Powersupply Boost Converter based on the TI TPS6120x
JohnR2010
Posts: 431
I recently had some time to put together a power supply circuit based on the TPS6120x Synchronous Boost Converter http://www.mouser.com/ds/2/405/tps61201-487633.pdf for a battery powered project I have been working on. So far I'm disappointed in the quiescent current usage of 100uA. The spec sheet says Quiescent Current Less than 55uA but that is not what I'm seeing.
My design calls for 3.3v and I'm driving the Boost Converter with two AA batteries. With no load on the boost converter it is pulling 90 to 100uA. I sure didn't expect that. My previous power supply designs have been based on Microchips LDO MCP1701A regulator with about 4 or 5uA quiescent current loss.
Is this normal? I have been told the boost converters are less efficient.
Thanks.
My design calls for 3.3v and I'm driving the Boost Converter with two AA batteries. With no load on the boost converter it is pulling 90 to 100uA. I sure didn't expect that. My previous power supply designs have been based on Microchips LDO MCP1701A regulator with about 4 or 5uA quiescent current loss.
Is this normal? I have been told the boost converters are less efficient.
Thanks.
Comments
Datasheet seems adamant that max quiescent current is 70ua, but that's for the test conditions which may not be the same. I wonder if it is possible to run a schottky from input to output so that you can shutdown the regulator completely so that it draws around 1ua and run the Prop on RCSLOW at the reduced supply. Even though boost converters are less efficient than buck they are still a lot more efficient than wasteful linear regs and quiescent current is not a parameter associated with efficiency, at least not in this application. The idea is to have something more energy efficient than a 9V battery regulated by a linear reg, and so two AA cells are about the same size as a PP3 but pack a lot more energy, that's efficiency. All the boost converter needs to do is to boost the battery voltage up to a working level and when it's not needed you can shut it down rather than leave it running. Well that's my understanding of your requirements and that's what I would be working towards.
http://www.ti.com/lit/ds/symlink/tps61291.pdf
Yep it can be shut down I don't have it designed that way but I sure could. I'm just having a hard time getting my head around the idea of having a boost converter and not having it turned on. I thought the attraction to a boost converter is it will allow you to get 3.3v out of a the battery's full voltage range all the way down to .7v. If I have it turned off that wont be the case. I have to be missing something??
I know you don't like my 9v down to 3.3v solution but from a power usage stand point it is out preforming the battery boost solution (if I keep the boost circuit on all the time). Yes I get more capacity out of two AA batteries compared to a 9v battery but it is about a wash with the extra current required for the boost circuit.
I see Tony is suggesting I do the same thing your suggesting. Let me ask this. When the propeller is running on the internal RC slow clock (~20kHz) will it continue to run reliably with say 1.4v VDD? If so I think I'm starting to get what you guys are suggesting. As part of my wake-up process I could pull a pin low wake up the boost circuit and then once I have stable 3.3v VDD switch back out to external crystal and rock on.
http://forums.parallax.com/showthread.php/129731-Prop-Limbo!-how-low-(power-voltage)-can-it-go!?highlight=limbo
If the eeprom needs 2v+, I would use 1M pull-up on enable so boost is on be default incase Prop have reset itself.
Read the fine print of the data sheet. Quescent current on boost converters is often specified as an equivalent load on the output. So for 2.5v input and 3.3v output, 100uA at the battery matches up with the datasheet 70uA equivalent load side current.
You can also drop the quiescent current of the boost regulator if you select a regulator with a "burst mode" where it switches in bursts at low loads.
I also agree that by-pass (or a by-pass diode) is a useful feature. Lets you shut down almost everything in the circuit and just use a few micro-amps to keep the prop running. AFIK, you can switch between RCSLOW and RCFAST with no delay. Further since the propeller boots above 1.4v, that means it will run at RCFAST above 1.4v as well.
Marty
Option 1:
- Use (2 AA Batteries) and a boost circuit when you can run the Propeller and supporting circuits directly off the battery during sleep periods. Make sure the boost circuit defaults to on (enabled pull high) so the prop can boot as the EEPROM voltage requirements are higher than the Propeller.
Option 2:
- Use (3 AA Batteries or more) and a liner regulator if a stable 3.3v for all supporting circuits is a requirement at all times even during sleep.
So my dilemma is I have to have at least 2.1v for the xBee that has to be powered 100% of the time. That means the circuit will stop working correctly down around 1.05v from each AA battery. Am I getting enough out of the AA batteries to make the voltage boost solution better than a 9v battery and my liner regulator?
Thanks!
Tony, I have been reading through TI TPS61291 spec sheet and in every reference design they show the battery directly connected to the Vin pin of the TPS61291 (no polarity reversal protection). Does the TPS61291 have polarity reversal built in? I can't find that it says that but the examples sure imply that it does. This chip looks perfect for what I need. Thanks.
They say to avoid pull-up in bypass/enable as even with a 1M it wastes 1.8uA, so look for a eeprom that works at 1.4V
http://www.ti.com/lit/ug/slvua29/slvua29.pdf
I think for this project I will still go with the 1M pull up as 1.8uA isn't going to kill me and I have other components besides the eeProm that need 3+v during boot. Once booted and the Prop completes IPL I will then put everything to sleep and pull the enable pin low putting the voltage boost into bypass mode.