LIN or DFM/FR Alternator question

[zylinx_za]

I have purchased a second alternator, custom H/O alternator.
It runs great. 270A, looks to be a housing used in many Toyotas.
Internal Voltage Regulator has a 4-pin square plug.
I cant seem to identify my regulator or find datasheets.
I want to build a arduino interface to sense the load.
the plug has 4 pins: [Ignition - Lamp - Voltage sense - Unknown]

This Unknown pin should be one of the following depending on the regulator he fitted:
DFM / FR: Field Monitor, block pulse signal indicating alternator load
COM: bi-direction communication to ECU probably LIN-2 19K2

On my setup this pin is floating currently, its not needed for the alternator to function and my car doesn't have an ECU(yes really).
When I hook up my o-scope to this pin I just get a steady 0V.
No signal at all when I switch on or off or load the alternator nothing... just a touch of noise.
I'm thinking ok maybe its LIN Bus as it would be silent with no master querying the bus.
Reading up on LIN Bus it says that it is high dominant so I was expecting a 12V pull-up. now I'm confused.

Can anyone give me some tips on identifying this 4th pin on my harness ?
Is it possible that LIN Bus slaves keep the line low when no master is on the bus ?
Any info or suggestions at all will be greatly appreciated

Attached are some images of what looks to be an identical alternator to mine.

 

[Diver300]

A LIN slave is supposed to have a pull-up. However, the LIN master has a stronger pull-up, so it could be that the pull-up has been omitted in the alternator, as it will work fine in a car without a pull-up in the slave.

LIN slaves should not keep the bus low, but if nothing is pulling it up, then it will just float.

It's also possible that the pulse signal needs a pull-up.

I suggest you put a 1 kOhm pull-up between the unknown pin and the ignition pin and see if there is any signal.

If it goes to 12 V and stays there, you could try sending each of the LIN identifiers. They are only one byte long, so there are only 256. You may get a slave frame sent back. The LIN master sends the break, the synchronisation byte and the identifier byte. The LIN slave responds with the data and a checksum. There is no gap between the identifier and the data, so on an oscilloscope it's difficult to see that there data and checksum has come from a different module, but the frame length is much longer if there is a response from the slave.

 

[zylinx_za]

Thank you Diver300

I went back today, threw on a pull-up bing bang boom like magic there was output.
Now trying to identify it, I thought this would be easier.

with the Ign. on and engine not running there is a square wave at around 530Hz

Engine running the square wave becomes sporadic. with pulses as short as 2Khz.
It almost looks like data but I'm not convinced as I don't see a synch frame or anything resembling 19K2 baud
Its also not a steady PWM signal that I would expect on DFM/FR.


If my engine RPM is very low then the signal drops low with no pulses (at the same time the alternator will stop charging)
If i raise the the RPM slightly ill start seeing pulses.
So this seems very much more likely to be a DFM type signal not a data bus.

I attached some pages from a PDF about DFM and I was confused because they seem to describe 2 different types of PWM used( Normal and Inverting ). regardless I'm not seeing either on my scope.



maybe its a 2KHz PWM and the duty cycle is being pinned at 0% and 100% quite erratically?

 

[zylinx_za]

Maybe its not DFM(Digital Field Monitor) but a more basic FR (Field Return) which could be different maybe less refined output? I will try research.

 

[Diver300]

I suspect that the signal represents whether the field winding is being supplied with voltage or is freewheeling with no voltage.

The field winding has a large inductance, so even if the voltage to the field winding is turned on and off quickly, the actual field current won't be changing much. If the field winding current does vary a bit, that will alter the voltage a bit, but the battery will smooth things out.

If there is a change in load or engine speed, the field winding current may have to change, but that change takes some time because of the inductance, so the regulator may supply voltage, or may let the field winding freewheel, for short periods until the field winding current is correct and the voltage is correct.

The result of this is that the "PWM" show as nice even waveforms on the document may be largely a fiction. In practice, small variations in load and engine speed will mean that there will always be variation.

Averaging the voltage / duty cycle over a few seconds would give a good idea of the average load, but the instantaneous duty cycle could well be just noise.

 

[Nigel Goodwin]

Some of the bottom images look very like RS232 style asynchronous serial data.

 

[zylinx_za]

Il try load the alternator properly and see if I can see a significant average difference in duty cycle.
Nigel Goodwin you think its data? interesting Il keep an open mind. If it were data it would indeed be async serial

 

[zylinx_za]

Did a little more research last night and found more evidence of it being DFM.

Here it mentions that the DFM output is just a reflection of the field excitation.
It mentions the pull-up resistor.
It even mentions what I observed when RPM was very low (insufficient phase amplitude) the line went to 0V

Here is some more info basically pointing to the same thing.

I even found documentation on a regulator IC that in pre-excite state it has a duty cycle of 14%

Well I went to my recording and checked and yeah it was 14% in the pre-excite stage, although the frequency didn't match.

My pre-excite frequency was 530Hz, I see most are 400Hz and most have a regulation frequency the same as the pre-excite frequency.
Well my regulation frequency seems to be 2Khz.


Anyways I will report back when I get to go do some testing later today.
Thanks for the help everyone.

 

[Nigel Goodwin]

Il try load the alternator properly and see if I can see a significant average difference in duty cycle.
Nigel Goodwin you think its data? interesting Il keep an open mind. If it were data it would indeed be async serial

Just that the scope diagrams look very like it.

 

[rjenkinsgb]

Put it through a low pass filter and in to an ADC, if you want a simple percentage field value.

However, you originally asked about sensing the load, which that will not give you; eg. at low revs, the field excitation will be higher than at high revs, for the same output current.

The simplest way of sensing output load is to put the main output cable through a hall effect current transducer. That gives you a voltage directly proportional to current.
eg.

HAS 400-S LEM USA Inc. | Sensors, Transducers | DigiKey

Order today, ships today. HAS 400-S – Current Sensor 400A 1 Channel Hall Effect, Open Loop Bidirectional Module from LEM USA Inc.. Pricing and Availability on millions of electronic components from Digi-Key Electronics.

www.digikey.com

 

[zylinx_za]

Put it through a low pass filter and in to an ADC, if you want a simple percentage field value.

However, you originally asked about sensing the load, which that will not give you; eg. at low revs, the field excitation will be higher than at high revs, for the same output current.

The simplest way of sensing output load is to put the main output cable through a hall effect current transducer. That gives you a voltage directly proportional to current.
eg.

HAS 400-S LEM USA Inc. | Sensors, Transducers | DigiKey

Order today, ships today. HAS 400-S – Current Sensor 400A 1 Channel Hall Effect, Open Loop Bidirectional Module from LEM USA Inc.. Pricing and Availability on millions of electronic components from Digi-Key Electronics.

www.digikey.com

Thanks rjenkinsgb I will try the LPF trick.
By 'load' I meant the saturation of the alternator field not the output current itself. so the percentage field value is perfect.

Great HE CT though thanks I might incorporate one in the future.

 

[zylinx_za]

So after more testing I'm not getting comprehensive results from this DFM / FR.
Ignoring the sporadic pulses, just getting an average RMS voltage. It seems possible its related to how hard the alternator is working.
I think my main issue is not being able to load the alternator consistently, its very difficult to get my voltage to drop when my engine RPM is at 1K or more.
and at 850RPM my alternator is not able to maintain voltage at all. so I have this very small window to play in.

850 RPM: DFM stays low 0V RMS [battery voltage drops into the low 13's]

950RPM: DFM will sit between 8V - 11V RMS [battery voltage steady 14.7]

1050RPM+ DFM will max out at 12V RMS [battery voltage steady 14.7]

not really sure what to think at this moment, don't feel like I will be able to get much useful info from this output.
It almost seems that the RMS voltage is more related to my engine RPM than any load I could apply (which also makes sense, more RPM less field required).
And maybe this is because I cannot load the alternator properly. I tried really hard to get my voltage to drop below 14.6V at 1000RPM+ to see if the DFM/FR voltage would drop below 11V. but no luck, the alternator can put out a ton of current at 1K RPM(engine).

Maybe in the future I will go to a friend with bigger loads, If I can just get the DFM to vary its voltage when a large load is applied and I am able to keep a consistent engine RPM of say 1100RPM I will be happy and feel like the relationship is tangible.

At this point it just feels like above 1K RPM nothing happens on DFM, and below 1K RPM its all wonky because the alternator can barely maintain the vsp.

 

[Nigel Goodwin]

Put a load on it then - it's not difficult - headlight bulbs!.

 

[zylinx_za]

Nigel Goodwin I did try my best loading it with over 1000W load but no luck, I'm not seeing a clear correlation.