sounds like your ignition was mucked up pretty badly before the conversion

what typenumber renix did you use, have you compared that to the chart posted elsewhere? (from V3C)

antiekeradio wrote:sounds like your ignition was mucked up pretty badly before the conversion

what typenumber renix did you use, have you compared that to the chart posted elsewhere? (from V3C)

I have not yet compared number, that was original renix that was there from factory, I guess that either coil or vacuum part of that has busted, spark was weak and spark strength was dropping sometimes 50% sometimes 30%, something was wrong with it.

But I think most annoying problem with it was that when cruising slowly it caused car to jump like kangaroo, maybe too much advance or something.

I did spent over 200 euros to garages that tried to adjust carb because that jumping, no one did not even think about ignition problem. I think that my carb is just fine, maybe testing carb + megasquirt setup next autumn when car needs MOT again

Good stuff, sounds as though its all going well. Have you got a knock sensor to check for detonation? Might be important as I had detonation on mine (due to excessive compression/bad fuel) but if you're playing round with a timing map would be a good idea.

What version of MS are you using? Is it Msns-E? I would be interested in your setup and how it decodes the Renix wheel. If you had weak spark in places with your old coil that would explain a lot of your previous driveability problems. Mine kangaroos a little bit on very light throttle around the 1400-1700 area which I reckon is Renix timing floating about (I'm running fuel only atm), but I still need to map the fuel a bit more.

You don't need to strip loads of stuff off to see the timing. Paint a mark on the front crank pulley to mark TDC and then if you get a good timing light, can shine it down the front of the engine and provided there isn't too much excess light around you'll be able to see the timing marks quite easily.

Hmm, I had difficulties to see timing marks on engine side as that air duct and one alternator belt hides it under them, marking on pulley was easier. But I think I let one buddy of mine to worry about that part with his timing light

I'm running Extra code version 29j currently and Megatune 2.25 Release version.

From here:
http://megasquirt.sourceforge.net/extra/setup-msns.html
I used Bosch module setup

And from here:
http://megasquirt.sourceforge.net/extra ... wheel.html
I used LM1815 based VR sensor conditioner setup ( scroll down, lot of)

Here is my setup file for megatune:
http://www.janiervast.com/projekti/tied ... 191409.msq

It is far from perfect as I decided to make new fuel map too so currently there is just generated fuel map.

I don't have TPS sensor connected currently as it stopped working long time ago, well this weekend I found out that TPS V+ wire was not delivering electricity at all, proble is somewhere middle of wire.
So I'm using +5V from TPS now, with new wire.

Wheel need to be set as 22-2 wheel.
Wheel decoder base teeth = 22
2nd trigger = off
Trig pos A = 3
Trig Return Pos A = 11

Spark Settings:
Trigger Angle = 65
Cranking timing = Trigger Return
Hold Ignition = 0
Spark output inverted = No
EXPERIMENTAL Oddfire = No
Fixed Angle = -10
Trim Angle = 0

Wheel decoder type need to be 24s9 style, 25 version seem not to work so well.

Here is my ignition table:

http://www.janiervast.com/projekti/tied ... 191952.vex

I don't know if I already told, but I did buy cheapest generic coil for electronic ignition, seem to work well

Here is those files above and some other various megasquirt stuff:
http://www.janiervast.com/projekti/tied ... artat_yms/

Some are rather old, perhaps outdated, but I'm lazy and have not updated those

I tested today also autotune feature of latest megatune, that seem to be very nice, you set it to autotune and drive a car around, megatune adjust fuel map on fly based to lambda readings and after you feel that have driven enough just hit alt-b to save map. What is best is that map is adjusted to megasquirts memory automatically and no extra steps needed.

Now if I just would have money for 2 tanks of gas, I would just cruise around tomorrow, however I don't have money for gas at all

I just did read from adding 390 ohm resistor across VR input, must try that tomorrow to see if I can get revs past 5000rpm.

It's FRANKENCIRCUIT!

Good luck Jani. I know MS can be a ballache if you're having to be a pioneer.

cheers

James

foggyjames wrote:It's FRANKENCIRCUIT!

Good luck Jani. I know MS can be a ballache if you're having to be a pioneer.

cheers

James

Thx, luck is needed

Hopefully after next weekend I will have little more refined setup and Frankencircuit would be history

390 ohm resistor across the VR input has helped a little, now I think that I need bigger resistor.

Car sounds quite bit different now, as before it did not pull too well from 1000rpm and sound were like it would been firing when exhaust valve is already opening (tsup, tsup, tsup...) now it is pulling so well from 3000rpm that I can use 3rd gear right from 1000rpm, could be of course because my fuel map is bit rich, but still sound is much more like what engine should be making in that situation.

Yesterday one van was following near my bumper so when I turned to slippery curvy road I step on it, was crazy as I started 2nd gear around 2000rpm and it was very quickly banging in exhaust as spark cut kicked in, rpms bit over 5000 going sideways, fantastic. Still rpm reading starts to spike around 5000rpm so that is why I get spark cut there, limiter set to 6200rpm and if it spikes above that spark cut will occur.
Oh yeah, there might have small flames too

When I get this rpm spiking sorted, my next thing will be to install switch for clutch, so I can get launch control and flat shift logic working, but let's see, maybe soon time to get older car and start doing turbo conversion.
I can't get all these trough MOT as car is too new, with stock parts I could get out of it with installing cat but with megasquirt maybe bit difficult. Of course I could source all stock parts and install MS hidden and using original wire harness, but cost and time would be bit too much.

Next car will be Volvo, maybe 300 or maybe 200-serie, injection model for easy MS conversion.

Hmm, hmmm, from logs it looks like that my ignition advance is opposite as what it should be, so it gives 56 degrees advance when throttle is floored, this may not be very good. But no detonation or other symptoms, maybe it is log that is wrong or maybe not, must test bit more.

So 26 degrees more advance at WOT than what should be, maybe Volvo made just little tolerance for their ignition map

I made funny pics

http://www.janiervast.com/gallery/v/alb ... _problems/

When rpm spikes up advance increases, no, no, no, it should decrease. Also rpm spike is bad thing, but luckily those does not occur in low rpm anymore, resistor helped there.

hi there!

the info your supplying here is great!! thanks for sharing!

I'll keep you posted on my ignition map hunt! I wont be getting much done till the summer though. However, Ive already got the maps for the B14.0 and the B14.1S. The only issue is the pressure it refers to, I guess this is the pressure at the end of the vaccum advance pipe on the distributor unit?

Shimon

I always like sharing more than taking, maybe not best filosophy for life but I'm not scientist so what can you except

Now I managed to put ignition map right way, feels almost same, but maybe bit more logical now.

Fueling need to be adjusted again as ignition map is different so fuel burns differently and lambda shows different values.

From 1000-2000rpm and map below 60kpa I can't run 14.7:1 mixture, I have tried now many ignition timings but it just does not work, richer mixture works better, I get very bad shaking and misfiring if I try to lean mixture to 14.7:1 level, so maybe it is ok to run bit rich there.
Also idle need to be bit rich and timing advance quite high.

In extra 29j code there is idle advance option so you can set idle advance to certain locked value, helps bit with idle rpm hunting. Also idling is not very smooth, but I think that with K-cam with stock head+valves and 2l engine there is no way to get smooth idle...
Also some ppl say that Volvo engine will never run very smoothly on idle.

Grip has been biggest issue now, there is some hard ice on road and it is very polished at some places so need to be very careful when accelerating, one intersection specially is wild as when I get 3rd gear in and I have around 60kph there is very slippery area, it is impossible to accelerate with 3rd gear there as wheels just starts spinning, same with 4th actually and 60kph @ 4th is very low revs.
Maybe there is just more low rpm torque

Actually I don't need to use 1st gear anymore, there is enough torque to use 2nd, also there is no load of power coming in at 3000rpm, it feels quite smooth from 1000rpm to 5000rpm and this is with lambda correction active so no cheating with rich mixture, very interesting how much stronger and better timed spark affects to engine.

One thing that I still don't understand is why there was no detonation when I had 55-56 degrees advance at WOT, this I need to investigate bit more, it can't be that ignition would be 20 retarded as then it would be over 5 degrees ADTC at idle and engine would not run then, I guess.

But if there is no detonation, would it be possible to use some insane advance at high throttle? As it felt at least bit more powerful on high rpm when I had that insane advance.

Could be indeed nice to have knock detection, but I have no proper parts for it, also it is not very easy to implement as you need to finetune circuit specially to redblock engine type and that would be lot of researching, finding correct frequenzies and finding correct resitsors and capasitors for that frequency range. Would be interesting to build that but currently it feels bit too complicated for me.

jtbo wrote:Also idling is not very smooth, but I think that with K-cam with stock head+valves and 2l engine there is no way to get smooth idle...
Also some ppl say that Volvo engine will never run very smoothly on idle.

Keep tuning Jani, mine purrs like a kitten when the carbs are set up right. The K cam is only 270 degree...not really enough to produce a lumpy idle.

cheers

James

foggyjames wrote:

jtbo wrote:Also idling is not very smooth, but I think that with K-cam with stock head+valves and 2l engine there is no way to get smooth idle...
Also some ppl say that Volvo engine will never run very smoothly on idle.

Keep tuning Jani, mine purrs like a kitten when the carbs are set up right. The K cam is only 270 degree...not really enough to produce a lumpy idle.

cheers

James

How rich is you idle set to? Do you have some emission data or lambda data available?

14.7:1 is very bad idle and from there I have gone bit of richer side, but I don't like to put tons of fuel going to engine just to get it idle properly...

I have so far kept my idle speed at 900rpm range, maybe rising it a bit would help, but I really don't like to get it very high, I like from this low sound that it makes around 900rpm

Sadly not so...I've been very lazy about getting my wideband bung welded in. The DP comes off this weekend though

cheers

James

foggyjames wrote:Sadly not so...I've been very lazy about getting my wideband bung welded in. The DP comes off this weekend though

cheers

James

Ah, I had that too, then I did pay to garage to do it and they did it all wrong, ah well.

Here is nice illustration what happens when my car thinks it is a kangaroo:


So, usually it goes like this, I do little engine braking and then come put just a bit of throttle so car would cruise, but as my lambda settings are not yet refined there is lambda correction that has leaned mixture so when I put more throttle I'm going to lean, not much, just a little bit, but enough, next you see there rpm sigging up and down, same does spark angle and gego is getting mixture back from lean, but that is very annoying 2 seconds what it usually takes, sometimes less, sometimes more.

Sometimes it just keep doing kangaroo dance even mixture is already back on rich also in this example so I usually cure that by hitting loud pedal, that is what you see there next, high mountain

It is so much on edge at 14.7:1 mixture that it will make this kangaroo dance sometimes even it would not bee really a lean and also does it sometimes when coasting, so that is why I'm thinking to make mixture tad rich and setting lambda correction to be active only from 2000rpm or so where this problem will not happen anymore.

Lot of learning and replacing old learnt information is still needed here

I already made correction to 1200rpm bin advance, now it is closer to 20 degrees as it was before 10, that helps a bit of this problem too.

For those not familiar with megasquirt/injection terms:

MAP = Manifold air pressure, ambient pressure is 100kpa, in manifold this is low when no throttle and high when you push more throttle.
Duty Cycle 1 = % of injector utilization, this should be below 85%
Spark angle = well, spark angle, that is what timing light would show, is degrees BTDC
Gego = mixture correction based Exhaust gas oxygen sensor information.
MAT = Manifold air temperature, often called IAT = Intake air temperature, mine is bit jumpy as I should replace standard bias resistor (2,4k ohm) with 33k ohm resistor. Problem from wrong resistor is that it is not too precise, it can jump several degrees as resolution is so bad, it has something to do with bits and all, but I just ignore all that
CLT = coolant temperature, I should replace this one's bias resistor too and make better temp map, it shows bit too low.
Oh yes, forgot, O2 = Lambda sensor data, I have narrow band lambda so that is jumping a lot, it is good to see if running lean (line near bottom of graph) or rich (line near top of graph), 1.0v is maximum rich and 0.0v is maximum lean, around 0,5v is only usable real value as it is stoich, Air/Fuel ratio of 14.7:1, in Megatune that is program to adjust Megasquirt settings there is AFR gauge that takes little time average and does same as any Air/fuel meter would do, shows value, very handy when tuning or autotuning, gives lot of information why car is acting bad/good.

Then there is many more that are not important (Read: I don't know what they are)

Maybe this is interest for someone? Well at least wrote something again

Hmm, I'm getting old, wrong link:


Well, both shows kangaroo, but this second one is from what I'm talking about