Searching \ for '[ee] single-cell boost converters for low-quiescen' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=single+cell+boost
Search entire site for: 'single-cell boost converters for low-quiescen'.

Exact match. Not showing close matches.
PICList Thread
'[ee] single-cell boost converters for low-quiescen'
2011\10\28@064713 by Mike Harrison

flavicon
face
I've spent the last couple of days trying to find a decent step-up converter for the following
requirement :
1.2V NimH cell, Vin down to 1V or lower
3.3V out
The load spends most of its time in sleep (20uA or so), so quiescent current  matters, ....BUT also needs to be able to supply about 80-100mA for short periods (few hundred mS). Haven't yet
determined this exactly but looks like being the 'killer' parameter!

Less than GBP2 (US$3) in 100 off, available ex-stock

I've looked through all the usual suspects on Mouser, Digikey & Farnell & found surprisingly little,
as most parts fail on either quiescent current,  maximum output current or price

A problem is many datasheets quote a 'headline' quiescent current figure taken from the OUT pin, so
this is multiplied by the step-up ratio and low-load efficiency e.g. Microchip MCP1624 is super
cheap, and quotes an Iq of 19uA, but in practice this ends up as well over 100uA from the battery.
Current candidates are Maxim Max1724 - 80mA, very low Iq but expensive

ST L6920 - plenty of output current, IQ <10uA but not clear if this is on the input or output, a
little expensive
OnSemi NCP1421 Iq about 10uA into OUT, good output current and cheap

TI TPS61221 Max Iout maybe marginal, low IQ, OK price

TI TPS61016 200mA out from 0.9V, 50uA total Iq, expensive

Not looked at LT as they are always too expensive!

I think I've covered all the major manufacturers - TI, NS, ST,Onsemi, Maxim, Allegro, Semtech,
Diodes, Torex, NJR

I wondered if anyone else had been through this and stumbled on something interesting from an
obscure manufacturer....

2011\10\28@072822 by RussellMc

face picon face
> I've spent the last couple of days trying to find a decent step-up converter for the following
> requirement :
>
> 1.2V NimH cell, Vin down to 1V or lower
> 3.3V out
> The load spends most of its time in sleep (20uA or so), so quiescent current  matters,
> ...BUT
> also needs to be able to supply about 80-100mA for short periods (few hundred mS). Haven't yet
> determined this exactly but looks like being the 'killer' parameter!
>
> Less than GBP2 (US$3) in 100 off, available ex-stock at Digikey USA.
Y_Digikey-GBP_MV


I have previously identified the L6920 as utterly marvellous - based
only on data sheet.
Despite what you have said it SEEMS ideal for your task.

$US2.11/100 Digikey in stock, lead free, ROHS compliant -  so well
inside your $ spec.

Page 5 of datasheet
www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATASHEET/CD00120294.pdf
seems (to me) to make it clear  that Iq battery is 9 uA typical and 12
uA max at 1.4V in 3.3V out.
(There is some slight confusion in figures but I'm reasonably sure
[tm] that that's what is meant. ie FB can be 1.4V as specified at Vo =
3v3 with feedback voltage sense resistors. Lack of these in diagram in
front page with 3v3 out is a mistake).

Min startup voltage is about 0.73V for 3v3 out and it will start into
a 100 mA load at Vin >= ~ 0.85 Volt.
(Graphs are usually of typical values but looks good).

Looks like a truly mavellous part.

Note that if values for FB divider similar to those in Fig 9 are used
then Idivider = 2.5/400k ~= 6.5 uA and this is no doubt not part of
quoted I1.
Ifb not quoted but probably small.

NB if you cam assert /SHDN then you get Idd of 0.1 - 1 uA.

Quite possibly Rfb can be > values shown.



___________

Rough sanity check of Iq:

At 3.3V out, 1,2V in 100 uA load fig 2 page 6 shows efficiency = 75% so

Wasted input uA = Iout x Vout /(Z x Vin)  x (1- Z)  uA     where 0<Z<1
= efficiency
= 100x 3.3 / (0.75 x 1.2) x 0.75 = 90 uA.

ie at 100 uA out Iin ~+ 360 uA of which 90 uA is 'lost'.

This is well up =on the 10 uA typical at no load but load is noticeable.


   Russell McMahon

2011\10\28@081344 by Spehro Pefhany

picon face
At 07:27 AM 10/28/2011, you wrote:


>I have previously identified the L6920 as utterly marvellous - based
>only on data sheet.
>Despite what you have said it SEEMS ideal for your task.
....

>NB if you cam assert /SHDN then you get Idd of 0.1 - 1 uA.

I see a bit of a gotcha there- the output voltage in "shutdown" won't
drop to ~zero, but rather input voltage less a diode drop. So, depending
on how the load behaves, the system load on the battery may not be ~100nA,
but could be considerably higher.

The kind of boost converter without synchronous rectification doesn't have
that feature, but efficiency will tend to be lower and they generally
require an external Schottky diode.

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
spam_OUTspeffTakeThisOuTspaminterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

2011\10\28@081758 by M.L.

flavicon
face
On Fri, Oct 28, 2011 at 6:47 AM, Mike Harrison <.....mikeKILLspamspam@spam@whitewing.co.uk> wrote:
> I wondered if anyone else had been through this and stumbled on something interesting from an
> obscure manufacturer....

Maybe you've heard of Microchip?
They have the MCP1640 which claims 19uA quiescent.


-- Martin K

2011\10\28@082408 by RussellMc

face picon face
I was thinking you may be able to run it in shutdown with  a large
enough cap to let it maintain voltage for "a while". I haven't tried
to see how practical that would be. You need a /SHDN driver. Depending
on how low a Vout you can tolerate a CMOS Schmitt inverter would be
near ideal.

>
>>NB if you cam assert /SHDN then you get Idd of 0.1 - 1 uA.
>
> I see a bit of a gotcha there- the output voltage in "shutdown" won't
> drop to ~zero, but rather input voltage less a diode drop. So, depending
> on how the load behaves, the system load on the battery may not be ~100nA,
> but could be considerably higher.
>
> The kind of boost converter without synchronous rectification doesn't have
> that feature, but efficiency will tend to be lower and they generally
> require an external Schottky diode

2011\10\28@083331 by Mike Harrison

flavicon
face
On Fri, 28 Oct 2011 08:17:18 -0400, you wrote:

>On Fri, Oct 28, 2011 at 6:47 AM, Mike Harrison <mikespamKILLspamwhitewing.co.uk> wrote:
>> I wondered if anyone else had been through this and stumbled on something interesting from an
>> obscure manufacturer....
>
>Maybe you've heard of Microchip?
>They have the MCP1640 which claims 19uA quiescent.

This looks similar to the MCP 1624, where as I mentioned the 19uA is from Vout, so it's multiplied
by the boost ratio and efficiency. Haven't tested  the 1640 yet, but on the 1624 it ends up over 100uA.

2011\10\28@083825 by Mike Harrison

flavicon
face
On Sat, 29 Oct 2011 01:23:26 +1300, you wrote:

>I was thinking you may be able to run it in shutdown with  a large
>enough cap to let it maintain voltage for "a while". I haven't tried
>to see how practical that would be. You need a /SHDN driver. Depending
>on how low a Vout you can tolerate a CMOS Schmitt inverter would be
>near ideal.
>
>>
>>>NB if you cam assert /SHDN then you get Idd of 0.1 - 1 uA.
>>
>> I see a bit of a gotcha there- the output voltage in "shutdown" won't
>> drop to ~zero, but rather input voltage less a diode drop. So, depending
>> on how the load behaves, the system load on the battery may not be ~100nA,
>> but could be considerably higher.
>>
>> The kind of boost converter without synchronous rectification doesn't have
>> that feature, but efficiency will tend to be lower and they generally
>> require an external Schottky diode.

Probably not practical as I'll have a standby draw of around 10-20uA for an RTC osc and Sharp memory
LCD - this is what the PFM mode is supposed to deal with, only kicking the coil when needed.

2011\10\28@112900 by Dwayne Reid

flavicon
face
Hi there, Mike.

I'm not clear on whether you have to maintain Vdd for the entire time or whether Vdd can be turned off when not needed.

My two currently-favorite low-voltage boost converters are NS LM2623 and Microchip MCP1640x, where x= A, B, C, D.  These suffixes determine whether the load is disconnected when the switcher is shut down.

Assuming that you need regulated Vdd all the time:

Question #1: can you use TWO boost converters?  Find a nice, cheap, low current device to run the processor and use a MCP1640B to power the high-current load.  Note that the 'B' variant completely removes power from the load when it is shut down.  The Microchip part is somewhere near $0.45 in hundreds.  Also note that its quiescent current is essentially zero when it is shut down.

Question #2: Why are you constrained to running from a single NiMH cell?  Space?  Cost?  Consider that if you did use more than a single cell, you have the option of running the processor from the un-regulated (raw) battery voltage while you are sleeping / quiescent.


I've used both of these parts in several projects.  The hardest project was handily solved by using the Microchip MCP1640B - this is for a device that has to run for a couple of years between battery changes and monitors the voltage on a standard POTS telephone line.  When the telco line is in use, the box wakes up and listens for DTMF tones.  Upon hearing the correct tone pair, it fires up a DC-DC converter which generates 30Vdc @ 50mA - 100mA for a few seconds.

I wound up using 3- "AA" cells in series - the PIC runs from either the raw battery voltage or the regulated 5V rail from the MCP part (diode summed) and the 5V-only DTMF decoder runs from the MCP converter.  That's why I chose the "B" variant - I needed to completely remove power from the DTMF chip while the system was sleeping.

Anyway - possibly something to consider.

dwayne


PS - nice capsule summary of the various DC-DC converter chips you researched.  I'm copying that into a file for future reference.  Thanks!

dwayne


At 04:47 AM 10/28/2011, Mike Harrison wrote:
>I've spent the last couple of days trying to find a decent step-up
>converter for the following requirement :
>
>1.2V NimH cell, Vin down to 1V or lower 3.3V out


-- Dwayne Reid   <.....dwaynerKILLspamspam.....planet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax
http://www.trinity-electronics.com
Custom Electronics Design and Manufacturing

2011\10\28@120149 by Mike Harrison

flavicon
face
On Fri, 28 Oct 2011 09:28:52 -0600, you wrote:

>Hi there, Mike.
>
>I'm not clear on whether you have to maintain Vdd for the entire time
>or whether Vdd can be turned off when not needed.

Need to keep an RTC running plus  an LCD (Sharp Memory LCD - super-cool 240x400 pixel, high contrast
and a couple of uA for static display).
{Quote hidden}

That may be an option, but the Microchip parts currently don#t look too good from the quiescent
current point of view.

>Question #2: Why are you constrained to running from a single NiMH
>cell?  
Customer requirement, for various reasons. It may be the case that self-discharge is at least as
much as the draw from the product....
I did tell them it wasn't ideal....

2011\10\28@123235 by Dwayne Reid

flavicon
face
At 10:01 AM 10/28/2011, Mike Harrison wrote:

> >Question #1: can you use TWO boost converters?  Find a nice, cheap,
> >low current device to run the processor and use a MCP1640B to power
> >the high-current load.  Note that the 'B' variant completely removes
> >power from the load when it is shut down.  The Microchip part is
> >somewhere near $0.45 in hundreds.  Also note that its quiescent
> >current is essentially zero when it is shut down.
>
>That may be an option, but the Microchip parts currently don#t look
>too good from the quiescent current point of view.

Actually, the quiescent consumption of the Microchip part is essentially zero while it is in shutdown.  That was my point: use something else for the stuff that has to remain operating all the time and enable the Microchip part only when you need the high current capability.

The MCP1640B is reasonably inexpensive.

Also, FWIW, the Duracell 'green' NiMH cells I've been using for the past couple of years have astonishingly-low self-discharge rates.  I'm seeing perhaps 60% to 70% of charge remaining at the end of a 1 year time period.  I'll gladly take the lower capacity (2000mA Hr for "AA") for the extra storage time.

dwayne

-- Dwayne Reid   <EraseMEdwaynerspam_OUTspamTakeThisOuTplanet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax
http://www.trinity-electronics.com
Custom Electronics Design and Manufacturing

More... (looser matching)
- Last day of these posts
- In 2011 , 2012 only
- Today
- New search...