astra 1.4 8valve lpg turbo.
Turbo Mini Forums

Im rambling...

But seing as your not using the peak and hold circuitry, you can probably omit it alltogether.

see note 3 on the diagram on the first post here. in your system current limit is provided by the dropping resistors and Q3 and 11 would provie the necesary flyback dampning.

http://www.msextra.com/forums/viewtopic.php?f=131&t=51751

Edited by evolotion on 31st May, 2019.

turbo 16v k-series 11.9@118.9 :)

Denis O'Brien.

Excellent info thank you denis , you make more sense drunk than i do sober !

i will try ot crear the cave of stuff in the way ,and hook the car up to the sun tester mra scopeness ,and see what i can take a picture of .

re the whole flyback pwm situation ,if you look back through the thread you can see that my perspective was always to try the resistors , and ,if troublesome ,then go to the pwm , in the end the resistors worked so i left it , however its a good idea to have it all on there in case i want to spend time trying , at the moment my only goal is to have a running daily car for the least time and money , so i can concentrate on the medusa project .

I did when first ding this injection ask , why am i not just using the pwm , i never got a understandable answer , so just took the safest appearing route .


i will also see today if i can get to auto tune the higher rev ranges rich enough.

i am imagining not ,at first i was thinking it was cutting out due to the boost making it impossible for the injectors to open on such a low voltage ,however , if i back off to no boost it still does not really fire , it may be the opening and closing durations have become so long that it cannot work well , however i would have thought that would be most effective during low rpm usage , and everything up to 4800 is perfectly normal in function .

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

here you go denis ..



from what i read the circuit is designed to hold the max flyback volts to 36 v ? i am using a 50v scale so that works out perfectly from the measurements.

by revving the car i got a piccy of a second weird trace..

close up ...








What the hell is that then ? i wondered if it was the second squirt being attempted by the ecu , but not going properly to ground ?





i took some voltage readings across the resistors..


resistor..............rpm.........volts ........... calced amps ...

1.....................1000...........0.18...........0.18
1....................1500............0.24............0.24
2....................1000............0.36...........0.18

i bypassed a 1 ohm resistor to give a total of 2.35 ohms pre the injectors instead of the total of 3.35 ohms i used above ...


this gave ..


1.................1000................0.20............0.2.




not sure if that makes any sense but anyway its recorded for posterity :)




regards
robert

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

quick update , driving the car again i think now it is a boost limit ,this fits with my measured data i took last october when i fitted the keihins , i soldered 4 1 ohm resistors in a series line , then put air pressure on the injector and moved down the ling with the 12v feed seeing how much pressure the injector could open with ,with varying resistance , looking at the data ..



.......Psi----------0----------30-------40------55

OHMS.4.............Y..............N...........N........N
OHMS.3.............Y..............Y...........Y.........N
OHMS.2.............Y .............Y...........Y.........Y
OHMS.1.............Y..............Y...........Y.........Y
OHMS.0............ Y..............Y...........Y.........Y

Amps were
4=2.24A
3=2.7A
2 =3.43 A
1=4.68 A
0=7.4A to 6.8A dropping .

this is with my 12v bench supply.

Temp rise i did for a 20s hold on each ,
4= no real change
3=3.5c
2=7c
1=20c


you can see that ,with 12 psi boost ,and 15 psi gas pressure , so a total of 27 psi pressure in the gas rail , anything in the 3 to 4 ohms resistance range is getting borderline ,so it make perfect sense that with 3.35 ohms its boost limited ..i have some .5 ohm resistors coming so i can try to drop ohms to 2.85 , and see how that does , fun fun fun !


regards
robert

Edited by robert on 3rd Jun, 2019.

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

I dont really understand what your showing on that trace, is that 2 channels? or persistance of vision and just one channel?

the flyback potential is set by the 36V Zener diode (D6 and D20) so what you know is your already above that value, and you could potentiallty have an issue where the flyback exceeds what the circuit can handle. however, on the scope it looks to be ding its job as it should.

is there any way you can go out on the timebase and see 2-3 injection events at once?

and now we know we are only working with 36v you could hook up a conventional scope if you have one.

turbo 16v k-series 11.9@118.9 :)

Denis O'Brien.

its one channel Denis , but the scope is triggering from the signal voltage cresting a certain value , and i wonder if that 2nd weird shape is the second squirt ?



i wont be able to do more sun tester scopage for a while i have to concentrate on medusa , i did have another thought ..

new thought ...


17.3 amps !

that's what i get if i have 1.25 ohms injector resistance in parallel on a channel ... so 0.69 ohms , then when cranking bypass my resistor pack and feed 12v straight to the injectors ,which is what i have been doing ,now , call me mister fluffy ,but if i am banging some sort of percentage of that through the channel ,when the inj open , is that not going be a chunky amount of flyback when they shut ....

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

new cranking test ameliorated my concerns ..


Cranking with no bypass ...0.24 amps dc...0.40 amps ac,

Cranking with bypass .......0.43 amps dc ...0.88 amps ac .

Cranking voltage was 10.9v


So whereas with the bypass ,and the inj constantly open i should have per channel ,. 1.25ohms/2= 0.625 ohms ,so with 10.9v ..17.44 amps ...in reality the short opening and v low frequency means it never gets to anything like that .i have it set to alternating so it won’t ever have all 4 open at the same time .


Next i did static revving amps test ...

Tickover...0.4A dc....0.78A ac .

6000 rpm ....2.2A dc ...1.8A ac


Then drove it , at max pre cut-out boost so about 10 psi ,and 4k rpm , i had 4.5 A dc and the ac went to 1.6A then lower as the revs rose .

All this was with the total resistor package , so 3.35 ohms on each channel .



noice.

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

quick update to say the car is now running very nicely and smoothly , also got 34 mpg the other day which is unheard of ..my new boost limit has its benefits lol.

interestingly , i can run it lean ,and so it makes less boost creep , and then i can rev it on without a cut .. now i have medusa running on the pwm i may pop that on the astra when time permits.


vunderbar .

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

ok well it appears either my engine is revving to 28k , or this ecu does not like pwm !





i tried all sorts of ms from 1.2 to 3.0 .and % from 30 to 40 ..no luck ,it fine on cruise but boost,not so much .

Edited by robert on 10th Aug, 2019.

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

cant see pic, but you may want to try (even temporarily) mounting a reverse biased (flyback) diode directly at each injector, and if the injectors are fed from 1 +12v feed, then install a capacitor on that to ground like you would find on the center pin of a ford EDIS coil. these two tricks should go a long way to cutting down interference.

turbo 16v k-series 11.9@118.9 :)

Denis O'Brien.

try this denis ..

Attachments:

240.59kb

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

what voltage amperage diode denis ? also i have a little grey box shaped thing with ford on it , looks like a cap from an edis is that the sortb of thnig you mean for the injectors 12v feed ?


like so ..

http://www.msextra.com/forums/viewtopic.php?f=71&t=60984

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

Ill pop this in here just in case it helps anyone Bearing in mind my microscopic grasp of circuits .

i just had my ms2 v3.0 board drop a injector driver circuit , and upon opening it i found this ..





and a strong smell .

so i replaced the chip , and the TIP42c chip in the same area , but i was still not happy with the comparison of readings between this channel and channel 2 , so after a quick call with a friend we worked out that the zener "D6" was bad , and replaced that too with a slightly different one ,

43v ,and 1.25 watt ...

all i had in the shed .....

it should be a "Diode Zener 36V 1W 5% DO-41"


now all runs well again .







here is the circuit ..







i can see how 42c would blow with the zener making only 70 ohms resistance but not the 127 Q9 chip . i have marked the chip that blew and the zener also .


i am using 4 cylinder 1.4 litre turbo ,with 13 psi boost ,pwm , lpg kei-hin 1.25 ohm injectors , and it has run perfectly for the last 4 years i think it is .


if there is a new problem occurring i imagine i will lose the other channel soon . maybe its just a bit stretched in its standard components for what i'm doing .


regards
robert

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

The diode limits the flyback voltage when the injector closes. If it no longer works, you will see a very high voltage peak which the TIP42C and TIP125 won't like.

Also, using a different Zener voltage value for the 2 channels will give you different closing times which will also give you different dead times. So you will have an imbalanced fuelling between the channels.

http://www.jbperf.com/

ah good info thank you jean ,

how different do you think ?

when i can i will solder in another higher voltage zener in the other channel , or perhaps just get the right one for this channel ,

i needed the car for tomorrows long journey to a hospital ,so did not have a lot of choice .

Edited by robert on 10th Oct, 2021.

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

Without testing it, it is impossible for me to say how much difference it will make. And it is possible that mismatch between injectors is actually more than the drivers mismatch but, again, that would need to be tested.

Just try to get the same values on both channels as soon as possible. And going to the higher value might not be a bad way to go, especially if you have pulse widths that are close to the dead time value.

http://www.jbperf.com/

thank you jean , i think my lowest pw is around 4 ms . i have ordered some of these.


https://www.ebay.co.uk/itm/121919177687


iin 36v so i have a choice .

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

humph . well , now thw other channel has popped its chip , but it was the tip42 this time , and the diode was fine . i swapped the 42 and the 127 to new , and also i have put on the 47uf caps across the centre pins of the flyback transistors middle pins ...just in case it smooth things off as i have read in the ms2 forum.now driving ok again , i also turned down the pwm settings from 50 % and 3 ms , to 40% and 2.5 ms.in case that helps.

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM