💬 Inexpensive 3.3V Boost Converter

[openhardware.io]

https://www.openhardware.io/view/278/Inexpensive-33V-Boost-Converter

[scalz]

Hi.

My humble thought if it can help you ;)

• be careful when designing booster/buck. I'm not sure how you routed it (looks like parts are on bottom), but the efficiency of your switching reg will depends on your design. Best is to fit datasheet recommendation ;)
• switching, power consumption of this booster is 300uA! It's <1uA but if it's disabled.
  And you can expect some switching on a sleeping node to keep the 3.3v voltage stable. Then efficiency on light loads <100uA is 20%, very bad, so the booster power consumption will be higher again..

Formula for efficiency is : power consumption (uA) = (Vout * I_mini_circuit)/(Vbat * Efficency_at_this_batt_voltage)

• Vout=3.3V,
• I_mini_circuit, let's say you have only the booster consuming (not possible) : 300uA
• Vbat says 1.2v for a fresh batt
• at 1.2V input and for <100uA general power consumption, efficiency is 0.2
• Result is 4125 uA! during light load, sleep mode..

[NeverDie]

@scalz

The schematic is from page 1 of the datasheet. So, yes, it fits the datasheet.

As for efficiency, one has to ask: compared to what? And what are the trade-offs? In this instance, the component cost is very low, so that's the trade-off. There are certainly more efficient boost converters out there, but at greater cost. Also, some of the alternatives are more difficult to solder, so that's a consideration also.

The application I'm considering it for is harvesting enough solar energy to power a TH sensor mote. I have no prior experience in doing that, so I'm planning to try a number of different boost converters to empirically test which is "sufficient." If the AAT1275 proves to be sufficient (and I don't know whether it will or it won't), then it will win on cost--at least compared to the other boost converters that I'll be testing.

Anyhow, if you (or anyone reading this) know of a more efficient or "better" boost converter, but which costs less, then I'd be interested to know. In fact, I'd be interested in any boost converter that costs less, even if it has worse efficiency. I don't think it will take much to power a mote that sleeps most of the time, but that hypothesis is one of the things I'll be testing.

[scalz]

well you're missing the point, that was infos only ;)

• i was not talking about schematics, i have not looked at it. I was talking about the routing. loop, feeedback, plane etc..
• lots of infos in datasheet, but also in appnotes. very important for routing, or application cases.
• ok i thought you were searching for efficiency! not in front of files, but i'm pretty sure that mcp1640 is a lot better. few cents more..
• but if you don't optimize for solar you may be disappointed. and if it's for indoor harvesting, it matters a lot including protocol and software.

well, always better to experiment, you're right. so good luck for your project :)

[Koresh]

@NeverDie said:

@scalz

The schematic is from page 1 of the datasheet. So, yes, it fits the datasheet.

As for efficiency, one has to ask: compared to what? And what are the trade-offs? In this instance, the component cost is very low, so that's the trade-off. There are certainly more efficient boost converters out there, but at greater cost. Also, some of the alternatives are more difficult to solder, so that's a consideration also.

The application I'm considering it for is harvesting enough solar energy to power a TH sensor mote. I have no prior experience in doing that, so I'm planning to try a number of different boost converters to empirically test which is "sufficient." If this proves to be sufficient, then it will win on cost (at least compared to the other boost converters that I'll be testing).

Anyhow, if you know of a more efficient or "better" boost converter, but which costs less, then I'd be interested to know. In fact, I'd be interested in any boost converter that costs less, even if it has worse efficiency.

I am sorry for interfering, I'll just try to add my 5 cent :) To my opinion, the most effective converter - no converter :) I came to this conclusion when I made calculations with the most effective contemporary convertors IC with coin batteries (2032, 2450, 2477). These batteries have very high output impedance after voltage drops below 1.8-2.0v and they have very low residual capacity in this case. Of course my conclusion can be wrong for other large batteries with other discharge curve. :smiley:

[NeverDie]

@scalz said:

mcp1640

Thanks! Looks like the mcp1640 would be worth trying also. What I like about MCP1640 is that the cost is comparable and yet both the startup voltage and the minimum operating voltage are both lower than for the AAT1275. I imagine that may be important for scavenging the most out of the available ambient light.

[NeverDie]

Here's a link to the the MCP1640 adjustable boot converter that I just made:
https://www.openhardware.io/view/279/Adjustable-Boost-Converter

Regarding the routing, you can see the tracelines pretty easily in the PCB renderings. I'm not aware of any mistakes that might affect efficiency, but feel free to point out any that you discover.

[NeverDie]

I just now released Version 002, which has improved routing placement.

[NeverDie]

I just now posted Version 3.

Changes from Version 2:

1. Changed inductor to be of same type as that used in my other boost converter projects.
2. Enlarged board to accommodate larger inductor.
3. Eliminated two unnecessary pin headers.
4. Cleaned up the top and bottom silk screens.

[NeverDie]

I assembled and tested this pcb, and it works. This project is now finished. :)

[tonnerre33]

Thanks you a lot for your work ;)

[NeverDie]

This post is deleted!