Electronics question please.

[tjbrace]

Encouraged by the responses to Nigel Mercier's recent question I am tempted to pose my question, so here goes:
The power supply to my autopilot has to come from directly from the batteries (not via the distribution board). This I have done, via a 25 amp MCB, as recommended. To retain ON/OFF control at the distribution board, I have a switch to activate an automotive relay in the A/P power line. All works fine, but I am aware that precious amp/hours are wasted maintaining the coil current so I have looked into a latching relay.
Durite have just the beastie, it takes a positive pulse to activate it and another positive pulse to deactivate it. I have found a circuit which uses a relay, resistor, and capacitor to generate a single pulse from a constant 12 volts at switch on time. But I need to generate another pulse when I switch off the 12 volt supply at the distribution board.
Could this be done from a charged capacitor? or some other technique?

Thank you, T

 

[Jeva]

Just use a 'momentary' switch. N.O. When you touch it, it will give you 12v at the latch.

 

[tjbrace]

I did think of that but I'm not sure other crew would think to press the switch to turn off the AP. Prefer to use the switch on the distribution board as at present.

 

[prv]

The power supply to my autopilot has to come from directly from the batteries (not via the distribution board).

Sez who?

Pete

 

[tjbrace]

NKE and subsequently Garmin

 

[bedouin]

I think you may be over-complicating things a bit. If you want a switch at the control panel then fit a conventional switch (just watch the voltage drop) - if you don't want it fed from the distribution board then run a separate feed for it. Personally I would still want the power to the autohelm to go through the main battery isolation switch which is at the distribution board anyway.

If you do want a relay then I would still go for a simple one with the coil controlled by a switch and have the coil energised whenever the autohelm is on - the current draw will be small compared to the power required for the unit itself. You could achieve a similar effect with a darlington pair with less current drain.

 

[prv]

NKE and subsequently Garmin

They're protecting themselves from the cost of support calls from idiots hanging a 20-amp actuator off the piddly little "aux" switch fed by 1mm² bell wire on a cheap all-in-one panel. If you use a correctly-rated switch and the right sized wiring accounting for volt drop, the autopilot is not going to know the difference.

Pete

 

[tjbrace]

I think you may be over-complicating things a bit. If you want a switch at the control panel then fit a conventional switch (just watch the voltage drop) - if you don't want it fed from the distribution board then run a separate feed for it. Personally I would still want the power to the autohelm to go through the main battery isolation switch which is at the distribution board anyway.

If you do want a relay then I would still go for a simple one with the coil controlled by a switch and have the coil energised whenever the autohelm is on - the current draw will be small compared to the power required for the unit itself. You could achieve a similar effect with a darlington pair with less current drain.

I do have a separate feed to the AP and it is taken from the switched side of the isolation switch. The isolation switches are closer to the batteries than the distribution board.
The Darlington pair is a great idea, could it take the possible 25 amps?

 

[bedouin]

I do have a separate feed to the AP and it is taken from the switched side of the isolation switch. The isolation switches are closer to the batteries than the distribution board.
The Darlington pair is a great idea, could it take the possible 25 amps?

You can certainly get Darlington Pairs rated up to 100A from people like RS or CPC/Farnell.

How much extra wiring would it take to run the supply to a switch near the distribution panel?

 

[lw395]

You can certainly get Darlington Pairs rated up to 100A from people like RS or CPC/Farnell.

How much extra wiring would it take to run the supply to a switch near the distribution panel?

Darlington pair drops about 1.2 volts though.
You can get FETs which will be better.

 

[pvb]

I think you may be over-complicating things a bit.

Surely an understatement? What's the problem with simply running wires to an on/off switch at the distribution board?

 

[tjbrace]

Aesthetics. The distribution board is standard Beneteau and, compared to my previous boats, looks quite nice. I would hate to carve it about to uprate a switch and the wiring for the switch to overcome the problems highlighted in your previous response. I'm sure they are just piddly little switches and cabling but they do all match at present.

 

[pvb]

Aesthetics. The distribution board is standard Beneteau and, compared to my previous boats, looks quite nice. I would hate to carve it about to uprate a switch and the wiring for the switch to overcome the problems highlighted in your previous response. I'm sure they are just piddly little switches and cabling but they do all match at present.

OK, so you have a relay to switch the current. A relay draws about 100mA, is that really too much for you to live with?

 

[tjbrace]

Darlington pair drops about 1.2 volts though.
You can get FETs which will be better.

What would be the loss in a FET? My inference is that a Darlington pair is worse than my existing relay coil for wasted energy.

 

[prv]

Aesthetics. The distribution board is standard Beneteau and, compared to my previous boats, looks quite nice. I would hate to carve it about to uprate a switch

Ah, that's fair enough then. I thought you were just taking the manual far more seriously than it deserved.

For what it's worth, my Smartbank battery system puts out a fair size current to close the relay, then it switches to PWM (rapidly switching on and off, basically) to reduce the consumption while holding it closed. That would be quite easy to do with something like a Picaxe.

Pete

 

[pvb]

The distribution boards on modern production Ben/Jen/Bav boats usually include relays to handle the loads. Another one won't make any appreciable difference to power consumption.

 

[tjbrace]

Thanks PVB, I don't think 100ma is much but I have had problems maintaining the domestic supply when sailing which results in too much engine. This winter I have renewed the batteries and I will be installing a D400 wind turbine.
The current AP power supply has evolved from using cb as the on off switch to the relay I have now. The CBs have been located in various places around the boat and always resulted in the power being left on accidentally with disastrous results to the battery life.

So what I have now should work ok, just trying for another improvement with a latching arrangement but it seems too difficult to produce the two pulses through electronic components. I've been trying to teach myself about transistors but the variations seem endless.
I guess your Picaxe suggestion is a microprocessor, but that would use power to run its CPU.
I'll keep looking, but what I have now should be ok for now.

 

[GHA]

What would be the loss in a FET? My inference is that a Darlington pair is worse than my existing relay coil for wasted energy.

These can take 16a (subject to temperature) and draw next to nothing -

http://www.st.com/web/en/resource/technical/document/datasheet/CD00002501.pdf

You could parallel up a few in a row....?

 

[tjbrace]

Very interesting, thank you. Two of the 16amp version should cope with the load. I may buy a few and experiment with paralleling them.

 

[Interlude]

Putting two directly in parallel may not help - they will have (small) differences in "on" resistance, which will result in unequal sharing of current - maybe small enough but will require careful wiring to equalize external resistance. See recommendations in paralleling batteries, i.e. external connections off + and - terminals of different devices.
Be nice to measure current in each, but inserting ammeter in one path will decrease current in the device being measured in parallel with the device not having the ammeter in series. Best to find a single solid state relay which can handle the current and which has negligible control current. Jaycar have a 100A 30V device that will drop 0.35 volts at 100A, needs 28 mA to control (at 28v control voltage, less at 12 v and can be as low as 3v. SY-4086. I have used this device to disconnect house load when inadvertently left on when house battery gets down to 11.5 volts or so.