Alternator Paralleler Circuit

[Gibbo]

D+ will still go through the same sequence with or without the internal regulator if the PDAR is still attached. ie: the charge lamp excitation current will pass throught the rotor (which will be grounded through the PDAR output stage) which will energise the stator which will produce an output which will be fed back to D+.

 

Er.........

 

I wonder if I was drunk when I wrote it?

 

Gibbo

[Smelly]

Hi,

 

Read post #101 carefully with regard to the voltage and power rating of Lamp 2.

 

cheers,

Pete.

 

Cheers Pete, everything has just come off the shelf at Maplins, although I forgot the diodes. Bought 2x3W 24v MES lamps for lamp 2 as per your suggestion.

 

I just need the relay and some decent sized cable fot the output and I'm there... Oh, and the diodes of course...

[Smelly]

Cheers Pete, everything has just come off the shelf at Maplins, although I forgot the diodes. Bought 2x3W 24v MES lamps for lamp 2 as per your suggestion.

 

I just need the relay and some decent sized cable fot the output and I'm there... Oh, and the diodes of course...

 

And a handy tip for DIYers such as I... make sure you've holesaws to fit the switches and lamps...

[smileypete]

Cheers Pete, everything has just come off the shelf at Maplins, although I forgot the diodes. Bought 2x3W 24v MES lamps for lamp 2 as per your suggestion.

 

I just need the relay and some decent sized cable fot the output and I'm there... Oh, and the diodes of course...

Diodes! Have realised that a 1N4007 is inadequate where it carries the coil current from the SW180/contactor, as it has a 1A rating, and the coil current will be just above this.

 

A much better choice would be the 1N5408. This has a 3A rating which is plenty. Probably worth using it for all three diodes, as it will have thicker leads and a bigger case which makes it more robust.

 

I'll amend the circuit diagrams today sometime.

 

cheers,

Pete.

[Gibbo]

Diodes! Have realised that a 1N4007 is inadequate where it carries the coil current from the SW180/contactor, as it has a 1A rating, and the coil current will be just above this.

 

Naaaaaah you're miles out.

 

The 1A rating on the 1N400X series is the continuous average rectified DC current meaning the peak is substantially higher than this.

 

On opening an SW180-2 from 14.8 volts coil voltage the diode will be required to conduct an average of about 600mA for 200mS with a maximum peak of 1.4 amps for about 50mS

 

That is absolutely miles within the rating. For 50mS, non repetitive it could easily handle 10 amps, no problem at all. The spec sheet itself states 30 amps for a single 8.3mS pulse.

 

Gibbo

[Smelly]

And don't forget, those 1N5408's are a whole 7p dearer... Is this an instance where size doesn't matter?

[smileypete]

Naaaaaah you're miles out.

 

The 1A rating on the 1N400X series is the continuous average rectified DC current meaning the peak is substantially higher than this.

 

On opening an SW180-2 from 14.8 volts coil voltage the diode will be required to conduct an average of about 600mA for 200mS with a maximum peak of 1.4 amps for about 50mS

 

That is absolutely miles within the rating. For 50mS, non repetitive it could easily handle 10 amps, no problem at all. The spec sheet itself states 30 amps for a single 8.3mS pulse.

I don't mean the quenching diode across the SW180/contactor coil, but the diode in the middle that carries the coil current continuously:

 

 

Might as well make all the diodes the same for a few pence more, so avoiding any problems with them getting mixed up.

 

cheers,

Pete.

[Gibbo]

I don't mean the quenching diode across the SW180/contactor coil, but the diode in the middle that carries the coil current continuously:

 

 

Might as well make all the diodes the same for a few pence more, so avoiding any problems with them getting mixed up.

 

cheers,

Pete.

 

Oh right, I was thinking of Chris's diagram.

 

Most 1N400X series are now specced at 1.5 amp contnuous:) (depends on the manufacturer). But you're right, 1N500x series makes more sense.

 

Gibbo

[chris w]

IT WORKS !!

 

 

The paralleler is now auto-switching off using the method of a couple of days ago, viz: Snibble's idea about separating the output transistor collector of the internal regulator from the field brush and using that as a simple comparator. Then the field brush is connected directly to the alternator controller (Sterling PDAR in my case).

 

Unfortunately, it didn't work the other day when I tried it. Back at the boat today, I took the PDAR lid off and discovered the output transistor fuse in the PDAR (5A) was blown. Replaced it and the whole thing works like a dream. I must have inadvertently touched something together the other day to cause the fuse to blow.

 

Now the ground of the relay is simply connected to the output transistor collector and hey presto (at about 14.4v) the relay drops out and the paralleler opens thus stopping the paralleling. The PDAR is driving the domestic alternator on its own.

 

I'll post the overall circuit when I have more time (tomorrow).

 

Chris

Edited May 12, 2009 by chris w

[Smelly]

Well I am pleased, however as an end user I'm skeptical as to the wisdom of alternator surgery...

 

However... well done experts for co-operating for long enough to come up with this I still reckon it's inherently a good idea!

[chris w]

Here's the final circuit for those who require auto shut-off AND have an alternator controller fitted on the domestic alternator.

 

Chris

 

[smileypete]

Hi Smelly,

 

Just a few more thoughts, when it comes to building and testing the circuit I posted, it could be best to do so in stages. I'd do this by leaving the diode in the middle out initially, then test each half individually in steps as follows:

 

1) First connect a 12V supply to the contactor coil fuse, and check Lamp 2 lights. Then bridge Lamp 2 and check it goes out and the contactor closes. After that un-bridge Lamp 2 and check the contactor opens and Lamp 2 lights up again.

 

2) Next connect a 12V supply to the latch relay coil fuse, and check Lamp 1 is unlit. Then press the pushbutton switch and check Lamp 1 lights up, and stays lit when the pushbutton is released.

 

3) Now connect the contactor coil fuse to the the ignition/accessory feed, and connect the contactor switching contacts between the alternators outputs. When the engine is running, bridge Lamp 2 and check the contactor closes. At some point one of the alternator warning lamps should come on, and it's the D+ connection on this alternator that is to be used for the latch relay feed.

 

4) Next connect the latch relay fuse to D+ of the above identified alternator. When the engine is running press the pushbutton switch and check the Lamp 1 lights, and stays lit when the pushbutton is released. Stop the engine and check that Lamp 1 goes out indicating the latch relay has opened.

 

5) Finally put the diode in the middle back (the right way round of course!) and test both parts together. When the engine is running press the pushbutton and check Lamp 1 lights and Lamp 2 goes out. Sometime later when one alternator takes over from the other, Lamp 1 should go out and Lamp 2 come back on. Cheer loudly when this happens!!!

 

6) As an extra check, when Lamp 2 comes back on the voltage across the SW180 contactor coil should be below 2.2V which is the minimum 'drop-out' voltage (the lowest voltage it can stay closed).

 

Hope this helps,

Pete.

 

Edit: Added to step 4.

Edited May 14, 2009 by smileypete

[chris w]

Sometime later when one alternator takes over from the other......

 

Pete.

This is where I have doubts that Smelly will see this happening. It didn't happen on my two alternators because the D+ doesn't go to ground when the other alternator rose above the first alternator's reg voltage (as we all thought it would). Instead the first alternator reg went into its really high speed switching mode which screwed things up.

 

For the relay we are talking in terms of 5mS timescales or 200Hz. The switching speeed of the reg is way above this.

 

Chris

Edited May 13, 2009 by chris w

[Gibbo]

This is where I have doubts that Smelly will see this happening. It didn't happen on my two alternators because the D+ doesn't go to ground when the other alternator rose above the first alternator's reg voltage (as we all thought it would). Instead the first alternator reg went into its really high speed switching mode which screwed things up.

 

But that's because you separated the alternators as soon as it happened isn't it?

 

Gibbo

[Sir Nibble]

D+ doesn't go to ground when the other alternator rose above the first alternator's reg voltage (as we all thought it would).

Chris

Exceruse me! I never thought it would. If this had been a battery sensed alternator then yes, but as a machine sensed unit it's entire life is dedicated to getting D+ to 14.2V and keeping it there. If the only current being generated is just enough rotor current to hold that voltage then the alternator will do it secure in the knowledge that it is a good little alternator that it's mummy can be proud of.

[Gibbo]

Exceruse me! I never thought it would. If this had been a battery sensed alternator then yes, but as a machine sensed unit it's entire life is dedicated to getting D+ to 14.2V and keeping it there. If the only current being generated is just enough rotor current to hold that voltage then the alternator will do it secure in the knowledge that it is a good little alternator that it's mummy can be proud of.

 

Yabutttt

 

When the paralleling relay is closed and the other alternator is above this alternator's reg voltage, the internal reg will shut down, it will be then be drawing next to no current ( a couple of milliamps) from the D+ terminal and the charge warning light will hold the D+ terminal above its regulation voltage. The reg will be completely off.

 

The reason it didn't stay off was because both relays (the paralleling relay and the control relay) both opened at the same time. Had the big relay been snubbed (thus increasing its opening time to about 50mS or more) but the smaller one NOT snubbed it would have worked without hacking into the internal regulator.

 

Gibbo

[chris w]

I fully understand what you are saying but I find it hard to believe though that the contactor would open as fast as the latching relay anyway - snubbing or no snubbing. The sheer comparative mass of the contactor, as compared to the relay, would imply that it would be slower. The small relay only needs to be a few mS faster than the contactor for the relay to have time to open and the paralleler to switch off.

 

That's why I remain to be convinced that snubbing will solve the issue. I guess Smelly is about to find out!

 

Chris

[Gibbo]

I fully understand what you are saying but I find it hard to believe though that the contactor would open as fast as the latching relay anyway - snubbing or no snubbing. The sheer comparative mass of the contactor, as compared to the relay, would imply that it would be slower. The small relay only needs to be a few mS faster than the contactor for the relay to have time to open and the paralleler to switch off.

 

That's why I remain to be convinced that snubbing will solve the issue. I guess Smelly is about to find out!

 

Chris

 

Well it all depends on the response time of the regulator but (just measured them):-

 

All from 14.4 volts - average of 5 measurements on each.

 

SW180-2 opening time.

 

No snubbing. = 8.9mS

1N4007 snubbing = 47mS

 

Auto cube relay opening time.

 

No snubbing = 1.14mS

1N4007 snubbing = 15.6mS

 

Gibbo

[chris w]

Well it all depends on the response time of the regulator but (just measured them):-

 

All from 14.4 volts - average of 5 measurements on each.

 

SW180-2 opening time.

 

No snubbing. = 8.9mS

1N4007 snubbing = 47mS

 

Auto cube relay opening time.

 

No snubbing = 1.14mS

1N4007 snubbing = 15.6mS

 

Gibbo

So with no snubbing, the cube relay opens nearly 8mS before the contactor. That's nearly 7 times the time it needs to open. So why is it not working without snubbing if the opening speed of the cube relay is the issue?

 

Even with snubbing on both, the time difference is about 32mS, more than twice the time that the cube relay requires to open with snubbing installed.

 

Chris

[Gibbo]

So with no snubbing, the cube relay opens nearly 8mS before the contactor. That's nearly 7 times the time it needs to open. So why is it not working without snubbing if the opening speed of the cube relay is the issue?

 

Even with snubbing on both, the time difference is about 32mS, more than twice the time that the cube relay requires to open with snubbing installed.

 

Chris

 

Dunno

 

Get your scope down to the boat and find out!

 

Gibbo

 

Edit: Oh wait, it's because your pDAR wasn't working.

Edited May 13, 2009 by Gibbo

[chris w]

Dunno

 

Get your scope down to the boat and find out!

 

Gibbo

 

Edit: Oh wait, it's because your pDAR wasn't working.

No... my PDAR was working at that point. The PDAR fuse blew only during the last set of experiments with the reg power transistor collector stuff.

 

Chris

[Gibbo]

No... my PDAR was working at that point. The PDAR fuse blew only during the last set of experiments with the reg power transistor collector stuff.

 

Chris

 

Oh right. I can't even remember what the diagram is now because (as per the norm with these things) it's changed so many times.

 

I still reckon it will work by sorting the snubbing out.

 

Anyway, I no longer care. Tomorrow a nice new shiny Lionheart goes back in the water so I'll be there for the weekend.

 

Gibbo

[smileypete]

This is where I have doubts that Smelly will see this happening. It didn't happen on my two alternators because the D+ doesn't go to ground when the other alternator rose above the first alternator's reg voltage (as we all thought it would). Instead the first alternator reg went into its really high speed switching mode which screwed things up.

 

For the relay we are talking in terms of 5mS timescales or 200Hz. The switching speeed of the reg is way above this.

 

I understood that when two alternators were paralleled, the lower voltage one would at some point cut out, and it's charge warning light come on and the tacho output would stop working.

 

Maybe Smelly could parallel the two alternators manually and see what happens to the charge warning lights. If it comes on for one of the alternators, then the circuit I posted should work OK.

 

cheers,

Pete.

[Smelly]

Hi Smelly,

 

Just a few more thoughts, when it comes to building and testing the circuit I posted, it could be best to do so in stages. I'd do this by leaving the diode in the middle out initially, then test each half individually in steps as follows:

 

1) First connect a 12V supply to the contactor coil fuse, and check Lamp 2 lights. Then bridge Lamp 2 and check it goes out and the contactor closes. After that un-bridge Lamp 2 and check the contactor opens and Lamp 2 lights up again.

 

2) Next connect a 12V supply to the latch relay coil fuse, and check Lamp 1 is unlit. Then press the pushbutton switch and check Lamp 1 lights up, and stays lit when the pushbutton is released.The Live CD provides a non-intrusive desktop running from CD, and a graphical installer.

 

3) Now connect the contactor coil fuse to the the ignition/accessory feed, and connect the contactor switching contacts between the alternators outputs. When the engine is running, bridge Lamp 2 and check the contactor closes. At some point one of the alternator warning lamps should come on, and it's the D+ connection on this alternator that is to be used for the latch relay feed.

 

4) Next connect the latch relay fuse to D+ of the above identified alternator. When the engine is running press the pushbutton switch and check the Lamp 1 lights, and stays lit when the pushbutton is released.

 

5) Finally put the diode in the middle back (the right way round of course!) and test both parts together. When the engine is running press the pushbutton and check Lamp 1 lights and Lamp 2 goes out. Sometime later when one alternator takes over from the other, Lamp 1 should go out and Lamp 2 come back on. Cheer loudly when this happens!!!

 

6) As an extra check, when Lamp 2 comes back on the voltage across the SW180 contactor coil should be below 2.2V which is the minimum 'drop-out' voltage (the lowest voltage it can stay closed).

 

Hope this helps,

Pete.

 

Lazy quoting I know but...

 

1) as lamp 2 is 24v will it light at all? Would testing the voltage across it be better?

 

2) I think the live CD bit reflects my trying to upgrade ubuntu as I'm typing this but i'll leave it in there anyways... The relevant bit makes sense

 

3) I know which one is my engine alt so do I need to?

 

4) makes sense

 

5) I bleedin hope so...

 

I'm not referring to the circuit diagram as I'm doing two jobs at once but hpefully this will help.

 

Cheers Pete

[Smelly]

I guess Smelly is about to find out!

 

So long as the f**king contactor arrives. I've made enquiries today; it's UPS's fault and they promise me it'll arrive tomorrow. My 50mm² cable should be here before the weekend as well, although if not I can still fit the relay and test from there. The only bit we won't find out is whether the snubbing works... I wonder whether i can cheat the d+ on the engine alt to see the domestic bank's voltage in the interim. Possibly a weeny cable from engine alt d+ to domestic alt d+?

 

One final variant... we're still stuck in the outback as the cat's finally managed to figure out how to get out of the cat flap when it's barricaded (and he's eaten my homework!). We'll have to rely on extra barricades tomorrow to keep him in & get back to Brum.

 

I've all the kit bar the contactor and cable, and wholly expect the cable to arrive in't next day or two.

 

Oh... Maplins didn't have the 1N5408s in stock so I gamble a quid on 5402's instead... will they do? They're all they had in.

 

edit. even if the snubbing doesn't work I'll still be able to leave it in manual mode and go from there...

Edited May 13, 2009 by Smelly