Etanol.nu

Hello,
BOSCH Mono Jetronic A2.2 is an alpha-n system, that means that the map(s) burned in ECU get engine conditions by rpm sensor (ignition coil) and load (position of throttle) (but also temperature...). Other systems are speed density which measures MAP (manifold absolute pressure) or MAF (mass air flow).

I borne an idea yesterday.
Similar to the guys who increase fuel pressure in that way that it is possible to go by E85 AND E05.

What's about to turn the TPS potentiometer(s) slightly backwards. The engine would think you step down gas pedal deeper and would enrich mixture.

I know the manual forbid to change anything and the TPS is not avalable as singular spare part.


Another more flexible way to achieve enrichment would be to modify the harness to TPS.
- If WOT (wide open throttle) tells the ECU a low voltage another serial resistor would lower voltage across the TPS range.

- If a high voltage is seen at WOT a seperated IC could provide the measurement voltage.
OEM voltage is 5V, the common cheap voltage regulator IC is LM7805 (there exists a better IC LM2937 which needs less voltage difference to supply (cranking <7V> don't forget to modify both.

What do you think?

»Horst, just some Kilometres away from 1st E85 fill up

_________________
» Citroën BX14 E.CO.2 Break: TU3M/KDY (iron cast) C/R=10.3:1, AX-Diesel gear box (+22%)
» Citroën Ami8i Break: V06/630 + MS EFI + cat

I don't think that is a good idea, and this for several reasons.

First, the air mass flow is not a linear function of the throttle position. So, if the output voltage from the TPS would be linearly tweaked (i.e. by a circuit with adjustable gain and bias), then a good conversion would only be reached at two specific throttle positions, not across the whole range. I have tried this, not on a Mono-jetronic system but on the signal from the air flow meter on a Motronic 1.3 (a Citroën BX with DFW / XU9J4 engine). I never got it to work well. Apart from the difficulty in obtaining a correct fuel mixture, the idle stabilization got upset and oscillated wildly under some circumstances. So i removed my amplifier and increased the fuel pressure instead.

Second, the TPS is also used to detect the full travel of the idle acutator motor. If now the TPS signal is tweaked in any way, there would be a risk that the ECU tries to run the actuator beyond limits, which leads to a risk of damaging the plastic gears in the actuator.

Third, the Mono-jetronic systems usually have a very wide lambda control window, to get full "flexifuel" performance of the system, normally only a very slight increase of the fuel pressure is necessary. This can easily be achieved by fitting a shim or washer (a plain washer for a M10 or M12 bolt, approximately 1,8 mm thickness, works well) on top of the fuel pressure regulator spring. Just make sure that the washer lies flat against the spring holder and grind or file of any grits at the edges of the washer. Also, make sure that the washer does not block the air hole in the spring holder.
Experience from VW 1,8 l engines with Mono-jetronic shows that about 50 % of them will run well on pure E85 without conversion, the rest needs the increased fuel pressure or any other suitable modification.

Mono-motronic has a double TPS potentiometer, one section givs full scale output for the whole throttle travel of 90 ° while the other gives full scale for the first 10-20 ° or so. This is used to get better resolution at small throttle openings, i.e. idle and light part load. The ECU's A/D converter has just 8 bits of resolution! Also, the software evaluates both signals across each other and will give a fault code if they are not in the correct relationship. Both signals are approximately 0 V at fully closed throttle and 5 V at fully open throttle.

_________________
Att mäta är att veta

Hi Horst,

In addition to Torbjörns exellent answer my experience with two BX14 is that it almost runs directly out of the box on E85 but at WOT and at cold start it gets too little also the stock (centrifugal) fuel Pump was not capable of creating a much higher pressure the stock 0,9 bar. I think I got 1-1,1 bar at most.

If you don not want to rely to the lambda adatption so it runs well in open loop you would theoretically need 1,8 bar.

So if you want/need to raise the pressure beyond that you need to ether replace the Fuel pump (this is what I did because the pump was bad anyway) or add an extra external (3-5 bar) fuel pump in series with the pump built into the tank.

The BX 16/19 pump can easily be mounted in the special bracket next to the fuel filter that already is there.

Many other cars (Opel/VW etc) use the same pump(shape) that is in the tank and these can create 3-4 bar easily.

greetings Aryan

_________________
Komplettera gärna data om dit piggybacksystem här: http://www.editgrid.com/user/aryan/E85_piggyback_system och här viewtopic.php?f=10&t=6352

Hello torbjorn and Ayran,
thank you for very detailed replies. really good content!


@non linear relationship TPS vs MAF:
Thanks, I have known that TPS is not linear to MAF, this should be just an approximation 8as increased fuel pressure is). As you both told O2 sensor is 'strong' enough to get suitable mixture (KDZ/KDY topic) in closed loop.
I just want to improve cold starts (autumn, winter).


@Fast idle stepper motor:
I hadn't thought on any troubles like them you tell.
If ECU is confused about 'huge' TPS then I expect that ECU would turn the actuator 'back' to closed throttle. No damage. I haven't dismantled such a unit, isn't it a 'simple' gear box like in electric window opener - driven by a stepper motor? Ok, when threath is screws out completely there could be a 'problem'.


@Washer in fuel pressure regulator
Thanks for detailed description!


@Fuel pressure:
This is my experience with the Suzuki Swift intank pump which was rather weak too.
In my converted BX14 (which was originally a BX16) is still the OEM BX16 (Magneti Garelli G6.10 ECU) inline fuel pump with very similar technical datas.


@Fuel pump characteristics:
recently I got a data 'sheet'/chart of a Monojetronic A2.2 fuel pump:

This looks wrong for me, haven't seen any pump chart like this (looks more like piston compressor), volume AND pressure increases!


»Horst


PS:
Sidenote, JFYI: Got also a simple sheet of MA3.0 Monomotronic Motronic of 1 litre engine (AX10/CDY...) which says that there is no vacuum regulated fuel pressure inside!

_________________
» Citroën BX14 E.CO.2 Break: TU3M/KDY (iron cast) C/R=10.3:1, AX-Diesel gear box (+22%)
» Citroën Ami8i Break: V06/630 + MS EFI + cat

Then I think it is much better to alter the temperature sensor resistance so the ECU thinks the engine is colder than it really is.



Yes Japanese cars have weak pumps that do not allow a raised fuel pressure, but Bosch systems do exept for this BX 14 in tank pump. I do not know what pressure the B16 normally works but I's suspect it is more than 1 bar? With the external pump next to the filter used in a BX19 that normally has 3,3 bar the pump should be more than sufficient for a raised pressure conversion. In My BX the O rings of the injector where a bit dry/old and of cause raised pressure could make them leak so it might be a good idea to change them.



Now I understand it.

It is not the pumps chart but the fuel pressure regulator chart! A pump that delivers more will apparrently increase the fuel pressure. The fuel pressure regulator is not as constant as we would think, that is good to know and something that I have suspected before.

»Horst



Yes this is natural, the injector is above the throttle so it doesn't "see" vacuum and the fp does not need to be compensated for vacuum either.

/greetings Aryan

_________________
Komplettera gärna data om dit piggybacksystem här: http://www.editgrid.com/user/aryan/E85_piggyback_system och här viewtopic.php?f=10&t=6352

In fact, the idle actuator on Mono-jetronic and Mono-motronic consists of an ordinary DC motor, not a stepper motor. It operates the pushrod for the throttle stop via a spiral gear and a nut-screw mechanism.
On Mono-jetronic, there is no position feedback within the actuator, just the TPS and a switch which detects when the throttle arm rests on the actuator's pushrod are used to give feedback.
Mono-motronic has additionally a hall sensor in the actuator.
A common problem is that the switch in the pushrod (or the sliding contacts at the pushrod) wears out so that the ECU never gets information that the throttle is released. I have heard of cases (but not seen with my own eyes) where this has caused the actuator motor to run with full speed towards the end position, damaging the plastic gears when it comes to a sudden stop.

_________________
Att mäta är att veta

Hello Aryan,

@faked temperature:
Yes this would be possible. It's easy to do with MegaSquirt because I know the enrichment - temperature table.
I'm afraid that the engine is sensitive in fuel overload.
A reverse engineering trick would be to make a bench with potentiometer as tempature gauge and a NE555 IC to trigger the ECU in crank mode to scope the injector pulse witdh.


@BX14/16/19 fuel pumps:
BX16 0.7 to 0.9 bar....~1300ccm/min
BX14 1.1bar.....~1200ccm/min
BX19 SPI ~1bar
BX19 MPI ~3bar


@pressure regulator chart:
Yeah - very good read!


@fuel - vacuum - pressure:
The Suzuki Swift TBI considers vacuum too (as Mono Jetronic) , but I don't know why.




Hello torbjorn,
thanks for clarification - a DC motor - how boring.
I'll check the unit, becasue idle speed is too high. The actuator works (tested). Maybe the increased compression is the reason (better efficiency).

»Horst

_________________
» Citroën BX14 E.CO.2 Break: TU3M/KDY (iron cast) C/R=10.3:1, AX-Diesel gear box (+22%)
» Citroën Ami8i Break: V06/630 + MS EFI + cat

This faking of the temp sensor ith an extra NTC in series works apperently very well, especially if you put the NTC close to a cooling fluid hose so the car get almost a normal signal when hot, of cause you can add a switch to turn it of when driving on petrol. With E85 there is no a problem of soaking the engine because that is hardly possible. E85 has a much wider A/F mixture that is burning than petrol.



A good list, I meant the BX 19 MPI which is the only BX 19 I've studied. I have seen 0,9 bar fp in the system in documentation for the BX 14 but 1,1 bar max pressure from the pump.



Very important to know though, I use a non stock China pump, so that might give me too much flow -> too high pressure at idle too.. Have to measure the pumps performance / the pressure and otherwise slow it down a little..



I think it considers vacuum to measure engine load and not to alter the fuel pressure? Vac + RPM & TPS + RPM makes that they can calculate the amount of air that goes into the engine twice and can do without an (expensive) MAF.



Increased compression does not change idle speed on the BX14, the ECU should regulate it anyway, it does on mine with 12,5:1!

/Aryan

EDIT FP BX14 & a spelling mistake of RPM

_________________
Komplettera gärna data om dit piggybacksystem här: http://www.editgrid.com/user/aryan/E85_piggyback_system och här viewtopic.php?f=10&t=6352

Check that the switch at the end of the pushrod is working and that it does not have a high voltage drop at "on" position.
Remember that this switch does not only control an input on the ECU, it also controls a solenoid valve for the vacuum ignition advance. So the current through it is probably 0,1 - 0,2 A.
That input of the Mono-jetronic ECU is not very well designed, it might detect a voltage as low as 0,5 - 0,7 V as "high". So, in case of problems with the switch or sliding contacts, it might happen that the ECU never gets the signal to begin controlling the idle RPM.

_________________
Att mäta är att veta

Hello Aryan,

@faked temperature:
typical automotive NTCs are not linear
T= 1 / [ a+ b (ln R) + c (ln R)^3 ]
http://www.ametherm.com/applications/nt ... t-hart.htm
Check the chart/table at http://www.bosch-motorsport.com/deutsch ... TC_M12.pdf
-20°C....15.5kOhm
0°C.......5.9kOhm
20°C.....2.5kOhm
40°C.....1.2kOhm
100°C....0.2kOhm
Enrichment is more or less (inverted) linear to temperature.

My idea would be just to add 'a' resistor (potentiometers) in series with the OEM NTC to increase resitance. There are some nice turnable switches with a number of positions.
http://www.neuhold-elektronik.at/catsho ... _ebene.jpg
Each position refers to a temperature range, e.g. -20°C....-10°, -10°...0°, 0°C...5°, 5°....15°
Each range could be 'felt' without themometer.

Thank you for the hint that E85 is easier to fire! 2 or 3 ranges would be enough.



@fuel - vacuum - pressure:
Definitely no load indicator for ECU.
If ECU needs load information:
- MAP sensor for manifold vacuum = 'speed density' (BX16 and SPI BX19)
- throttle position/angle = 'alpha-n' (BX14)
- MAF (MPI BX19)

Maybe the fuel pressure is decreased by manifold vacuum in that way that at idle less pressure difference between fuel injector and ambient pressure causes less flow --> longer pulse witdh.



Hello torbjorn,
@idle actuator:
Thank you for pointing me to the switch. I haven't noticed it before (never looked for).
Checked wiring diagramme: As you said vacuum ignition valve and this switch are connected - but in a strange way:
» +12V feed one contact of the vacuum ignition valve
» Ground is lead to the actuator switch.
» Pin3 of ECU is connected to the other contact of the vacuum switch AND the other contact of the actuator switch.

I estimate that ECU pin3 provides Ground if engine managment thinks OR the actuator switch is on.


»Horst

_________________
» Citroën BX14 E.CO.2 Break: TU3M/KDY (iron cast) C/R=10.3:1, AX-Diesel gear box (+22%)
» Citroën Ami8i Break: V06/630 + MS EFI + cat

Pin 3 of the ECU is ONLY an input. The solenoid connected between +12V and that pin also works as a pull-up resistor for the input.
One should also notice that there is a diode in the ECU, connected with the cathode to +12 V and the anode to pin 3, this is used for clipping the voltage spike that the solenoid's inductance would else cause at switch-off.

Full details for the temperature sensor used in, among others, most Bosch engine management systems, can be found here (the subcontractor delivering the NTC discs to sensor manufacturers): http://www.epcos.com/inf/50/db/ntc_06/LeadlessDisks__B57220__K220.pdf
In the Mono-jetronic ECU, a resistor of 36,5 kohm is connected in parallel to the sensor, and the pull-up resistor to 5 V is 2,87 kohm.

The method of connecting a resistor in series with the temperature sensor during cold starting and then shorting it out is commonly practiced here in Sweden, we call it "cold start resistor". But it is a somewhat crude method, in particular for vehicles that are to be driven by not so technically interested drivers.

_________________
Att mäta är att veta

Hi Horst,



Yupp I know, but with E85 you do not want linear enrichement, below 10°C you want it to rather to increase logarithmic. So I think a suitable extra NTC in series is not too stupid.

At -20°C you need a lot of fuel to be able to start, I think you are basicly starting on the petrol part of E85 than, I have no absolute proof for this theory but exhaust gas analizes have found more than 90% pure unburnt ethanol just after coldstart @-7°C so I think that says enough.



Yes of course that is a good way to get a feel for what is neccesary.



Easier to fire a overrich mixture: yes, easier o fire a cold engine: no




Maybe misunderstood you, but did you mean that the BX14 or other monopoint injection systems with its injector above the throttle butterfly has a vacuum compensated fuel pressure? I'm quite sure that a BX14 has not vac compensated fuel pressure and I would be supprised if any other car with a simular mono point injection had, even if you have a interesting idea; lower fuel pressure at idle/high vac could help getting the shortest pulses longer and thus making it easier to control the smallest injection amounts. In some modern cars the fuel pressure / pump is regulated electronically to give the injector a greater dynamic range.

greetings Aryan

_________________
Komplettera gärna data om dit piggybacksystem här: http://www.editgrid.com/user/aryan/E85_piggyback_system och här viewtopic.php?f=10&t=6352

Yes, indeed, monopoint injection systems do not use to have any vacuum correction for the fuel pressure regulator. As the regulator is placed in the upper part of the throttle body, its ambient pressure will be the same as the ambient pressure at the injector.
On systems where the pressure regulator is fitted on the outside of the throttle body (for instance, Chrysler TBI systems found on V8 engines in Dodge vans from the late 80's), there is a vacuum hose from the regulator to the volume between the air filter and throttle body, without that hose the regulator would not be able to take the pressure drop of the air filter into account.

_________________
Att mäta är att veta

Hello,
I 'cross post' (quote) my little brand new experience with Ethanol + A2.2 Mono-Jetronic partially ( http://etanol.nu/forumrecover/viewtopic.php?p=12739 )


@too high idle speed:
As told I have problems with too high idle (1400-1600 at warm engine, no cam mods).

I put in just a screw driver like a wedge between throttle plate lever and the pushrod which drives idle speed (air). ECU drives the idle speed motor in that way that the idle falls down again to ~1500rpm.
When I remove the idle motor plug (4 pins) and remove the screw driver idle fall down to ~700rpm. Sound more relaxed!
The throttle is forwared up to the TPS-switcg. If I connect the idle motor plug again idle raises up to well known 1500rpm.
--> BOSCH ECU feedback works but control the wrong idle speed.
The crazy idle is also produced if I change the ECU (KDZ older type).
--> I guess have to check the harness and the TPS.

»Horst

_________________
» Citroën BX14 E.CO.2 Break: TU3M/KDY (iron cast) C/R=10.3:1, AX-Diesel gear box (+22%)
» Citroën Ami8i Break: V06/630 + MS EFI + cat

Hi Horst,

Is it possible that the ECU "thinks" that the engine is colder than it is and therefor keeps idle much higher?

Have you tried reading out the error codes?

/Aryan

_________________
Komplettera gärna data om dit piggybacksystem här: http://www.editgrid.com/user/aryan/E85_piggyback_system och här viewtopic.php?f=10&t=6352