If the 15µA quiescent current figure is right, that's 131 mAh per year.
The fact is they do have an IC, there is nothing in the LT range that comes close,
The problems of this "invention" IMO are:
1) That it cuts your power supply from 100% straight to 0 without notice.
2) That you no longer can know how much juice there's left in your batteries.
3) For not much gain in total mAh (minus those 131mAh per year) as you can see in the discharge curve of a batt.
The output actually goes down from 1.5V to 1.3V. If you're designing this into a product then you can have a battery indicator.
I'm guessing LT simply haven't made any effort in that voltage range due to lack of demand.
I'm guessing LT simply haven't made any effort in that voltage range due to lack of demand.
If there was a valid application for this Batteroo chip it would already have existed. The closest thing with presumed valid application are fixed 1.8v output parts like the chips Batteriser used in their patent applications.
I'm guessing LT simply haven't made any effort in that voltage range due to lack of demand.
If there was a valid application for this Batteroo chip it would already have existed. The closest thing with presumed valid application are fixed 1.8v output parts like the chips Batteriser used in their patent applications.Removing the Batteroo's optimisation for its battery cell boosting role from this chip would not be hard, and then you end up with a chip that outperforms anything TI, LT, AD, ST, Microchip and Maxim can currently make. To say that a company like LT doesn't make their chips better because people like spending over 3 times the money for a lower efficiency and lower capacity chip that needs a bigger and more expensive inductor does not even begin to make sense.
The high efficency is probably because of the agressive burst mode:
Unlike most other converters, the Batteriser stepup seems to run at a fixed duty cycle (or maybe fixed peak current) regardless of the load current. When the target voltage is reached, it shuts down until a much lower threshold. This results in a rather large output voltage ripple. Most other manufacturers instead try to keep the the ripple in PFM/burst mode to an acceptable limit.
The ic may have a high efficency, but 1.5V output voltage with a rather large, load dependend ripple makes ist useless for most applications using the voltage for some analogue circuits.
The high efficency is probably because of the agressive burst mode:
Unlike most other converters, the Batteriser stepup seems to run at a fixed duty cycle (or maybe fixed peak current) regardless of the load current. When the target voltage is reached, it shuts down until a much lower threshold. This results in a rather large output voltage ripple. Most other manufacturers instead try to keep the the ripple in PFM/burst mode to an acceptable limit.
The ic may have a high efficency, but 1.5V output voltage with a rather large, load dependend ripple makes ist useless for most applications using the voltage for some analogue circuits.
That would be a good topic for another EEVblog vidjeo.
I'm guessing LT simply haven't made any effort in that voltage range due to lack of demand.
If there was a valid application for this Batteroo chip it would already have existed. The closest thing with presumed valid application are fixed 1.8v output parts like the chips Batteriser used in their patent applications.Removing the Batteroo's optimisation for its battery cell boosting role from this chip would not be hard, and then you end up with a chip that outperforms anything TI, LT, AD, ST, Microchip and Maxim can currently make. To say that a company like LT doesn't make their chips better because people like spending over 3 times the money for a lower efficiency and lower capacity chip that needs a bigger and more expensive inductor does not even begin to make sense.The high efficency is probably because of the agressive burst mode:
Unlike most other converters, the Batteriser stepup seems to run at a fixed duty cycle (or maybe fixed peak current) regardless of the load current. When the target voltage is reached, it shuts down until a much lower threshold. This results in a rather large output voltage ripple. Most other manufacturers instead try to keep the the ripple in PFM/burst mode to an acceptable limit.
The ic may have a high efficency, but 1.5V output voltage with a rather large, load dependend ripple makes ist useless for most applications using the voltage for some analogue circuits.
It would be great to have a version of this same chip with a higher output voltage. Something that could run down to 0.6v and could step it up to 3.3v or 5V at over 200mA, and yet only consume 15uA when there is no load. That would allow the battery to be permanently connected to the converter, and it would not need to be switched off. The 5V circuit will have its 5V continuously, so it can have conventional soft power ON.OFF switching. It would not be technical difficult (unless the chip is designed for under 3.3V maximum - which it might be to maximize efficiency). If Batteroo, or the company who designed this chip released a higher output voltage version, it would probably have better specs then anything else on the market. We have no idea about issues such as reliability.
The part you're looking for is a MCP1640. I agree with many who say that the best kept secret of the Microchip (of PIC processor fame) is the strength of their analog product line.
The MCP1640 has 19uA Quiescent current, operating voltage down to 0.65V, 1000mA current. Not quite as good but close. There's also a cheaper 175mA version...
Edit: Looks like the MCP1640 is specced at 800mA typical 'input current limit'. Which means that at as an example 1.2V in, you'll only get 100mA at 3.3V.
To me it is no surprise that a chip such as the BTR004 has not been done before.
The requirements have dictated a narrower area of operation than how a general purpose chip would have been designed - and I feel these conditions allowed for improvements in other areas, such as efficiency.
Herein lies the fundamental reason why no such chip had been produced before. Aside from some obscure
situation, the application could only ever be for a single cell booster, which puts it straight into the 'let's suck as much as we can from a battery" arena. Some quick assessment of batteries and battery operated devices would have shown this really didn't make a lot of sense. Sure, the converter could do the job, but was it a job worth doing?
It took Batteroo and crowd funding to take a shot at it.
The IC Batteroo is using does seem to be a customised IC with the internal fixed voltage divider for the 1.5V output. If it is a custom IC, there is no need for any proper datasheet to be made. To have a block diagram for the Batteroo webpage, someone might take an old datasheet for a similar chip and edit it.
And so what if the block diagram shows two outputs from the driver blocks? That is completely fine. If you use a triangle, that is what is going to happen, and it is just a block diagram. As a block diagram, it is as informative as many other block diagrams I have seen in professionally produced datasheets.
Some of the things I do not recall being tested in this forum include:
- Max output current vs battery voltage
- Burst mode vs voltage - is there a voltage at which burst mode ceases?
- Temperature vs load (using the reverse substrate diode as a temperature sensor).
- Is the chip thermally protected?
- Reverse substrate diode temperature versus current (since when a batteroo turns off in a multi battery device, current passes through the reverse substrate diode and it is near impossible to thermally protect it)
- Can it switch on under full load? (The way the Batteroo is used, the device usually switches on at no load)
- What is the minumum startup voltage?
- Efficiency vs battery voltage for a full sweep of load current from microamps to 100% load. (Would be a big task)
- Does the device latch up if battery voltage is applied as the substrate diode is conducting in a reverse direction?
- Switching resistance (by measuring the coil to -ve voltage when the coil switch is on) at different batt. voltages.
- In burst mode, does the pulse width vary with battery voltage?
- For the AA cell (that has 2 inductors), is it running as a 2 phase converter?
- RF radiation
On the Batteroo technology page, they claim the Batteroo can switch on at 0.6V I think, but I assume they mean that as long as the device is on, burst mode can still work fine from 0.6V. If it did actually turn on at 0.6V, that would be impressive.
Doing all these tests would be a massive job, so I cannot blame anyone for not bothering.
Efficiency vs battery voltage for a full sweep of load current from microamps to 100% load. (Would be a big task)
The MCP1640 has 19uA Quiescent current, operating voltage down to 0.65V, 1000mA current.
The MCP1640 has 19uA Quiescent current, operating voltage down to 0.65V, 1000mA current.You have fallen in the common trap: Never stop reading at the first page of a datasheet.
If you look a the footnotes of the specs it tells you the actual truth:
"IQ is measured from VOUT; VIN quiescent current will vary with boost ratio. VIN quiescent current can be estimated by: (IQPFM * (VOUT/VIN)), (IQPWM * (VOUT/VIN))."
So for 5V out and 1V in, Iq drawn from the battery is around 0.1mA! That is much higher, but still ok for many applications.
MCP1640 is a great ic because it is cheap and works well. I have used it in many projects. Typically I use it as a soft power switch: The on/off button pulls the enable pin high for a short time before the microcontroller takes over. This reduces the current when powered down to around 1uA.
If I need a low power microcontroller powered on permanently, I prefer a 3V source like a CR2032 or 2x AAA batteries without any voltage regulator at all.
Some measurements on the Batteriser have been done in this thread:
https://www.eevblog.com/forum/projects/batteroo-testing/
Based on those measurements and observed waveforms it doesn't look like PWM mode at all to me. Instead the converter seems to keep operating in burst/PFM mode over the full load range.
The MCP1640 has 19uA Quiescent current, operating voltage down to 0.65V, 1000mA current.You have fallen in the common trap: Never stop reading at the first page of a datasheet.
If you look a the footnotes of the specs it tells you the actual truth:
"IQ is measured from VOUT; VIN quiescent current will vary with boost ratio. VIN quiescent current can be estimated by: (IQPFM * (VOUT/VIN)), (IQPWM * (VOUT/VIN))."
So for 5V out and 1V in, Iq drawn from the battery is around 0.1mA! That is much higher, but still ok for many applications.If you look at fig 2.1 in the datasheet, it shows that with 1.2V in, 5V out, the quiescent current at 20 degC is just over 20uA. I am hoping it is right. I have ordered some to try. 0.1mA would be pretty pathetic. I did look at Note 3 and it doesn't seem to make much sense. I guess I will find out when I test it.
QuoteSome measurements on the Batteriser have been done in this thread:
https://www.eevblog.com/forum/projects/batteroo-testing/
Based on those measurements and observed waveforms it doesn't look like PWM mode at all to me. Instead the converter seems to keep operating in burst/PFM mode over the full load range.If you look at this post from that thread, you will see the sawtooth wave from the burst mode at 100mA. At 500mA, there is no sawtooth. That means at 500mA, it has to be PWM mode.
https://www.eevblog.com/forum/projects/batteroo-testing/msg1101667/#msg1101667
The 1 A waveform just looks crap. Not sure if this is the AAA battery.
Look at the caption below fig 2-1: "FIGURE 2-1: VOUT IQ vs. Ambient Temperature in PFM Mode."
Yes, the marketing departmant achieved their goal: They fooled you to think the ic is much better than it actually is.
Somewhere in the datasheet there is a description of the internal operating mode:
The ic starts up using a free running clock controlling the switching transistor. Once the output voltage is high enough, it switches over to normal operation mode, powering all of the internal functions from the output voltage. Therfore all internal supply currents are sourced from the output voltage. Once the ic is running, the input voltage can go much lower than at startup (0.35V if the output current demand is low).
The MCP1640 has 19uA Quiescent current, operating voltage down to 0.65V, 1000mA current.You have fallen in the common trap: Never stop reading at the first page of a datasheet.
QuoteIt would seem odd if Fitipower were the designers of the BTR004... If you're Batteroo, would you pick an IC designer that noone has ever heard of, for your one shot at getting the custom IC that you need for your "world-changing invention"?
Would you put one over that Battero will not just do exactly that?
After all if their goal was really getting a proper custom IC for their 'world-changing invention' instead of the scamming merry-go-round that they took us and their investors on, they would've heed Dave's advice way back when... that it's just not feasible?
I don't understand this current wave of conspiracy theories. Yes, obviously Batteroo is marketing their product with wildly inflated claims, and is probably using dishonest marketing methods like bought votes and "customer" statements. There are many good reasons to distrust them.
But the battery sleeves are working boost converters, right? They are apparently designed around a custom IC which is made by an OEM who has a track record in boost converter ICs. Where's the problem here? Again, the claimed benefits are wildly inflated, but why would the miniaturized boost converter itself be "not feasible"?
Time to move on and let Batteroo fade away in silence, I'd say...