Charge splitting diodes.

Extra 10% compounded for each letter in the word "MARINE". Every Chandlers has a sign on the other side of the counter that says "Rip those Mugs Off"
 
The diodes are usually mounted on a substantial aluminium heat-sink. This must cost something, though I agree nothing like the prices being charged. As a matter of interest I bought one of those ebay diodes and wired it across the battery isolating switch on my boat - to prevent blowing the alternator diodes if the switch was inadvertently opened when the engine was running.
 
A marine charge splitter is generally an assembly incorporating diodes and a heat sink, possibly a few terminals and some epoxy potting compound.

A suitable rectifier from a trade source might be :http://uk.rs-online.com/web/p/rectifier/6518642/

£55 plus VAT.
If you buy a 100Amp diode, you may have to de-rate it substantially for temperature unless it is very well cooled.
Then you want a good heatsink, as it might be dissipating 100W.
Having it blow up and take out your alternator could be an expensive way of saving a few quid.

It is a small market, you will have to flog a lot of units to pay even my mortgage.
 
Why are these so expensive?

100A Schottky Rectifiers on ebay from a fiver.

Apart from the heatsink needs already mentioned, one aspect worth considering is max voltage tolerance. Most marine diode splitters are rated at around 48v max. The £5 ones you've seen are, I believe, rated at 15v max, and anyone with a "smart" regulator could well have the alternator running at over 15v sometimes.
 
You could look at something like:
http://uk.rs-online.com/web/p/rectifier/0192663/

If you have a nice big bit of metal to keep it cool.
But you may have issues connecting fat enough cables to something like that if you need to keep the losses down.
Packaging, thermal stuff and mechanics always cause half the problems.
If it was someone else's boat (and bill!), a nice bought in lump with M8 studs for connection suddenly looks good!
 
Apart from the heatsink needs already mentioned, one aspect worth considering is max voltage tolerance. Most marine diode splitters are rated at around 48v max. The £5 ones you've seen are, I believe, rated at 15v max, and anyone with a "smart" regulator could well have the alternator running at over 15v sometimes.

There is a choice and I am not qualified to judge but some are rated over 1000v. My alternator has an adjustable voltage output so forward drop is not a problem (as long as it is not too current dependent).
 
I wouldn't use diodes anyway. VSRs are the modern way IMNSHO and have several advantages - voltage drop being one of them.

I would not say a VSR is modern I bought ours out in 1980 :)

Blocking diodes were around pre 1970, and the volt drop problem was a problem back then, in the 70's Lucas made a split charge relay controlled by alternator warning light.

Price wise remember the chandler is looking for 25%, the wholesaler is looking for 25%, which means your £60 + VAT unit is £30 ish when sold by manufacture. Heatsink, diodes, potting, terminals etc are around £6 - £8, labour and overheads £8, so cost price around £18 with packaging and but with shipping, profit may be £8 - £12. But then as there made in China the above is rubbish.

For cost start looking at galvanic isolators :eek:

Brian
 
Vetus and X-split use MOSFET's due to the very low/zero volt drop.

Has any one considered using MOSFET's in a DIY diode splitter.

I thought about it, but the voltage drop in mosfets is only zero at zero current.
Although the RDS(on) can be small it is not zero.
Most of the big ones have a parasitic diode due to their structure.
It is a lot easier to increase the alternator voltage a tad, by sticking a diode in series with the regulator.
I did consider adding three diodes to take the three phases of the alternator straight to another battery.
But in practice, a splitter diode pack on a modern Bosch alternator with the regulated voltage tweaked up did what I needed. It gave me a boost charge setting, which I never got around to automating before I sold the boat.
The splitter diode pack cam with the boat.
 
Vetus and X-split use MOSFET's due to the very low/zero volt drop.

Has any one considered using MOSFET's in a DIY diode splitter.

Remember MOSFET's have a reverse biased diode across them when you design your circuit. This means you can turn them on & off in one direction, but they always conduct the other way.


Brian
 
Apart from the heatsink needs already mentioned, one aspect worth considering is max voltage tolerance. Most marine diode splitters are rated at around 48v max. The £5 ones you've seen are, I believe, rated at 15v max, and anyone with a "smart" regulator could well have the alternator running at over 15v sometimes.

Now I am out of my depth (as usual) but surely the maximum voltage difference is that between the alternator and the battery? Worst case 6 volts?
 
Now I am out of my depth (as usual) but surely the maximum voltage difference is that between the alternator and the battery? Worst case 6 volts?

Between alternator output and ground.

Plus this is not a smooth dc voltage, but a series of pulses ( around 27 per alternator rev) above the battery voltage. These are produced by your rectifier diodes in the alternator, so you have a nominal say 14.0 volt, but pulses that go up to 16 volt. It's these pulses that charge the battery, or the voltage in the pulse does.

Brian
 
Between alternator output and ground.

Plus this is not a smooth dc voltage, but a series of pulses ( around 27 per alternator rev) above the battery voltage. These are produced by your rectifier diodes in the alternator, so you have a nominal say 14.0 volt, but pulses that go up to 16 volt. It's these pulses that charge the battery, or the voltage in the pulse does.

Brian

………but still not a pd of >15v.

As a drift, to overcome the dangers of diodes blowing due to the removal of a load, would not a simple filament lamp across the alternator output suffice?
 
As has been mentioned many times, the voltage drop need not be an issue. If you are able to take your "sense" wire to the battery, the voltage drop is removed from the loop.
This can produce the right voltage at the service battery but can produce too high a voltage on the starter battery. Consider a heavily depleted large service bank - and an almost full smaller starter battery. The voltage drop across the diode feeding the service bank increases with the current so may well reach 1.4 volts or more. The alternator senses the voltage at the bank and raises the output by 1.4 volts to say 15.8 volts. The voltage drop across the diode feeding the starter bank is only 0.7 volts because the smaller service bank is taking a much smaller current. So the starter bank is sitting at 15.1 volts for maybe several hours! This is almost an equalizing charge - so not very good for any sealed, Gel, and AGM batteries. Modern split diodes use newer technology with virtually zero voltage drops, but they are much more expensive.
 
………but still not a pd of >15v.

As a drift, to overcome the dangers of diodes blowing due to the removal of a load, would not a simple filament lamp across the alternator output suffice?

I think 16 volt is, this is conservative value, when I was building mains chargers I would fit 45 volt diodes and thyristors rated at 2 -3 times average charger output amps. This gave a reasonable service life, 20 + years. You cannot run electronics on a 100% duty cycle, they don't work that way.

MOSFET's are very much in this criteria, how you drive them is critical, they are prone to thermal runaway, so how you install and cool them is critical. They are not nuts and bolts that you just screw together.

Why not fit a VSR and leave the alternator permanently connected to the engine battery ? When I designed our first VSR 35 years ago it was for a reason, to maintain alternator connection to battery, and avoid diode volt drop were just two.

Brian
 
I think 16 volt is, this is conservative value, when I was building mains chargers I would fit 45 volt diodes and thyristors rated at 2 -3 times average charger output amps. This gave a reasonable service life, 20 + years. You cannot run electronics on a 100% duty cycle, they don't work that way.

MOSFET's are very much in this criteria, how you drive them is critical, they are prone to thermal runaway, so how you install and cool them is critical. They are not nuts and bolts that you just screw together.

Why not fit a VSR and leave the alternator permanently connected to the engine battery ? When I designed our first VSR 35 years ago it was for a reason, to maintain alternator connection to battery, and avoid diode volt drop were just two.

Brian

Points taken. There appear on ebay diodes boasting 100A and 1600V so I guess that's a bit better. They are connected as a three phase bridge so would need to see the connections. Murphy would suggest they are pointing the wrong way.

I actually have two alternators;- a 40A connected permanently to the starter battery and a 130A for the domestics. This has a voltage out adjustment.

Now;- how about the light?! Also should you fuse the output? That implies the fuse would blow for over-current……………thus causing the diodes to blow?:cool: So is it a no-win situation? Or would the lamp idea save the diodes?
 
Top