Dynojet PC-5, with Autotune, and full exh. system Tune...

Roger.
As we have been discussing this for a good long while now, I simply cant help seeing some bias here. I mentioned that the stock bike had an upward of 18 afr as we spend some time driving on the dyno. You mentioned back then that was impossible. I saw it..... Here we are again, and you say it is an error. I say it is there, weather you like it or not ( not trying to pick on you....I am really not) but I trust the dyno, and the results it yields. You seem to put all your eggs in the gs-911 basket, but I will trust the dyno any day over the gs-911. I will admit to that you know more in depth, and the inner functions of the BMSK, than me.... hands down, but I think it is wrong to discredit dyno results that has been so clearly illustrated.

In this case i have to agree with mr Erling, Dyno guy i went too has been doing this stuff for 20 years. His DynoStar dyno has some of the lowest HP and Torque figures in the area making i a lot closer to reality than some of the big name guys over here. (Friend of mine got about 15 bph more on one of his compititors dyno's... Making 135 bhp on the wheel out of a nearly stock CB1000R )

Gentlemen,
I'm not singling out the dyno operators, who I have no reason to mistrust. All I'm commenting on is the AFR test method. The tailpipe AFR probes can't do a good job at the start of an short inertial dyno burst. The exhaust flow is low and they are too near fresh air. The engines would likely stall if the AFRs were that high.

The better test method is an upstream probe such as an innovate (or other) LM-1, LC-1 or LC-2. (I'm not basing any of these AFR comments on GS-911 data.)

Think of It this way: you're running down the road at 2000 rpm in 4th gear. Your stock bike would be running a Closed Loop AFR of 14.7:1. At that point you crank the throttle to WOT. The BMSK goes Open Loop instantly and changes fueling (much faster than a tailpipe probe can respond). It picks its new fueling point based on TPS of 100% and RPM 2000. I would estimate that point is 12.8-13.2:1, and it adds an acceleration enrichment. I have measured this on F800S, R1200GS, R1150, and R1100. They all instantly go to an AFR between 12:1 and 12.8:1 approximately.

This is why I say that those high AFR numbers are in error.

RB

I know this looks impressive.
IMO 2hp seems like sfa difference really, yes it's in a positive direction, but-

Would the same dyno and bike re-produce the same benefits in the positive direction if you keep the same programmed PC-V map and then-
changed it's elevation?
humidity?
fuel?
temprature?

Hmmmm.... I'm thinking that losing 8kg's off my fat arse might just be an even bigger improvement than a 2hp gain at a controlled altitude, humidity and enviroment.

I'd also take a decent suspension package over 2hp gain any day.

The dyno graph posted doesn't really show what happens under 3250rpm's very well either and if any thing it trends into negative territory. It's under these RPM's what really counts in rideability for an ADV bike. Or am I missing something?

I think I will call this a difference of opinion then..... . The 8gs does exactly have issues with a terrible throttle response from idle through a few K rpms. What I am seeing, I conclude is consistent with a roll of from a very high AFR, to a lower AFR. I will say this Roger...... Reading back through this thread, I think it is fantastic that we have all been able to express our different opinions, and do so in a civil matter, with all of us behaving like adults, and no one getting upset. I think this thread represent the best of ADVR culture, in a great collective way. As we have different opinions, none are rejected, and bashed, and for the great contribution, Roger, I commend you. You could have bailed long time ago, but are still offering your opinion. All too many times do you see a thread going to hell, because of those opinions, and peoples egos, as their bikes are being picked on. I would like to thank all who have been contributing to this thread, and has done so in a civil matter......

Noted, and yes. I would like to think that the results would reverse them self if he reversed the install. But mind you, we are not just doing this for the power. I only gained about 5 HP on mine with all the extra junk, but in my case it was to make a better handling bike, as mine was impossible off idle. So bike control was the ticket for me, and a smoother response through out the band. The fact that it yielded a few more HP.... I will take it too.....

Likewise. I feel very committed to studying and experimenting on BMW motorcycles. This is very rewarding. You've tried several experiments on behalf of everyone reading this thread and have been a gentleman about it.

What I really would like to see is a closed loop PCV like the PC III USB. When Dynojet came out with the PC III w/Wideband for the 1100 and 1150, Harley riders were up in arms, demanding one but didn't get it.
RB

I did mine because the bike kept stalling in slow speed hardly any throttle turns (like on roundabouts) Was super infuriating, and was finaly done with it after it happened 5 times in one 2 hour ride with some friends!

I already had Bas at hyperpro fix the suspension

The dyno operator has the ability in input temperature, elevation (?) and humidity. However, many, if not most, dyno pulls only last a few seconds and the machine only gets rolling at above 2000-2500 rpms.

A more interesting test would be to put the bike in each gear at several rpms between 2000-4000 against a load controlled dyno, turn the throttle a fixed amount and see what the rate of acceleration was. This is more representative of what we want to know as riders, IMO. However, this is the type of costly testing that only a company can usually afford.

Roger,

While I commend your work, you are either unaware, or choose to ignore how a skilled dyno operator produces a quality dyno tune. For starters, they don't simply stuff a sniffer up the tail pipe. If proper bungs aren't already on the headers, in the proper location, then Rivnuts are installed where they belong, to provide sample ports for wide band sensors. To produce the fuel/spark maps, the engine is run through as many throttle position, rpm, MAP as possible, under varying loads, so the correct adjustments can be made to the fuel/spark tables. The final pull that you see printed out, is done at WOT, in a single gear, in a few seconds, as you say. What you don't see are the hours of data collection done to reach that point.

Is a generic, across the board enrichening device helpful to correct an engine tuned on the lean side....absolutely! However, it cannot possibly match the performance of a QUALITY dyno tune. With it, you can run the engine rich where either low speed manners, or under heavy load/high throttle settings for power. But keep it near Stoich at steady state cruise, and preserve fuel economy, as well as keep emissions low.

Devices like the Xied are great bang for the buck, but to say they outperform a proper dyno tune, coupled with a remap of the fuel tables, is wishful thinking, IMO.

Jeff

Really wonderful discussion going on, good for all of us. Since this discussion started due to idle hop, throttle response, off throttle control, low speed throttle issues my question is still the same, how does a wot 5th gear run relate?

Regarding the graph just posted how many of us can actually feel 2hp more in a 76hp machine and how many ride in 5th gear at wot? The reason the bike is acceptable to ride now is the low speed throttle control is better probably due to an increase in fuel in these areas all of which has zero to do with the posted dyno results.

I understand what Jeff is saying but is it possible to do anything with the spark tables on a BMSK ECU? Ideally we would but to my knowledge we cannot.

Does anyone have a dyno they can post showing the PC V, Techlusion, and Stock runs from idle to 4,000 RPM?

Isn't a PC V without Autotune changing the fuel map by a set percentage based on the ECU map? Is that different than the Techlusion and if so how?

Just trying to understand.

Terry

"Isn't a PC V without Autotune changing the fuel map by a set percentage based on the ECU map? Is that different than the Techlusion and if so how?"

It changes the fuel table relative to the stock map, yes. Without AT or a air/fuel run on a dyno, or data logging via a wide band sensor, GS911,etc...it's just guesswork. Without a measurable reference of some sort, I don't see how one can accurately tune a bike, no matter the method.

The difference with map based tuning, like the PCV, you can change the fuel in specific regions where it's needed, and leave it alone, everywhere else.

Too bad Dyno-Jet doesn't come out with BMW version of their Power Vision product. Been using it on my Harley, and it's the ultimate, for the DIY home tuner.

Jeff

"I understand what Jeff is saying but is it possible to do anything with the spark tables on a BMSK ECU? Ideally we would but to my knowledge we cannot. "

I can't answer that, and I should not have mentioned spark, as I'm not familiar enough with the BMS-K. All of my home tuning experience is with both the PCV and Power Vision on H-D touring bikes. That said, the principle is the same. Fortunately for Harley tuners, virtually the entire ECM is open to manipulation, via a variety a products like Dyno-Jet, TTS Mastertune, hell, even Harley themselves, sells a race tuner.

Jeff

Thanks Jeff, I appreciate it. There is no doubt a good Dyno operator with enough time can get an acceptable map with the BMSK. However, this map will be after the ECU has made it's choice based on closed loop inputs. In order to use the PC V the O2 sensors are disconnected thus putting the ECU into a different map. The question becomes whether that map is stable. The BMSK is adaptive. I have dual wideband sensors allowing me to record each cylinder. I have disconnected one cylinder Lambda input to the ECU completely. The bike ran like crap in the beginning but started getting rideable as the miles went by. I recorded this with a GS-911 and A/F R logger. The ECU was adapting the disconnected cylinder to match the good one.

Some ECU's are open source, those yield excellent results when all inputs including timing can be manipulated. The BMSK is a totally different animal.

Terry

Terry,

Great info, thanks! Pretty amazing that it'll run on one lung, so to speak. Must have a fall back plan, in case one O2 sensor craps out, it defaults to using the good one as a reference? Wonder raw hat happens with both out?

Unfortunately, with relatively low production numbers, I doubt that folks like Dyno Jet will crack the code of a BMS-K, like they have with Harley's. While the H-D Delphi system isn't open source, there's enough money to be made, to make hacking it worth the expense.

Jeff

Good discussion, many good points.

We're saying much the same thing. My point was that developing a fuel map requires that Wideband samplers are placed in locations where they can accurately measure the exhaust oxygen content. Too many plots of AFR that I've seen obviously do not reflect what is going on. When an AFR curve starts at 18 or 20 to 1, AFRs where the engine won't run, it puts a lot of the data in question.

My other point, what you're saying too, is that it takes hours of skilled measurements of AFR and spark advance trials to develop a proper set of maps.

Autotune is a neat concept with the obvious flaws that you have to know what AFR targets to input, and you have to evaluate cell to cell discontinuities manually. Dynojet seems to promote it as the equal of Adaptive Closed Loop operation, which it is not. The reason I say that is Autotune is concerned with fuel map development (the map design) and Adaptive Closed loop control is concerned with maintaining the map design as actual engine conditions vary.

Another consideration that is usually neglected in the discussion but is on the table here now is that ECU map development usually means Fuel and Spark but the PCV on BMWs only seems to allow fuel map changes. When You see +30% in a cell you have to wonder what spark advance adjustment is required.

Lastly, without O2 sensors connected, the Motronic and BMSK both revert to a Limp Home fueling pattern, together with stray Adaptives that Terryckdbf mentioned, that can make it difficult to get a good tune. In the end, even a poor tune that adds fuel is probably an improvement over EPA restricted fueling.

Regarding the BMW-AF-XIED, which is quite different from the Harley IED devices designed by nightrider, the concept is that BMW has tuned the intake and exhaust plumbing, and built high quality fuel and spark maps and backed them with Adaptive Closed Loop control. A small shift in fueling (4-8%) while keeping BMW's quality maps, seems to deliver significant benefits at a low cost. BMW's main constraint is keeping the catalytic converter doing its thing. The idea of lambda shifting is to relax that constraint thereby getting more of the motor's natural performance.

A good example of what a motor can do without the EPA constraints is the R1100 series. It is one of the few electronic fuel injected motors that BMW shipped in two flavors: Open Loop without a catalytic converter and Closed Loop with a cat. The Closed Loop version has its AFR centered around 14.7:1, many of the bikes surge and balk at light throttle and low rpms. The Open Loop bike is much more driveable and surge free, its AFR is 5% richer at 14.0:1. So to the question, "What would BMW do to make the Oilheads run better if they didn't have to meet EPA specs?", we have the answer: Add 5% to the fueling.

RB

Gents........ Great to see the discussion continue. I am traveling,and will spend the next two weeks in Wichita,KS, so I will be away from the bike. I am getting anxious to try to open the intake,and have found a supposedly good dyno operator in el,paso. The makers of msd ignition. As soon as I have a chance, I will give it a run with, and with out the air box cover on.

Hi guys!

Super interesting discussion, very educational!

I've continued to experiment with the PCV and AT. I've changed target AFR from typically 13.2-13,5, to 13,5-13,8 broadening the "cruising range". From about 3250 rpms to 5000, for throttle 5-40%. At 2% I've kept 13,2 and over 40% throttle it's 13,2-13,5. The bike just runs better and better and I see the trims the AT develops to reach the new AFRs. Seems to work good enough. Probably unnecessary rich and increased fuel consumption, bit the bike runs great!

I see the challenge however. DJs intial AFRs are for max power. I have no way of knowing which target AFRs that are best, regardless of if I go for power, economy, driveability or a mix. I probably loose more economy than I really would like to, gaining some power I don't really need.

A slight enrichment all over, simple and low cost, would improve driveability and keep most of the economy. Very nice.

Does anyone have the "perfect" AFRs for this lovely engine?

Best regards,
Mats P

You could try leaner mixtures in the 0-2% columns, 13.5-13.8. At high throttle angles you could go to with a 12.9 to 13.5. I suggest this based on the measured AFR numbers I've seen on my 1150 with 8% overall enrichment.

Consider resetting the BMSK Adaptatiion Values now so that they don't get accidentally reset later, messing up your Autotune.
RB

Thanks Roger!

How do I reset?

DJs target AFRs, does anyone know how "good" they are and how much effort they have put into determining them?

Your suggestions makes sense, is this common knowledge, educated guesses or do you know the subject better than DJ?

Best regards,
Mats P