JPI EDM LF operation

[0TreeLemur]

I'm curious about how to best use the LF function on their JPI EDM's.  Not sure which models have this function.  I have a model 900. 

Since installing a refurb carburetor on our C this last summer, it doesn't detect "Leanest" until the engine almost dies.   Then when I screw the mixture back in, it pops and sputters.  With the old carb, it didn't seem to get that close to dying when it detected the leanest cylinder.   I don't like making the engine pop and sputter.

Also, I've heard mention of a mode that shows all the temps relative to each other rather than the full range.   How do you make it do that?  What's the utility?

Finally, my EDM somehow showed the "delta T" for a while but now it doesn't automatically do that, I have to press "Peak" to get it to show that.  I liked it better when it automatically did it when in the LF mode.  Not sure why it stopped.  I haven't done anything that I know of to change that behavior.

Can someone who is proficient with the JPI LF function share their procedure?  I'd sure appreciate it, and I'm sure others will benefit too.

Thanks!

 

[Ragsf15e]

1 hour ago, 0TreeLemur said:

I'm curious about how to best use the LF function on their JPI EDM's.  Not sure which models have this function.  I have a model 900. 

Since installing a refurb carburetor on our C this last summer, it doesn't detect "Leanest" until the engine almost dies.   Then when I screw the mixture back in, it pops and sputters.  With the old carb, it didn't seem to get that close to dying when it detected the leanest cylinder.   I don't like making the engine pop and sputter.

Also, I've heard mention of a mode that shows all the temps relative to each other rather than the full range.   How do you make it do that?  What's the utility?

Finally, my EDM somehow showed the "delta T" for a while but now it doesn't automatically do that, I have to press "Peak" to get it to show that.  I liked it better when it automatically did it when in the LF mode.  Not sure why it stopped.  I haven't done anything that I know of to change that behavior.

Can someone who is proficient with the JPI LF function share their procedure?  I'd sure appreciate it, and I'm sure others will benefit too.

Thanks!

 

I have a dumb question… are you definitely in ROP function and not LOP? I’m on a -930, so slightly different, but the one showing near the select button is not the mode you’re currently in.  Instead It selects the mode displayed above the button.  Clear as mud?

They may have updated firmware too because i have to press a button after ROP peak detection to show me the delta as well.  

[Greg Ellis]

I second what @Ragsf15e says.  Make sure you are coming at it from the ROP side.  This will show the first cylinder to go leanest.  If you come at it from LOP side you won’t see the leanest cylinder until the last cylinder goes LOP.  In my airplane that would be close to the engine quitting.  My plane gets very rough just getting one cylinder LOP.

[0TreeLemur]

11 hours ago, Ragsf15e said:

I have a dumb question… are you definitely in ROP function and not LOP? I’m on a -930, so slightly different, but the one showing near the select button is not the mode you’re currently in.  Instead It selects the mode displayed above the button.  Clear as mud?

They may have updated firmware too because i have to press a button after ROP peak detection to show me the delta as well.  

 

1 hour ago, Greg Ellis said:

I second what @Ragsf15e says.  Make sure you are coming at it from the ROP side.  This will show the first cylinder to go leanest.  If you come at it from LOP side you won’t see the leanest cylinder until the last cylinder goes LOP.  In my airplane that would be close to the engine quitting.  My plane gets very rough just getting one cylinder LOP.

That's a not a dumb question @Ragsf15e, that's a great question.  If it is set to LOP that would cause the bad stumble.   Not 100% certain but I think that when I select LF, it reports "ROP" at the bottom of the screen.  I'll verify.  Thanks.

What about my other questions?   How do you get it to consistently report delta T without having to hit a button each time?  It used to do that, but stopped for some reason.  The only programming I've done on the thing in forever is adjusting the ff K factor.

[jlunseth]

The ROP and LOP lean side functions are misnomers. The JPI does not actually know from sensor information whether the engine is operating lean or rich of peak. Rather, it assumes that all leaning is done from the rich side. 

ROP simply means that the unit will find the first cylinder to peak regardless of whether you are “leaning” from the rich side or the lean side. LOP simply means the unit will find the last cylinder to peak. The other thing the LOP function does is give you the hanging icicle look if you like that.

Which Delta T are you asking about? The unit displays or is capable of displaying several. Difference between hottest and coldest EGT is the one people tend to fixate on. Difference between hottest and coldest EGT is an irrelevant number if that is what has disappeared. There are always differences, they are caused by many things including difference in probe placement from one cyl to the next. Differences are normal but large changes are not, so if the difference in EGT between the hottest and coldest is normally, say, 100 dF, and all of a sudden it is 300 dF that should wake you up, but then so should the straight up EGT display because now one of the EGTs is significantly hotter or colder than it was before.

The function that displays the CHTs and EGTs relative to one another is Normalize. It is a push and hold of the second button from the left. A quick tap of the same button turns on the leaning function. I am just the guy that pushes it I don’t look at the labels much anymore. The CHT/EGT bars become flat with each other notwithstanding that the values are different from one cylinder to the next. I am pretty sure that is the LF button, just look up Normalize in your manual and it will tell you which button and how to push it. The button serves more than one function depending on how you push it.

[0TreeLemur]

7 hours ago, jlunseth said:

Which Delta T are you asking about?

After it finds the peak EGT of the leanest cylinder, it shows how many degrees it changes from the peak.

[Aerodon]

First, there are several versions of the software installed in an EDM700, so you may not have the 'icicle' display and LOP installed?  See if you can see what version is installed during the start sequence?

Second, the delta T is on CHT's not EGT's.  The difference in EGT's is not at all important, what counts is the peak.  

Third, use the normal 'ROP' peak function to record which cylinder peaks first and what the fuel flow is?  Then use the 'LOP' function to find out what the last cylinder to peak, and what fuel flow.  The difference is the famous GAMi spread.

Finally, I think you are chasing you tail with a carbureted engine.  I tried hard on my 180 hp 172 to use the JPI but it would always start running rough before finding a LOP cylinder, so I never could get the LOP function to work.  I was also not happy with the fuel flow, it seemed about 1 GPH higher than my previous 180hp C172.  We did a top overhaul with 4 matching cylinders and it runs much better, at about 1 GPH less.  Two of the 4 old cylinders were original, and the two replacements were the worst - must have been 'low cost / poor condition / return to service' kind replacements.  Much happier with the new cylinders.

 

Aerodon

 

 

[Pinecone]

For a carb engine, try running just slightly below WOT.  The slightly cocked throttle plate can cause turbulence to even out the misture distribution to the cylinders.

Also try a touch of carb heat.  That can also help.

Otherwise, go old school, lean until it is rough, then just rich enough to smooth things out. 

[0TreeLemur]

19 hours ago, Aerodon said:

First, there are several versions of the software installed in an EDM700, so you may not have the 'icicle' display and LOP installed?  See if you can see what version is installed during the start sequence?

Second, the delta T is on CHT's not EGT's.  The difference in EGT's is not at all important, what counts is the peak.  

Finally, I think you are chasing you tail with a carbureted engine. 

I've got the EDM900 with ROP and LOP options.   Yes, EGT's

I've flown 9 hours in the past week giving me lots of time to experiment.   I've discovered that the rate at which I lean the mixture is critical.  There is something like an "optimal" rate where it detects the first cylinder to peak, and if I immediately start to enrichen the mixture it doesn't stumble too bad.

12 hours ago, Pinecone said:

For a carb engine, try running just slightly below WOT.  The slightly cocked throttle plate can cause turbulence to even out the misture distribution to the cylinders.

Also try a touch of carb heat.  That can also help.

Otherwise, go old school, lean until it is rough, then just rich enough to smooth things out. 

Yes.  I always put in just enough throttle plate to smooth it out a tad when it starts to roughen.  Learned that here on MS.

Thanks to both of you for your replies.

[kortopates]

I am afraid what your trying to do with a C model really isn’t possible. if you set aside the monitor and just do slow, very slow, mixture sweeps, download the data and look at it on Savvy you’ll see you are only going to be able to get first and maybe a second cyl to peak, but highly unlikely you’ll see more than that before it’s so rough you’ll think the engine will quit.
The “optimal rate” to lean is slowly for accuracy such that you’re seeing a 1/10 GPH change about every second (data point). there is too much hysteresis to go fast and going too quick is just garbage in and garbage out with unrepeatable results.
when you see what’s going on in the data, then start experimenting with reduced throttle but still WOT MAP, and carb heat - but you need carb temp to be repeatable; you’ll see you can make some improvements to how far lean you can get before needing to change your underwear

[Shadrach]

I think the lean find function is gimmicky at best. It's unnecessary and can actually muddy the water for someone new to the intricacies of advanced engine setting. 

The pilot needs to know the first cylinder to peak (leanest for ROP operation) and the last cylinder to peak (richest for LOP operation), that is all. The lean find function complicates the process with additional and unnecessary steps and feedback that might be confusing. 

[0TreeLemur]

12 hours ago, Shadrach said:

I think the lean find function is gimmicky at best. It's unnecessary and can actually muddy the water for someone new to the intricacies of advanced engine setting. 

The pilot needs to know the first cylinder to peak (leanest for ROP operation) and the last cylinder to peak (richest for LOP operation), that is all. The lean find function complicates the process with additional and unnecessary steps and feedback that might be confusing. 

In the case of the JPI EDM900 at least, it actually seems pretty good at detecting the first cylinder to peak.   It also memorizes the EGT on that first to peak cylinder when that happens so I don't have to, and shows the degrees ROP as I richen it back up.

I've found that the process is pretty sensitive to the rate at which I lean the engine though.   Too fast or too slow it runs real rough.   Just right and it skips a few beats.   Lean too fast, it almost dies, lean too slowly, and it pops & sputters longer than I like it to.   

This past summer we put on a remanufactured carburetor, which is more sensitive to the rate of leaning than the old one was, for some reason.

[Shadrach]

13 hours ago, 0TreeLemur said:

In the case of the JPI EDM900 at least, it actually seems pretty good at detecting the first cylinder to peak.   It also memorizes the EGT on that first to peak cylinder when that happens so I don't have to, and shows the degrees ROP as I richen it back up.

I've found that the process is pretty sensitive to the rate at which I lean the engine though.   Too fast or too slow it runs real rough.   Just right and it skips a few beats.   Lean too fast, it almost dies, lean too slowly, and it pops & sputters longer than I like it to.   

This past summer we put on a remanufactured carburetor, which is more sensitive to the rate of leaning than the old one was, for some reason.

Maybe I’ve just been doing it so long the lean find seems arduous.
I know which cylinder is leanest and I know which cylinder is richest.  I also know about where they both peak...So...

My SOP for ROP is to lean to 1370 on #3 then watch CHTs and fine tune from there.

My SOP for LOP is to pull mixture to slight deceleration and enrich #2 to 1415 then watch CHTs and fine tune from there. Simple and fast with minimal fiddling.

[Ragsf15e]

On 11/15/2022 at 7:31 PM, Shadrach said:

Maybe I’ve just been doing it so long the lean find seems arduous.
I know which cylinder is leanest and I know which cylinder is richest.  I also know about where they both peak...So...

My SOP for ROP is to lean to 1370 on #3 then watch CHTs and fine tune from there.

My SOP for LOP is to pull mixture to slight deceleration and enrich #2 to 1415 then watch CHTs and fine tune from there. Simple and fast with minimal fiddling.

I like my lean find as well.  Mostly for the reasons @0TreeLemur said, but also because my gami spread is close enough that my last to peak (richest) can be different based on conditions.  If I cruise down low at high mp and without ram air, maybe it’s #2, up real high with ram air on, likely it’s 3.  One other nice thing about LF is it allows the jpi data analysis program to quickly analyze that point and you can see if things are changing.  Maybe a plug Or timing issue slowly developing?

In general, you’re right that it’s unnecessary if you just watch the temps closely and have an idea how it will end, but extra toys/options are always nice, no?

[Shadrach]

6 hours ago, Ragsf15e said:

I like my lean find as well.  Mostly for the reasons @0TreeLemur said, but also because my gami spread is close enough that my last to peak (richest) can be different based on conditions.  If I cruise down low at high mp and without ram air, maybe it’s #2, up real high with ram air on, likely it’s 3.  One other nice thing about LF is it allows the jpi data analysis program to quickly analyze that point and you can see if things are changing.  Maybe a plug Or timing issue slowly developing?

In general, you’re right that it’s unnecessary if you just watch the temps closely and have an idea how it will end, but extra toys/options are always nice, no?

If your GAMI spread is that close, you’re splitting hairs using one cylinder of over the other anyway. Being able to state precise degrees from peak is academic at that point. Airspeed, FF and CHT are giving more useable feedback.

[Ragsf15e]

4 minutes ago, Shadrach said:

If your GAMI spread is that close, you’re splitting hairs using one cylinder of over the other anyway. Being able to state precise degrees from peak is academic at that point. Airspeed, FF and CHT are giving more useable feedback.

Yeah, fair enough. I just like to make sure they are all at least slightly past peak, but at the higher altitude I usually fly, it wouldn’t hurt either way.

[0TreeLemur]

Since I created this post, we've sold our C and bought a J.   In our carbureted C running at 75% power, using the LF function to get 80 to 100F ROP usually resulted in a ff of about 10.3-10.6 gph.

In our J, with its fuel injected engine and well balanced injectors, the LF function is not needed.   I just lean it out quickly until it starts to stop producing power, the richen it up until it runs smoothly.   Results in about 9 gph at 75% power LOP.

Message: The LF function is a lot more valuable when you have a carburetor, or I suppose when you have a fuel injected engines but imbalanced injectors and need to run ROP.

[Ragsf15e]

On 2/11/2023 at 11:02 AM, 0TreeLemur said:

Since I created this post, we've sold our C and bought a J.   In our carbureted C running at 75% power, using the LF function to get 80 to 100F ROP usually resulted in a ff of about 10.3-10.6 gph.

In our J, with its fuel injected engine and well balanced injectors, the LF function is not needed.   I just lean it out quickly until it starts to stop producing power, the richen it up until it runs smoothly.   Results in about 9 gph at 75% power LOP.

Message: The LF function is a lot more valuable when you have a carburetor, or I suppose when you have a fuel injected engines but imbalanced injectors and need to run ROP.

I think you’ll find it useful in your J as well if you ever fly higher.  Down low you can just pull to 9gph ff / 75% as you indicated and be pretty good.  Up higher where you’re limited by the MP, it’s nice to know exactly where peak is.  If you get high enough, you’ll probably want to run ROP just for want of more power, but 9-11000’, LOP still works ok, but there’s a noticeable speed difference between peak and 30 LOP.  Either setting is likely cool enough on the CHTs as you’re only making 60-65% power.  I like to set mine just barely LOP up there and then see how that works for chts and speed.  If you’re impatient, you can add 4-6kts by going back to 100ROP.

Its nice to be able to be relatively precise with it when power is limited.

[0TreeLemur]

1 hour ago, Ragsf15e said:

I think you’ll find it useful in your J as well if you ever fly higher.  Down low you can just pull to 9gph ff / 75% as you indicated and be pretty good.  Up higher where you’re limited by the MP, it’s nice to know exactly where peak is.  If you get high enough, you’ll probably want to run ROP just for want of more power, but 9-11000’, LOP still works ok, but there’s a noticeable speed difference between peak and 30 LOP.  Either setting is likely cool enough on the CHTs as you’re only making 60-65% power.  I like to set mine just barely LOP up there and then see how that works for chts and speed.  If you’re impatient, you can add 4-6kts by going back to 100ROP.

Its nice to be able to be relatively precise with it when power is limited.

power=speed=1/time=$