AF-XIED Beta for Husqvarna TR650

[mag00]

Check out this good article on wall-wetting. It is a major source of transient fuel.

http://www.megamanual.com/ms2/xtau.htm

You're moving in the right direction Roger, the MicroSquirt development for the TR650 is going to be the final product to fix the bikes. There is no support from husky, and not enough numbers for the other tuner companies to get involved.

One thing I have not seen mentioned enough is Exhaust temp numbers. You claim no that tuning by O2 is better than a dyno tune, I disagree. Many of the hot rodders tune with Exhaust temp. I still think the dyno is better, but using all the readouts.

With the pinging, combustion temps are a factor. While wall wetting is part of the process, stoic combustion is the hottest combustion temp. Fattening the mix, surely could reduce pinging, as the combustion temp is dropped. Overall engine temp is a combination of residual heat, which wall wetting can help dissipate.

On the higher mile bikes, after running rich for 10k or so, carbon will build. This takes up combustion chamber volume, thus increasing combustion pressure, increasing heat. Lean is a cooler combustion as well, as it is not stoic, which is the full burn of all gas and air, nothing left over, raw gas or unburnt air, both of which reduce the combustion temp. With higher compression, higher heat, and higher heat retention in the carbon deposits, premature detonation is a risk. Hot spots on the piston is a risk.

I have recently read a few articles from fairly knowledgeable sources. MOST agree that to 02 feedback system is not a good thing for EFI and is a problem. It is only there for emission standards, not performance. The argument they provide is early non o2 efi and the results. Also, I do not know of a race engine using o2, but they don't typically care about mileage either.

So, let's get on with the micro squirt development. It is alot more fun than half assing a bandaid fix.

And just before hitting post, I was thinking that many think a lean engine runs hot. Well, maybe it is running stoic (or closer to it) more than a fat/rich burning mix??? But that's were an exhaust temp probe might help determine what is going on.

a

 

[engineerk9]

You're moving in the right direction Roger, the MicroSquirt development for the TR650 is going to be the final product to fix the bikes. There is no support from husky, and not enough numbers for the other tuner companies to get involved.

One thing I have not seen mentioned enough is Exhaust temp numbers. You claim no that tuning by O2 is better than a dyno tune, I disagree. Many of the hot rodders tune with Exhaust temp. I still think the dyno is better, but using all the readouts.

With the pinging, combustion temps are a factor. While wall wetting is part of the process, stoic combustion is the hottest combustion temp. Fattening the mix, surely could reduce pinging, as the combustion temp is dropped. Overall engine temp is a combination of residual heat, which wall wetting can help dissipate.

On the higher mile bikes, after running rich for 10k or so, carbon will build. This takes up combustion chamber volume, thus increasing combustion pressure, increasing heat. Lean is a cooler combustion as well, as it is not stoic, which is the full burn of all gas and air, nothing left over, raw gas or unburnt air, both of which reduce the combustion temp. With higher compression, higher heat, and higher heat retention in the carbon deposits, premature detonation is a risk. Hot spots on the piston is a risk.

I have recently read a few articles from fairly knowledgeable sources. MOST agree that to 02 feedback system is not a good thing for EFI and is a problem. It is only there for emission standards, not performance. The argument they provide is early non o2 efi and the results. Also, I do not know of a race engine using o2, but they don't typically care about mileage either.

So, let's get on with the micro squirt development. It is alot more fun than half assing a bandaid fix.

And just before hitting post, I was thinking that many think a lean engine runs hot. Well, maybe it is running stoic (or closer to it) more than a fat/rich burning mix??? But that's were an exhaust temp probe might help determine what is going on.

a

I also have an exhaust cavitation plate which improves cylinder scavenging

 

[Roger 04 RT]

...

I do agree that your AF-XIED unit is possibly the simplest & most effective of all the ecu "enhancers"

...

Thanks but in fairness to the developer, Steve at Nightrider, it is his product. My role was to persuade him to branch out beyond Harley's, to provide the specs for BMWs (which were quite different) & now the TR650, and to do the field testing.

 

[engineerk9]

Thanks but in fairness to the developer, Steve at Nightrider, it is his product. My role was to persuade him to branch out beyond Harley's, to provide the specs for BMWs (which were quite different) & now the TR650, and to do the field testing.

I'm very pleased that you did, cheers

 

[Roger 04 RT]

... the MicroSquirt development for the TR650 is going to be the final product to fix the bikes. There is no support from husky, and not enough numbers for the other tuner companies to get involved.

One thing I have not seen mentioned enough is Exhaust temp numbers. You claim no that tuning by O2 is better than a dyno tune, I disagree. Many of the hot rodders tune with Exhaust temp. I still think the dyno is better, but using all the readouts.

...

The megamanual on the Megasquirt web site is a wealth of information. For my 1150 I've seriously considered replacing the Motronic with a Microsquirt, and an associate is doing it for his 1150. That said, it is a LARGE project, requiring that VE is calculated or measured for each RPM/TPS pair in the fuel table, the development of AFR targets for each fuel table cell, and then developing the Spark Advance matrix. If you took this on, I'm sure you'd get a lot of support from me and the TR650 riders who have LC-2s.

What I've seen in the LC-2 fueling data suggests that the basic fueling of the TR650 is pretty good when lambda-shifted 4-8% to the rich side. As you suggested, this does reduce the EGT from its peak value at stoic. So a side benefit is a cooler exhaust. However, the MM ECU doesn't appear to hand throttle tranisients as well as the Motronic or BMSK.

My specific criticisms on Inertial Dyno tuning are: most dyno cylinders don't create a large enough load, there is no load at the start of the "pull", and the measured AFR numbers are usually junk. Performing the test that Geeza did overcomes all these issues.

EGT tuning is very familiar to me. It is how we manage aircraft performance. It is usually a relative measurement requiring the measurement of peak EGT by trial and error. In other words a direct measurement of temperature without a measurement of peak EGT may not have meaning. The Innovate product line allows EGT monitoring and I've considered adding a probe to each cylinder. I think before I did that though, I would weld an O2 bung into each pipe and measure AFR at each cylinder simultaneously.

The main benefit of the Dyno is it's fast and repeatable. If I were going to use an Inertial Dyno I would make sure it had a static load option so that immediately prior to the pull, the engine was in stable Closed Loop, as it is on the road. Many, many of the Dyno AFR plots I've seen strongly suggest that the bike is in Overrun Fuel Cutoff immediately prior to the pull, as shown by the extremely lean AFR readings, which often is not even back to 14.7:1 until 3000-4000 RPM. Also the probes are so far back in the exhaust that (as with engineerk9's plot) the readings are in error.

All good food for thought.

 

[Roger 04 RT]

Heres a good article of EGT vs Wideband AFR tuning. http://www.sdsefi.com/techegt.htm

 

[engineerk9]

You're moving in the right direction Roger, the MicroSquirt development for the TR650 is going to be the final product to fix the bikes. There is no support from husky, and not enough numbers for the other tuner companies to get involved.

One thing I have not seen mentioned enough is Exhaust temp numbers. You claim no that tuning by O2 is better than a dyno tune, I disagree. Many of the hot rodders tune with Exhaust temp. I still think the dyno is better, but using all the readouts.

With the pinging, combustion temps are a factor. While wall wetting is part of the process, stoic combustion is the hottest combustion temp. Fattening the mix, surely could reduce pinging, as the combustion temp is dropped. Overall engine temp is a combination of residual heat, which wall wetting can help dissipate.

On the higher mile bikes, after running rich for 10k or so, carbon will build. This takes up combustion chamber volume, thus increasing combustion pressure, increasing heat. Lean is a cooler combustion as well, as it is not stoic, which is the full burn of all gas and air, nothing left over, raw gas or unburnt air, both of which reduce the combustion temp. With higher compression, higher heat, and higher heat retention in the carbon deposits, premature detonation is a risk. Hot spots on the piston is a risk.

I have recently read a few articles from fairly knowledgeable sources. MOST agree that to 02 feedback system is not a good thing for EFI and is a problem. It is only there for emission standards, not performance. The argument they provide is early non o2 efi and the results. Also, I do not know of a race engine using o2, but they don't typically care about mileage either.

So, let's get on with the micro squirt development. It is alot more fun than half assing a bandaid fix.

And just before hitting post, I was thinking that many think a lean engine runs hot. Well, maybe it is running stoic (or closer to it) more than a fat/rich burning mix??? But that's were an exhaust temp probe might help determine what is going on.

a

Are you considering using this instead of the MM ecu ?

 

[engineerk9]

The megamanual on the Megasquirt web site is a wealth of information. For my 1150 I've seriously considered replacing the Motronic with a Microsquirt, and an associate is doing it for his 1150. That said, it is a LARGE project, requiring that VE is calculated or measured for each RPM/TPS pair in the fuel table, the development of AFR targets for each fuel table cell, and then developing the Spark Advance matrix. If you took this on, I'm sure you'd get a lot of support from me and the TR650 riders who have LC-2s.

What I've seen in the LC-2 fueling data suggests that the basic fueling of the TR650 is pretty good when lambda-shifted 4-8% to the rich side. As you suggested, this does reduce the EGT from its peak value at stoic. So a side benefit is a cooler exhaust. However, the MM ECU doesn't appear to hand throttle tranisients as well as the Motronic or BMSK.

My specific criticisms on Inertial Dyno tuning are: most dyno cylinders don't create a large enough load, there is no load at the start of the "pull", and the measured AFR numbers are usually junk. Performing the test that Geeza did overcomes all these issues.

EGT tuning is very familiar to me. It is how we manage aircraft performance. It is usually a relative measurement requiring the measurement of peak EGT by trial and error. In other words a direct measurement of temperature without a measurement of peak EGT may not have meaning. The Innovate product line allows EGT monitoring and I've considered adding a probe to each cylinder. I think before I did that though, I would weld an O2 bung into each pipe and measure AFR at each cylinder simultaneously.

The main benefit of the Dyno is it's fast and repeatable. If I were going to use an Inertial Dyno I would make sure it had a static load option so that immediately prior to the pull, the engine was in stable Closed Loop, as it is on the road. Many, many of the Dyno AFR plots I've seen strongly suggest that the bike is in Overrun Fuel Cutoff immediately prior to the pull, as shown by the extremely lean AFR readings, which often is not even back to 14.7:1 until 3000-4000 RPM. Also the probes are so far back in the exhaust that (as with engineerk9's plot) the readings are in error.

All good food for thought.

Similarly, EGT is how we tune our Vee 12, 14,000 kw medium speed diesel engines (600rpm) - we also take cylinder pressures - however tuning individual cylinders by EGT is simplest & effective - the temperature probes are immediately after the exhaust valves - we've had 4 valve heads & turbo charging since 1928 - run BMEP's much higher than motorcycle engines, etc etc - slow speed 2 stroke diesel engines being the most thermally efficient machines developed by man - that's me underneath a ship's propeller, I get to "play" with some "big-boys" toys

 

[engineerk9]

The megamanual on the Megasquirt web site is a wealth of information. For my 1150 I've seriously considered replacing the Motronic with a Microsquirt, and an associate is doing it for his 1150. That said, it is a LARGE project, requiring that VE is calculated or measured for each RPM/TPS pair in the fuel table, the development of AFR targets for each fuel table cell, and then developing the Spark Advance matrix. If you took this on, I'm sure you'd get a lot of support from me and the TR650 riders who have LC-2s.

What I've seen in the LC-2 fueling data suggests that the basic fueling of the TR650 is pretty good when lambda-shifted 4-8% to the rich side. As you suggested, this does reduce the EGT from its peak value at stoic. So a side benefit is a cooler exhaust. However, the MM ECU doesn't appear to hand throttle tranisients as well as the Motronic or BMSK.

My specific criticisms on Inertial Dyno tuning are: most dyno cylinders don't create a large enough load, there is no load at the start of the "pull", and the measured AFR numbers are usually junk. Performing the test that Geeza did overcomes all these issues.

EGT tuning is very familiar to me. It is how we manage aircraft performance. It is usually a relative measurement requiring the measurement of peak EGT by trial and error. In other words a direct measurement of temperature without a measurement of peak EGT may not have meaning. The Innovate product line allows EGT monitoring and I've considered adding a probe to each cylinder. I think before I did that though, I would weld an O2 bung into each pipe and measure AFR at each cylinder simultaneously.

The main benefit of the Dyno is it's fast and repeatable. If I were going to use an Inertial Dyno I would make sure it had a static load option so that immediately prior to the pull, the engine was in stable Closed Loop, as it is on the road. Many, many of the Dyno AFR plots I've seen strongly suggest that the bike is in Overrun Fuel Cutoff immediately prior to the pull, as shown by the extremely lean AFR readings, which often is not even back to 14.7:1 until 3000-4000 RPM. Also the probes are so far back in the exhaust that (as with engineerk9's plot) the readings are in error.

All good food for thought.

Would you care to comment, I'm running Nippon Denso iridium spark plugs ??

 

[mag00]

Are you considering using this instead of the MM ecu ?

Not unless mine goes south.

 

[Jon Knutson]

Geeza also ran multiple WOT test runs in 4th gear on his TR650. The procedure he used was:

--Stabilize and cruise at 2000 RPM in 4th gear.
--Fully open the throttle until 6000 RPM is reached.

This is like a test run on a dyno but much better. The improvements are:

1) While cruising at 2000 RPM in 4th, there is a load on the engine caused by the bike being on a road and having to overcome road and wind resistance. (On a dyno, the bike is sitting still and the inertial cylinder presents no load except when the rear wheel is accelerating. Also, the dyno load is often much less than the resistance caused by the mass of bike and rider, plus the increasing wind resistance.)

2) The Wideband AFR sensor is ahead of the catalytic converter and is more accurately recording the AFR immediately following the exhaust valves.

The results above are for Geeza's bike running at an LC-2 setting of 13.8:1, which would be the same as AF-XIED setting 8. Here are some key points:

--When the throttle is first opened the AFR spikes to about 17:1. This is not good in that it creates a momentary power lag. This may be a natural tendency of the TR650 or it may be due to the POD mod or some other intake tract changes.

--Quickly after beginning WOT, the AFR reaches about 12.4:1 and then gets slightly leaner at about 13:1. This is acceptable and the early richness is probably due to the MM ECU trying to catch up on the fuel dried off the intake walls from the instantaneous throttle opening. Note that there is no sign of leanness in the mid-RPMs. (And if the LC-2 was set to 14.7:1 (stock), the AFR in the mid-RPM's would still be an acceptable 13.9:1.)

--After a short time at 13:1 the mixture steadily richens until 5000 RPM in a predictable manner, reaching about 11.8:1.

--Finally, from 5000 to 6000 RPM it reaches an AFR of about 11:1. Likely it is richer than best power, but many engines are fueled in this manner so that at peak HP, there is extra fuel for cooling the combustion chamber and exhaust valves.

Following the WOT test you can see some other interesting points:

--A Deceleration Lean-Off to about 15:1

--A small throttle movement acceleration enrichment to about 13:1

--During Cruise operation the ECU works with the LC-2 and quickly and predicably returns to target AFR of 13.8:1

RB

Could this lean spike just be a lag in reaction of the EFI to throttle position?

 

[engineerk9]

Could this lean spike just be a lag in reaction of the EFI to throttle position?

Could this lean spike just be a lag in reaction of the EFI to throttle position?

Managed a test run today - set at # 8, ecu reset, 4th gear, 2,000 rpm, the slow & steady (10 seconds) increase to WOT - pulled like a train, huge improvement in drivability

Slight popping on deceleration

Now to do more rides

 

[Roger 04 RT]

Could this lean spike just be a lag in reaction of the EFI to throttle position?

One reason I suggested more tests is that it is "possible" that the throttle is momentarily relaxed just as the rider goes to WOT. That said, I've seen the lean spikes fairly often in the LC-2 logs.

 

[Roger 04 RT]

Managed a test run today - set at # 8, ecu reset, 4th gear, 2,000 rpm, the slow & steady (10 seconds) increase to WOT - pulled like a train, huge improvement in drivability

Slight popping on deceleration

Now to do more rides

Good progress. After 3 tanks of fuel, try setting 9.

 

[DeLewis]

Two questions if I may:

1. In order to 'read' the setting on the AF-XIED, do I count the number of times the yellow LED flashes prior to it remaining illuminated? Or should the count include the LED remaining in the steady illuminated state? (Probably a stupid question but my inclination - which is most likely incorrect - is to count the final illuminated state when 'reading' the setting).

2. Fuel consumption: my device is currently set to 9 and my fuel consumption is consistently around 44MPG or in metric about 5.4l/100km. My riding consists of mostly 60km/hr commuting. I guess my riding style is fairly 'spirited', generally accelerating pretty hard to get ahead of traffic, etc, but nothing like flat-out all the time. Compared to the figures I have seen under the 'fuel consumption' thread, mine seem to be way over the average - perhaps 20% over average.

However, I note that someone on ADV recently posted a adventure bike comparison article which included a TR. The fuel consumption that the reviewers recorded for the TR was almost exactly the same as mine.

Naturally, I understand that the AF-XIED will add a predictable margin of fuel. Nor do I mind some increase in fuel consumption. Overall the bike runs much better: smoother, much cooler running and more power. Finally the question: do my consumption figures seem unreasonable given all the factors mentioned above?

I will try notching back to setting 8 and I may try the bike for a while to get a baseline fuel consumption figure.

Any thoughts welcome.

 

[engineerk9]

Two questions if I may:

1. In order to 'read' the setting on the AF-XIED, do I count the number of times the yellow LED flashes prior to it remaining illuminated? Or should the count include the LED remaining in the steady illuminated state? (Probably a stupid question but my inclination - which is most likely incorrect - is to count the final illuminated state when 'reading' the setting).

2. Fuel consumption: my device is currently set to 9 and my fuel consumption is consistently around 44MPG or in metric about 5.4l/100km. My riding consists of mostly 60km/hr commuting. I guess my riding style is fairly 'spirited', generally accelerating pretty hard to get ahead of traffic, etc, but nothing like flat-out all the time. Compared to the figures I have seen under the 'fuel consumption' thread, mine seem to be way over the average - perhaps 20% over average.

However, I note that someone on ADV recently posted a adventure bike comparison article which included a TR. The fuel consumption that the reviewers recorded for the TR was almost exactly the same as mine.

Naturally, I understand that the AF-XIED will add a predictable margin of fuel. Nor do I mind some increase in fuel consumption. Overall the bike runs much better: smoother, much cooler running and more power. Finally the question: do my consumption figures seem unreasonable given all the factors mentioned above?

I will try notching back to setting 8 and I may try the bike for a while to get a baseline fuel consumption figure.

Any thoughts welcome.

What I found when I first fitted the Booster Plug that the fuel consumption improved, on my Birdsville adventure 4,600klms my average was 26klm/litre, 3.8ltr per 100kls - now that's good
With the AF-Xied once "settled" I wouldn't expect the fuel consumption to dramatically increase, unless I start to use the improved horsepower !!

Some of the other posts seem to indicate something like 2% increase which in the scheme of things is insignificant
More importantly engine life improved 100%, reliability improved, rideability improved, performance improved etc etc

 

[engineerk9]

One reason I suggested more tests is that it is "possible" that the throttle is momentarily relaxed just as the rider goes to WOT. That said, I've seen the lean spikes fairly often in the LC-2 logs.

I understand that, my practice is to start at zero throttle with my wrist high, then roll it around, hopefully eliminating that slight "back-off" prior to WOT - l take time (a few seconds) going from minimum too WOT which maybe gives the ecu time to adjust

 

[nwrider1]

When I last rode my '14 Terra back in December, I was on setting #7 and was still having some running issues -- I believe I posted it back then. While the bike was in storage, the battery was disconnected. I reconnected the battery and changed the setting to #8, confirmed the number of flashes on the LED and then started and let it run until there were 5 bars on the temp display. During a 20 mile ride I had no issues of any sort. Not sure about the power delivery as it's been too long since I rode it last. However, there's no doubt in my mind that it's an improvement. I'll leave it there for another few hundred miles before making another change. Just turned over 1500 miles on the ride.

I also have a '13 Strada with the Wuka Booster Plug. I was starting to have some running issues with it after several thousand miles of excellent running, but we'll see if having the battery disconnected for an extended period has any impact on it's performance. I should get a chance to ride it tomorrow.

 

[LED]

I installed an AF-XIED yesterday. Setting #7 and have done pod mod (not that it may make any difference). Only ridden around 80 km since then but believe I've noticed smoother idling and running, no deceleration pops and more linear throttle response. Looking forward to putting more km on and seeing what effect it has had on fuel consumption.

 

[Roger 04 RT]

I installed an AF-XIED yesterday. Setting #7 and have done pod mod (not that it may make any difference). Only ridden around 80 km since then but believe I've noticed smoother idling and running, no deceleration pops and more linear throttle response. Looking forward to putting more km on and seeing what effect it has had on fuel consumption.

After you get a few tanks of fuel on setting 7, try setting 8.