Page 3 of 4

Re: Distributor question

PostPosted: Thu Nov 03, 2005 2:38 pm
by type26owner
Hi Martin,
I was trying to apply some rational reasoning into the timing issue. No were did I mention any values which I thought would be okay. It would have been nice if you had asked me first before appyling your own numbers and then judging the situation by those. :wink:
The distributor is a mechanical bodge, albeit a relatively good one.

Here we are in agreement. It's prone though to exhibit quite a lot to stiction and hystersis if it's lubrication neglected.

Rohan,
You see the logic here obviously. You're the special case were a racer does care about extracting 100% of the power from the engine at all times.

I've been doing quite a bit of experimenting with the timing curves but I lack a dyno to measure any torque values. When the carburetion is delivering a steady mixture and no matter what the operating conditions and 10 degrees retarded the engine still runs just fine. At least I find this to be the case in the retarded direction though it's noticable it's down on power at WOT. The challenge is to get the engine to not ping or detonate with the lower octane fuel available today and stay close to the MPT. It helps a bunch to have a knock sensor!

distributor timing

PostPosted: Thu Nov 03, 2005 3:33 pm
by sk178ta
My distributor was worn and the clamp b***ered so I went for the Bestek complete electronic kit. Subsequently a tight bend followed by a steep hill, nearby, now required a drop into second gear when previously it stormed up in third. So I had the carbs. (dellortos) and timing set up on a rolling road by J.P. Autotechnics. They commented " timing advance seems a little low on and above 2500 rpm". Besteks pre-settings are 14 degrees advance @1000 rpm, 21 @2000, 26 @3000, and 30 degrees from 4000rpm upwards.
What do you think and why am I down on power? There are no gains higher up the rev. range.
Jim

Re: Distributor question

PostPosted: Thu Nov 03, 2005 4:52 pm
by type26owner
Jim,
Bummer! Don't know what to tell you to do exactly. A wild guess is it's a carburetion problem. Knowing the parameters of a Weber and surmising the Dellorto is similiar it suffers from the lack of much vacuum signal from the auxiliary venturi too. Suggest you do my rpm flow test to establish when the mainjet circuit starts to kickin. Likely you've got a flatspot.

Everyone,
There's a fair amount of folks out there who are probably running their timing by 8 degrees retarded more than what they think. The 4" steel front crankshaft pulley that Dave Bean sells is for a Cosworth I suspect. The timing mark is in the wrong place for the Lotus engine. The engine runs fine but is down on a little power is all. If you've got one of those pulleys then you must set the engine to TDC and remark the pulley!

Re: Distributor question

PostPosted: Thu Nov 03, 2005 5:36 pm
by M100
type26owner wrote:Hi Martin,
I was trying to apply some rational reasoning into the timing issue. No were did I mention any values which I thought would be okay. It would have been nice if you had asked me first before appyling your own numbers and then judging the situation by those. :wink:
The distributor is a mechanical bodge, albeit a relatively good one.

Here we are in agreement. It's prone though to exhibit quite a lot to stiction and hystersis if it's lubrication neglected.


Keith, the basis of my argument was seat of the pants experience over a generic set of figures that might not be directly applicable to the Twincam especially given its strange combustion chamber (any idea what reference engine they did use in that book?)

With the (at least) seven different advance curves that the factory used and the fact they fall into two distinctive groups; the rapid advance to 2000rpm with a flatter slope to 6000rpm used on the early cars, and the steady slope *from* 2000 that flattens out completely at 5000 on the later ones, with a spread in total advance at 5000rpm of 14 degrees shows a huge number of variations have been adopted. Ok so there were changes in the cams and the compression ratio, but given the variations in fuel and the temperature ranges the Elan went to I doubt that few if any of them were the optimum solution then, even less so now.

With no other changes I do know I can feel by the seat of my pants 2-3 degrees base timing error on my "other Lotus" In terms of peak power measured on a dyno it drops by around 8% (corrected back to back tests) Similarly moving the timing only a few degrees at lower revs totally transforms the car (this was done to simulate the potential of a granny mode) But its a modern engine and its timing sensitivity might be much greater than the twincam.

Given the number of advance curve sources floating around I think I might be able to add yet another source. Somewhere I have a book that lists the Lucas distributor specs across their whole range. I know some twincam specific ones are missing (there is I think only one set of figures for a 41189) but most of the other twincam advance curves are in there.

Re: Distributor question

PostPosted: Thu Nov 03, 2005 7:26 pm
by type26owner
Martin,
I wonder why there are so many variants too? Perhaps it was to fix another problem area like the carburetion for example. I get ordered to do damage control for other folks engineering stuff frequently. Supposedly they were experts but when put to the test they demonstrated it wasn't so. I would use that grain of salt to question whether the engineering was done correctly at the factory or not. Blindly trusting this stuff is just asking for the KoolAid. :lol:

Re: Distributor question

PostPosted: Fri Nov 04, 2005 3:03 pm
by type26owner
Jim,
To give you a benchmark for performance of a well tuned twinkcam. There's this hill on my way to work which has about a 5% grade. Lots of people know it if you've done the Broines reservoir tour with Mike Ostrov before. Anyway at the bottom I slowed the car down in fourth gear to 2000 rpms and went WOT. It pulled cleanly up the grade with no problem. At 4500 rpms the torque increase is a thrill to experience everytime. I love the sound of a twinkcam at full song!

Smacked into two deer out here last year with my Elan though on two different occasions and that kinda puts the fear in a guy. :cry:

distributor

PostPosted: Sat Nov 05, 2005 9:40 am
by sk178ta
It`s a bit far from Derbyshire but I get your drift.........but what is WOT?
Jim

Re: Distributor question

PostPosted: Sat Nov 05, 2005 1:03 pm
by stuartgb100
Could be wide open throttle, maybe!

Here's a link to an interesting technical article on tuning Lucas distributors:

http://www.teglerizer.com/mgstuff/lucastuning.pdf

Hate to admit it, but this site has some rather good articles.

Re: Distributor question

PostPosted: Sat Nov 05, 2005 3:41 pm
by type26owner
Stuart,
There is an order of events here which must be obeyed to get the maximum performance from the twinkcam. First the carburetion must be optimized to achieve the desired best power mixture at all times. The AFR (air/fuel ratio) determines the burnrate of the mixture which remains constant if it does not change value. The ignition timing is then advanced to position the peak pressure at the best angle of the crankshaft. The timing is slaved to the burnrate period. Since I've discovered that Lotus muffed the Weber jetting and got the carburetion all wrong, it follows likely none of their timing curves are ideally correct either. I'm forced to optimize this stuff on my own now and it simply can't be done without a dyno to measure the torque at every rpm. :cry:

Having the normal idle mixture at 22:1 forces the timing to start at the wrong amount of advance to begin with. Get that wrong and the whole damn curve is skewed I'm guessing. Since I've set the idle mixture to 12:1 I find I need less static advance but this is just based on rotating the dizzy to get the strongest, smooth idle. Sitting at the MBT at idle should make it purr.

All in all it's not a hopeless cause afterall. I bought a digital air/fuel meter which will record ANY 5vdc signal. It's obvious to me one can cobble together a simple device which would measure the relative force of the torque. As the reaction of the engine against it's compliant rubber mounts it leans over and this motion can be measured with linear displacement transducer. With this info I can quantify the affect of the timing to achieve the MBT.

What a pain to have to do this stuff over again that Lotus should have correctly to begin with if money were no barrier! However, I can well imagine the stressful work environment at the time to get on with it and ship the customer the cars to pay the bills. I see shoddy work all the time because there simply was no time to devote to get it right. They'd have gone out of business pronto!

Re: Distributor question

PostPosted: Sun Nov 06, 2005 9:43 am
by stuartgb100
Keith,

I started this thread in the hope of finding a single, clear solution to the advance curve issue ...... something that is clearly not to be achieved!

Nevertheless, there's been much to learn, and in that respect it has been a useful experience ... for me, at least.

I'd certainly like to explore the AFR possibility, so any links etc that you could post would be much appreciated. Perhaps a few of us could 'club' together to cover the initial cost (if anyone's interested please PM me).

I'm sorry you don't have a personal dyno lying around ..... perhaps this link will get you started?

http://www.niksula.cs.hut.fi/~mdobruck/ ... 0/dyno.htm

Please do let me know how you get on (vbg).

Regards,

Stuart.

Re: Distributor question

PostPosted: Sun Nov 06, 2005 2:53 pm
by type26owner
Stuart,
I started this thread in the hope of finding a single, clear solution to the advance curve issue ...... something that is clearly not to be achieved!

We have a common goal and I did not mean to trounce upon your enthusiasm to understand this stuff either. However, I'm doing it carefully using all my engineering skills to sort out and understand all the unknowns along the way. I'm just not there yet but it's been fun to push the envelope a bit for my own good for a change!
I'd certainly like to explore the AFR possibility,

http://www.tuneyourengine.com I'll help anyone on this forum to get through the learning curve part of using one of these devices. It renders a Colortune to the level of a novelty toy. No way will a Colortune do what this state-of-the-art digital meter will do for you. Because trying to tune a DCOE is hopeless without one of those wideband O2 sensors I've started posting the technical stuff I've learned in that user forum instead of here.

The reason Mr Vizard did it that way is because the Colortune cannot be observed while the car is in motion. The real world problem with a real dyno is the car is not pushing through the air and pulling along it's real amount of mass either real time. My scheme to add a transducer potentially blows doors off the data one gets from a stationary dyno. Bet I can make it with at least the same resolution too. I have not figured out a simple way how to range through and find the timing whichs give the highest MBT at any particular rpm. For me to do this I'll need to see the extremes of the envelope so that means I'll have to know what gear I'm in too. I'll have to invent another sensor for doing that. I already have the 5vdc power supply to run all these sensors.

Re: Distributor question

PostPosted: Mon Nov 07, 2005 5:32 pm
by M100
type26owner wrote:The reason Mr Vizard did it that way is because the Colortune cannot be observed while the car is in motion.


Of course it can, you just need to be brave (stupid?) and hang on real tight :-) Somewhere in one of my books there is a picture of a Lotus single seater during the 1960's driving round a circuit with the mechanic hanging on while observing the engine for some problem. I've heard of cine camera's being used for monitoring Colortune on road cars many years ago. A wideband O2 meter makes it much easier though!

Keith have you thought about getting a GPS based data logger that will enable you to accurately measure vehicle performance changes? Something like the Race Technology DL1
http://www.race-technology.com/WebPage2 ... 1Home.html

Re: Distributor question

PostPosted: Mon Nov 07, 2005 8:24 pm
by type26owner
Martin,
I should just buy the LM3 unit from InnovateMotorsports. It has dual axis accelerometers built into the unit. The point is I need reliable feedback to plot the timing graph at 500 rpm intervals to stay at the MBT at all times. The question is will the accelerometer provide high enough resolution to map out the timing and will the time response be fast enough. I won't know for sure until I try it but I have my doubts since I have some prior experience using these type of devices before. Can always add the transducer on later if needed. When I first bought the LM-1 these units were not available yet. I still need the gear selection sensor though.

The real tricky part to mapping out the timing is it has to be done in a single day's effort at the racetrack. My tack will be to set the dizzy up with a know advance curve and run through the rpm range with a particular static setting on each recording run. Changing the static setting setting by increments of two degrees on each recording session will build up a dataset that one can look at a particular rpm and find where it pulled hardest. The timing curve is generated by connecting all the dots or the advance points where the MBT was achieved at each rpm. I can't see how it can be done any better than that.

Re: Distributor question

PostPosted: Tue Nov 08, 2005 1:33 pm
by steveww
I can't see how it can be done any better than that.


Try digital mapped ignition :wink:

For some insteresting advance curves have a look at www.TDCperformance.ca

BTW They have a Lotus twinc version on the way. The chap who runs this company has a Lotus Cortina that is the test bed.

Re: Distributor question

PostPosted: Sat Nov 12, 2005 4:08 pm
by type26owner
Unfortunately the accuarcy of the LMA-3 accelerometer unit is only 1_part_in_50. Guessing I'd need at the very least 1_part_in 100 to do it with the precision required. That won't be adequate to map out the MBT since I probably can't cross over the maximum value of the bell curve by advancing the timing by much before causing the engine to ping or detonate.

Steve, the timing curve is independant of the ignition system used. No matter what type of igntion they all will want to be set to the same ideal MBT maxed out curve.