Alternator refurb?

I have been sorting through my collection of lucas ACR alternators, taken off and replaced on various motors and just hoarded.

I'm wondering what the more common point of failure is.

I have rigged a lead with an incandescent 12V bulb to the field terminal and connected these alternators up, all exhibit a bright warning light when connected and the 12V supply feeding the large output terminal with the negative taken to the alternator earth.

None of them dim even when spun at 2600rpm with a drill.

Taking one apart I find the field is being energised with 100mA via the brushes and slip ring and the field coils are not shorted to the armature body.

The rectifier checks out good after de soldering the three stator connections, however with the field energised and testing the stator coils for ac there is next to no volts being generated, yet the stator coils are not shorted to earth. I surmise the insulation of the stator coils has failed to a dead short.

Any other theories?

Are you using the correct wattage bulb? Not sure if it is critical, though. Normally 2.2W?

Older alternators needed a faster 'kick-in' speed for the alternator to 'boot' up. Try a higher wattage bulb in series with the field connection.

If you have disconnected all the diodes, then you have also disconnected the common 'earth' leg used by the field side. I suggest you reconnect at least the earthy side of the diode stack.

I tested the system with the alternator in the MGB and it worked, using a jump pack as the battery,extinguishing the test light at 900rpm.

Are you sure? As far AFAICS the field circuit is from the IGN spade, through the two brushes and field coil to earth, with the regulator shunting the field coil and earth.

I measure it at 42 Ohms, which suggests about 3.5W but as I say it works okay when used as a lead for the MGB which has a Lucas of about 30A so probably a 16ACR.

Is that 900 rpm engine speed?

You need diode set (2) and (3) to normally power the field coil. If either are missing there is no circuit and the bulb current passes through the rotor and regulator irrespective of any voltage in the stator coils (1).

The diagram differs from the norm and has some incosnsistencies. A typical regulator is powered only from the field diodes (3) and the slip rings are at either end of the field (rotor) coil.

Yes so about 1200rpm on the alternator

Yes for the alternator to drive the field once it is generating but I was pushing 75mA through the warning lamp and field coil to earth. spinning the rotor and expecting to see an ac output from a stator coil and the common terminal. I saw next to nothing.

I have now tried an LRA292 alternator from an unknown source, it has a huco regulator and brush box , once I cleaned the slip rings and freed up the brushes that kicked in but the torque reaction when it started recharging the jump pack knocked it out of my hand, so that one should be good.

I need one to replace one in my old tractor that seized up starter and alternator over winter through not being started.

As precaution I have ordered a new lucas but will try and see what is wrong with my other 4.

<snip>

Thanks for those mate, I'll follow it up.

Depending on the resistance of the load?

Understood.

Understood.

Not for charging 12V batteries?

Interesting, thanks.

Cheers, T i m

<snip>

That could either be a flap off the cooling fan, as per many lawnmowers or a position sensor and servo connected to the throttle via an Arduino [1]. ;-)

Cheers, T i m

[1] When I designed and built my electric racing (endurance) motorbike I used a basic voltmeter module for all sorts of things. B1V, B2V, B1+2V, I (voltage drop over the longest battery cable) and MPH. The speed was obtained by sensing the teeth in a(n unused) gear internal to the rear hub (I had scrounged parts from a local motorcycle shop).

The teeth were 'read' by a magna tic sensor I fitted through the casing, the output of that fed to a F to V converter and the V fed to my module and read in MPH. I calculated the loaded circumference of the wheel, what frequency that would equate to in f/mph and then used a bench oscillator to calibrate the voltmeter to mph. Going though a circuit radar trap showed it was within .5 mph and theirs only showed in whole digits. ;-)