D.I.Y. Rolling Road on-the-real-road ; how would you go about it? serious Dellorto subject creep in this thread

[anthony1956]

I have the kit, Kane road sensor just arrived. Hopefully I have sorted the carb pump jet dribble, so next step is static testing. 

Then, the exciting bit, how do do this on the open road, or maybe a quiet section of it - like the road behind our house after 7pm.

Something that perplexes me is that with an ECU there are loads of spaces to be filled in and the injectors will be supplied with fuel to match all the way to one million rpm or whatever is the max on that car.

A twin Dellorto DHLA-C 45 engine doesn't have any such programming space available. Yes one can play with jets sizes and emulsion tubes and so on, but I am not going to change any of these at this time. My objective is to make the best of what Geoff (r.i.p.) built. ok I have stuck a six-speed and 4:1 open diff in it, but the engine is the same as it was. 

So having made all those variables fixed, I can tweak the mixture and the air compensation with a Carbtune Pro manometer all at tick-over and maximise the emissions efficiency using the gas analyser. All of which seems to be more than most try. I will be also using Gunsons Colortune on all 4 cylinders, but this is a new adventure.

So with my new open road sensor say I take readings at various RPMs for 3,000, 4,000, 5,000, 6,000 and 7,000 say (just guessing). While these will be interesting to see. What is there to tweak? The timing? 

Answers on a postage stamp please.. 

Regards Perplexed Anthony now armed with far more skills than I ever wanted.

 

supplementary question: why don't we all use gas analysers? Emissions, which are easy to comply with (the image below shows "pass" results for a 1987 BDR in Switzerland),  is only the beginning of the story of what can be done with them; and it really cannot be more difficult in the UK. Ok so none of this for those who prefer to have their cars managed by professionals. It's for the rest of us.

 

Edited September 10 by anthonym

[OldAndrewE]

39 minutes ago, anthonym said:

So with my new open road sensor say I take readings at various RPMs for 3,000, 4,000, 5,000, 6,000 and 7,000 say (just guessing). While these will be interesting to see. What is there to tweak? The timing?

When you say take readings I assume you mean AFR.  The engine load and throttle position will also be variables. Assuming your readings are in the acceptable range then OK, if not then you are into changing jets/tubes I think

[anthony1956]

That's very interesting, thanks. 

My thoughts about this rolling road stuff is for a forty year old car that's been perfectly ok so far, aside for it being interesting (and I do find it so) there doesn't seem to be any driving force to do this. It's a nice to have, but that's all it is. I have had suggestions my engine might be destroyed because something is wrong.. which aside from major user error (mine!) is simply excessive in its attempted persuasive power, in my opinion.

I find with the guidance available from the folk on here, it's unlikely I will destroy my engine. No bent valves so far (thanks Andrew R) A month ago I would NEVER have even considered changing a cam belt. No longer. 

 AFR: Yes the numbers are shown in the image I have now added in my o p. 

And in any case we have Manometer readings as well, a much less expensive way of getting reasonably good settings per cylinder. 

Gunsons Colortune I have not tried yet; that excitement awaits.

 

So this is Techtalk: The Blatchat School of Engines

 

p.s. my current sub plot started with "my car won't start" and it would have been very useful to have a standard check list somewhere on here to run through - sure I had one in my head, but .... I could probably now add a few things to check! 😉  Maybe there is one? Dis I just re-invent the "f.a.q."?

Edited September 9 by anthonym

[OldAndrewE]

3 minutes ago, anthonym said:

Gunsons Colortune I have not tried yet; that excitement awaits.

The tests with the Colortune will presumably be to check/verify the gas analyser results at idle 

[anthony1956]

Well now. Yes is the short answer. The gas analyser relies on the results from the average of all four pots. Each Colortune provides insight in to each pot, like but not the same as the Manometer. 

The weakness of the Gas Analyser is it works on the average and we know what those are like. For example the AFR results shown in my image above are the average when one carb is in fact WAY wrong owing to pump jet dribbling. The Manometer does indicate something is afoot, whereas the Gas Analyser just says "yes ok for emissions test". 

and not forgetting the idle settings stand all the way up to 3,000 rpm through the six progression holes.

edit: is that a "no" really?

My carbs have 160 Main Jets, which means wide open throttle is preferred !

Of course while I want to give effect to what Geoff created decades ago, I accept that in the end we now have technology he could not then (past) compete with, but that is for then (future) and not for now. 

Edited September 9 by anthonym

[anthony1956]

59 minutes ago, OldAndrewE said:

engine load and throttle position will also be variables

I suppose throttle position is usually measured by a T.P.S. 

Engine load.. what's that?

[OldAndrewE]

2 minutes ago, anthonym said:

I suppose throttle position is usually measured by a T.P.S. 

Engine load.. what's that?

What I am saying is that you can achieve a certain rpm at different throttle positions depending on the load on the engine (what work it is having to do).  To achieve say 3000 rpm on a level road will need a certain throttle position but up a steep hill will require a wider throttle opening.  The carbs will be working differently in the two situations

[anthony1956]

ok gotcher. Need a strategy for that; plenty for different load sites (in the real world sense 😉 around here. Maybe there's some standard set .. where to start any idea?

do rolling roads offer resistance to emulate loads?

Edited September 9 by anthonym

[Miker7]

Question how fast does your gas analyser react? My Gunsons gas analyser takes something like a minute (according to the manual) to settle on a reading. I suspect yours is faster but still seconds. If you're driving how can you keep your engine steady state without a rolling road holding the engine back? Lambda sensors react in factions is a second.

In Speeduino and megasquirt world their is a bit of real world tunning done. That uses a combination of feedback loop from a lambda sensor and brain power / seat of the pants. A suggestion was to find a series of long hills so you could hold the car at a series of settings (throttle position, speed, gear) for a fixed period of time (seconds) to give the software a chance.

Once you had a base safe ignition map and fuel tune you drove it to get more logging.

Without the ECU logging lambda which changes in factions is a second, and assuming your gas analyser is multiple seconds to react I don't think you can do this.

[rj]

1 hour ago, Miker7 said:

Question how fast does your gas analyser react? My Gunsons gas analyser takes something like a minute (according to the manual) to settle on a reading. I suspect yours is faster but still seconds. If you're driving how can you keep your engine steady state without a rolling road holding the engine back? Lambda sensors react in factions is a second.

In Speeduino and megasquirt world their is a bit of real world tunning done. That uses a combination of feedback loop from a lambda sensor and brain power / seat of the pants. A suggestion was to find a series of long hills so you could hold the car at a series of settings (throttle position, speed, gear) for a fixed period of time (seconds) to give the software a chance.

Once you had a base safe ignition map and fuel tune you drove it to get more logging.

Without the ECU logging lambda which changes in factions is a second, and assuming your gas analyser is multiple seconds to react I don't think you can do this.

That's why you normally use a wideband lambda sensor rather than a gas analyser.

[anthony1956]

1 hour ago, Miker7 said:

series of long hills so you could hold the car at a series of settings (throttle position, speed, gear) for a fixed period of time (seconds) to give the software a chance.

loads of those here

It seems fast enough, I'll get some time next time out (when it stops raining). Of course there's no cat to slow things down, it feels like I twiddle, look up and it's responding

Interesting to hear the Gunsons experience; no regrets now.

Edited September 9 by anthonym

[anthony1956]

2 minutes ago, rj said:

That's why you normally use a wideband lambda sensor rather than a gas analyser.

MY R500 has one of those fitted; seemed like a good idea. What is it doing for me?

[rj]

It's used for AFR measurements.

In reallity the lambda sensor determines the remaining oxygen content in the exhaust emission.

[anthony1956]

Just now, rj said:

It's used for AFR measurements.

In reallity the lambda sensor determines the remaining oxygen content in the exhaust emission.

yes I know that.

What are the benefits of having one in one's car?

[anthony1956]

evening 7wow 

 

[7 wonders of the world]

A wideband used on a rolling road will accurately measure enablint a plot of AFR over a full range of rpm (normally 2500 to max rpm) at various throttle positions with a repeatable load, thus enabling accurate comparison between runs, you can also make timing adjustments and monitor the effect of AFR to, naturally power and torque curves are available for each run enabling direct comparison of the effects for any changes made.

You will need to have some form of tps even in mechanical form a pointer and scale on the throttle linkage type set up obviously you will need to run bonnetless in order to see this.

You only have the option of running at constant rpm runs this is usually easier on a long flat road using left foot braking to maintain

Your post regarding the colurtune adjust being accurate  up to 3k is not correct, the colour tune adjustments you make are only on idle as once off idle you move onto the progression holes over which the idle screw has no control. these are controlled directly by the idle jet and tube combination, prior to moving onto the main jet air jet and emulsion tube.

[anthony1956]

9 minutes ago, 7 wonders of the world said:

Your post regarding the colurtune adjust being accurate  up to 3k

What I meant was setting it at idle determines what it's going to do in progression.

Anyway I haven't tried Colortune yet, what can you give me by way of suggestions for first time use?

I don't think my road use is going to be as accurate and measure as you would think is good. However, it will match how I drive the car. As far as I can tell all we are looking for is not running too lean.

Edited September 9 by anthonym

[7 wonders of the world]

That's not correct the idle setting is adjusted via the 4 idle screws these alter the orifice thro which the idle fuel flows - the screw simple impeds the flow.

Once off idle the progression hole transition the fueling up to the mains, these progressions holes are directly fed and controlled by the physical idle jet size inconjunction with the idle emulsion tube the only way to change progression fueling is to change either of these physical items.

[anthony1956]

5 minutes ago, 7 wonders of the world said:

That's not correct the idle setting is adjusted via the 4 idle screws these alter the orifice thro which the idle fuel flows - the screw simple impeds the flow.

 

yes of course it is.

I can see you think I don't understand. I do. And I have been tuning these things now for several weeks. Definitely an expert novice.

They are not idle screws, they are mixture screws for the idle mixture, followed by air comp screws. All of which has been screwed by pump jet dribble.

Edited September 9 by anthonym

[7 wonders of the world]

Just for clarity I will put it another way.

Your could that idle jets size 40 and achieve perfect idle settings.

Likewise you could have idle jets of size 60 and achieve the same idle settings and readings..

However the light throttle running and transition once on the progression holes will be quite different.

 

[anthony1956]

cool. 

I'm working with what I have until I see what Geoff (my car's builder in 1987) designed is maxed out. 
In any case she goes like a rocket and makes a wonderful noise (which is why I never sold her). 
My purpose is to reverse engineer Geoff's thinking processes. 
The rolling road stuff is just a bit of fun and now I am curious, but there are limits how far I am willing to go. For example not willing to drive bonnet off etc. 
So within the constraints I set will lie the experiences I have with this car. 

All that said, she has lasted just fine for nearly 40 years so unless I do something real stupid, like bend valves when changing or correcting cam belts (just survived that with continuous 24/7 revilla support), she should be ok. We all remain perplexed how i managed to get the crank to skip two notches backwards (hence a new belt).

In fact I will be very interested to hear your list of possible causes of pump jet dribbling, please? It's caused me no end of grief and ruined my mixture setting session the other day; she's overdue for emissions test, which is easy enough to do usually.

Edited September 9 by anthonym

[7 wonders of the world]

The air compensation screws are to enable balancing of air flow on idle.

Did you have check and/or reset each throttle butterfly individually in their respective bores prior to fitting?

Have your sorted the pump jet issue..? Worth checking the orifices don't rely on the stamped numbers as I previously mentioned many get reamed or redrilled

[anthony1956]

Yes understood air screws.

I did not fit the butterflies these were professionally done when the carbs were rebuilt/refurbed..

I don't think anything has been re-drilled etc. In any case how would I discern this?

The pump jet dribble MAY now be sorted, I just need the rain to stop as I am not allowed by wicked neighbour to make car noises near the building. 

If it is then two possible things cured it, one or both, correcting the float height using blow and set instead of 15mm of ruler (did that several hundred times) and maxxing out the adjusting nut up its thread. Mind crushing detail is in my blog, but the latter sentence saves reading it.

I have asked the carb rebuilder if he would have replaced the float chamber needle valve parts, but no reply yet. In any case they work when blown through and I lift the float by hand. And the fuel height is 27mm i i r c 

Edited September 9 by anthonym

[7 wonders of the world]

Re jet sizing you can get a set of fine sized needles rather like a selection of drill bit shanks to check each jet. Think feeler blades in principle.

Redrilling saved the cost of new jets and reaming was common to add a little fuel to one cylinder to reduce spitting.

Jet gauges here on ebay

Item 304958193391

You can also get jet drills and jet reamers too.

 

[anthony1956]

cool, jet gauges will do nicely, not least as the jets are VERY difficult to read. I won't be drilling etc