of the same type.
Perhaps, 50mA, 100mA, 250mA, 500mA, 1A, and 2A discharge curves.

I have done that, but I do not expect to publish anything before the coming weekend, then I will add it to my comparator: http://lygte-info.dk/review/batteries2012/CommonAAcomparator.php

It is possible to stretch a battery 4 times (or maybe more) with a batteriser, but it requires a product that stops working below 1.3V.
 
The following users thanked this post: samgab

Offline Brumby

  • Supporter
  • ****
  • Posts: 12281
  • Country: au
Re: Batteroo testing
« Reply #491 on: January 11, 2017, 07:58:37 am »
It is possible to stretch a battery 4 times (or maybe more) with a batteriser, but it requires a product that stops working below 1.3V.

I can't see anyone getting anywhere near that.........

I would like to see what data leads you to that conclusion.
« Last Edit: January 11, 2017, 08:00:34 am by Brumby »
 

Offline HKJ

  • Super Contributor
  • ***
  • Posts: 2892
  • Country: dk
    • Tests
Re: Batteroo testing
« Reply #492 on: January 11, 2017, 08:12:14 am »
It is possible to stretch a battery 4 times (or maybe more) with a batteriser, but it requires a product that stops working below 1.3V.

I can't see anyone getting anywhere near that.........

I would like to see what data leads you to that conclusion.

You need a product that stops working when the voltage drops below 1.3 volt, I do not not believe many product are that badly constructed.

As I wrote above I will publish my data in the weekend, both test with batteries and with a power supply (I used a Keithley 2460 with 4 terminal connection as battery) . The result will be as curves with hundreds of samples (Easy to do with computer control). You can see more about how I do stuff on my website.


 
The following users thanked this post: Kean

Offline samgab

  • Frequent Contributor
  • **
  • Posts: 423
  • Country: nz
Re: Batteroo testing
« Reply #493 on: January 11, 2017, 08:52:10 am »
It is possible to stretch a battery 4 times (or maybe more) with a batteriser, but it requires a product that stops working below 1.3V.

I can't see anyone getting anywhere near that.........

I would like to see what data leads you to that conclusion.

You need a product that stops working when the voltage drops below 1.3 volt, I do not not believe many product are that badly constructed.

As I wrote above I will publish my data in the weekend, both test with batteries and with a power supply (I used a Keithley 2460 with 4 terminal connection as battery) . The result will be as curves with hundreds of samples (Easy to do with computer control). You can see more about how I do stuff on my website.

Fantastic, can't wait to see that data, charts etc. The comparitor will be great because we'll be able to directly compare it to any number of NiMH, Alkaline, etc cells; even those li-ion "AA" cells with built in charger and DC/DC buck converter.

Incidentally, that comparitor is a fantastic tool, I just wonder, how difficult would it be to make it so there are three selections rather than two? If it is really difficult, no worries, but if it's relatively simple, that would be a great improvement, to be able to compare three different batteries/cells side by side... Anyway, it's still really useful as it is. I only ask because I have no idea how the web design for that is done.
 

Offline HKJ

  • Super Contributor
  • ***
  • Posts: 2892
  • Country: dk
    • Tests
Re: Batteroo testing
« Reply #494 on: January 11, 2017, 09:38:47 am »
Incidentally, that comparitor is a fantastic tool, I just wonder, how difficult would it be to make it so there are three selections rather than two? If it is really difficult, no worries, but if it's relatively simple, that would be a great improvement, to be able to compare three different batteries/cells side by side... Anyway, it's still really useful as it is. I only ask because I have no idea how the web design for that is done.

There is a few issues with that. Each curve is its own image and I stack them when you select more currents/batteries. This means I would need 50% more data space. I always regenerate all the curves when doing a new review, the 50% extra generation time would be irritating. The layout also works best with two sets.
What I would more like to add was some sort of filters, this would be useful, especially in the LiIon comparator (Due to the many batteries in it), but I am always low on time and postpone it.

 

Offline dexters_lab

  • Supporter
  • ****
  • Posts: 1889
  • Country: gb
Re: Batteroo testing
« Reply #495 on: January 11, 2017, 11:51:35 am »
Huh  :-//
I thought the Batteriser gave a constant 1.5V out regardless of input voltage?

none of mine are fixed, they all exhibit the same variable output behaviour

Frank, can you confirm if your AAA batterisers are a fixed output voltage or does the output voltage change depending on input?

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16531
  • Country: 00
Re: Batteroo testing
« Reply #496 on: January 11, 2017, 01:42:54 pm »
Huh  :-//
I thought the Batteriser gave a constant 1.5V out regardless of input voltage?

none of mine are fixed, they all exhibit the same variable output behaviour

Are you sure you're seeing a true voltage, that the lower voltage isn't a measurement error because of output ripple?
 

Offline HKJ

  • Super Contributor
  • ***
  • Posts: 2892
  • Country: dk
    • Tests
Re: Batteroo testing
« Reply #497 on: January 11, 2017, 01:49:41 pm »
Are you sure you're seeing a true voltage, that the lower voltage isn't a measurement error because of output ripple?
As I wrote earlier: It is more than just a boost converter.
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16531
  • Country: 00
Re: Batteroo testing
« Reply #498 on: January 11, 2017, 02:20:55 pm »
As I wrote earlier: It is more than just a boost converter.

Yes, I just wanted to be sure there's no measurement error.

I don't see why the output would be a curve, that makes no sense in a boost circuit.  :-//





« Last Edit: January 11, 2017, 02:22:46 pm by Fungus »
 

Offline Zbig

  • Frequent Contributor
  • **
  • Posts: 927
  • Country: pl
Re: Batteroo testing
« Reply #499 on: January 11, 2017, 02:27:01 pm »
As I wrote earlier: It is more than just a boost converter.

Yes, I just wanted to be sure there's no measurement error.

I don't see why the output would be a curve, that makes no sense in a boost circuit.  :-//

It seems they've made some attempt to address the battery gauge issues after all.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf