BMW M5 Forum and M6 Forums banner

41 - 57 of 57 Posts

·
Registered
Joined
·
58 Posts
I think a lot of people are interested in the results of your work. If you can find a solution as elegantly as you did with the additional control devices on the B7/1, this will be a great victory for all of us. Good luck, Oliver. I will come to your city on September 3 in the evening. Cool me a glass of beer!

930038
 

·
Registered
Joined
·
9 Posts
@Redox
For shure cold beer is not a problem...

M5_Board_DG_V1_1.jpg

Splash screen of the DG reapir kit 2012... This kit I fitted in April 2016 at the ALPINA headquarter in Buchloe.


ALPINA _Digital_Gauges_NG_Series1.jpg

Next generation of display electronics, will replace the old design and housing...

Back to the sensor, I can confirm the original M30 flywheel deal, now shipment is on the way to me.

Regards,
Oliver
 

·
Registered
Joined
·
286 Posts
Discussion Starter #43
Count me among those intensely interested in your work on this Hartig sensor. Looking forward to further updates.

As previously discussed, you'll have a ready market with the Porsche 931/924Turbo world, too.
 

·
Registered
Joined
·
9 Posts
As previously discussed, you'll have a ready market with the Porsche 931/924Turbo world, too.
It is the same idea and construction (but NLA), a double coil inside, one about 500mR, the bigger one about 9R.
RAFI has an automotive group for sensors, the 5WKxxx on the sensor looks like Siemens numbers. The Porsche ignition control unit was also made by Siemens. The sensor was very critical about the air gap, if I see the resistance it is very clear, therefore I have to check the conditions on the flywheel. The "missing" or different cog is only electrically different mechanically it is a teeth for the starter, but it is the start point the unit is counting. Because of the two coils inside I will monitor - record the waveforms. May be a small re-engineering of the control input section will clearify the rest.
Porsche and ALPINA crossed the ways not only at this time, some years later (1989) on the B10 BITURBO it was the same way. The Porsche knocking control unit was the boost pressure control unit on the B7/5 engine.
 

·
Registered
Joined
·
9 Posts
I made some more investigations today, the best working sensor was the Delo sensor from GM type Opel ASTRA G, the corresponding part is a BOSCH inductive rotantional speed sensor ...138.
This sensor has a winding resistance about 860R, this is too much, because it will not perfect work with the ignition control box (this is a question of the air gap, the resulting voltage of the sensor and the circuit which generates the information for the next step in the ignition box. (I mean how the circuits were designed to work inside the specs.) The other problem is, there is only one coil inside. Parallel I studied the Opel schematics of this ECU, for shure the same, even the BOSCH datasheet will clarify the same problem.

The function of this more modern technology sensors is explained here:
***
Design and function The soft-iron core of the sensor is surrounded
by a winding, and located directly opposite a rotating toothed pulse ring with
only a narrow air gap separating the two.
The soft-iron core is connected to a permanent magnet, the magnetic field of which
extends into the ferromagnetic pulse ring and is influenced by it. A tooth located
directly opposite the sensor concentrates the magnetic field and amplifies the
magnetic flux in the coil, whereas the magnetic flux is attenuated by a tooth
space. These two conditions constantly follow on from one another due to the
pulse ring rotating with the wheel. Changes in magnetic flux are generated at the transitions
between the tooth space and tooth (leading tooth edge) and at the transitions
between tooth and tooth space (trailingtooth edge). In line with Faraday's Law,
these changes in magnetic flux induce an AC voltage in the coil, the frequency
of which is suitable for determining the rotational speed.

The sensor generates one output pulse per tooth. The pulse amplitude is a function of
the air gap, together with the toothed ring's rotational speed, the shape of its teeth, and
the materials used in its manufacture. Not only the output-signal amplitude increases
with speed, but also its frequency. This means that a minimum rotational speed is
required for reliable evaluation of even the smallest voltages.
A reference mark on the pulse ring in the form of a large "tooth space" makes it possible
not only to perform rotational-speed measurement, but also to determine the
pulse ring's position. Since the toothed pulse ring is an important component of the
rotational-speed measuring system, exacting technical demands are made upon it to
ensure that reliable, precise information is obtained.
***

This older system is working with the same idea of the old BOSCH M1.3 ECUs.
The crankshaft was controlled with two sensor one for speed, the second for the reference mark.
If you check the schematics on the HARTIG box there are three lines, one common, than two coil signals. So the reference is missing in your condiseration.

Opel_Astras_G_Crankshaft_b34.jpg

BOSCH_inductive_rotational_sensor.jpg


After my flywheel simulator is finished I can monitor both signals and we can see more.
 

·
Registered
Joined
·
286 Posts
Discussion Starter #46
So redox was right that he was missing his reference signal.

Thank you for the updates. It is so exciting to hear someone with real understanding tackling this problem, and explaining it! I knew it was solvable, but my engineering skill set doesn't extend to this.
 

·
Registered
Joined
·
58 Posts
A reference mark on the pulse ring in the form of a large "tooth space" makes it possible
not only to perform rotational-speed measurement, but also to determine the
pulse ring's position. Since the toothed pulse ring is an important component of the
rotational-speed measuring system, exacting technical demands are made upon it to
ensure that reliable, precise information is obtained.
It should be noted that the teeth on the flywheel are used by the starter. So there can be no space. I spoke with a Porsche specialist. On a 924 turbo, a magnet was inserted into one of the teeth in a flywheel.

Thus, on the sinusoid of the sensor signal, one wave was different. Apparently, this was the beginning of the countdown. However, I examined the Alpina flywheel very carefully and did not find no one magnetized tooth. Maybe he was demagnetized? And that was the beginning of my problems?

On the other hand, this only matters for the Hartig ignition system. If we have not three ignition coils, but one, and an ignition distributor, the order of work of the cylinders is determined by the distributor in any case. The ignition coil continuously delivers a spark to the distributor, which carries it in the right order through the cylinders.

Why should the system know the position of the first cylinder? The distributor knows this. And the task of the control unit is only to synchronize the supply of sparks with the speed of rotation of the crankshaft.

Everything that I wrote here is probably wrong. But try to refute! Agree, it looks logical. ;)
 

·
Registered
Joined
·
9 Posts
Why should the system know the position of the first cylinder? The distributor knows this. And the task of the control unit is only to synchronize the supply of sparks with the speed of rotation of the crankshaft.
The two signals before TDC (reference mark on the flywheel and the reference mark from the 1th cyl.) have a difference about 25°KW it is for plausibility check, that the reference mark is valid (yes or no) The reworked teeth has a different material and was like evaporated on the flywheel. The plausibility check is done if you compare the amount of counted teeth between the two TDC signals, they should always be similar.

As I wrote mechanically you cant see it but physically it is different.

I will provide a flowchart with some more explanations. This system was also used in the AUDI quattro, if you read the technical training from 82 there are some resistor values explained.
 

·
Registered
Joined
·
58 Posts
The two signals before TDC (reference mark on the flywheel and the reference mark from the 1th cyl.) have a difference about 25°KW it is for plausibility check, that the reference mark is valid (yes or no) The reworked teeth has a different material and was like evaporated on the flywheel. The plausibility check is done if you compare the amount of counted teeth between the two TDC signals, they should always be similar.
Maybe this is true for an atmospheric engine? We are dealing with a turbo engine. I assumed that the ignition timing is engaged in vacuum sensor in the control unit.

I could not find any differences in the teeth of the flywheel. No one.
 

·
Registered
Joined
·
286 Posts
Discussion Starter #52
Well, it doesn't seem to differ visually, although I don't have a standard flywheel to compare. I can work on that.
From the press photo it would appear the ring gear is unique. https://www.m5board.com/attachments/hartig-press-photo-jpg.74479/
My assumption was always that the difference in ring gears was embedded magnets, as the teeth themselves couldn't vary or you would have starter/wear issues.
If it would help the cause I would consider shipping my spare flywheel over for testing/investigation
 

·
Registered
Joined
·
9 Posts
Maybe this is true for an atmospheric engine? We are dealing with a turbo engine. I assumed that the ignition timing is engaged in vacuum sensor in the control unit.
I could not find any differences in the teeth of the flywheel. No one.
The engines "ignition map data" are in conjuction with the engine speed, air-, water temperature and the pressure- underpressure here you´re right. But the reference mark has nothing to do with a charged or naturally aspirated engine . The ignition timing itself works only if the engines reference mark to TDC is correct. The reference mark was a non magnetic tooth in the flywheel.

The scenario I described in post 45 explains a modern flywheel and how it is working. It was for a better understanding of the physical conditions inside the sensor. At least it is the same way, only with a different tooth.
The whole flywheel material is magnetic, if you put the sensor on the tooth you will have the effect I described in the upper scenario, but one tooth is different. The magic trick of this sensor is the second coil inside and as I said the air gap was very critical (I am not wondered if I see the resistance of the coils) therefore adapter rings were made for each gear box.
My information comes from my longtime friend who was working at ALPINAs headquarter more than 40 years. He was the former testbench leader and made most of the engine maps, even for the racing cars.

Anyway I also organized the original ALPINA flywheel ring today and one more working Hartig control unit.
So we will see, but believe me the Tektronix Scope will show you what I try to explain.
 

·
Registered
Joined
·
58 Posts
Oliver, I can say only one thing: no one has yet delved into this story as deeply as you. And by the way, about the clearance. Indeed, I noticed that for the Dello sensor I was trying to apply, the clearance was critical.
 

·
Registered
Joined
·
9 Posts
We made the first schematics (with some ideas) Redox will arrive very soon. A lighter automatic flywheel is at ALPINAs headquarter for a rework. I had contact with the German service guy who is still servicing the boxes and he gave me specs about some signal translation. From the German Porsche forum I got the construction specs about the original sensor. Now some more brainstorming is on the way, but I know how the sensor was constructed and how he was talking to the Hartig box.
 

·
Registered
Joined
·
286 Posts
Discussion Starter #56
Again, so exciting to hear intelligent discussion on this issue.
Also again, I will offer that I have an early spare flywheel here.
Also note in pictures in post #29 that my once car had a replacement sensor installed (not Rafi).
If either of these are any help to your process, I could send them over.
 

·
Registered
Joined
·
9 Posts
Again, so exciting to hear intelligent discussion on this issue. Also again, I will offer that I have an early spare flywheel here. Also note in pictures in post #29 that my once car had a replacement sensor installed (not Rafi). If either of these are any help to your process, I could send them over.
Hi Mike323i,
Thanks for your reply and the offer about your flywheel. Today morning I got an information from ALPINAs headquarter that the rework of the flywheel is in progress. So I think it will arrive very soon, but if we could need another one I would contact you.
Redox visit me at my home on Thuesday and I explained my ideas and the final solution. (He also got the beer I promised) :p
After we decide us for the right upcoming sensor (its a mechanical question paired with the best signal recognition) we will start writing the software for the translator. The analog part (preliminary CAD design and schematics) is near finished. I think we will include a short calibration test, were the controller is testing the signal blanking under more critical conditions. (I mean less signal quality) After finding the best distance, the test will give a yes/no response with a green LED. After this test the sensor distance is configured and the rest of the signal process is done inside the CPU. It takes now a little bit time to build the right simulator, for the flywheel and adapt our development tool.
 
41 - 57 of 57 Posts
Top