Date: Sat, 12 Jun 2010 16:41:48 -0400
Reply-To: David Beierl <dbeierl@ATTGLOBAL.NET>
Sender: Vanagon Mailing List <vanagon@gerry.vanagon.com>
From: David Beierl <dbeierl@ATTGLOBAL.NET>
Subject: Re: How to modify Bosch regulator for remote sensing.
In-Reply-To: <4C13E69F.3030107@cox.net>
Content-Type: text/plain; charset="us-ascii"; format=flowed
At 03:57 PM 6/12/2010, mark drillock wrote:
>One question I have is how high can the regulator be fooled into
>driving the output voltage of the alternator at the alternator? If
>during charging the wiring connection between the regulator sense
>point and the alternator output leg is somehow lost, what happens?
>In gas Vanagons the loss of the connection to the battery is a very
>real possibility and the ECU could still be on the alternator output
>side, so if the regulator could force the output voltage very high
>without knowing it I wonder if something bad might happen. In the
>stock configuration the regulator sees the charging voltage at a
>point internal to the alternator so a loss of the connection to
>battery does not mean a loss of the sense connection as well. Any data on this?
Excellent point. Something bad could definitely happen. On this
unit and I presume all of them, if you float the input the alternator
will go to full output which with a small load might be a very high
voltage. The sense input draws 8-9 ma at 14.5v, so to failsafe it
would need a suitable resistor across where the internal bus was
cut. I'll figure out what suitable is, but it shouldn't be critical.
Now if the sense lead is *grounded* that will definitely drive the
thing to full output.
I haven't addressed the situation yet of the alternator output no
longer driving the battery, but the sense input still connected there
-- effectively it's the same as floating the input, since it will be
immediately carried down to the loaded battery voltage. Looks like
the systematic solution to that would be to get the 30 terminals of
the ECU/Fuel relays off the direct alternator wire and feed them from
up forward somewhere, either the fuse/relay panel 30 terminal or
direct from the battery.
Another solution would be to use a diode to bridge the cut instead of
a resistor. That way the internal regulator voltage should never go
more than a diode drop above the regulator set point. Not ideal,
would want a warning probably; but not instant disaster either. And
put a quarter-amp fuse in the sense lead, so it would blow if the
lead was grounded.
Have I missed anything?
Yours,
David
|