Maker Pro
Maker Pro

Single cell boost help.


Not a bad try, but the max-current-out vs. input voltage graph on
page 6 of the spec sheet shows only 300mA at 3.3v out with 1.25volts
input. Its 0.3 ohm n-FET and 0.4 ohm p-FET are just not quite hefty
enough, same as the TI part I suggested.

The manufacturers like to spec large current output capability and
low voltage capability, but the datasheets indicate you can't get both
at the same time!

I can imagine paralleling chips & other methods to achieve the goal,
but if the prof's purpose is to assign a converter challenge, I'd think
it bad manners to interfere.
Using the MAX1797 (typ. Iout = 0.55A), try this - *temporarily* put
*two* 1.2V batteries in *parallel*. Will you get 3.3V out then?

Not likely to make much difference -- NiMH impedances are _low_.

Best,
James Arthur
 
I think some caps are allowed as long as i dont have a bank of them
fully charged. I just found out we are allowed to buy boost modules but
after looking at a few they are way too expensive for my design.
 
L

linnix

I think some caps are allowed as long as i dont have a bank of them
fully charged. I just found out we are allowed to buy boost modules but
after looking at a few they are way too expensive for my design.

Then build the booster to charge as much as you can
while the motor is not active. Down convert while
running the motor. Forget about buying, you need
to get into the circuit to build it. Be prepare to
burn out some parts, the point of this exercise is
to see how far you can push on the chips.
 
I think I'm am allowed to use anything as long as i don't charge it
before the run. Would switching the voltage high before bringing it
back down hurt my efficiency? I really thought long and hard before
asking. My classmates and and I are truly stumped, next week is block
integration no one i know has anything reliable. Thanks for your reply
-Rich
 
Here's a little more detail. Sorry about being to vague earlier. For my
Jr. design i have to build the power block to our solar powered hen.
The power block has to power the main control block which has back
lighted LCD and a motors and sensors block that consists of a 2
parallax servos and 3 sharp IR sensors with 3.3V. The "hen" gets its
power from a 100W flood lamp through a solar panel that provides 3V at
50mA. The "hen" get one hour to charge a depleted nimh battery and find
a foil target in a electrical tape bordered box without coming in
contact with one of the other 5 "hens". The motors block is also
required to have a fast mode where the provide 3.3V is boosted to 5V
for high speed.

The main controller block should use a low amount of current around
50mA peak. While the motors run 250mA in there slow mode (i don't even
want to think about fast) and the sensors run 3x30mA (I'm sure they can
be multiplexed). Finally my block uses runs about 3mA for my PWM micro
that sends battery level.

After some research i have found the LT1700 its DC to DC controller
that uses 2 mosfets. I will give it a try after some more research, I
am reluctant to order since all these attempts are costing me arm and a
leg. I will also look into super caps, 10F cap sounds fun.
Thank you all
-Rich
http://www.linear.com/pc/downloadDocument.do?navId=H0,C1,C1003,C1042,C1031,C1115,P1925,D1410
 
L

linnix

I think I'm am allowed to use anything as long as i don't charge it
before the run. Would switching the voltage high before bringing it
back down hurt my efficiency? I really thought long and hard before
asking. My classmates and and I are truly stumped, next week is block
integration no one i know has anything reliable. Thanks for your reply
-Rich

Let's separate the booster in two parts. First part is to oscillate to
high voltage (as high as the parts can handle, probably 10s of V).
Second part is to regulate it down. High voltage will minimize
resistive lost, especially at the high current input stage. With
a good switching regulator, there should not be too much lost due
to higher voltage.

You would need the micro to oscillate/monitor the supercap voltage,
then turn on the motor for the movements. You should keep a
separate bootster for the micros/sensors. So, you only need 250mA
for the booster. 250mA sound possible with commerical units. But
as you are finding out, manufacturers don't always tell you everything.
 
Here's a little more detail. Sorry about being to vague earlier. For my
Jr. design i have to build the power block to our solar powered hen.
The power block has to power the main control block which has back
lighted LCD and a motors and sensors block that consists of a 2
parallax servos and 3 sharp IR sensors with 3.3V. The "hen" gets its
power from a 100W flood lamp through a solar panel that provides 3V at
50mA.

One suggestion/optimization, if you have the time at this point: your
NiMH cell will charge faster and extract substantially more energy from
its solar panel if you use a maximum power point controller (MPPC) to
optimally load the solar array. As an example, Stephen Woodward
published a simple Design Idea entry in EDN magazine some time ago.
The "hen" get one hour to charge a depleted nimh battery and find
a foil target in a electrical tape bordered box without coming in
contact with one of the other 5 "hens". The motors block is also
required to have a fast mode where the provide 3.3V is boosted to 5V
for high speed.

The main controller block should use a low amount of current around
50mA peak. While the motors run 250mA in there slow mode (i don't even
want to think about fast) and the sensors run 3x30mA (I'm sure they can
be multiplexed). Finally my block uses runs about 3mA for my PWM micro
that sends battery level.

After some research i have found the LT1700 its DC to DC controller
that uses 2 mosfets. I will give it a try after some more research, I
am reluctant to order since all these attempts are costing me arm and a
leg.
I will also look into super caps, 10F cap sounds fun.

Supercaps are fun, but their 2.5v ratings are pretty limiting.

That looks like a good choice. Beware of the start-up requirements
-- see the datasheet's "Start Up Load Current," Fig. 4.

Best,
James Arthur
 
One suggestion/optimization, if you have the time at this point: your
NiMH cell will charge faster and extract substantially more energy from
its solar panel if you use a maximum power point controller (MPPC) to
optimally load the solar array. As an example, Stephen Woodward
published a simple Design Idea entry in EDN magazine some time ago.

Correction: "Maximum-power-point-tracking solar battery charger,"
Woodward, Stephen, Electronic Design, Ideas For Design, Sep. 14, 1998.

http://www.elecdesign.com/Articles/ArticleID/6262/6262.html

Also, MPPT seems to be the standard abbreviation, not MPPC.

http://www.redrok.com/electron.htm offers a few more links.

Best,
James Arthur
 
K

Kevin White

Hi,

This app note from Maxim addresses the step-up conversion of very low
voltages by having a separate converter provide the drive for one that
provides the main power. This way an external FET can be used with
very low on-resistance requiring higher gate drive.

This design can produce 2.5W at 5V with inputs down to 0.8V. It can be
easily adapted to 3.3V out by changing the feedback reistor on the
output stage.

kevin

http://www.maxim-ic.com/appnotes.cfm/appnote_number/1029
 
F

Fritz Schlunder

Hey guys, I am haven't quite a bit of trouble boosting a single nimh
cell to 3.3V at 300mA. At first Maxim looked great and user friendly.
So i order a few samples and built a few boost using the MAX1797,
MAX1724 and MAX1674. I am getting 3.3V out but as soon as the current
increases the voltage drops to 2V. I was just wondering someone could
give me some pointers because im out of ideas. I was thinking about
using mosfet driver boost regulator instead of boost ICs. But my search
has been unsuccessful. Thanks for your time.
-Rich


What kind of battery contacts are you using to connect to your AAA battery?
What is the brand of battery you are using and do you have a datasheet for
it (which we could see)?

What you are trying to achieve is probably possible (1W of useful energy
from a mostly discharged AAA NiMH cell), although extreme vigilance will be
needed in trying to achieve maximum efficiency.

You aren't using an ordinary spring contact based battery holder are you?
The spring contacts so often used on hobbyist battery holders look like they
should be of low electrical resistance, but looks can be strongly deceiving.
If you uncoil the spring, you may find the steel wire is surprisingly long,
perhaps 30 centimeters or so. Typically the wire diameter is roughly that
of 22 AWG copper wire, however, the steel they make them from has
dramatically higher resistivity to that of plain copper, perhaps in the
neighborhood of ten times as much. All this results in adding a good 100 to
200 milliohms of resistance or so directly in series with each battery (when
used in multi-cell holders). In your application, every milliohm counts, so
you must use a battery holder with superior contacts.

Aside from this, you also need to make sure to use adequately large wiring
for the given wiring distances. On breadboards or other non-permanent
prototype connection schemes, it is not unusual to end up using surprisingly
long wire runs perhaps over one meter using thin wire (IE: 22 AWG). This
can easily end up inadvertently adding tens of milliohms or so of extra
resistance which will negatively effect circuit capability.

This utility may help in selecting the proper wire sizes for practical
resistances in your application.

http://www.mogami.com/e/cad/wire-gauge.html

Additionally, your power switch needs to be of good construction with very
low resistance. A small primitive switch could easily add a few tens of
milliohms, which again will compromise circuit capability. Ideally you
would use a gold plated switch with large contact surface area.
 
B

Bob Monsen

I think I'm am allowed to use anything as long as i don't charge it
before the run. Would switching the voltage high before bringing it
back down hurt my efficiency? I really thought long and hard before
asking. My classmates and and I are truly stumped, next week is block
integration no one i know has anything reliable. Thanks for your reply

Look up "beam" robots... you have the same problem, which is a voltage
source which won't power your motors. They charge a cap, and then burst
the motor for a bit. The *average* power is thus decreased, and the
momentary power while the thing is going can be arbitrarily large.

You won't need supercaps for this, just normal caps. In fact, smaller caps
will allows the voltage to recover much faster, so you might be able to
keep the robot moving, thus working around the additional current required
to restart the motor if it stops.... Note that the actual power you can
pump into the motor is the same regardless of the cap size; all the size
does is select the frequency (vs pulse width) of the supply.

However, I have little johnson motors that will easily run down to 1V. A
gearbox could decrease the torque load on the motor sufficiently to allow
it to move, albeit slowly. You can get gearboxes online.

--
Regards,
Bob Monsen

There are two ways of resisting war: the legal way and the
revolutionary way. The legal way involves the offer of alternatinve
service not as a privilege for a few but as a right for all. The
revolutionary view involves an uncompromising resistance, with a view
to breaking the power of militarism in time of peace or the resources
of the state in time of war.
Albert Einstein (1879 - 1955)
 
Top