Roger Long wrote:

"Stephen Trapani wrote:


Or, of course, the capacity of the batteries.


And, in this case, it would have had to be pretty impressive capacity.
After the fiasco with the oil sensing switches, I'm all for the
simplicity promoted above. What's driving this now is the fact that
the pumping capacity I want (enought to have a fighting chance of
surviving a burst stuffing box hose), combined with the hose lengths
forced by the design of the boat, means that any float switch on the
market will go into an endless cycle of pumping out the hose.

****ing big centrifugal pump (or even two) with big hose with its float
switch mounted above that for a little, positive displacement diaphram
pump with a small hose?

The little pump doesn't affect the reliability of the big one but should
be enough to cope with the runback from the big hose.

Alternatively, back to the relay idea:
Take big pump and upper float switch (NO), wire in series, pump -ve to
batt -ve, other side of switch to +ve supply. Wire the relay (NO)
contacts accross the upper float switch. Wire a lower float switch (NO)
in series with the relay coil and feed it from the pump motor +ve.
Return the other side of the relay coil to the pump -ve (which is also
the battery -ve) Wire RC snubbers accross both the relay coil and the
motor as near as practical to them and a reasonably tolerable buzzer
accross the motor. Feed from an appropriate fuse directly off the
battery with a high dB output pizeo beeper accross the fuse. Put the
sounders and the relay (in a socket) somewhere where you can get at them
if you have to. e.g. back at the panel. You need four wires down to
the bilge, (Upper switch +ve, common junction of upper switch -ve &
lower switch +ve & motor +ve, lower switch -ve [to relay coil], motor -ve)

Now lets consider the possible failure modes:

1. Upper switch fails open: No alarm, bilge can flood silently till boat
sinks.

2. Upper switch fails closed: Continuous buzzer, bilge is emptied until
eventual flat battery or pump burnout

3. Motor OC: Continuous buzzer if enough water in bilge until the boat
sinks

4. Motor SC (or jammed so fuse blows): Continuous extremely loud
unignorable high pitched tone untill fault cleared and fuse replaced,
beeper smashed or disconnected :-) or boat sinks

All the above are the same as for a simple pump setup with one switch
and no relay. The only difference is you have two audiable alarms.

5. Relay fails open or lower switch fails open: Circuit reduced to
simple switch and pump, you hear a double (or more) buzz due to runback
restarting the pump, boat does not sink.

6. Relay fails closed or lower switch fails closed: Continuous buzzer,
bilge is emptied until eventual flat battery or pump burnout or untill
you pull the relay out of its socket which puts it back to the simple
single switch case.
**** This is the only increased risk ****

The snubbers have a resistor + a capacitor in the same package.
Resistors almost always fail by going open circuit which would make the
snubber ineffective but not cause an immediate circuit failure or the
capacitor might short, in which case the resistor would limit the
current until it went open and the circuit would keep working. If you
are fairly (justifiably?) paranoid, you might use two snubbers in
parallel in place of a single one in each circuit position.

Finally for backup security fit the second pump which only needs a
simple switch a little higher in the bilge than the first pump's upper
switch with pizeo sounder accross its fuse and another accross the pump
(+ another RC snubber) off your other battery. This catches any fault
with the first pump that haven't sounded its alarm.

(NO stands for Normally Open contacts)

N.B. *DONT* PUT THE LOWER SWITCH FOR THE FIRST PUMP IN SERIES WITH ITS
MOTOR (as per your JPEG) OR YOU DOUBLE THE CHANCE OF A FAILED SWITCH
STOPPING THE PUMP WORKING WITH NO ALARM.

Hope that helps, Roger.
Have I missed anything Larry?
--
Ian Malcolm. London, ENGLAND. (NEWSGROUP REPLY PREFERRED)
ianm[at]the[dash]malcolms[dot]freeserve[dot]co[dot]uk
[at]=@, [dash]=- & [dot]=. *Warning* HTML & 32K emails -- NUL:
'Stingo' Albacore #1554 - 15' Early 60's, Uffa Fox designed,
All varnished hot moulded wooden racing dinghy.

On Mon, 16 May 2005 10:30:55 GMT, "Roger Long"
wrote:

Like this, right?

http://home.maine.rr.com/rlma/Switch.jpg

I can appreciate the effect. I made a box in high school with a coil
in it and a switch made from a nail dangling in a ring. It had
textured aluminum foil sides wired to the coil and said "Mystery Box"
on top. It was a hoot when I left it in the faculty room.

Do you think I can get away with a single contact relay? The sealed
fog light relays that are readily available seem like they should be
sized and just right for this application.


I am sure you can get away with the one pole switch relay - but the
float switch then takes the motor current make and break and I suspect
the usual float switch won't feature 20 Amp contacts - but an auto
relay will. So there's a life trade-off.
If you can source a two pole relay, let the float switches handle the
modest relay current, and the second relay switch then handles any
reasonable - even unreasonable - bilge pump load.

Brian Whatcott Altus OK

I found LD1A-12F headlight/horn relays for six bucks at the auto
store and available here for less:

http://www.jameco.com/Jameco/catalogs/c252/P138.pdf

where I've ordered sockets with coil suppression diodes. The relays
are completely sealed and appear to be as watertight and ignition safe
as any marine unit. 30 Amp, 220 watts. It'll be a simple matter to
carry a spare and they can serve as the shut off switches if the pump
jambs.

I got the parts for the snubbers. Does the presence of the diode in
the socket change anything?

Many thanks for your help on this. It's what this group should be for
instead of worrying about who's pretending to be Peggie.

--

Roger Long

I'm not sure that there needs to be any snubbing in this arrangement.
For one thing, a 30 amp contact should easily handle the surge of 12 volt
bilge pump circuit being broken.
But mainly:
the relay connects across the hi level switch (after that switch has closed)
then when the hi level switch opens, the relay stays closed
and after the lo level switch opens (stopping the pump)
the relay opens.
So the relay never sees any starting or stopping load.

wot sort of switches are you using? I've had a terrible run with Rule. The
most reliable one I have sounds like it has a ball inside that runs along
and makes contact when the body tilts.

"Roger Long" wrote in message
...
I found LD1A-12F headlight/horn relays for six bucks at the auto
store and available here for less:

http://www.jameco.com/Jameco/catalogs/c252/P138.pdf

where I've ordered sockets with coil suppression diodes. The relays
are completely sealed and appear to be as watertight and ignition safe
as any marine unit. 30 Amp, 220 watts. It'll be a simple matter to
carry a spare and they can serve as the shut off switches if the pump
jambs.

I got the parts for the snubbers. Does the presence of the diode in
the socket change anything?

Many thanks for your help on this. It's what this group should be for
instead of worrying about who's pretending to be Peggie.

--

Roger Long

Sorry to hear about your experience with Rule since I have two of
them. I'm whipping myself soundly for buying mercury switches after
preaching against them but they didn't have any clean ones in the
store. I'm glad to hear that the rolling ball ones are reliable.
Johnson makes one which I'll try to get a pair of them for the bottom
set.

I understand that the Rule's go because the repeated flexing breaks
the wires going to the switch. Is that consistent with your
experience?

I really appreciate the advice on the snubbing because incorporating
it was going to required a whole enclosure for the switch set up.

--

Roger Long



"barry lawson" wrote in message
...
I'm not sure that there needs to be any snubbing in this
arrangement.
For one thing, a 30 amp contact should easily handle the surge of 12
volt
bilge pump circuit being broken.
But mainly:
the relay connects across the hi level switch (after that switch has
closed)
then when the hi level switch opens, the relay stays closed
and after the lo level switch opens (stopping the pump)
the relay opens.
So the relay never sees any starting or stopping load.

wot sort of switches are you using? I've had a terrible run with
Rule. The
most reliable one I have sounds like it has a ball inside that runs
along
and makes contact when the body tilts.

Roger Long

Roger Long wrote:

I really appreciate the advice on the snubbing because incorporating
it was going to required a whole enclosure for the switch set up.

Minimum Snubbing, put a snubber across the pump motor. If you solder
stranded wires onto the resistor and the capacitor and twist and solder
the two components in series then put in a n old 35mm film canister and
pour in epoxy resin (glass microbubbles would be fine as a filler), you
will have a sealed unit you can connect to the motor terminals.

Why not post a link to the new circuit diagram for some peer review of
the final design?


--
Ian Malcolm. London, ENGLAND. (NEWSGROUP REPLY PREFERRED)
ianm[at]the[dash]malcolms[dot]freeserve[dot]co[dot]uk
[at]=@, [dash]=- & [dot]=. *Warning* HTML & 32K emails -- NUL:
'Stingo' Albacore #1554 - 15' Early 60's, Uffa Fox designed,
All varnished hot moulded wooden racing dinghy.

Why would the pump motor need snubbing in this design more than in the
usual set up? Or, is snubbing one of those things that is always a
good idea but seldom done?

Good idea on the potting.

I've been constantly updating the jpg behind the original link which
is:

http://home.maine.rr.com/rlma/Switch.jpg

The socket I'm getting for the relay has a "coil suppression diode".
I assume that is intended to do the same sort of thing by preventing
current from flowing in reverse when the field collapses through the
coil.

--

Roger Long



"Ian Malcolm" wrote in
message ...
Roger Long wrote:

I really appreciate the advice on the snubbing because
incorporating it was going to required a whole enclosure for the
switch set up.

Minimum Snubbing, put a snubber across the pump motor. If you
solder stranded wires onto the resistor and the capacitor and twist
and solder the two components in series then put in a n old 35mm
film canister and pour in epoxy resin (glass microbubbles would be
fine as a filler), you will have a sealed unit you can connect to
the motor terminals.

Why not post a link to the new circuit diagram for some peer review
of the final design?


--
Ian Malcolm. London, ENGLAND. (NEWSGROUP REPLY PREFERRED)
ianm[at]the[dash]malcolms[dot]freeserve[dot]co[dot]uk
[at]=@, [dash]=- & [dot]=. *Warning* HTML & 32K emails -- NUL:
'Stingo' Albacore #1554 - 15' Early 60's, Uffa Fox designed,
All varnished hot moulded wooden racing dinghy.

Roger Long wrote:
Why would the pump motor need snubbing in this design more than in the
usual set up? Or, is snubbing one of those things that is always a
good idea but seldom done?
Snubbing provides a path for the current flowing in a coil to decay into
rather than sparking accross the opening switch contacts. The bigger
the coil the more it needs snubbing. Worst case, the arc can actuallly
weld the contacts so they dont switch no more. Lack of snubbing on
motor loads is probably a major factor in the notorious poor reliability
of float switches.



Good idea on the potting.

I've been constantly updating the jpg behind the original link which
is:

http://home.maine.rr.com/rlma/Switch.jpg

Hmm, if EITHER switch fails open, you loose the pump. If you move the
pump +ve to the JUNCTION of the two switches (dont change anything
else), then only the top switch is critical. Also you could then
manually override the switches to run the pump by pulling the relay and
stuffing a (pre prepared?) jumper wire down the appropriate contacts of
the socket. Otherwise it WILL work the way its ment to, the discussion
is now just "Can we 'gold plate' it at no/little extra cost?".

The socket I'm getting for the relay has a "coil suppression diode".
I assume that is intended to do the same sort of thing by preventing
current from flowing in reverse when the field collapses through the
coil.
Yes and No. it does the same thing by letting the existing current in
the relay coil flow round and round chasing its own tail until it
dwindles to nothing. If you ever want an electric 'palm tingler' try a
9v battery, and a fairly chunky relay connected so as to open its own
coil currrent circuit (no snubbers/diodes). Typically it will develop a
couple of hundred volts accross the contacts. Oddly enough, this is the
first thing about relays an average to bright 10 year old will discover
.. . . :-)

I'm off sailing for a couple of days, back Friday. Probably get to spend
this afternoon upsidown is someon else's bilge & lockers looking at
'lectrickery. . .
--
Ian Malcolm. London, ENGLAND. (NEWSGROUP REPLY PREFERRED)
ianm[at]the[dash]malcolms[dot]freeserve[dot]co[dot]uk
[at]=@, [dash]=- & [dot]=. *Warning* HTML & 32K emails -- NUL:
'Stingo' Albacore #1554 - 15' Early 60's, Uffa Fox designed,
All varnished hot moulded wooden racing dinghy.

Gawd that's clever!

However, just as I'm rediscovering how much fun electricity can be,
I've come up with a non electrical solution.

The float switches should be protected from sloshing and debris so I
have to build an enclosure anyway. If I just put a PVC pipe siphon in
it like this:

http://home.maine.rr.com/rlma/Siphon.jpg

it should work just like the electrical circuit.

Anybody need a couple of headlight relays with sockets and snubber
components?

--

Roger Long



"Ian Malcolm" wrote in
message ...
Hmm, if EITHER switch fails open, you loose the pump. If you move
the pump +ve to the JUNCTION of the two switches (dont change
anything else), then only the top switch is critical.