Distributorless 5K Engine

[Banjo]

Yeh, Taz did something similar, but instead of using COPs, he mounted quad coils on the side of the engine.

Quote

With the custom cas I built, it has a 'dual pulse' on cylinder 1 and singles on the other 3 so the ecu will now know which cylinder is actually firing. Doing this you can run quad coils like this as well as sequential injection.... firing the injectors individually as the fuel is required essentially. 

Real nice neat setup !   I believe the Nissan CAS is under that spun aluminium lid on the dizzy. 

Cheers Banjo

[Banjo]

Well here is us guys putting together our Megasquirts, olde Haltechs, & the Jaycar programable ECUs; & now Speeduino, on little rigs in our garages & cars.

So how about this guy, having a play also, doing his PhD on tuning engines in real-time, on the fly.

Just look at his setup !

I want his DC engine dyno.

https://hackaday.com/2015/01/28/raspberry-pi-learns-how-to-control-a-combustion-engine/

P.S.  I can't even work out what kind of 4 cylinder engine it is, as there is so many wires & plumbing over & around it !   Actually, he describes it as a 2 litre 4 cylinder GM EcoTec engine, when I listened again carefully.

Cheers Banjo 

Edited July 31, 2018 by Banjo

[Banjo]

Having lots of fun with this system, & just implementing the crankshaft trigger sensor at the moment, which I will document in this thread shortly.

Have been reading about "GMR" crankshaft position sensors, developed by Mitsubishi, which are said to be ever better than a Hall Effect sensor. 

Quote

The sensitivity and the MR ratio of the GMR element are respectively more than 10 times higher than the sensitivity of a Hall element and the MR ratio of a ferromagnetic MR element. 

MR, the last one listed, in the table below, is I believe, what we more commonly refer to as a VR [Variable Reluctance] sensor)

http://www.mitsubishielectric.com/en/about/rd/advance/pdf/vol121/vol121_tr3.pdf

 

Wouldn't mind getting hold of one of these GMR sensors, & seeing if it can't be used in this project for the flywheel tooth sensor.  Does anyone one happen to know, what current Mitsubishi model engines, use GMR sensors ?

Cheers Banjo

Edited August 8, 2018 by Banjo

[Banjo]

For the flywheel tooth sensor experiment, I found some Honda Civic Hall Effect sensors cheap on ebay.  A crankshaft & a camshaft sensor (2 sensors) for $ 20.00.

https://www.ebay.com.au/itm/2PCS-Car-Auto-Camshaft-Crankshaft-Position-Sensor-For-Honda-Civic-2001-2005-L4/253646611587?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

Managed to find the connections for the crankshaft sensor on the internet last night. Hoping that the camshaft sensor connections are identical.

https://troubleshootmyvehicle.com/honda/1.7L/how-to-test-the-crank-sensor-1

I hooked it up last night on the bench, and waved it by hand past the flywheel teeth, with the sensor switching an LED, and surprise, surprise, it appeared to work perfectly.  I was a bit concerned the teeth on the flywheel, may be too close together, for the sensor to effectively switch off, between teeth.  The purpose made toothed wheels on cars, seem to have the teeth widely spaced.

 

So will fit the flywheel back on, the test engine, make a stable bracket to position the sensor, run the engine, hook up a CRO to the Honda Crankshaft Position Sensor, & see how clean the resulting pulse train looks.

The other sensor at the top of the pic, is the crankshaft/flywheel trigger Hall Effect Sensor.  It triggers off two rare earth magnets, I embedded in the flywheel, exactly 180 deg apart.

P.S.  Does anyone know where I can get the harness connectors for these sensors, here in Oz ?  They sell them in the USA, as a repair item, complete with wires, but neither Amazon or the USA spares companies, will ship or post them to Australia.  Might have to head to the wreckers, with a pair of wire cutters in my back pocket !

https://www.amazon.com/Position-Sensor-Harness-Repair-2001-2005/dp/B078JZZJWF/ref=sr_1_1?s=automotive&ie=UTF8&qid=1533860684&sr=1-1&keywords=Honda+Civic+crankshaft+sensor+connector+harness

Cheers Banjo 

 

Edited August 10, 2018 by Banjo

[Banjo]

I had a good weekend experimenting with the Honda crankshaft sensor, on the 5K engine, with the engine rotating.  I was delighted to find it worked perfectly, & that my concerns that the teeth on the flywheel ring gear were possibly too close, to effectively prevent the Hall Effect device/sensor from turning off, were not realised.

This is the area, where I mounted crankshaft trigger sensor.  It is accessible from the underside of the engine.

'

The Hall Effect sensor fitted through the engine backing plate, with air gap circled. On the RHS of pic is the Hall Effect sensor & its cable.  On the LHS of pic is the remains of a previous experiment, where the radial inserted rare earth magnets, let go after a few months operation. I'm still find little bits of smashed rare earth magnet, attached to various items at the back of the engine !

Before fitting the two (2) rare earth magnets to the flywheel, I used the pilot hole to drill the hole in the engine backing plate so the centre of the magnet & the centre of the Hall Effect sensor lined up perfectly.  The flywheel bolts onto the crankshaft in six (6) different positions, (60 deg apart) so you will always be able to find a point suitable for the trigger point position range required for any particular ECU requirements.

''

The trigger sensor accessablity from under the rear of the engine.

The CRO traces of the ring gear teeth sensor, & flywheel trigger sensor, indicated they were both working well.

All these tests were done with the spark plugs removed, & the starter motor used to crank the motor.

12 Volt flywheel ring gear teeth clean square wave pulses.

Crankshaft trigger pulses.

Next test is the fit the Honda sensor permanently in a cut-out of the bell housing, a little bit higher than its temporary position, so it is easily accessible from within the engine bay, behind the oil filter area.

Honda Crankshaft sensor in temporary position for tests. (It's almost like it was physically designed for this K Series engine application !

My design criteria is, that all sensors must be accessible, & easily replaced from the outside of the engine, without any major “removals”.

I’ll then refit the bellhousing & rear engine mount, & run the engine at operating revs up to say 3K-4K rpm, and check that the pulsed outputs from both sensors, still remain stable & clean.

Only the non critical synch pulse to do now, somewhere off the camshaft train, & I’m then ready to do some tests on an external ecu & ignition system connected to these sensors, but not actually running the engine.

Lots of fun, & a good learning experience.  The timing & the absolute accuracy & reliability of these sensor signals are critical to whether the ecu works well on not.  Doesn’t matter how good the ecu is, if the signals are poor, the results will be less than ideal.

What do they say . . . “rubbish in, then rubbish out”

Cheers Banjo

Edited August 14, 2018 by Banjo

[Banjo]

 

Hi Dave,

Quote

Yah but you can use the 60-2 on the crank. Then just reduce the VR trigger in the electronic distributor to 1 pulse per 720deg. This does not need to be 100% accurate. So gear slop isnt that big of a concern. And use this as your reset. This will get you full sequential:)

The problem with using a missing tooth system is, that my teeth are the flywheel ring gear, & I don't think the starter motor pinion, would take too kindly, to all of a sudden encountering 2 off teeth missing.

However, the Speeduino system does cater for what I am envisaging, with a trigger system call Dual Wheel, as described below, in the extract from their manual. This allows the Speeduino to simply continuously count the full complement of teeth of the flywheel, & reset it with an external signal from the camshaft line somewhere.

Quote

38 CHAPTER 3. SUPPORTED DECODERS
3.3 Dual Wheel
3.3.1 Overview
A dual wheel trigger is one where there is a primary multi-tooth wheel combined with a secondary single pulse to
provide location information. The primary input should contain no missing teeth. Both pulses can run at either cam
or crank speed, but sequential operations requires that the secondary pulse is located on the cam. The design of the
secondary trigger can vary (Eg a single short tooth, half-moon wheel etc), provided it only provides a single pulse per
revolution.
Tooth #1 is defined to be the first tooth on the primary wheel AFTER the pulse on the secondary wheel.
3.3.2 Applications
Dual wheel triggers are standard fitment on a number of Euro make cars, particularly those from VW and Audi. They
are also a popular aftermarket fitment due to their simplicity and ease of fitment.

Quote

Fields:
Primary base teeth: This is the number of teeth on the primary input wheel. If the primary wheel is located
on the cam (or is otherwise running at cam speed), divide it’s teeth by two for this setting
Trigger Angle: This is the angle in crank degrees AFTER TDC (ATDC) of the first tooth on the primary
input, following the single pulse on the secondary input.
Trigger edge: Whether the trigger will be taken from the leading (rising) or trailing (falling) edge of the primary
input
Secondary trigger edge: As above, but for the secondary input

Currently, just building a "convertible open top K40 gutted gearbox", so I can run the engine on the test rig, at 3K - 4K rpm, and have a look at the pulses on the CRO again, & be able to strobe the degree wheel on the flywheel.

An hours work with a hand held grinder, with a metal cutting disc & grinder !

Cheers Banjo

 

Edited August 16, 2018 by Banjo

[Banjo]

I got around to fitting the gutted K40 gearbox, & attaching the flywheel ring gear Honda crank Hall Effect sensor in its final position.

Was able to run the engine up to 3K - 4K rpm, & view the output of the Honda crank sensor switching perfectly off the flywheel teeth.

Currently got the front of the engine off, fitting a new Rollamaster dual timing chain & new sprockets & chain tensioner. The camshaft sprocket will be fitted with a rare earth magnet, to switch a "sync" Hall sensor, that will mount off the back of the timing chain backing plate, near the side/back of the mechanical fuel pump, which will eventually disappear, when the 5K goes full EFI.

Also received my Speeduino kit this week, which is already assembled & communicating with the TunerStudio software.

i did find a great place to mount the strobe light permanently on the back hump of the gearbox frame.  Lines up perfectly with the timing arrow.

 

Soon the fun will begin !

Cheers Banjo

  

 

Edited September 6, 2018 by Banjo

[Banjo]

Quote

Currently got the front of the engine off, fitting a new Rollamaster dual timing chain & new sprockets & chain tensioner. The camshaft sprocket will be fitted with a rare earth magnet, to switch a "sync" Hall sensor, that will mount off the back of the timing chain backing plate, near the side/back of the mechanical fuel pump, which will eventually disappear, when the 5K goes full EFI.

Well that good idea, did not work well, as it was too tight in the area I proposed.  Finished up making a little additional mod to mount a magnet, that required no drilling or changing the engine, except a hole in the timing chain cover.  It worked perfectly, & I now have a reliable & steady sync pulse (1 pulse every 720 degrees of engine rotation).  The little magnet mounting plate can be fitted in any one of 10 positions on the sprocket, so that a perfect spot can be assigned.

Just got to build a little fixed logic decoder, to get the timing pulses I need for the ECU, so the ECU only has to worry about all the engine sensor inputs & calculations.

P.S.   The hole in the centre is for a crankcase ventilation inlet, so that filtered air takes a much longer scavenging path, through the engine.

You might have noticed that the Hall Effect sensor I used, has a little red LED built into the back of it, which makes it very quick & easy in real life, to ensure that the sensors are producing pulses. Should be good for trouble shooting.

Cheers Banjo

Edited September 11, 2018 by Banjo

[Banjo]

Well it's the weekend, & I've some time to work on programming the crankshaft positional sensors, I've installed, to create a sequential ignition system, which is totally independent from the ecu, which does all the engine performance "timing calculations".  So far, the results are exceeding my expectations, & electronically, it is not that difficult.

I'm seriously thinking of implementing another form of "synchronisation sensor", rather than the Hall Effect sensor, I've detailed above, that is fitted to the camshaft timing chain sprocket.

Don't get me wrong, the timing chain sprocket sensor works very well, but is time consuming & a bit difficult to implement.  Because the synch signal has to produce a pulse once every two revolutions of the crankshaft, it must be produced from the camshaft timing. The easiest way is to simply use the existing dizzy points, or with an optical/Hall Effect sensor fitted therein.   However, this project is about getting rid of the dizzy totally, so that's not my option. The other point on the camshaft, where a signal could be obtained is the fuel pump's eccentric "operating lobe", on the camshaft.

This could be a very easy modification, if it works. Remove fuel pump.  Make little plate to cover hole in the block. Drill hole in plate. Fit Hall Effect sensor that will detect the tip of the eccentric lobe.  Hey presto !  Now I haven't got a fuel pump, so my idea is to just fit an electric one, so I can see on the road, in real-time, how reliable a signal would be from this source.

What type & brand, of electric fuel pumps have Rollaclub members fitted, that have proved reliable ?

Of those of you who have had fitted an electric pump;  did you fit it in the engine bay, or down in the boot somewhere, so the whole fuel line from rear end to carby is under pressure, rather than suction ?

Very interested to hear anyone's thoughts on electric fuel pumps fitted to N.A. Rollas.

Cheers Banjo 

 

  

Edited September 29, 2018 by Banjo

[altezzaclub]

We use these on the rally car & on Steve's road Corona with DCOEs.  Problem free, try to soft-mount them, fitted the rally car in the boot by the tank so it pushes, but Gramps' is in the engine bay so it sucks.  Doesn't seem to make a difference...  although I like them to push so there is no chance of air being sucked in.

https://www.ebay.com.au/itm/new-Fuel-pump-low-pressure-E85-Petrol-Diesel-Ethanol-Biodiesel-Facet-type/322545930611?epid=853553484&hash=item4b193c5173:g:iuAAAOSw03lY6Jh5

 

[irokin]

Noisy bastard things though if you're sitting in traffic not flowing any fuel.

[coln72]

Used "Facet" type fuel pumps on both my rollas with no issues. 

Had one hard mounted under the parcel shelf, but got moved very quickly.

 

Soft mounted under boot floor in the end.

[coln72]

2 minutes ago, irokin said:

Noisy bastard things though if you're sitting in traffic not flowing any fuel.

Thats what the radio is for ;)

[Banjo]

Tah All !  It's only for an experiment, so $ 16.00 incl. free postage, won't break the bank.  Not too worried about the noise, but have noted the suggestion of an "isolated" mounting.

How do they work ?  Are they some sort of rotary type ?   Looks like an electrical transformer !

My only experiences with electric fuel pumps, are with those early SU tick, tick, tick, ones, with a diaphram, that slowed down to a very slow tick, when you were idling.

Cheers Banjo

Edited September 29, 2018 by Banjo

[irokin]

Its just a solenoid with a couple of valves.