DPRG
DPRG List  



[DPRG] Remedial Circuitry

Subject: [DPRG] Remedial Circuitry
From: Earl Bollinger earlwbollinger at comcast.net
Date: Sat Dec 6 06:42:01 CST 2003

Well, that looks pretty good to me at this early in the morning.

The resistor dividers should be setup for what you expect the maximum no
load voltage to be. But you may need to throw in some reverse biased
Schottkey fast-switching diodes on the ADC input side to limit the back
EMF that can be generated off the motors. This may be a problem if your
ADC's have a problem with negative going voltage spikes. Some ADC's are
0v to +Vcc and some ADC's can handle -Vdd to +Vcc voltages just fine.
You need to check your ADC's specs on that. On some ADC's you might have
to use a ZENER diode to limit the maximum voltage a ADC could see, if
the motors generate a lot of forward voltage spikes (such as braking in
reverse). Some bypass caps may be needed to smooth out the voltage
levels on the ADC inputs, to prevent the weird erratic readings being
caused by the motors. 

Depending on the motors and motor controllers, the back EMF or RFI that
could be generated, could cause some noise on the ground line that could
get into the computer system logic circuitry. Thus you might need to put
in a diode between the logic ground and the motor ground connection to
help block the back EMF or RFI off the motors from getting into the
logic circuits.
If it is noisier, you might have to run a inductor-capacitor "PI" low
pass filter on that common ground connection (like the ones in your car
radio audio systems to filter the engine noises out of the audio
system). Of course I think it's just us "old guys" anymore that remember
the loud motor whine coming out of the radio in the old days, with the
car's engine running. I remember when my car was OK, but I could pick up
the "unfiltered" car next to me from several hundred yards away on my
radio.
The motor power system and the logic circuit system should only have a
single common ground point of contact between the two. Usually the motor
system has the chassis ground and the logic system floats with only a
single common ground point of contact. It does get interesting when the
computer system requires a good shield or box of it's own, and you need
to keep them separate from the chassis/motor ground.
A lot of times that simple single common point of contact for ground is
all that is needed to filter the noise out of the computer system. But
if it doesn't then it's easy to add in filters as needed.

I assume your using opto-isolators to isolate the motor system from the
computer system. I have had motor controllers blow out on me in the
past. Sometimes the motor controller blows out and allows high voltage
to appear on the logic lines leading to the computer system, which is a
definite bad thing to have happen. The computer fuse may pop right away
but the high amp fuse on the motor side won't pop. Of course, debugging
is easy, replace all the semiconductors and low voltage caps, reprogram
and it' back up again. :(

You can't go wrong having fuses in the system. These larger systems
start to get downright dangerous. Especially at 3AM and your tired, and
the probe slips or you drop that screwdriver or bolt, "oops". Of course,
it does really irritate you when you blow your last fuse at 5AM on a
Sunday morning, and now you have to wait for the Radio Shack or Auto
Parts store to open. I consider it to be very important to try and use
common automotive type fuses (not the special electronics custom fuses)
that you can get easily at auto parts stores or Radio Shack, et cetera.
Although you can "kludge" up a temporary fuse out of many things, such
as a strand of copper wire, you have to decide how important your
computer system is if the temporary fuse doesn't pop if you have another
"oops". I almost forgot they just opened the new 24hr Walmart
SuperCenter, about a mile from my home, they have automotive fuses too.
:)  I'm saved.

Isn't it interesting, how it always seems to take three fuses to pop
before you realize there is something wrong?
The first fuse pops, maybe it was an old one and went bad.
The second one pops, humm, maybe it was a fluke.
The third one pops, drats! Stupid %^#%#^#%#%#% darn #&#&#&#& infernal
machine!



-----Original Message-----
>From: dprglist-admin at dprg.org [mailto:dprglist-admin at dprg.org] On Behalf
Of Karim Virani
Sent: Saturday, December 06, 2003 2:02 AM
To: Tom Gralewicz; dprglist at dprg.org; Earl Bollinger; SOLID Corp.
Subject: RE: [DPRG] Remedial Circuitry

Thanks Robert, Scott, Earl, Tom and Ralph for your input.  I understand
now  that reconnecting variously charged SLAs in parallel was the
primary problem with my concept.  So I'm working from your various
suggestions to plug 3 chargers into the existing circuit.  Also adding a
couple of voltage dividers to sense battery levels.  Please see the
diagram (below) that I adapted from Scott's original.

The chargers are all the harbor freight model that Tom pointed out.   

http://www.harborfreight.com/cpi/ctaf/Displayitem.taf?itemnumber=45005  

I went ahead and bought 3 during the post-Thanksgiving sale at the
Richardson store.  I'll be using them at the 2A setting.  The chargers
plug into the "O" docking connectors.  From my research, cyclic chargers
are designed to put out around 14.7v, standby chargers less.  Assuming
these are cyclic chargers (they don't specify their voltage output), and
rounding up, I'd expect a max of 15v from each charger.  When running on
batteries I'm interested in the region from 12.7v-fully charged to
12.0v-emergency shutdown.  Thus the voltage divider values for level
monitoring.  Do these calculations seem right?  I think I went in the
right direction when figuring worst-case actual resistor values.

How do I verify the actual voltage from these chargers before hand?
They're supposed to autosense when to shutoff, and don't show a voltage
when the circuit is open.  The manual says they need to detect at least
.7v from the battery before attempting to charge.

Assuming they shutoff on their own - I should be able to sense the drop
to 12.8v in the ADC.  Can I have the ADCs monitoring while the chargers
are active?  I don't know nuthin bout noise and filtering, etc.

I didn't put much in the way of protection - these chargers are supposed
to be circuit-breaker protected.  Do I need more fuses?

I don't understand why the diode is needed - I just copied it from
Scott's diagram.  On a low battery it'll drop the voltage too much for
the DC/DC converter down the line.  Can I drop it?

        
      motor power
            |
           fuse
     56k    |
  +-/\/\/---+-------O
  |         |       |              logic power (dc/dc converter)
  |        -=-      +                  |
  |        -=-   charger3             fuse
  |        -=-      -                  |  
  |         |       |                  -
  |         |       |                  ^    
  |-ADC.M   +-----O-+                  |  30k
  |         |       |         O--------+-/\/\/--+
  |         |       |         |        |        |
  |        -=-      +         +       -=-       |
  |        -=-   charger2  charger1   -=-       |--ADC.L
  |        -=-      -         -       -=-       |
  |         |       |         |        |        |
  |         |       +---------+        |        |
  |         |       |                  |        |
  +-/\/\/---+-------O---------+--------+-/\/\/--+
     10k    |                             13k
          __|_
          \ \ \

I'd appreciate any further thoughts,

Karim
_______________________________________________
DPRGlist mailing list
DPRGlist at dprg.org
http://nimon.ncc.com/mailman/listinfo/dprglist



More information about the DPRG mailing list