Best step up 1.5V->5V?

[Bob Ammerman]

I am looking for step-up switcher with following characteristics:

1: Input voltage: 1 alkaline cell, the lower the starting voltage and sustain voltage the better.

2: Output voltage: 5V

3: Current: max of 500ma (yes, I know this will kill the input cell pretty quickly)

4: Efficiency: not critical, 80% nice

5: Through hole mounting

6: Reasonable cost: In 100's total cost of circuit, including switcher, inductor, cap(s), diode under $2.00 if possible

7: Soft failure would be nice: ie: if it can't produce a full 5V on output given the source impedence then generate what it can.

Can be IC or discrete solution.
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Mike Hagen]

Good Luck on that shopping list!

Bob Blick has a really cheap and neat switcher,  it has been on the Pic
List before.  May not go down to one input cell?

Maybe he will chime in here or you can Google it?




Bob Ammerman wrote:

> I am looking for step-up switcher with following characteristics:
>
> 1: Input voltage: 1 alkaline cell, the lower the starting voltage and sustain voltage the better.
>
> 2: Output voltage: 5V
>
> 3: Current: max of 500ma (yes, I know this will kill the input cell pretty quickly)
>
> 4: Efficiency: not critical, 80% nice
>
> 5: Through hole mounting
>
> 6: Reasonable cost: In 100's total cost of circuit, including switcher, inductor, cap(s), diode under $2.00 if possible
>
> 7: Soft failure would be nice: ie: if it can't produce a full 5V on output given the source impedence then generate what it can.
>
> Can be IC or discrete solution.
>  

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Bob Blick-4]

500 mA and 80% efficiency. Ouch. I don't think those two will be so easy.

Attached is a converter that meets neither spec but does decent amounts
of current. Next email will be one with good efficiency.

Cheers, Bob

Mike Hagen wrote:

> Good Luck on that shopping list!
>
> Bob Blick has a really cheap and neat switcher,  it has been on the Pic
> List before.  May not go down to one input cell?
>
> Maybe he will chime in here or you can Google it?
>
>
>
>
> Bob Ammerman wrote:
>> I am looking for step-up switcher with following characteristics:
>>
>> 1: Input voltage: 1 alkaline cell, the lower the starting voltage and sustain voltage the better.
>>
>> 2: Output voltage: 5V
>>
>> 3: Current: max of 500ma (yes, I know this will kill the input cell pretty quickly)
>>
>> 4: Efficiency: not critical, 80% nice
>>
>> 5: Through hole mounting
>>
>> 6: Reasonable cost: In 100's total cost of circuit, including switcher, inductor, cap(s), diode under $2.00 if possible
>>
>> 7: Soft failure would be nice: ie: if it can't produce a full 5V on output given the source impedence then generate what it can.
>>
>> Can be IC or discrete solution.
>>  

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Bob Blick-4]

Here's one with good efficiency but not much current, maybe 50 mA.

Cheers,

Bob


Mike Hagen wrote:

> Good Luck on that shopping list!
>
> Bob Blick has a really cheap and neat switcher,  it has been on the Pic
> List before.  May not go down to one input cell?
>
> Maybe he will chime in here or you can Google it?
>
>
>
>
> Bob Ammerman wrote:
>> I am looking for step-up switcher with following characteristics:
>>
>> 1: Input voltage: 1 alkaline cell, the lower the starting voltage and sustain voltage the better.
>>
>> 2: Output voltage: 5V
>>
>> 3: Current: max of 500ma (yes, I know this will kill the input cell pretty quickly)
>>
>> 4: Efficiency: not critical, 80% nice
>>
>> 5: Through hole mounting
>>
>> 6: Reasonable cost: In 100's total cost of circuit, including switcher, inductor, cap(s), diode under $2.00 if possible
>>
>> 7: Soft failure would be nice: ie: if it can't produce a full 5V on output given the source impedence then generate what it can.
>>
>> Can be IC or discrete solution.
>>  

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Apptech]

> Here's one with good efficiency but not much current, maybe 50 mA.

Bob knows all this. Others may find it useful:

Add a suitable FET and a bit of gate waveform shaping and you can get as
much power as desired.

The CES2310 FEt that I have mentioned previously would work OK there.
Zetex do some very low Vgson (low gate voltage ) FETs.
There aren't a vast number that are happy at 1 volt.
A good trick is to use Vout to provide the gate drive voltage thereby
allowing a far wider range of FETs to be used.



        Russell


--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Alan B. Pearce-2]

>5: Through hole mounting

I believe this one will kill your search. Most such chips come in the micro
miniature packages.

Linear Technology and Maxim both make suitable chips, but none in a TH
package that I am aware of.

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Alan B. Pearce-2]

>The CES2310 FEt that I have mentioned previously would work OK there.
>Zetex do some very low Vgson (low gate voltage ) FETs.
>There aren't a vast number that are happy at 1 volt.

But do they come as TH parts (requirement 5 from the OPs list)?

>A good trick is to use Vout to provide the gate drive voltage
>thereby allowing a far wider range of FETs to be used.

I guess a couple of bootstrap components are not going to eat into the
couple bucks budget he gave, that much.

But I am wondering how well Bob Blicks circuits would meet the budget. There
is a reasonable number of components on those.

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Apptech]

> But I am wondering how well Bob Blicks circuits would meet the budget.
> There
> is a reasonable number of components on those.

Around $US0.20 in China in 10ks for his PFM boost converter.

Digikey in 100's, somewhat more..

Still under $2 by quite a long way I think.

The boost converter could drive an arbitrarily large through hole FET with a
bit more playing. A D-Pak is probably essentially through hole for this
application - ie can be easily hand soldered. Failing that a TO220 FET. Use
output to provide gate drive voltage and it should work well enough. FET
needs to switch after a fashion at Vgs= 1 volt for startup, but need not run
well there.




            Russell

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Olin Lathrop]

Bob Ammerman wrote:

> I am looking for step-up switcher with following characteristics:
>
> 1: Input voltage: 1 alkaline cell, the lower the starting voltage and
> sustain voltage the better.
>
> 2: Output voltage: 5V
>
> 3: Current: max of 500ma (yes, I know this will kill the input cell
> pretty quickly)
>
> 4: Efficiency: not critical, 80% nice
>
> 5: Through hole mounting
>
> 6: Reasonable cost: In 100's total cost of circuit, including
> switcher, inductor, cap(s), diode under $2.00 if possible
>
> 7: Soft failure would be nice: ie: if it can't produce a full 5V on
> output given the source impedence then generate what it can.

A long time ago I did a discrete circuit for running a string of LEDs from
two AA batteries.  This was really a boost switcher with the feedback coming
from a low side current sense.  It would be a voltage regulator if the
feedback came from the output voltage instead.

It is basically a free running oscillator with the feedback killing the
oscillation when the output gets above the regulation threshold.  The
accuracy isn't that tight since the reference is a B-E junction, but this
could be changed of course.

I sortof remember this circuit or versions of it with different values
working down to 900mV.  There is very little left in a alkaline battery when
it gets down to 900mV.  Still, you'd have a easier time if you can use two
batteries in series.  If it's a space issue, how about two AAA in series
instead of a single AA?

Anyway, I put the schematic of the 20mA LED driver circuit at
http://www.embedinc.com/pic/light2/light2.pdf.  Parts of this circuit are
overkill for a 20mA output, but when you want up to 500mA input the extra
gain stages and the like are probably a good idea.  You will need a
different switching element that can take the higher current.  Figure peak
currents will be in the 1 to 1.5A range.

I did make boards of this circuit a long time ago, and can probably find a
few left over ones if you want.  This was meant for others build easily and
is all thru hole.  I can send you some if I can find them.

I just looked around and found 2 blank boards, and 2 where someone had
installed the parts.  If I remember right, one of those never worked (the
guy barely knew which end of a soldering iron to hold), and the other I
think worked fine.  I'd be happy to send you the whole mess if you want to
experiment with them.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
[phone].  Gold level PIC consultants since 2000.
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Bob Ammerman]

For:

5. Through hole mounting

Read:

5. Hand solderable by an old man who drinks too much coffee.

Also, for

 6: Reasonable cost: In 100's total cost of circuit, including switcher,
inductor, cap(s), diode under $2.00 if possible

Read:

 6: Reasonable cost: In 100's total cost of circuit, including switcher,
inductor, cap(s), diode under $4.00 if possible


Does that help open up the possibilities??

-- Bob Ammerman

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Apptech]

> Anyway, I put the schematic of the 20mA LED driver circuit at
> http://www.embedinc.com/pic/light2/light2.pdf.  Parts of this circuit are
> overkill for a 20mA output, but when you want up to 500mA input the extra
> gain stages and the like are probably a good idea.  You will need a
> different switching element that can take the higher current.  Figure peak
> currents will be in the 1 to 1.5A range.


It's interesting that it both output voltage and output current limits.
Note that the output current limit adds about 50/Vout % inefficincy and
should be byassed if not needed.
(eg for a 5V vout the current sense takes ~~ 50/5 = 10% of the output
energy.

IF you can find an acceptable way to start it you can use a packet of HC
gates to do most of what Olin is doing here with discretes. (Q4 still needed
and Q5/6 may be). HC family operate down to 2 volts and run off their own
boosted output.(eg HC series Schmitt). Not as compact as discrete but much
easier to get OK results.Starting with one cell can be by start button
bootstrapping a capacitor to 2 volts or, as some do, pulsing a start
inductor to provide initial Vdd.


        Russell McMahon


 

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Alan B. Pearce-2]

>For:
>
>5. Through hole mounting
>
>Read:
>
>5. Hand solderable by an old man who drinks too much coffee.

OK, under those conditions I would say that any 1206 and larger SM
components, and SO- ICs are viable (although a head mounted magnifier is
advisable). I certainly have little trouble with these. Doing closer pitch
ICs than SO- series ones I do find awkward. having suitable tweezer
soldering iron certainly helps removal of passive components during
development.

So, for 1206 sized components, you can all R & non-polar C values, tantalums
come in packages that start at around 1206 (I think the A size might be
closer to 0804, but really you only get these in low value/low voltage) and
get larger as value*voltage climbs, over a range of 5 or 6 packages.
Inductors vary depending on just what you require, but people like Coilcraft
have ranges that correspond to 1206 type sizing, which would probably be
satisfactory for what you want, but if you need real low R in the L then you
will end up with larger packages where they can wind them with thicker gauge
wire.

Transistors in SOT-23 I find satisfactory to solder as they have well spaced
leads. You can get a whole heap of jelly bean types in this size package
that would be well suited to Bob Blicks circuits.

You may be able to find a suitable IC in an SO-8 package, but IIRC most of
the ones designed to operate off of a well run down single cell are in tiny
packages, designed for cell phone type mass market use, or putting in LED
torches where space is a real premium. It is probably worth working your way
through the selector guide at Linear Technology to see if something comes
into the right package and price range. Maxim also have a selector guide.

Have fun.

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Olin Lathrop]

Apptech wrote:
> It's interesting that it both output voltage and output current
> limits.

This is really a regulated current supply since it's job was to drive a
string of LEDs.  The output voltage limit was only to prevent damage in case
the output was ever left open circuit.

Since Bob wants voltage regulation, he can dispense with the current limit
altogether and make the voltage limit more accurate.  Something like a 4.3V
zener into transistor base with a little resistance to ground is probably
good enough, and will be a lot less temperature sensitive and more accurate
than what I had.  My voltage limit was only there to keep the voltage below
damaging levels, so it didn't need much accuracy.

> IF you can find an acceptable way to start

That's the tricky part.  I originally was trying to create something that
would run from a single cell.  If I remember right, I got this topology with
maybe different value to run from around 900mV.  Since I wanted it to work
from rechargable batteries, I didn't think 900mV was acceptable.  In the end
I used two batteries in series.  But Bob specifically said alkaline, which
start at around 1.5V with very little of the total energy left at 1V.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
[phone].  Gold level PIC consultants since 2000.
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Apptech]

>> IF you can find an acceptable way to start

> That's the tricky part.  I originally was trying to create something that
> would run from a single cell.  If I remember right, I got this topology
> with
> maybe different value to run from around 900mV.  Since I wanted it to work
> from rechargable batteries, I didn't think 900mV was acceptable.  In the
> end
> I used two batteries in series.  But Bob specifically said alkaline, which
> start at around 1.5V with very little of the total energy left at 1V.

I'm using the AN6601D which i've mentioned before in several made-in-China
products (not the BOGO light).
It starts on 0.6V. Runs down to 0.3V or so. Costs about $US0.08 in 10k. DIP
8 and SOIC 8.
Only 40 mA output and only about 5V- output - designed for 1 White LED or
several in parallel. It's big brother is the AN6201 but it is in SOT89? and
SOT23. The former is high current but a 3 pin brain dead topology. The
latter has a sense pin of sorts. 6201 5 pin and 6601D can be voltage or
current regulated with 1 external transistor.

One of these as a boot strap and drive rail supply for something else would
allow 1 cell operation, <<1 V startup and more.
Minimum configuration is IC, 0.1 uF, inductor. Add 1R and 1Q  for current
regulation. Add 2R and 1Q for voltage regulation.
eg this plus an MC34063 and a small amount of glue would be very very useful
and still very cheap.

Using the An6601D and current regulation and a white LD I can maintain
current to within about 5% for 1.1V to 1.5V Vin.
Cuts off at 1.1V for rechargeable battery protection BUT would go lower with
much the same results.

Don't know of a western supplier, alas. Shenzhen knows them well.



       Russell

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Brian B. Riley]

Bob, in the higher current less efficient one, is R4 correct as "22 R"  
or should it also be "220 R" like R1 ?


On Sep 4, 2008, at 11:29 PM, Bob Blick wrote:

> 500 mA and 80% efficiency. Ouch. I don't think those two will be so  
> easy.
>
> Attached is a converter that meets neither spec but does decent  
> amounts
> of current. Next email will be one with good efficiency.
>
> Cheers, Bob
>
> Mike Hagen wrote:
>> Good Luck on that shopping list!
>>
>> Bob Blick has a really cheap and neat switcher,  it has been on the  
>> Pic
>> List before.  May not go down to one input cell?
>>
>> Maybe he will chime in here or you can Google it?
>>
>>
>>
>>
>> Bob Ammerman wrote:
>>> I am looking for step-up switcher with following characteristics:
>>>
>>> 1: Input voltage: 1 alkaline cell, the lower the starting voltage  
>>> and sustain voltage the better.
>>>
>>> 2: Output voltage: 5V
>>>
>>> 3: Current: max of 500ma (yes, I know this will kill the input  
>>> cell pretty quickly)
>>>
>>> 4: Efficiency: not critical, 80% nice
>>>
>>> 5: Through hole mounting
>>>
>>> 6: Reasonable cost: In 100's total cost of circuit, including  
>>> switcher, inductor, cap(s), diode under $2.00 if possible
>>>
>>> 7: Soft failure would be nice: ie: if it can't produce a full 5V  
>>> on output given the source impedence then generate what it can.
>>>
>>> Can be IC or discrete solution.
>>>
> <1_cell_up_pwm.gif>--
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Bob Blick-4]

On Fri, 05 Sep 2008 10:16:18 -0400, "Brian B. Riley" <brianbr@...>
said:
> Bob, in the higher current less efficient one, is R4 correct as "22 R"  
> or should it also be "220 R" like R1 ?

It's 22 ohms (needs the extra current to drive the base of the next
transistor).

Cheers,

Bob

--
http://www.fastmail.fm - Access all of your messages and folders
                          wherever you are

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist