Topic: O2 sensor tuning

[drumstyx]

A thread here recently got me thinking -- I finally got my 750K8 starting decently enough, and I was not enjoying the process of plug chops, so I took the plunge and got an air/fuel meter setup. $20 in parts for the probe, and....quite a bit for the meter setup, and here we are!

Preliminary testing in the garage shows I'm slightly rich at idle, at about 13:1, and very, very lean at medium-high rpm with throttle only about 1/8 open (in my garage, so no load)...hitting around 19:1. This makes sense, being a K8, and having an aftermarket exhaust and air filter (though the filter, as advertised, should flow the same as OEM). Pilot screws were out 2 turns when I tested, and I believe stock is 1.5, so the slightly rich idle helps it start and take throttle off idle better.

I'm starting this thread to show off my mounting setup, but also to put some numbers to the tuning, and if I feel like it, I'll post pics of the plugs along with numbers as I go.

Main jet kit should be arriving today, along with needle shims. First step is to bump the mains from 110 (stock) to 120s, and take it for a road test while recording. The analog cable to measure RPM alongside A/F ratio is coming tomorrow, as well as a couple throttle position sensors to see if maybe I can make them work and get fancy data for you all 😁

Wish me luck!

[ef9tuning]

It will be very interesting to see the data you come up with! Seat looks great on the bike, where did you get it from?

[69cb750]

Preliminary testing in the garage shows ....

Can you test on the open road ? (throttle is open more under load).
Thanks for posting, on my list for two years.

[Erny]

Interesting topic for me (as owner of K7). Just I have 4into4 (stock) pipes...

Nevertheless, interested in measurement setup, results under load and tuning.

[drumstyx]

It will be very interesting to see the data you come up with! Seat looks great on the bike, where did you get it from?

Seat is from Ebay about 4 years ago -- I actually wanted a stock style seat, but it seems stock style reproductions for the K7/K8 don't (or didn't at the time) exist. Both the original seat pan and the one from the parts bike I had were completely rusted out, so my plan was to get a stock-repro cover, buy this seat just for the seat pan, and rebuild it/get it rebuilt with new foam to be like the original. I've been too lazy to bother, and the stepped design looks "close enough" to fool casual onlookers. According to my records it cost $264USD shipped to my door in Canada, so at least it wasn't *too* terribly spendy.

Preliminary testing in the garage shows ....

Can you test on the open road ? (throttle is open more under load).
Thanks for posting, on my list for two years.

That's on the docket -- Since I was pretty sure it was going to be lean somewhere in the range, I figured I'd jump right into the carbs before I took it on the road, and I had to wait for a jet kit...good thing Sirius Consolidated (the only jet kit I could find without intensive searching) is only a few hours away, and Fedex will have it here today. I hope to have time between raking and other fall chores today. All that said, the data will be MUCH more useful when I have at least the RPM sensor hooked up, and if I'm able to jerry rig a throttle position sensor (just picked up the cheapest 2 TPS's on amazon for 5 bucks a pop, so we'll see how that goes...)

Interesting topic for me (as owner of K7). Just I have 4into4 (stock) pipes...

Nevertheless, interested in measurement setup, results under load and tuning.

I wish I had 4-into-4 pipes, not only because of the originality (damn things are nearly $2000 these days...) but because tuning individual carbs is tricky with my setup. My probe uses standard 1/4" copper line on a compression fitting, so I'll be making another probe end that's more like 24" as opposed to the current 15", to hopefully guide it into each individual header pipe. Of course, I could take off the collector or even just the muffler to make it easier, but technically that affects the flow, and I'm going for as precise a tune as possible.

[2wheels]

Interesting test setup.  I have questions.  It looks like a copper tube going up into the muffler, how far up does it go?  why the black rubber hose spooled up on the output side?
Is it a wide band sensor or narrow band?

[drumstyx]

UPDATE: Took er for a spin with the recording function going -- it's pretty useless without SOME data to link time to what's going on, RPM, throttle position, even speed would clue me into what I was doing at the time. Despite thinking through what I was doing methodically (Hold throttle at 1/8 for 5 seconds, that sort of thing) I can't really read the pretty graph. This weekend, I'll hook up the RPM sensor, and HOPEFULLY the throttle position sensor!

Interesting test setup.  I have questions.  It looks like a copper tube going up into the muffler, how far up does it go?  why the black rubber hose spooled up on the output side?
Is it a wide band sensor or narrow band?

Yep, 1/4" copper tube, goes up about 14-15", to get a nice warm air-free exhaust reading. I'm intending to build a longer probe for individual headers, but that'll take some careful bending and trial and error -- since this is only the beginning, I thought I'd just start simple.

The black hose is to prevent fresh air from feeding back into the pipe section. Any hose would do, and in fact, I intend to fancy it up a bit with a nicer copper coil, like the real-deal here: https://www.zippersperformance.com/758-777/. See, the original idea for this came from here: http://www.nightrider.com/biketech/diy_exhaust_probe.htm . I sucked butt at welding a plate/washer to exhaust pipe (seriously, it's so hilariously bad, I have a full MIG setup and no idea how to use it...thanks dad! LOL), so I went to the hardware store for some browsing, and came up with this.

The coil ends right in front of the exhaust for good reason too -- borrowing yet another idea from here: https://www.amazon.ca/Innovate-Motorsports-3728-Exhaust-Cast-Stainless/dp/B000CO9MF8 (same company that makes the LM-2 meter itself), I figured the venturi effect, having exhaust pass over the tube, would make the probe draw exhaust better from the copper line inside. I don't know if it actually does anything, and while moving, as long as the outlet is pointing anywhere roughly rearward, you'd get a venturi effect just from the wind (come to think of it, the exhaust itself would get a venturi effect from the wind at speed...interesting engineering thought...)

Sensor is a Bosch LSU 4.9 wideband, pretty much the standard for this application as I understand it.

[2wheels]

Very interesting.
You get all that in place and next thing will be a add fuel injectors.
Plus an arduino (very little cheap computer) to control it all.  If you are interested in that kind of thing.
With our long winters in South Ontario there is lots of time.

[drumstyx]

Very interesting.
You get all that in place and next thing will be a add fuel injectors.
Plus an arduino (very little cheap computer) to control it all.  If you are interested in that kind of thing.
With our long winters in South Ontario there is lots of time.

Man, if I could get a spare old set of carbs or a set of throttle bodies that fit and an ECU for a decent price I'd be so up for that. I'm already planning on speed and TPS sensors....an MAF sensor wouldn't be out of the question either lmao!!

[PeWe]

That sniffer look useful.
https://www.zippersperformance.com/758-777/
A little bit pricey for me.
I have thoughts to make one. Find a kit like this for a lower price would be perfect.
I have a Daytona Wego 3 A/F kit. It can record the measurements with rpm reference.

[Deltarider]

Our oldstyle bikes need nothing more sophisticated than a CO-meter. Not that long ago I have managed to revive an old one from the seventies and, having 4-4 exhausts, I'm really pleased with it. I'd wish I could show you a wiring diagram, so you can build one yourself, but in spite of my efforts, I haven't found one yet. Anyway, in principle it works by combining a thermal conductivity cell and a Wheatstone bridge (the main components being some big resistors and two potmeters). What I appreciate most, is that the gauge is analogue, which is a joy to read, compared to hectic digits that need interpreting every time they change, which unfortunately they do all the time.
Don't be surprised your bike will not run right (driveability!) on less than 6% CO*. Later models like the CB650 may run somewhat leaner.
* which equals an A/F ratio of 12,25

[drumstyx]

Snow today, so no test run on the road, but I have fun news -- a throttle position sensor has been rigged up! It's a sensor for a Nissan 350z, which has a little lever that I tied some safety wire to, and the other end tied to the idle stop. This technically introduces a curve from the rotational-linear-rotational motion conversions, but I honestly think it'll be close enough to be useful. The 5v supply is from a cheap-as-dirt usb car adapter, so the variation in the 5v alone is enough to not technically be perfectly accurate.

The fun part about this, though, is that it's simply held in place with some pipe strapping, which means it's pretty portable.

While I'm waiting for better weather, I'll clean up the wiring and take some pics along the way. Lame update I know, but I just had to tell someone about the TPS!

[drumstyx]

Juicy update -- I have a graph!

Legend:
Violet: AFR - 7.5-22
Black: RPM - 0-10230
Red: Throttle position - 0-100%
Green: Speed - 0-200km/h

Of course, this is just a snippet, since I'm not familiar enough with Logworks to zoom out on the time -- every time I try to, it crashes the app. This is a snippet of pulling onto the highway heading home, chosen because it shows a snapshot of 100% throttle for a couple seconds.

Setup: #35 pilot (stock), #120 mains (stock is 110), 1.5 turns out on fuel pilot screws.

Info I'm gleaning: I was overzealous in bumping the mains before even doing a road test. It's VERY rich at full and even half throttle. Of course, momentary richness is to be expected from the accelerator pump, but this is too much for sure. Idle is lean, so either needs jets bumped, or screws out more. Cruising at small throttle openings is rather lean though (second attachment), so I'm thinking that might also indicate the need for a shim or two under the needles.

Tech stuff for nerdy folks like me: Speed is implemented with an arduino and a GPS module. I thought it'd be fun at the very least, and quite possibly useful to know how much load the engine might be under, given that speed implies wind resistance. I just had 3 extra analog channels to play with, so I figured I might as well!

EDIT: Ignore the random RPM drops to 0, I didn't stall, just that the signal is apparently a bit too strong for the logger -- an attenuation circuit is recommended in the manual in this case, but I haven't gotten around to building it.

[69cb750]

Try #115 mains, 1 turn out air screw, aim for 13.5 idle, cruise, 11.5 wot

[drumstyx]

Try #115 mains, 1 turn out air screw, aim for 13.5 idle, cruise, 11.5 wot

'78, PD carbs, so it's a fuel pilot screw . 2 turns out seemed to make idle around 13.2 when I had everything stock (35 pilots, 110 mains), so I'll adjust them out a bit when I swap in the 115s and add the needle shims.

[drumstyx]

Another decent day, another couple data points.

Put the carbs back to stock with with 110 mains, but 2 turns out on the pilot fuel screw. This fixed the idle AFR right up to 13.2-13.8....pretty much right where I want it.

But god damn, it's still RICH AS HECK at anything over ~15% throttle. I'm not just talking about a snap, though that floods and causes a gurgle/stumble, but even just holding at 1/4 throttle (or even 16-17% throttle, as shown in the graph) AFR sits around 9-9.5. One graph here shows holding throttle at 100% during a run up the freeway, and AFR is around 10, so if I understand correctly, even 110 mains are too rich! Sounds crazy...am I going crazy? PD carbs don't have the o-ring I've heard about under the main jet holder, so that's no issue. Is it possible my "like oem" filter is just more restrictive? Will 105's be small enough? Should I maybe look to a different air filter?

[69cb750]

Sounds crazy...am I going crazy?

Ideas - run test without filter, check results, check needle position, raised needles add fuel at 15%, original needles?, check #1, #4 plug color.

[drumstyx]

Gave the 105's a shot -- definitely made a pronounced difference, but I'm still seeing 10-11AFR at moderate throttle openings all the way up to WOT. On the plus side, it's consistent through that range -- again, with the exception of snaps activating the accel pump, it's pretty solidly 10.5-11 throughout the range over a certain throttle opening.

What I get from this is that either this air filter/airbox is probably too restrictive for the delkevic exhaust, or this filter is much more restrictive than the OEM one. I always assumed a better flowing exhaust would make things lean, but thinking about it, it's possible a better flow makes more vacuum, and thus draws more fuel.

Dropping to maybe 100 or 95 mains would get things there, but that's very obviously choking the system to cope with a restrictive air filter.

All that to say, I'll be looking for recommendations for a free-flowing air filter. The K&N one looks to cost $95 from fortnine here in Canada....YIKES!

[scottly]

What do you mean by "oem like" filter? BTW, the A/F at WFO should be 12.8-13.2, at least above 6000 RPM.

[drumstyx]

What do you mean by "oem like" filter? BTW, the A/F at WFO should be 12.8-13.2, at least above 6000 RPM.

"oem like" filter is to say a filter that is advertised to flow similar to the original filter -- direct replacement. In my case, it was either an EMGO or Hiflofiltro filter (don't remember which, I bought it years ago during the rebuild...only now actually getting around to tuning)

I suspect it may be a combination of a slightly more restrictive filter (or possibly truly oem-flowing), and a significantly better flowing exhaust causing more vacuum. Not to mention a very freshly rebuilt engine (less than 200km on the fresh odometer).

[drumstyx]

Sounds crazy...am I going crazy?

Ideas - run test without filter, check results, check needle position, raised needles add fuel at 15%, original needles?, check #1, #4 plug color.

I can't remember if I ended up with the original or the aftermarket kit needles in there. The needles I have on hand, NOT in the carbs currently, are marked "15A". The ones in there, I forget what the exact marking is, but if I recall correctly, it's a 3 or 4 character marking, and includes a "D", either at the beginning or end.

I can't build up the nerve to do an actual road (load) test  without an air filter. I bit the bullet and ordered a K&N high flow filter. I intend to hook up a vacuum sensor to put numbers to the change, so I'm not just throwing parts blindly.

[drumstyx]

Some reference info I found: http://forums.sohc4.net/index.php/topic,129596.msg1452489.html#msg1452489

The spreadsheet is especially useful and interesting. I also found some threads pointing to the EMGO filter being overly restrictive, so it's not just me. I expect a HUGE difference with the K&N, but unfortunately won't be able to road test it for a while...snow and salt are here to stay it seems. Any ideas for a DIY dyno, or at least some resistive rollers LOL!!