**LOW** Power applications

[Giorgis]

May work for you but I think you're missing the most valuable data.
Again you're being intentionally vague about what this is all about,
but BY FAR the most interesting data from a black-box recorder is the
minutes (even half hour or hour if you're concerned about equipment
failure, correlated with "normal" operations) before the crash! Not the
minutes after the crash.

I know you're looking at power-hungry boards for doing this but truly
the CPU horsepower is minimal: recording a dozen parameters a hundred
times a second should not take more than a few milliamps! And you have
the 12V vehicle battery (at least up to the crash), right?

Tim.

The Air-bag was an example of the type of data. I have to supply my own
battery power. The crash will not be so severe as to destroy the
recording device.

I am thinking the whole thing can fit in a packet of sigarets now.

George

 

[Ken Smith]

The Air-bag was an example of the type of data. I have to supply my own
battery power. The crash will not be so severe as to destroy the
recording device.

I have repaired equipment that went through the crash of an aircraft. I
have also fixed things that have jumped out of the back of a truck as it
went over bumps. From this I learned few things that you may find useful.


(1) There is no way you can make a box or chassis stiff enough that it
won't flex when hit like this. You are far better off making one
very tough than very stiff and allowing it to flex a little.

(2) No connection that is made by a spring's force will remain in
contact. The time they open for is made briefer by increasing the springs
pressure but you should include a large enough ceramic capacitor across
the Vcc to hold the power up when it happens.

(3) Large value surface mount capacitors are easy to crack. Don't use
2225 capacitors. Instead use some smaller ones in parallel.

(4) Those small surface mount 10 turn pots come clean off the PCB if the
shock has any high frequency components.

(5) Small silicone-rubber O-rings make great little shock mounts. They
don't change much with temperature and it is quite easy to get a 0.010"
travel out of them. This can be enough to kill the 10KHz component in the
shock.

 

[Ken Smith]

May work for you but I think you're missing the most valuable data.

One mans noise is another mans data. He may only need to know about how
hard the head long crash was. It such a case, just one number is needed.

[....]

I know you're looking at power-hungry boards for doing this but truly
the CPU horsepower is minimal: recording a dozen parameters a hundred
times a second should not take more than a few milliamps!

I suspect that recording it would take much less than a milliamp.
Digitizing it would consume more power than storing. That would be the
place to try to minimize the power. If averages rather than snap shots
are needed, perhaps a home made delta-mod ADC would be better than any
other way:


2.2M -------------------
---/\/\----------! Cancel
! !
+---!!-- 0.01u ! CMOS logic
2.2M ! ! !
----/\/\--+--!-\ ! !
! >-+---------! Bit in
Vcc/2 --!+/ -------------------


The op-amp can be very slow since it only needs to work at a few 100Hz.

The "Bit in" input of the CMOS logic could be only active when the bit is
being latched so that a partly biased CMOS circuit would not draw a lot of
power. Since the op-amp integrates, the O/1 boundary doesn't matter so no
comparitor is needed here.

 

[linnix]

You can power the output circuit from the connected RS-232. This would
require that you not use the MAX232. I'm thinking along these lines:

I have tried charge pump circuits from cable signals (DTR and RX).
They work for low speeds and duty cycles, but fails under certain
conditions. The Max232 is more reliable, although for the price of
more PCB spaces.

If Vcc is at least 3V, this will work. With the RS-232 cable not plugged
in, the circuit floats and only the leakage in C1 matters. The P-MOSFET's
leakage is taken out of the picture.

My Vcc goes down to 1.8V, when the BOD kick-in and shut it down.

 

[Ken Smith]

I have tried charge pump circuits from cable signals (DTR and RX).
They work for low speeds and duty cycles, but fails under certain
conditions. The Max232 is more reliable, although for the price of
more PCB spaces.

Two points:

The Max232 you can't get is by its nature not reliable at all.

These were circuits you designed. Did you consider the posibility that
you did something wrong.

My Vcc goes down to 1.8V, when the BOD kick-in and shut it down.

I assume the micro isn't expected to do RS-232 when shut down. The 1.8V
level is a bit low to use directly but there are still ways to deal with
it.

 

[linnix]

[...]
The trick is power management. At 10mA, the Max232 is drawing several
time more power than the micro at idle. I am redesigning my circuit to
turn it on only when a cable is connected (A2D the DTR signal). The
Max232 (started out at Maxim, second sources from many) is about $1,
but it requires 4 1uF caps (a penny each). Siplex actually has a part
with build in caps, but they wants $3 each.

You can power the output circuit from the connected RS-232. This would
require that you not use the MAX232. I'm thinking along these lines:

I have tried charge pump circuits from cable signals (DTR and RX).
They work for low speeds and duty cycles, but fails under certain
conditions. The Max232 is more reliable, although for the price of
more PCB spaces.

Two points:

The Max232 you can't get is by its nature not reliable at all.

Why? I have no problem getting the Generic Max232. I got 5 TI and 10
Sipex on hand for testing. I don't see any problem getting enough
Sipex, TI or ST Max232 for productions.

These were circuits you designed. Did you consider the posibility that
you did something wrong.

Yes, it is possible. I just don't want to waste more time when there
is a easy alternative. The charge pumps have 3 transistors to store
energy from incoming signals. It is fine for prototypes (work Ok at
9600 baud at less than 50% duty cycles with good Vcc). I just don't
want to risk it for production.

I assume the micro isn't expected to do RS-232 when shut down. The 1.8V
level is a bit low to use directly but there are still ways to deal with
it.

It will try to transmit data on the way down (i.e. dropping Vcc) as
much as possible.
And here could be a problem, the micro is transmitting most of the
time. There is not enough receving signal transitions to trigger the
charge pumps.