4K EFI Test Rig - IAC Valve

[Banjo]

Hi Kayzz,

               You'll be pleased to know, I've been able to turn a K series "Bosch" distributor housing, into a Camshaft driven trigger wheel, & CAS signal.

The technique that Altezzaclub suggested of retrieving the "tooth wheels" from the 4AGE dissy, & "turning them down to fit the narrow K series dissy; was a good one, but not having a lathe, I settled to fit an aluminium disc to the large K series Bosch dissy, which was more doable.

It was not without some setbacks along the way.  The disc could not be fitted in the large top section of the dissy, as the sensors would have to be fitted upside down, in the base of the narrow section, where adjustment of air gaps etc, would be difficult. 

I therefore turned the disc down, & was only able to fit 18 off magnets around the disc. The rare earth magnets are beautiful, but working out the right distance between them, & the best air gap; so they will turn off, can be frustrating. I eventually reverted to "assymetrical" rare earth magnets, that have their north & south poles on the same face of the magnet, rather than at each end.  I've had it running on the bench, driven by a variable speed electric motor, & the results are excellent. The oscilloscope traces of the two streams of pulses, show crisp clean & sharp pulses.

Have no doubt it will work well, on the 5K engine, when I get around to hopefully fitting it, on the upcoming weekend.

Here are a few pictures I took along the way, as a picture tells a thousand words; so they say.

 

The only proviso I have at the moment, is whether there will be any "jitter" in the RPM speed update to the ECU, as the holes in the aluminium disc for pressing in the magnets, was not carried out on a CNC machine; but by my steady hand.

Also there could be some "timing scatter" from the CAS signal, as it mounted on the cam drive, which could have a couple of degrees of slop, in worn chain, sprockets & gears etc., to the crankshaft.

But that's another subject, which I believe I have a solution for. Let's get it running on an engine, & then the strobe light steadiness; will depict, how stable the RPM update, & CAS signals are.

Cheers Banjo

 

  

Edited February 12 by Banjo

[Ke555555]

On 2/1/2025 at 5:00 PM, altezzaclub said:

I was thinking the Hall Effect sensor on the pin at 2o'clock, like this

Chop the plate down to just an arm holding the post and mount the sensor through the wall near it. You may not even need the post, just the end of the arm swinging past might be enough to activate the Hall sensor. Banjo will know!

That gives you a cam position signal to tell the ECU that #1 is the next cylinder to fire.

The crank position signal is the hard one, the more teeth you have to read the more resolution the computer has. So the 4AGE uses 24, but others use 36 and some use less. Finding what is essentially a gear wheel to mount above the cam sensor will be the hard part, either find a steel one from a distributor, or make one out of an alloy disc and put magnets in it. You can see the factory one in that image of the distributor I posted above, lots of gear teeth close together.

That disc gets mounted on the shaft further up, and that Hall sensor goes in the side or through the lid.

Manufacturers used this idea for a while in the 80s and 90s, a cam & a crank sensor built into a distributor and a rotor and plug leads on top. Those are the dizzys you want to look inside at a wreckers, the later ones did away with a distributor altogether and ran Coil Over Plug systems with remote sensors.

Here's a Camry one from then, I've never looked at one, but it has the 24tooth wheel in there, the part 19235 picture.

https://www.amayama.com/en/catalogs/toyota/camry/3-sedan-right-v30-1990-2407/engine-fuel-system-and-tools-1/distributor-34

4AFE motors, the most common engine back then, have them too. Wouldn't it be nice if this shaft and gears fitted easily.

$40 buys a 2JZ one, somewhere in the world.

https://wardautoracing.com/products/2jz-ge-distributor

Otherwise, its alloy disc and magnets time.

What if you come straight down the top 

[Ke555555]

Does the gap size matter? 

[Banjo]

Couple of questions !  What is the model number of the Hall Sensor You are using ?   Is it a model that detects north or south poles of magnets; or is it one that detects ferrous metal targets ?  Does the Hall Effect sensor have an LED light, built into the rear, where the cable enters the Hall Sensor.

The setup You have there, maybe permanently switched on, depending on how sensitive the Hall Sensor is, & whether the "steel base", from which the pin protrudes is too close to the sensor, whilst it rotates.

Nomally, You connect the DC supply voltage across the twi wires, that supply the Hall Sensor, with a DC voltage between 5-12V DC.  Then rotate the dissy, & see whether the LED turns on & off, when the vertical pin passes the sensing face of the Hall Sensor.  If You have model number for the Hall sensor, then list it here, & I can look it up, & see what type it is.

Be careful when connection it up; as some Hall Effect sensor colour codes are a bit unusual. I've got some here, where the "brown" wire is +12Vdc, "black", is the output wire; & blue is the -ve wire.

Cheers Banjo 

[Ke555555]

Does the gap size matter? 

1 hour ago, Banjo said:

Couple of questions !  What is the model number of the Hall Sensor You are using ?   Is it a model that detects north or south poles of magnets; or is it one that detects ferrous metal targets ?  Does the Hall Effect sensor have an LED light, built into the rear, where the cable enters the Hall Sensor.

The setup You have there, maybe permanently switched on, depending on how sensitive the Hall Sensor is, & whether the "steel base", from which the pin protrudes is too close to the sensor, whilst it rotates.

Nomally, You connect the DC supply voltage across the twi wires, that supply the Hall Sensor, with a DC voltage between 5-12V DC.  Then rotate the dissy, & see whether the LED turns on & off, when the vertical pin passes the sensing face of the Hall Sensor.  If You have model number for the Hall sensor, then list it here, & I can look it up, & see what type it is.

Be careful when connection it up; as some Hall Effect sensor colour codes are a bit unusual. I've got some here, where the "brown" wire is +12Vdc, "black", is the output wire; & blue is the -ve wire.

Cheers Banjo 

It’s a Hall sensor that detects metal part number HAL-502

[Banjo]

The Cherry ZF sensor You have purchased, is one of the best available, & usually very expensive here in Australia.

It does not however, come with a built-in LED to indicate it is working & producing an output, when You place the sensing face close to a ferrous metal object like your vertical pin. They are generally utilised to sense ferrous teeth on a trigger disk, as depicted on the spec sheet.  As your ferrous pin, has a curved surface facing the sensing surface of the Hall Sensor, it may well need to be a bit closer to the pin, than the 1.5mm, the spec sheet indicates for the gap.

The only way, is to test it. As the Cherry unit does not have a built-in LED, You are going to have to hook up an external one, across the output of the Cheery Hall Sensor, as per my sketch below.  You will need a small 6 or 12 volt battery to power it.  The frequecy response of the Cherry Hall Sensor of 15kHz, makes it idea for this application.  Most commonly, these sensors are used to pick up ferrous 30-60 teeth on a trigger wheel, attached to the crankshaft pulley. However, in an application like distributor body, where the distributor rotates at half the speed of the crank; & only has one tooth per rotation to sense, it is more than suitable for your application.

So get yourself an automotive LED bezel lamp, suitable for 12 volts, at an auto store. These will already have a resistor, inside the lamp, in series with the LED, as the LED only needs 2-3 volts across it; (depending what colour they are), to light up.

Detach the Hall Sensor from the dissy body, & then place various sized bolt heads, washers etc., close to the sensing face, & get a feel for how close it has to be, before the lamp is turned on fully.

Then put it back into the dissy, & move it as close as necessary, to get the LED light to come on reliably & consistently.

Let us know how You go.

Cheers Banjo 

 

Edited yesterday at 01:34 AM by Banjo

[Banjo]

I've had some success, in the next stage of the development of a Trigger Wheel, & CAS signal, from within a K Series dissy.  The Bosch K Series dissy was used, because they are physically larger in diameter, than the more common K series dissy body.  The rings of rare earth magnets worked, but did present some issues.

There are a couple of incremental rotary encoders used in industry; & the education industry, that provide an identical output to the rare earth magnets & Hall Sensors, but are a "drop in" item.

Rotary encoders, can provide outputs, that can be used to indicate the rotational speed of the encoder; the direction of travel of the encoder; & the actual angular position, at any time. 

An industrial grade absolute encoder, can be quite expensived, but the small "educational" incremental ones, for teaching concepts; are less than $ 30.00 ea. As can be seen in the three pics below; they are relatively easy to mount, in the dissy. 

If You want to read more about encoders, in general, then head across to this link.   https://en.wikipedia.org/wiki/Rotary_encoder

These encoders come in a variety of models, with different numbers of pulses per single rotation.  Because we are interested in angular movement of the distributor, & everything ECU wise is related to degrees BTDC etc.; I chose, & ordered; an encoder that provided 360 pulses per rotation. The encoder has two of these outputs, slightly off-set angularly, so that it is possible to work out which direction, the encoder is rotating in; although that is of no interest to us, in this dissy application, as the dissy only rotates in a clockwise direction. (more on that second output later) 

However, most ECUs accept pulse trains from trigger wheels, that produce pulses of 24/36/60/72 pulses per single rotation, of the crankshaft.  As the dissy only rotates once per two (2) revolutions of the crankshaft, then we are only interested using 180 of those dissy encoder pulses, & we set the ECU up, as if these pulses are being derived, from the crankshaft.

I then decided to see if I could divide the 180 pulses per "effective crankshaft revolution", down to one of these above common ECU trigger wheel numbers, for a crank shaft rotation. Pulse chains in electronics, can only be easily divided by whole numbers. 180 divided by those trigger wheel numbers are (180/24 = 7.5; 180/36 = 5; 180/60 = 3; & 180/72 = 2.5;  180/4 = 45. We can only easily divide by whole numbers in electrionics, so dividing the pulse train by 3 or 5 produced the equivalent of a 60 or 36 tooth crank trigger wheel. Most ECU manufacturers, advise that crankshaft trigger wheels between 24 to 60 teeth, provide best resolution. So it didn't take long to then produce these switchable trigger outputs, to an ECU. Oscilloscope grabs of the primary (360) & divided by 3 & 5 pulses, are below.

I had left my single CAS magnet & Hall Effect sensor in the base of the dissy body, so do have a working system.  However, the rotational encoder sits up pretty high, so I cannot create my low profile dissy, with my "jam jar" lid. 

Could I possibly, use that second pulse chain of the encoder, to produce a single CAS signal ?  Unfortunately; not with this particular encoder, but could be possible with a absolute positional encoder, where each of the 360 pulses per rotation, creates a unique number. We could only look for just one number, & use that pulse as the CAS pulse. However, it may be simpler than that, as an absolute rotary encoder has a single pulse per revolution, that resets the counters. That may make a perfect CAS signal, if brought to the outside of the encoder.  All I've got to do, is find an "absolute" rotary encoder, in a small footprint, at a reasonable price.

Simon down in Tassie, is sending me up a K Series Bosch dissy, so I can then try this theory out, in practice.  It may be; that ultimately, we may be able to even accomodate this angular encoder, into one of the "narrower", olde K Series dissies.

All good fun !

Cheers Banjo 

 

 

 

 

Edited yesterday at 12:54 PM by Banjo