HypoJets and O-tubes on the way...

[simonknee]

The 25mm is from the top of the inner deck plate to the surface of the fuel. Check out my images earlier in the thread. You take out a main jet and measure down the hole. By "inner deck" I am referring to the one you take the main jet out from.

Keith describes this in his white paper. He sells a clever optical tool. Or you can use a screwdriver like me. Or a vernier calliper of the type that has a protruding depth gauge. With all these methods you need a nice bright little light to illuminate when you break the surface of the fuel.

[Galwaylotus]

I should have looked back before my last post. Thanks, Simon. The photos illustrate it perfectly!

[simonknee]

Keith's bits are not here yet but I did get out of London for once.
It's a bit scary out in the countryside so I did a high speed dash back from Banbury on the M40:

Left hand side is a 4000-5000 (erm and a bit) rpm in 4th gear.
You can see that is a bit rich when I am cruising 20-40% throttle but goes just about right when I put my foot down at 50-90% throttle.
There is no hesitation when I put my foot down at any of these throttle openings.

Right hand side is when I hit the 50mph section on the M40 around Northholt all the way to the Westway elevated section.
This is looking much like my normal sub-30mph pootle round town but again a bit rich when down on the progression circuit.
However again I have no hesitation with going WOT.
There is the slightest hint of a hunt around 2.5K cruising. Not shown in the data but I can feel it.

[simonknee]

Ooo....

Ahh...

Top to bottom.
Hypo 105 dry jets
Obert - E-tubes (like an F7 with the above fuel level holes plugged)
120 wet jets (not from Keith) (also have some 125 and the existing 1
.022 hypo (idle) jets
.020 hypo (idle) jets
Idle jet jackets.

You can see why they are called hypo-jets as Keith uses a section of a hypodermic needle to provide accurate metering. The hypo-jets slide into the jackets which in turn has four air holes of varying diameter. Only one will line up with the single hole in the hypojet and hence allows for tuning of the air flow. With these parts I have 8 combinations of mix available on the idle jets.

[simonknee]

Here we go...

I the interests of science (ahem) I am resisting the temptation to put it all in at once.
So hypojets first I think especially since I am pootling around town at present.

Out with the 45f9 and in with the H203 (this means the .020" wet jet set to reveal the 3 largest hole in the jacket)

Immediately I had to up the idle mixture to 3/4 of a turn (I have the coarse thread version on my 40DCOE18)
Then I increased this further to 7/8 open and ended up with an idle AFR of approx 12.2
Then off I went to Battersea to fill up on V-power!
I could see that I did not have a particular improvement in mixture control.
You can see the swing on the purple graph line.
It was noticeable that I did not get as many very lean spots compared with the 45f9
However it was also going a bit too rich.

So at the Shell garage I leaned out the idle mix slightly to an AFR of 12.7 (probably around 13/16 of a turn).

Doesn't go so rich but you can see the lean spikes on purple graph.
These occur when I lift off the throttle and hence only the idle mix screw holes are fueling the engine.
This is one reason why setting a 14.7 (best lean) idle will result in problems - super lean!

So lets go a bit richer, H223:

I left the idle mix screws the same and this resulted in an AFR at idle of 12.
This was only a short run as I wanted to get back to watch the Opening Ceremony (go Hawking!)
However the graph immediately looks smoother with no nasty lean spikes.
Felt good too!

So this morning I leaned the mix back a bit for the run to work:

Ended up with an idle AFR of 12.4
Ooooo look at that!
No nasty lean spikes and much more tightly bunched around the ideal afr.
Feels smoother to drive and easier to move off too.

One issue has crept in and that is the idle sometimes sticks at 1050rpm rather than the 850-900 I have set it at.
Since I have a TPS I know that the throttle is in the correct place.
This lazy idle is a known issue and I will have to go a read some posts in Keith's forum about this.
Probably to do with the position of the plates with respect the progression holes.

I aught to do some more testing before I put the custom Obert E-tubes in but not sure I can wait!

[bill308]

Hi Simon. Thanks for all the data and effort. I'm trying to understand the data you presented.

On your frequency of occurance plot, I see a lot of data spikes that seem to fall on 250 AFR multiples, perhaps periodic?

Is this noise or is something else happening?

What is the vertical scale on this plot and what are the increment values?

Thanks again,
Bill

[holywood3645]

I read this thread with interset. I wish i understood it better.
James

[simonknee]

Hi all,

Yes I do realise that chucking these graphs at you easily mystifies. I will try and add more generic explanations as I go along. It's not that bad really!

First lets look at what the Logworks Software calls Channel Statistics -
It shows the data collected for just one sensor - most useful for AFR which I always colour purple.
e.g.

The vertical scale represents the number of times a particular AFR value occurred during the run.
It gets rescaled to the largest value and so doesn't have any values on it.
This is because a long run has tens of thousands of values and short run only hundreds but we want to compare the overall effect regardless of how long the data is collected for.
I think the spikes are a bit of a bug with the software since it does this on sample data from what they call a Vintage 340 tri-power Cuda TransAm-dry!! I ignore the spikes like a good statistician should

So looking at this you see that this was 4 minute run with the minimum value at 10 and the max at 20, hence a spread of 10 and an average value of 14. It's a bit of an indistinct lump.

For maximum power we want to achieve 12.5. Anything lower is too rich and we are wasting fuel. Maximum MPG is at 14.7, anything greater and we are too lean and also wasting fuel! This graph from the Weber tuning manual shows the relationship of AFR (air to fuel ratio) to HP and AFR to MPG:

So in making modifications to the tune I am trying to reduce the quantity of values below 12 and above 15. That's why my latest graph with HypoJets installed pleased me greatly:

Whilst the actual spread is still 9 look at how few values actually contribute to this - I could probably claim a spread of 5 on this! The data is also tapering in a volcano shape to the peak. Usually I achieve an Ayers rock lump with a peak sticking out at 12.5.

I have to be careful not to be misled by peaks around 12.5. Why? Well the only value that you can precisely set on the carbs is the idle. This is what the idle mix screws set. Like your RPM the carbs will always return to this value at idle. I use the AFR meter to deliberately set this to 12.5. So I am bound to get a nice bunch of data at this value. This makes it easy to think everything is going well when it aint. There is a way around this and in the first image of this post I chose 4 minutes of data when I was never at idle (you can slice and dice the data in the software). This means I get to see what is really going on when actually driving. The later graph does include some idling. However if you mentally round off that peak it is still very good and is all tapering to the ideal of 12.5.

The other thing here is that I am not going above 2500 rpm since I'm only testing the Hypojets (idle jets) and don't want the main jets to cloud the picture.

So finally, for now, it is worth mentioning that idle jets are poorly named since they govern the mixture way beyond idle. Low speed cruising and town driving may barely ever trouble the main jets. You can test this with a method I first read in Keith's tuning white paper. Take out your main jets. Put the carb covers back on and then go for a drive. As long as you only use light throttle the car will run fine. Gradually accelerate and somewhere around 2000-3000rpm the engine will cut out (fuel stops flowing). All of this will have been done on the idle jets only and you will have found the range yours operate over. To my shame I have not done this yet but in my defense you need a quiet road and I haven't get one of those round here!

Does this help James??

[CBUEB1771]

Simon,
I just want to let you know this is the best thread I have followed in a long time. The data from your instruments are very illuminating and I do appreciate your cautions on interpretation. I look forward to the future reports. I too am very interested about the frequency of occurrence spikes at roughly 0.25 AFR increments.

[holywood3645]

Simon
Thanks for taking the time to explain, I know it took some effort on your part. Yes! It?s helps me (and I?m sure others) a lot, but I need to get a good book on 40DCOE151 and do some studying. It appears this is not just a simple change them out and hope they work.
I have baseline settings from QED for their Super-Sprint 1700cc 420S spec motor, and was hoping to substitute Franks modified parts for some of the items QED recommend. The carb?s I have were jetted for a stock spec twin cam and I have changed out the Choke, Main nd Air Correction per QED recommendation below, then use Keith Franks hypoJets and o-tubes

CHOKE From 30 To (QED recommended) 34
MAIN From .115 To (QED recommended) .140
AIR CORRECTOR From .200 To (QED recommended) .155
Emulsion Tube ?? (QED recommended) F16
Idle Jet ?? (QED recommended) 45F8

Frank suggested O5 Obert, 22 thou hypojet, and figure the main should be around 130
I?m sorta at stalemate now as far as ordering the remaining QED suggested parts and see how they work or just ordering Keith Franks parts and rolling the dice.
I have the job half done and not sure how should proceed.
I will continue to follow your posts here, and also follow the yahoo group?s posts and hopefully it will all be clear soon.

Keep up the good work, all very intersting

James

[bill308]

Simon,

Is it possible to access the raw data in ascii real numbers so that it could be dumped into a spreadsheet for some further data reduction?

Bill

[bill308]

Russ,

Thanks for correcting my observation of 250 to 0.25 AFR.

The time histories are great showing coherent AFR, throttle position, and rpm.

It would me interesting to look at frequency of occurance data for a steady state idle condition and at seperate transient set of data where there is almost no idling.

Bill

[simonknee]

James,

Is this currently running. If so how does it go?
What are your answers to the questions I posted to Jeremy a little while back?

[simonknee]

Bill,

It will export a .csv file with the columns of the data as recorded.

You can also download the Logworks software for yourself at:
http://www.innovatemotorsports.com/support.php
It's completely free - Innovate would like you to buy an LM-2 or similar.

The software lets you highlight portions of the timeline and thus limit the values included on the graph or plot.
I would give you the two images but I've left the SDcard with all my logs on in the car.
I'll post some tomorrow.

What I would like (and will suggest to Innovate) is that the software could filter the AFR data based on another of the logged data. e.g. only graph AFR at times where RPM is 1000 or greater.

What I have not done is a proper set of control tests. The roads of London are not conducive to this you can imagine. So this will have to wait until later. So I console myself with is the fact I am tuning for the actual driving that I do.

[jimj]

I`ve been reading this thread avidly hoping to comprehend even part of it but this phrase;

"It would me interesting to look at frequency of occurance data for a steady state idle condition and at seperate transient set of data where there is almost no idling"

made me realise that I don`t speak carborettor (or carboretter) but keep up the good work.
Jim