Alternator Paralleler Circuit

[smileypete]

That's mine, it switches on but the voltage isn't getting high enough to see whether it switches off yet (got 14.2ish when the Sterling ran to absorption today). The relay on the right is the SW180.

Hi Smelly,

 

My only suggestion is, that it should be made exactly like this:

 

 

I'm more than happy to help at the weekend, to build and test the circuit as above, in a straight forward step-by-step way.

 

cheers,

Pete.

[Semitrad]

I have been following this thread with interest and have a circuit design in my head which uses very few components and fulfils all the criteria. Sadly I seem to be unable to post the circuit on here! Perhaps someone could draw and post the circuit if I describe the layout and operation. The only components used besides the contactor are a diode, momentary switch and a thyristor.

 

12V from the ignition is applied via the momentary switch to the thyristor gate and through the diode to the contactor coil, the other side of the contactor coil is connected to earth via the thyristor. The contactor coil/diode junction is also connected to the alternator W terminal.

 

Operation is fairly obvious, a full 12V is initially applied to the contactor, the W output keeps it latched and when W output stops the thyristor switches off, switching off the contactor.

 

Would it work? What do the experts think?

[WotEver]

I have been following this thread with interest and have a circuit design in my head which uses very few components and fulfils all the criteria. Sadly I seem to be unable to post the circuit on here! Perhaps someone could draw and post the circuit if I describe the layout and operation. The only components used besides the contactor are a diode, momentary switch and a thyristor.

 

12V from the ignition is applied via the momentary switch to the thyristor gate and through the diode to the contactor coil, the other side of the contactor coil is connected to earth via the thyristor. The contactor coil/diode junction is also connected to the alternator W terminal.

 

Operation is fairly obvious, a full 12V is initially applied to the contactor, the W output keeps it latched and when W output stops the thyristor switches off, switching off the contactor.

 

Would it work? What do the experts think?

 

Hi Semitrad,

 

This right?

 

Gibbo?

 

T

 

edited to change the image for one where the logo didn't obscure 0V!

Edited June 5, 2009 by tony@hdheaven.co.uk

[Semitrad]

Yes, that's it exactly, many thanks. Although it may need some tweaking it would dispose of all those imprecise bulbs and chattering relays.

[Gibbo]

Yes, that's it exactly, many thanks. Although it may need some tweaking it would dispose of all those imprecise bulbs and chattering relays.

 

Looks pretty, Except it won't work

 

The "W" terminal gives out AC which drops to either very close to zero volts or just below it. The thyristor will switch off as soon as it does so.

 

Gibbo

[blueprince]

so many responses to such a simple problem, this isn't a make your best guess technology.

[Sir Nibble]

Looks pretty, Except it won't work

 

The "W" terminal gives out AC which drops to either very close to zero volts or just below it. The thyristor will switch off as soon as it does so.

 

Gibbo

Also, closing the switch will create a short circuit!

[Gibbo]

Also, closing the switch will create a short circuit!

 

You mean via the gate on the thyristor? Yes you're quite right. I didn't even look at that! Or have you spotted another one?

 

Gibbo

 

so many responses to such a simple problem, this isn't a make your best guess technology.

 

So how come you think a hull doesn't need to be earth bonded yet everyone else, including the RCD, ABYC, ieee etc say it should be? Who's making the "best guess"?

 

Gibbo

[WotEver]

Who's making the "best guess"?

 

Hey, I'm not saying this here is best, but it's another take...

 

 

What do you think, Gibbo? The push switch is changed for spno and the circuit's self-latching. The unknown for me is whether W would ever drop low enough for long enough to switch off the thyristor, because I simply don't know that much about alternators. (Although I do know how much voltage to expect on the W terminal)

 

It's only got 4 components including the switch, so it scores for simplicity if nothing else

 

I guess the SPNO switch defeats Chris's original intention of not forgetting to turn it off, but this came to me as I was driving home last night after posting Semitrad's circuit for him.

 

Cheers,

Tony

[Gibbo]

Hey, I'm not saying this here is best, but it's another take...

 

 

What do you think, Gibbo? The push switch is changed for spno and the circuit's self-latching. The unknown for me is whether W would ever drop low enough for long enough to switch off the thyristor, because I simply don't know that much about alternators. (Although I do know how much voltage to expect on the W terminal)

 

It's only got 4 components including the switch, so it scores for simplicity if nothing else

 

I guess the SPNO switch defeats Chris's original intention of not forgetting to turn it off, but this came to me as I was driving home last night after posting Semitrad's circuit for him.

 

Cheers,

Tony

 

Right that's a better. R1 is probably much too high in value for most thyristors. It will also need another R between the gate and the junction of R1 and C1, then it will (ideally) need a diode between R1 and the junction otherwise W will discharge C1 just as well as it will charge it. But that then means that R1 can be dispensed with.

 

I still have the feeling that even if the reg shuts down completely, the remnant magnetism in the rotor will be plenty sufficient to keep the W terminal activated enough to keep the relay energised.

 

Gibbo

[Sir Nibble]

You mean via the gate on the thyristor? Yes you're quite right. I didn't even look at that! Or have you spotted another one?

 

Gibbo

As soon as voltage on W falls below the voltage at the diode anode there will be an increasingly viable path to -ve through the alternator rectifier. Phut! no more diode.

 

Hey, I'm not saying this here is best, but it's another take...

 

 

What do you think, Gibbo? The push switch is changed for spno and the circuit's self-latching. The unknown for me is whether W would ever drop low enough for long enough to switch off the thyristor, because I simply don't know that much about alternators. (Although I do know how much voltage to expect on the W terminal)

 

It's only got 4 components including the switch, so it scores for simplicity if nothing else

 

I guess the SPNO switch defeats Chris's original intention of not forgetting to turn it off, but this came to me as I was driving home last night after posting Semitrad's circuit for him.

 

Cheers,

Tony

Ok, so you turn on the switch and the W output supplies the relay and gates the thryristor to connect the -ve until the switch is turned off. So what is the thyristor there for? what will that circuit do that W-switch-relay-earth won't do?

 

I think it's worth mentioning here that the W output is not half wave, it's a full ac and will only appear to be half wave because the alternator +ve and -ve are more or less at the +ve and -ve peak.

Edited June 5, 2009 by Sir Nibble

[Gibbo]

As soon as voltage on W falls below the voltage at the diode anode there will be an increasingly viable path to -ve through the alternator rectifier. Phut! no more diode.

 

Well spotted. I didn't even notice that one.

 

Gibbo

[smileypete]

Hi,

 

I have redone the diagram of my latest circuit, to make it a bit easier to understand :

 

 

Previous comments still apply:

 

The basic idea is to overcome the problem that the latch relay and contactor need a lot more 'oomph' to close initially, than hold closed.

 

So with this circuit, when the push switch is pressed, the latch relay and contactor get full battery voltage across them. Though the latch relay has a lamp in series (which I'll come to in a bit), this will be cold and so at low resistance when the push switch is pressed.

 

When the push switch is released, the output voltage from W should be enough to hold the latch relay and contactor closed. As the other alternator takes over, the output from W will drop and allow the latch relay and contactor to open.

 

The lamp is to ensure that as the output voltage from W drops, the latch relay opens before the contactor, otherwise the contactor may 'chatter'. If the latch relay does not close initially or stay closed, the power rating of the lamp should be increased gradually, to 2W, then 3W etc.

 

If a cancel function is needed, an additional push switch can be connected across the latch relay coil. This will cause the latch relay to open when this push switch is pressed.

 

The pair of diodes D1 and D2 can also be formed by using a bridge rectifier with push on terminals, like a KPBC3504 (£2 each).

 

cheers,

Pete.

Edited June 5, 2009 by smileypete

[WotEver]

Ok, so you turn on the switch and the W output supplies the relay and gates the thryristor to connect the -ve until the switch is turned off. So what is the thyristor there for? what will that circuit do that W-switch-relay-earth won't do?

 

D'uh!

 

Good point - at this stage I'll bow back out

 

T.

[Gibbo]

D'uh!

 

Good point - at this stage I'll bow back out

 

T.

 

Aye but........

 

If you rearrange it as I suggested, keep W feeding the gate but take the power feed for the relay from the ignition feed the thing then works.

 

We're still left with the problem that I think W will still throw out close to normal voltage even with the reg turned fully off due to remnant magnetism.

 

Gibbo

[WotEver]

We're still left with the problem that I think W will still throw out close to normal voltage even with the reg turned fully off due to remnant magnetism.

 

At how much current though? Would the contactor coil (I'm thinking of Pete's circuit here) be sufficient to pull that voltage down? That coil draws an amp or so.

 

T

[Gibbo]

At how much current though? Would the contactor coil (I'm thinking of Pete's circuit here) be sufficient to pull that voltage down? That coil draws an amp or so.

 

T

 

That's the big question. It varies wildly between alternators, even of the same make and model. Even sometimes with the same alternator on different days.

 

Gibbo

[smileypete]

That's the big question. It varies wildly between alternators, even of the same make and model. Even sometimes with the same alternator on different days.

If an alternator excites via a small 'charge warning' lamp, but not without it, I'd expect any output from residual magnetism to be pretty small.

 

Edit: Also the W/tacho output is just one phase, not all three.

 

cheers,

Pete.

Edited June 5, 2009 by smileypete

[Sir Nibble]

Right, time for me to throw my hat into the ring. Sorry, I can't post a diagram, my office connection won't let me.

Remove the engine alternator regulator. Solder a lead to the -ve brush holder as if you were fitting an external reg, and connect another cable to the D+ connection instead of the yellow lead from the alternator. Now take the D+ wire you have just added to one of the common terminals on a 2 pole relayalong with the warning light, and a lead from alternator D+ to the n/c terminal, . The n/o contact connects to B+. Connect the W terminal to one side of the winding, and the other to the wire added to regulator -ve brush.

Warning light current now flows to regulator (and +ve brush) through rotor and regulator to earth, alternator spins and excites generating an output to W energising relay and transferring regulator +ve to B+ fully supplying the field and making the regulator battery sense. When the battery reaches acceptance voltage the regulator begins to switch and the lead connected from the -ve brush to relay -ve starts to become increasingly +ve until at around half output the relay drops out and returns the regulator/field input to D+. The second contact set switches the paralleling contactor.

This set up should be fully auto on/off.

 

Take out all the insults, and this thread is an interesting technical discussion.

[RLWP]

Right, time for me to throw my hat into the ring. Sorry, I can't post a diagram, my office connection won't let me.

Remove the engine alternator regulator. Solder a lead to the -ve brush holder as if you were fitting an external reg, and connect another cable to the D+ connection instead of the yellow lead from the alternator. Now take the D+ wire you have just added to one of the common terminals on a 2 pole relayalong with the warning light, and a lead from alternator D+ to the n/c terminal, . The n/o contact connects to B+. Connect the W terminal to one side of the winding, and the other to the wire added to regulator -ve brush.

Warning light current now flows to regulator (and +ve brush) through rotor and regulator to earth, alternator spins and excites generating an output to W energising relay and transferring regulator +ve to B+ fully supplying the field and making the regulator battery sense. When the battery reaches acceptance voltage the regulator begins to switch and the lead connected from the -ve brush to relay -ve starts to become increasingly +ve until at around half output the relay drops out and returns the regulator/field input to D+. The second contact set switches the paralleling contactor.

This set up should be fully auto on/off.

 

Take out all the insults, and this thread is an interesting technical discussion.

 

How can this work? You haven't included any semi-conductors, op-amps, thermionic valves or anything but wires and a relay.

 

You crazy, far-sighted fool you.

 

Richard

[MoominPapa]

How can this work? You haven't included any semi-conductors, op-amps, thermionic valves or anything but wires and a relay.

 

You crazy, far-sighted fool you.

 

Richard

A prize for anyone who can produce a vaguely plausible circuit using a coherer or a barretter or possibly a uniselector.

 

MP.

[Gibbo]

How can this work? You haven't included any semi-conductors, op-amps, thermionic valves or anything but wires and a relay.

 

Can't possibly work. Not enough (wrong) algebra.

 

Gibbo

[smileypete]

Right, time for me to throw my hat into the ring. Sorry, I can't post a diagram, my office connection won't let me.

Remove the engine alternator regulator. Solder a lead to the -ve brush holder as if you were fitting an external reg...

If hacking the engine alternator isn't a problem, another way is to tap into the 3 stator phase outputs, feed them into a diode trio, and connect the other side to the domestic batts.

 

If the diode trio heatsink is sized right, a 14.4V engine alt will give full output until the domestic batts reach 14v or more.

 

For the diode trio, part of a 35A bridge rectifier could be used for each phase, giving 100A rating in total. Not quite as good as a relay solution, but only £10 in parts.

 

cheers,

Pete.

[Gibbo]

Right, time for me to throw my hat into the ring. Sorry, I can't post a diagram, my office connection won't let me.

Remove the engine alternator regulator. Solder a lead to the -ve brush holder as if you were fitting an external reg, and connect another cable to the D+ connection instead of the yellow lead from the alternator. Now take the D+ wire you have just added to one of the common terminals on a 2 pole relayalong with the warning light, and a lead from alternator D+ to the n/c terminal, . The n/o contact connects to B+. Connect the W terminal to one side of the winding, and the other to the wire added to regulator -ve brush.

Warning light current now flows to regulator (and +ve brush) through rotor and regulator to earth, alternator spins and excites generating an output to W energising relay and transferring regulator +ve to B+ fully supplying the field and making the regulator battery sense. When the battery reaches acceptance voltage the regulator begins to switch and the lead connected from the -ve brush to relay -ve starts to become increasingly +ve until at around half output the relay drops out and returns the regulator/field input to D+. The second contact set switches the paralleling contactor.

This set up should be fully auto on/off.

 

Take out all the insults chris, and this thread is an interesting technical discussion.

 

I like that but there's something telling me that with some alternators there will be no quench diode across the rotor at the time of switchover between D+/B+ and vice-versa.

 

Gibbo

[Semitrad]

The sticking point in this auto-switchoff circuit seems to be whether the W output falls to zero or not. If it does, even for a fraction of a second, this is enough to switch off a thyristor, which of course will not conduct again until "restarted". It may not be enough to delatch a relay. Can an alternator expert tell me if this happens?

 

OK, my circuit needs a diode in the W feed same as smileypete's and the thyristor may need to be biased correctly but it is still simple and positive in operation. (if it works! I'm not a circuit designer)