Welcome to Netrider ... Connecting Riders!

Interested in talking motorbikes with a terrific community of riders?
Signup (it's quick and free) to join the discussions and access the full suite of tools and information that Netrider has to offer.

Another Tiny Step in my Fuel System Education

Discussion in 'Modifications and Projects' at netrider.net.au started by PatB, Oct 25, 2010.

  1. Well, MrsB was out yesterday, so I wagged gardening duties and spent half an hour soldering together my Jaycar mixture display kit.

    Dead easy construction, with a dozen or so components, excluding the 10 LEDs for the readout. A good introduction for someone who's never built an electronics kit before and a piece of the proverbial to replicate from various circuit diagrams on the web should you have a little experience and a reasonable stash of electronic odds and sods.

    Only difficulty I encountered was that Jaycar had provided a 100 MegOhm resistor instead of a 1 MO one. However, given that it was simply to protect the chip from excessive currents from the oxygen sensor and given that there should be effectively no current from the oxygen sensor, making the exact value of the resistor unimportant as long as it's big, I bunged it in anyway, on the basis that it should be OK. Worst that was going to happen would be that I'd have to dig another resistor out of the bits box and stick it in instead.

    Oh yes, and the fact that my eyes ain't what they were and I was having to use a magnifying glass to read resistor colour codes. Not too much of a problem given that I was using a multimeter as a double check.



    Anyhow, with everything glued together, I hooked it up temporarily to the bike, with a power feed from the headlamp fuse, an earth under a sidepanel screw and the signal wire crimped to the O2 sensor I fitted in last week's thrilling installment. Ignition on, fire up and wait for the O2 sensor to get up to temperature. I'm using a cheapo single wire unit with no integral heating element so it takes a few minutes.

    Once things were nice and hot, the readings from the LEDs looked reasonably sensible. At idle, there's quite a bit of flicker up and down the display, but given a few revs I get a nice, steady green a bit to the rich side of centre. There's a variable resistor on the unit to get this central, which I'll tweak before I start looking for proper data points.

    I suspect that the flicker at idle is a result of being a four-stroke single. Fresh combustion gases are only in contact with the sensor for about half a crankshaft revolution or a quarter of the full cycle, leaving one and a half revs when there's goodness knows what in the downpipe, composed of old exhaust gas, fresh, unburned fuel and air from valve overlap and whatever might make its way up from the tailpipe in the time available. The latter two would give rise to a lean reading, and the idle flicker observed seemed to favour the lean end of the scale and seemed to coincide with firing frequency. With more cyclinders, I suspect the effect would be less pronounced. It certainly seems to go away with more revs.

    Anyhoo, that was as far as I got before having to hurriedly hide the evidence of non-domestic type activity. Next step is to put the meter in some sort of reasonably neat housing and wire a suitable connection plug into the bike's loom. Having done that, I can get mixture readings at various throttle settings and engine speeds.

    I'm looking at doing quarter, half, three-quarter and full throttle at 1000 rpm steps from 1000 to 8000 (or less if that's too scary). That'll give me 32 data points. Not exactly a comprehensive fuel map, but probably all that my simple Engineer's brain and my goldfish like attention span will be able to cope with.

    Having done so, I can start looking at the effects that various modifications make. The baseline mixture readings will then, hopefully, allow me to tweak things like jetting and needle position to return fuelling to something close to standard.

    You will note that I will not be looking for absolute readings (which you simply can't get, with any accuracy, from a $20 Ebay narrow-band oxygen sensor. The purpose of getting baseline readings is to allow me to then do a comparison after modification that will tell me if things are leaner, richer or the same as stock. Using the instrument as a comparator should be fairly accurate.

    That's the idea anyway. Once I've proved to myself that the O2 sensor method works, I'll feel in a better position to drop a few hundred bucks on a Megasquirt kit and the associated ironmongery to go to the next step.
     
     Top
  2. great to see you've given it a shot!

    i played around with narrow bands & the jaycar kits in the past, but have since graduated to a wide band for propper tuning. there really is no better tool.

    your narrow band set up will be good enough for "ball park" tuning, and that sounds like what you're after, yeah?

    also, be ready for the narrow band sensors, especially the cheapies. if you're readings become way out of left field, replace the sensor before making drastic changes to the tune of the bike.... as dead o2 sensors have cost people engines in the past.
     
     Top
  3. Not bad. :)

    One suggestion - if you want an accurate absolute mapping, why not take the bike to a dyno shop or emissions testing place and calibrate your narrowband kit using their wideband equipment's readings? (Cost, time, too-pedantic, I guess?) :)
     
     Top
  4. Thanks for the feedback gentlemen.

    I fully agree that wideband is the way to go, but the gear is more expensive and this whole project is very budget constrained. I figure that the narrowband set up will be as good as using a Colourtune, with which I've had good success in the distant past on cars, but which can't be read when you're using the road as your dyno, as I intend to do in this case.

    At this stage, I'm looking to perform various mods to the bike (open up airbox, different filter, maybe a pipe change) and use the narrowband sensor reading to return the mixture to standard. Eg. If WOT at 6000 rpm gives me LED number 5 lit with a standard airbox and LED number 7 lit with the airbox open, it tells me I need to up the main jet until I get numer 5 lit again under the same conditions. Similarly, comparisons at smaller throttle openings should give me a guide to which way to shift the needle. Depending on how sensitive the sensor and readout are, it should be possible to get within a smidge of standard mixtures this way.

    I'm not looking to do anything particularly original at this point. I'm just looking for a better (well, more objective, anyway) way to do what others do by seat-of-the-pants methods (which I know myself to be crap at) and prove to myself that I can get jetting somewhere close.

    As for an unreliable O2 sensor, I've considered the possibility (Ebay's finest, unreliable?:shock::D) and figure that, because I'll be looking for back to back comparisons, I should be OK unless the sensor fails in the short period between runs. I also won't be blindly relying on it. If I make a mod that should improve airflow and the sesor tells me things have gone rich, I'll be sceptical of the instrument rather than automatically jetting down. Besides, I haven't bought any smaller jets :LOL:.

    Cost, time and a deep rooted desire to do as much as possible for myself, basically. I want to see how far I can go with the most basic of equipment and limited of budgets. Past experience suggests a good prospect of reasonable success.
     
     Top
  5. Sounds good PatB!

    How much have you got invested at this stage?

    I'm working on a wideband sensor kit (still), so far I'm about $250 down including the sensor.
     
     Top
  6. Round figures, about $50 (sensor, exhaust boss and readout), plus the big torch (which I needed anyway) and a stick of Bunnings (not much) silver solder. Once I've got it all hooked up and housed neatly, there'll be a bit more for odds and sods like heatshrink, cable ties and a 3-pin plug/socket. Say $60 all up. Less than a Colourtune anyway, for something which should do much the same job more conveniently.

    I do want to go wideband eventually, but the need for it is a way down the track at the moment.
     
     Top
  7. The mixture "hunts" at low speed because the ECU is in closed loop mode. It constantly adjusts the mixture slightly rich and lean of stiochiometric in a feedback loop with the o2 sensor.

    Regards, Andrew.
     
     Top
  8. It's a DR650. It wouldn't know an ECU if one slapped it in the headlamp with a wet fish :wink:.
     
     Top
  9. If you want to do a gross check to see if your O2 sensor is working you can always hook up a multimeter and turn off your fuel or put the choke on to get it to go lean or rich.

    on a car I had to do it with propane gas and unplugging an engine breather. (with the system running open loop)
     
     Top
  10.  Top
  11. Link won't let me in without registering.

    I intend that the DR will become fuel injected as an educational exercise. I've got quite a lot of the necessary hardware already and just need to justify the expense of a Megasquirt kit. Then, of course, I'll need to hook up a high pressure fuel system which, on the DR, looks like being the hardest part.
     
     Top