DC-DC converter or LDO regulator

CMatic · Joined: 11 Jan 2012 Posts: 69

I need advise if I should select a DC-DC converter or use a LDO regulator?

I have two identical boards with PIC16F1825, it has an LCD and DS18B20 temp sensor. I am running them continuously with (2xAAA) 3V battery. One board has a MCP1640 DC-DC converter (200mA) and a board which is directly connected to the 3v (2xAAA). They both are running identical program and yet here are the results:

1) Board without any regulator has been running continuously for nearly 5 weeks. The circuit is consuming about 0.975mA @ 2.60V

2) Board with MCP1640 is still working but the LCD is not displaying temperature anymore, even though I am measuring the battery current shows 1.30mA @ 2.04V

So, it seems that MCP1640 is consuming more current and thus draining the battery far quicker then the other board which is still working.
My question then is why should I use a LDO regulator or use DC-DC MCP1640 to maintain the 3V supply voltage. Any insight will be very much appreciated.

temtronic · Posted: Tue Sep 27, 2016 5:35 am

I just downloaded the PDF of the 1640 and it seems like a nice device but..
re: your LCD. Is it rated to be run at 3 volts ? the spec sheet will say......

re: project design: I'm assumming you're running at 3 volt,so have you followed uC's design for the 1640 for 'best performance'?

re: project design: Any real reason you can't use a 'D' size battery ? Battery size/capacity should be based on true life desired for the project. I have an LCD based clock/controller here with 2 AA alkalines in parallel that's been running for almost 3 decades,so QUALITY of battery selection is important.

re: battery config. using 2 AAA in parallel to power the 1640 as a 1.5 > 3 V boost would be best as it will supply a constant 3 volts for the life of the batteries. If in series for 3 volts, over time the voltage will drop and 'funny' things can happen..

I come from the 'old school' where you install a BIG battery to ensure long life of the product.

Jay

Ttelmah · Joined: 11 Mar 2010 Posts: 19552

The MCP1640, is a boost inverter. It has to give a higher voltage 'out' than 'in'. Now you talk about your circuit using 0.975mA at 2.6v, but if you are feeding the MCP from two AAA cells, you will have about 3 to 3.3v (assuming alkaline cells), and if you are trying to set the regulator to an output below this, it'll stop oscillating, and just be wasting power.....
Look at 'note 2' in the data sheet.
Your switcher just won't be working, and may well be wasting significant power (most circuits do this quite well if not working the way they are meant to be...).
They do allow the chip to be 'non used', for this mode, but it is meant to then be switched off and bypassed. 5.2.1 in the data sheet. Even worse, they tell you that the switcher should not be enabled, until at least 0.15v of boost is required, or the result will be high frequency oscillation on the supply line (a sure way of wasting a lot of power....).

Assuming you have tested your circuit at 2.65v, you should be configuring the MCP to deliver this voltage, and running it off the cells _in parallel_, not in series, so giving about 1.5v to 1.65v into the switcher. In fact the easiest thing would be to use something like a single C cell.

CMatic · Joined: 11 Jan 2012 Posts: 69

Thank you temtronic and Ttelmah for your suggestions. I missed reading the section 5.2.1 regarding oscillations and surely it will waste lots of power. And connecting the batteries in "parallel" is a great idea as I have never seen any circuit do so.

With my goal to make the batteries last a very long time,
Do you still see its better to use MCP1640 or I was thinking about using LDO regulator like TI TPS781?

Thank your insights.

Ttelmah · Joined: 11 Mar 2010 Posts: 19552

A switcher can do things that a LDO can't.

Can boost voltage.
Can operate over a larger range.
Efficiencies can be better (with good design).

However downsides:
More likely to give noise problems.
Generally more complex.

On paralleling batteries, a surprising amount of kit you use, does do this. Things like LED torches often do. Things like e-bikes commonly do. Big things like Tesla cars do. A lot of small kit using one or two cells driving LCD's, actually use the batteries in parallel to give them a voltage below what they want, and then a boost inverter to generate the required supply.
It is much easier to design a circuit to always work 'one way' (so always be reducing voltage or always be increasing voltage - designs for a switch mode that can go 'up and down', require things like Cuk inverters, or multi-stage - buck boost). The chip you have found is really designed to always be used to boost the voltage, so run from 1.5v to give (say) 3v.

Quite a lot of kit uses 'composite' supplies, with a switcher, giving a voltage (say) just 0.5v above what is required, followed by a LDO. Advantages are the noise performance of the LDO, combined with most of the efficiency of the switcher.

You need to think about your battery layout and choices. You currently have the 'worst' situation, of having a battery that starts above the supply you want, and can then droop below what you want before being considered exhausted. Much better to switch to a larger single cell (or parallel cells) and run using a boost circuit, or switch to a higher voltage battery and use a buck converter.

CMatic · Joined: 11 Jan 2012 Posts: 69

Ttelmah thanks for explaining this concept it really makes a lot of sense to me. I will absolutely use the 2 AAA batteries in parallel and may consider to use the LDO after all along with the MCP1640. Thanks

Gabriel · Joined: 03 Aug 2009 Posts: 1067 Location: Panama

I might be missing something but is it not generally a bad idea to parallel non rechargable batteries?
(Due to diferences in voltages, one battery will try to charge the other)
G.
_________________
CCS PCM 5.078 & CCS PCH 5.093

Ttelmah · Joined: 11 Mar 2010 Posts: 19552

Not really. They'll only do that, if they are at different states of charge. Even primary batteries will charge a little if they are a little out of balance.
What is important is to ensure that the batteries are always changed together.

Where there is big danger is if a cell is inserted backwards.

You'll find mice like the Microsoft wireless, and the Logitech ones, do exactly this with 2*AA cells.

temtronic · Posted: Wed Sep 28, 2016 2:37 pm

Also semi trucks generally use 4x12 volt batteries in parallel. What IS important is to use IDENTICAL batteries and replace them as SETS. That's the best way to ensure long life and reliable power.

Jay

Gabriel · Joined: 03 Aug 2009 Posts: 1067 Location: Panama

Hmmm was under the impression that alkaline and lithium batteries should not be paralleled Unless their internal resistance was matched.... atleast the non rechargable type.

In my bussiness lead acid batteries go parallel all day long... but these are rechargable and more "forgiving" chemistrys.

I didnt know it was cool for other chemistrys.

G.
_________________
CCS PCM 5.078 & CCS PCH 5.093

temtronic · Posted: Wed Sep 28, 2016 8:05 pm

I've got two non rechargable alkaline AA batteries in parallel, driving a 40 year old Radio Shack LCD clock. Batteries have been soldered in for at least 29 years. Silly thing keeps great time too, better than the PC it's next to !

I do know if you do parallel batteries it's best if they come from the same batch..better chance the chemistry is 'matched'.

Jay

Gabriel · Joined: 03 Aug 2009 Posts: 1067 Location: Panama

29 year old battery and not leaking?

Breed it before it dies!
_________________
CCS PCM 5.078 & CCS PCH 5.093

guy · Joined: 21 Oct 2005 Posts: 297

No one remarked about using an LDO - correct me if I'm wrong, an LDO can only be used if Vin-Vdrop > Vout. If Vin goes below 3V (or even slightly higher) Vout will also drop below 3V and no sense in using an LDO.
So the question of LDO vs. switched regulator is only relevant to buck regulators, not boost.

drolleman · Joined: 03 Feb 2011 Posts: 116

This is where wives tales come from. I put large gen sets in. and I put battery box / inverters with them. The huge genset batteries are connected to a small lead acid battery. The small battery never have a problem after 12 years, when everything gets replaced.
If I use the internal charger it goes to 14.2v the battery only will last 2 years. If I hook them to the genset witch holds at 13.7v, no problems. After working with a engineer that worked with Delco battery division for over 20 years, he stated as long as all the cells are in good condition. You can mix and match any lead acid battery. But if one cell goes down, it can take the rest with them.
So every month the maintenance person separates the batteries and load tests them individually, and replaces only batteries with dead cells. It is rare to have a battery fail, if it is held at 13.2 - 13.7v. Above that you are working for the battery manufacturers.