Reflections on a new alternator

[Keeping Up]

I recently upgraded from an 18-year old Motorola alternator to a new A127 clone. I had to do this when the Motorola died yet again, just one month after being fitted with its 4th set of electronics at ridiculous expense; I’d been putting it off as the upgrade also needed the polarity of my Adverc to be changed, but it was clearly time to bite the bullet so while broken down in Birmingham I hopped on a train to Wolverhampton and visited Adverc who were amazingly helpful.

 

Comparing the performance of the two units, gave me the following reflections:

 

Physically the units look similar except that the A127 fan has twice as many blades as the Motorola one did. Does that improve the cooling I wonder? It certainly has a most satisfying whistle now.

 

Both alternators were rated at 70 amps. The Motorola provided 70 amps, but only if the engine was revved to at least 1800 rpm; the A127 provides 70 amps from 1500 rpm upwards

 

The Motorola gave only about 10 amps at engine tickover speed (800 rpm); the A127 gives 25 amps. At speeds slightly above tickover the A127 consistently gives considerably more output than the Motorola used to.

 

I then changed the pulley for a smaller one and naturally got even better results.

 

All these features were fairly unsurprising, I had expected a modern unit to behave better than an 18-year old unit, but there is one other difference which did surprise me. When the external controller kicked in, the output from the Motorola would ramp up slowly (taking perhaps half a second) whereas the A127 jumps instantly to the higher level of output. I’d never noticed it before, I just accepted the relatively slow movement of the ammeter needle as being normal, but I was startled by the sudden jump from the needle with the new alternator. Why should that be, I wonder?

 

The slow ramp-up of output from the Motorola, may be the explanation for the odd behaviour which I had last year after increasing the size of the battery bank, when the alternator output started oscillating between a high and low level with about a one-second period when it was being driven by the Adverc, until I made a few changes to the wiring.

 

Being fed up with being incapacitated by a dead alternator, I will now purcgase an eBay clone that can sit in my engine bay as a spare; with luck that will mean I never need it!

[Nickhlx]

A pleasing conclusion - how much was the Adverc ?

 

Was it an option to get a slightly larger capacity item in the same case size, or is that size appropriate for the bank ( battery, I meant !! )

 

Nick

Edited September 16, 2009 by Nickhlx

[Arnot]

A pleasing conclusion - how much was the Adverc ?

 

Was it an option to get a slightly larger capacity item in the same case size, or is that size appropriate for the bank ( battery, I meant !! )

 

Nick

You can convert an A127 to battery sensed for a fraction of the price of an Adverk...

 

Regards

 

Arnot

[Keeping Up]

A pleasing conclusion - how much was the Adverc ?

 

Was it an option to get a slightly larger capacity item in the same case size, or is that size appropriate for the bank ( battery, I meant !! )

 

Nick

The Adverc I've had for 15 years. They can usually change it over from "positive" to "negative" operation, as needed when changing from a Motorola to an A127, for a nominal fee.

 

I could have increased to a larger size, in fact that would have been easier in some ways because they had the larger units in stock but had to search around for a different supplier to get me a 70 Amp unit. However I didn't want to increase the current because that would also have involved a fair bit of re-wiring (with a few logistical complications along the way), plus a new splitter diode and a new ammeter.

 

You can convert an A127 to battery sensed for a fraction of the price of an Adverk...

Indeed. But I like the way the Adverc works.

[Arnot]

Indeed. But I like the way the Adverc works.

As it happens, so do I and they seem to be very well made. At very least I seem to have few customers with problems with them apart from the occasional pulsing problem.

 

I do however think that they are a bit more sophisticated than the avarage narrow boat can benefit from. I certainly would not suggest that anyone with one should change it but if considering an alternative regulation system for an alternator, especially one like the A127 clone, there are probably better ways of spending the money.

 

Regards

 

Arnot

[madcat]

Dear Arnot, tell us your plan for converting the A127 to battery sensed please, you know cats are curious.

[Arnot]

Dear Arnot, tell us your plan for converting the A127 to battery sensed please, you know cats are curious.

You will need to get hold of this regulator which can be ordered from any good auto electrical specialist. It is intended for fitment to tractors and has an orange lead which senses a thermistor to lower the voltage when the batteries are getting hot.

 

If you insert a resistor of about 220 to 240 ohms instead of the thermistor and connect the other end to the battery positive, you get battery sensing for about £20 including all parts. If you take the orange wire up to the control area and instead fo the fixed resistor, use a 100 ohm pot in series with a 180 ohm resistor, you have a variable regulator for your alternator, this time for about £30.

 

Simples...

 

Regards

 

Arnot

[madcat]

Thankyou Arnot, I may well experiment on a spare alternator next winter if I get my essential jobs done. Not sure if there is a spare 127 in my lair or not as it is quite difficult to find things since I had to put the contents of my shed in the garage.

[Gibbo]

Alternatively....................

 

Remove the yellow lead going onto the spade on the A127 reg. In its place connect the spade that normally connects to the "ind" terminal. Unscrew the charge warning light, wrap it in tin foil, replace it.

 

Battery sensing for the cost of the silver paper from a fag packet.

 

 

Gibbo

[Guest]

Alternatively....................

 

Remove the yellow lead going onto the spade on the A127 reg. In its place connect the spade that normally connects to the "ind" terminal. Unscrew the charge warning light, wrap it in tin foil, replace it.

 

Battery sensing for the cost of the silver paper from a fag packet.

 

Gibbo

Now you are talking my language.

[Arnot]

Alternatively....................

 

Remove the yellow lead going onto the spade on the A127 reg. In its place connect the spade that normally connects to the "ind" terminal. Unscrew the charge warning light, wrap it in tin foil, replace it.

 

Battery sensing for the cost of the silver paper from a fag packet.

 

 

Gibbo

I assume that was tongue in cheek...

 

Regards

 

Arnot

[Sir Nibble]

My battery sense circuit is to take the yellow lead off and connect the regulator to both the warning lamp and one n/o contact of a relay, the other contact to battery pos. The relay winding is connected to the yellow lead off the reg and -ve.

[Arnot]

My battery sense circuit is to take the yellow lead off and connect the regulator to both the warning lamp and one n/o contact of a relay, the other contact to battery pos. The relay winding is connected to the yellow lead off the reg and -ve.

And of course this would bring about an improvement in charging voltage (if I have interpreted it correctly) as for that matter would Gibbo's idea. However...

 

IMHO

 

Your system would have the rotor current flowing through it and thus when the field was "on" there would be a slight voltage drop across the sensing lead lowering the voltage at the regulator. Once the regulation point was reached and the field switched "off" the drop would disappear and the voltage would rise slightly. A form of positive feedback. I have not actually tried this idea but suspect that it would lead to the voltage "pulsing" when the engine was running at constant speed, especially when the battery was almost fully charged and the alternator output was minimal. Just as a matter of interest, have you tried it? I would be interested to know what the results were.

 

Gibbo's system follows similar logic in that it separates the regulator feed from the tri-diode and connects it to a battery supply that reflects the drop across the main rectifier. In this case however the wiring that is used for sensing not only carries the field current but also any other loads on this supply such as the instrumentation and posibly the tunnel lamp, horn and others. The outcome would probably be similar.

 

For true battery sensing, it is essential that there is a wire that connects the battery positive terminal itself directly to the sensing input of the regulator without any other current flowing through it. It is a control loop and any other connection to it that takes current will introduce an error.

 

For this to be possible on any alternator it is necessary to have a regulator circuit that has separates the regulation sensing lead and the rotor supply and most of the integrated brush box/regulators don't do this (at least for the A127). There is much discussion of the Adverk, the Sterling DAR and others of this ilk but essentially they all share a common factor, in that the sensing wire is used only for that purpose.

 

After that, the establishment of the appropriate regulation voltage can be multi stage, digital, timed or varied in any number of ways that may confer limited benefits, but, it is the battery sensing that gives rise to the major improvement in charging voltage.

 

The system I suggested is one I have used many times for upwards of 20 years in various guises and is well proven in practice. I offer it free of charge to anyone who wishes to use it - enjoy!

 

The use of a regulator designed for battery sensing brings about another quite important benefit. Without any exception I know of, they all have circuitry that does not allow the voltage on the alternator output to exceed the battery voltage by more than a volt or two and many will actually shut down if no sensing feed is detected. This is important because if, on a battery sensing system, the main charging or sensing wires become detached, the feedback at the alternator does not reflect the true battery voltage and it will continue to operate at the maximum output. This will either boil the batteries or fry the rotor windings on the alternator, possibly both.

 

Regards

 

Arnot

[Sir Nibble]

And of course this would bring about an improvement in charging voltage (if I have interpreted it correctly) as for that matter would Gibbo's idea. However...

 

IMHO

 

Your system would have the rotor current flowing through it and thus when the field was "on" there would be a slight voltage drop across the sensing lead lowering the voltage at the regulator. Once the regulation point was reached and the field switched "off" the drop would disappear and the voltage would rise slightly. A form of positive feedback. I have not actually tried this idea but suspect that it would lead to the voltage "pulsing" when the engine was running at constant speed, especially when the battery was almost fully charged and the alternator output was minimal. Just as a matter of interest, have you tried it? I would be interested to know what the results were.

 

 

 

Arnot

I anticipated the volt drop issue you describe and factor in a fat cable for that reason. In practice it seems not to make any difference, in theory when the current is at it's greatest and voltdrop is high the thing is not ready to regulate anyway and as it begins to do so the volt drop becomes less and less of an issue. I suppose it will change the shape of the charge curve slightly at the beginning of acceptance, but I've not detected it.

I've only tried it on the test bench and eventually fitted a digital regulator from cargo, which unfortunately is +ve regulation only (so needed a little alternator work) but it is rock solid at it's set (adjustable) voltage and has a blue anodised case with LEDs on!

 

I had one of those tractor regs floating about for ages and never got around to playing with it to ascertain what resistors would be needed to utilise that, now you have come up with that information I shall file it for future reference. I agree that your system is certainly the best idea, especially incorporating adjustability.

[Gibbo]

I assume that was tongue in cheek...

 

Regards

 

Arnot

 

Well yes and no. It's rough arse I admit. But it does indeed work. Better than may be thought and can usually be done in less than 5 minutes.

 

As you detail in your later post there are one or two problems but not as bad as you make out:-

 

Losing battery sensing in this "system" means the alternator shuts down so no clever protection is needed. Losing the connection between B+ and the batteries makes a big mess. But it usually does anyway.

 

It is true that the rotor current passing through the sense wire messes up the regulation a bit but think about the results of this before dismissing it. When the batteries are low, the current through the wire is high, the voltage drop therefore high, the reg therefore "sees" a lower voltage than the batteries are actually at. This results in the reg wanting to up the output of the alternator. But it can't because the alternator is already hard on. As the battery voltage approaches the regulation voltage the rotor current reduces thus the voltage drop reduces and the reg gets closer to the desired regulation voltage. The actual difference as a result of the voltage drop becomes less and less the closer the system gets to needing to be well voltage regulated. The regulation during the transition from bulk to acceptance can be a bit ropey but that period doesn't last very long. As soon as the reg starts to actually regulate it settles down again.

 

It is true that other loads on the engine panel can mess it up a bit but these loads are usually very intermittent such as a tunnel light, horn etc. A few minutes charging at a higher voltage as a result of such loads isn't really an issue.

 

The biggest problem with it is that when the ignition is first turned on the alternator is drawing full rotor current which normally it wouldn't be. This (obviously) reduces the available power for starting the engine but that really shouldn't be an issue as we're only talking a few amps.

 

And as I've said before, a better solution is to sort the cabling out so there aren't any losses thus battery sensing isn't needed anyway

 

Gibbo

[Keeping Up]

As it happens, so do I and they seem to be very well made. At very least I seem to have few customers with problems with them apart from the occasional pulsing problem.

Re Adverc. What are the symptoms of the pulsing problem? And do you manage to cure them?

 

I'm asking because I have been getting some strange symptoms and I'm working with Adverc to try and identify the cause. When the batteries are fully charged (or nearly so) they naturally take very little current and so become rather ineffective as voltage stabilisers. If I then switch on a significant load (20 amps or more), the ammeter needle vibrates around zero at maybe 10 or 20 times a second. I suspect that it is caused by the Adverc trying to control the field current to stabilise the voltage, but, with an inevitable delay around the system and the fact that any change in alternator output will inevitably cause a noticeable change in voltage after this delay, the system is effectively acting as an oscillator with the output voltage overshooting the desired voltage in each direction as the Adverc alternately attempots to increase and decrease the field current.

 

My previous (motorola) alternator, which as I mentioned above was for some reason very slow to react to requests for more/less current, used to exhibit similar symptoms but with a much slower oscillation period (up to a second)

 

I could be wrong, it may be the internal regulator trying to turn itself on anf off, I don't know yet.

[nb Innisfree]

Not the same prob but may be connected.

 

We have used an Adverc for a few years and after it seemed to stop working I replaced it after a few weeks with a Sterling Pro Digital. When I started to alter my wiring to suit the Sterling I found a loose connection! Adverc was still ok but I decided to carry on with both as a comparison. I noticed when acceptance voltage was reached the Adverc started to control as normal but instead of current increasing with voltage there was only a voltage increase, Adverc advised me that it was due to my batts getting tired.

 

Anybody had experience of this?

Edited September 23, 2009 by nb Innisfree

[Doorman]

I recently upgraded from an 18-year old Motorola alternator to a new A127 clone. I had to do this when the Motorola died yet again, just one month after being fitted with its 4th set of electronics at ridiculous expense; I’d been putting it off as the upgrade also needed the polarity of my Adverc to be changed, but it was clearly time to bite the bullet so while broken down in Birmingham I hopped on a train to Wolverhampton and visited Adverc who were amazingly helpful.

 

Comparing the performance of the two units, gave me the following reflections:

 

Physically the units look similar except that the A127 fan has twice as many blades as the Motorola one did. Does that improve the cooling I wonder? It certainly has a most satisfying whistle now.

 

Both alternators were rated at 70 amps. The Motorola provided 70 amps, but only if the engine was revved to at least 1800 rpm; the A127 provides 70 amps from 1500 rpm upwards

 

The Motorola gave only about 10 amps at engine tickover speed (800 rpm); the A127 gives 25 amps. At speeds slightly above tickover the A127 consistently gives considerably more output than the Motorola used to.

 

I then changed the pulley for a smaller one and naturally got even better results.

All these features were fairly unsurprising, I had expected a modern unit to behave better than an 18-year old unit, but there is one other difference which did surprise me. When the external controller kicked in, the output from the Motorola would ramp up slowly (taking perhaps half a second) whereas the A127 jumps instantly to the higher level of output. I’d never noticed it before, I just accepted the relatively slow movement of the ammeter needle as being normal, but I was startled by the sudden jump from the needle with the new alternator. Why should that be, I wonder?

 

The slow ramp-up of output from the Motorola, may be the explanation for the odd behaviour which I had last year after increasing the size of the battery bank, when the alternator output started oscillating between a high and low level with about a one-second period when it was being driven by the Adverc, until I made a few changes to the wiring.

 

Being fed up with being incapacitated by a dead alternator, I will now purcgase an eBay clone that can sit in my engine bay as a spare; with luck that will mean I never need it!

 

Didn't this mod effect the reading from your Tacho..?

[Keeping Up]

Didn't this mod effect the reading from your Tacho..?

Yes it did.

 

Luckily my Tacho is easily re-set, there's an adjuster on the rear. I knew the ratio of the pulley sizes, so I ran the engine and then backed off the adjustment by the same percentage. It's probably not all that acccurate, but it's near enough and one day I'll borrow a strobe and set it precisely.

[alan_fincher]

Luckily my Tacho is easily re-set, there's an adjuster on the rear. I knew the ratio of the pulley sizes, so I ran the engine and then backed off the adjustment by the same percentage. It's probably not all that acccurate, but it's near enough and one day I'll borrow a strobe and set it precisely.

Or you can use the trick of putting a magnet on the crankshaft pulley, and using it to drive a cheapo cycle computer with an appropriately set wheel size.

 

(There are good previous threads on this topic).

 

We already had the computers, so it cost us nothing, but suitable ones can be found at around a fiver.

[Keeping Up]

... suitable ones can be found at around a fiver.

A fiver? I'm not spending that sort of money.

 

It's near enough for now (can't be that far out, it shows the same as before on tickover and at max revs). Eventually I'll be able to borrow a strobe and do it precisely - but I really don't need to know it that precisely anyway.

[Gibbo]

Not the same prob but may be connected.

 

We have used an Adverc for a few years and after it seemed to stop working I replaced it after a few weeks with a Sterling Pro Digital. When I started to alter my wiring to suit the Sterling I found a loose connection! Adverc was still ok but I decided to carry on with both as a comparison. I noticed when acceptance voltage was reached the Adverc started to control as normal but instead of current increasing with voltage there was only a voltage increase, Adverc advised me that it was due to my batts getting tired.

 

Anybody had experience of this?

 

There have been several threads (arguments) on this subject between me (who has researched batteries and battery charging for several years) and another ex-poster (who played with a meter and his batteries once). He maintained that this cannot happen. I know for a fact it can and usually does.

 

Increasing the charge voltage above the gassing voltage doesn't always increase the measured charge current. Sometimes it reduces it. It always reduces the actual energy that is going into real charging. The rest of the energy goes into electroysing water. But it helps to desulfate the batteries.

 

What you are seeing is actually quite normal. Though the effect does sometimes increase as the batteries age. On the other hand it sometimes decreases. It can be worked out what is happening but it takes a long time as there are so many variables involved. Don't worry about it, there's nothing odd going on.

 

Gibbo

[nb Innisfree]

There have been several threads (arguments) on this subject between me (who has researched batteries and battery charging for several years) and another ex-poster (who played with a meter and his batteries once). He maintained that this cannot happen. I know for a fact it can and usually does.

 

Increasing the charge voltage above the gassing voltage doesn't always increase the measured charge current. Sometimes it reduces it. It always reduces the actual energy that is going into real charging. The rest of the energy goes into electroysing water. But it helps to desulfate the batteries.

 

What you are seeing is actually quite normal. Though the effect does sometimes increase as the batteries age. On the other hand it sometimes decreases. It can be worked out what is happening but it takes a long time as there are so many variables involved. Don't worry about it, there's nothing odd going on.

 

Gibbo

 

Thanks Gibbo

 

Using Pro Dig at the mo which is behaving similarly so I can relax and get on with the good life

[jakub]

Alternatively....................

 

Remove the yellow lead going onto the spade on the A127 reg. In its place connect the spade that normally connects to the "ind" terminal. Unscrew the charge warning light, wrap it in tin foil, replace it.

 

Battery sensing for the cost of the silver paper from a fag packet.

 

 

Gibbo

coming back to this thread...

but why wrap the lamps in tin foil?

[Tony Brooks]

coming back to this thread...

but why wrap the lamps in tin foil?

 

 

So it presents close to zero ohm resistance between the ign. switch and the alternator on that cable. It, in a very simple way, is the equivalent of running a cable from the battery to the said terminal on the alternator and then using a relay controlled by the ignition switch to isolate the lead when the engine is stationary and thus not charging.

 

I am sure this reflects Gibbo's sense of humour.