Engine build progress
- Thread starter TTM
- Start date
TTM
Well-known member
After a winter that felt insufferably long I resumed doing some small works on the car, first being the addition of a secondary oil cooler a couple of weeks ago. Oil pressure at tick over when the engine is fully warmed-up seems to be slightly higher than before, so it seems to work. For a reason I will expand on in the coming weeks, today I replaced the spring in the wastegate. I had been running a 1.0 bar (blue spring) since I installed the Tial several years ago, with the 2.5 engine. I am now running a 0.7 bar (yellow) spring. Since I could now run less than a peak boost pressure of 1.0 bar I figured I would add timing advance, so I switched to the standard, maximum advance map on the Vitesse chipboard, and set the FQS back to position 0, as I had been running to FQS#4 (=less advance) because of chronicle and significant knock activity. So I started to drive on the wastegate spring, running 0.7 bar max. Did minor adjustment to the fueling, everything was fine, the engine was feeling so much together, smooth and all, though not exactly "fast", so I figured I would progressively increase the duty cycle of the EBC (=increase boost) until I would reach this knock threshold I had been suffering from since day 1. So I added boost, 0.8, 0.9 then 1.0 bar and guess what? I could not provoke any knock. Surprised from this finding-on I began to drive like a complete nutter, pedal to the metal at each occasion, supposing that behaving a bit violently would eventually bring back knock, to no avail [
] Now the question to the experts : how could the stiffness of the wastegate spring have any relation to knock activity, if any? A wild guess is that the boost solenoid I use might be working more efficiently with a softer wastegate spring, and the stiffer spring might have been causing invisible boost spikes until peak boost was reached, causing some havoc during combustion as boost rose? I plan to spend some more time on the road tomorrow in the hope that what I experienced today wasn't just an accident, but today was probably the most enjoyable time I've had driving the car since I bought it. That knock might have eventually disappeared has suddenly bled a lot of frustration!
] Now the question to the experts : how could the stiffness of the wastegate spring have any relation to knock activity, if any? A wild guess is that the boost solenoid I use might be working more efficiently with a softer wastegate spring, and the stiffer spring might have been causing invisible boost spikes until peak boost was reached, causing some havoc during combustion as boost rose? I plan to spend some more time on the road tomorrow in the hope that what I experienced today wasn't just an accident, but today was probably the most enjoyable time I've had driving the car since I bought it. That knock might have eventually disappeared has suddenly bled a lot of frustration!Does a 1 bar spring open at all before the boost pressure hits 1 bar? I'd guess it probably wouldn't and the crossover pressure with your turbo probaly isn't great enough to shift the spring either so I would suspect your wastegate was clamped shut until the boost hit 1 bar. If this is the case it would almost certainly cause the boost to overshoot to some extent. With the 0.7 bar spring the wastegate will open up a bit before 1 bar boost is hit and damp it out a bit I would have thought so it doesnt overshoot. If your datalogger didn't detect the boost spike it might be due to how much it is averaging the signal.
George Elliott
New member
Thom, great to hear you fixed the knock, its an unpleasant and painful noise. I dont KNOW what fixed it, but i have a couple of thoughts which others may elaborate on, some is experience, some is theory, some is a measure of understanding/learning about how these engines operate. My SPS engine had covered 1900 miles when I first drove it, it now has 12000 miles and is not so tight. I warm and cool the engine each trip, and change the oil every 6k with Silkolene, the engine uses no oil, which amazed me at first. As the engine started to free up, it would occasionally knock in a 3 gear-sequence drag using WOT. My car has a ball bearing turbo, and Tial W/G. Since fitting a 3" exhaust (and Stage one I/C) this problem has gone away. For the engine to knock, it is either : Lean AFR, High Compression (or boost pressure), Ignition too advanced. OR : a combination of some or all of these. The spring you mention is interesting, because although it does not control the boost, it must influence the movement of the wastegate opening / closing - ie make it quicker or slower. The FQS setting will adjust the ignition timing. The other factor is the Fuelling which is controlled by the TPS and AFM. (You have a vitesse unit I am guessing??). Tom and Graham (Barks and GPF) had / have a thread running which is way above my head, but it described some of the magic involved in fuelling by the ECU. Basically, as someone said, its amazing it works. The other factor to bear in mind is the milliseconds of time separating the firing of the cylinders, - the engine management has to be on the ball to avoid an out of spec combination of Boost/Ignition/fuel to occur for very small periods of time. My guess, taking all this into account, is that your boost was momentarily spiking with the stronger spring, the fuelling is not capable of anticipating the spike, and the ECU is not retarding the timing and the result is Knock. This is made worse in my experience by a 2.5" exhaust and standard D/P, and the other challenge for the engine management is the SPS ball bearing turbo which boosts so much quicker than the KKK. The is trying to manage boost at rpm levels where it used to manage lag. I dont know how much of this you agree, or how much is correct, but some I have proved on my own car, which is why i'm posting a reply. George 944t
TTM
Well-known member
George, yes, the engine was probably developing "localised boost spikes" (at the lack of a better term) that I have never been able to spot. Interesting to hear that the larger exhaust helped in your case, but that doesn't really surprise me. I'm sure your SPS turbo performs at least as well as mine, and as a 3.0 engine will pump a fair bit more air than a 2.5 then yes a larger exhaust is on the must-do list. It always was actually, but I have some significant upgrade in the works meaning the exhaust upgrade will have to wait a bit. Patrick, I compared logs from today with older ones and cannot spot a difference in how boost builds, so probably the Zeitronix lacks resolution or perhaps having it read boost from the usual line to the intake manifold doesn't offer enough precision. Oh well, I already said in the past that I thought I had solved this issue yet it always came back, so I think this will have to wait until some more driving in various weather to draw any conclusion.
Interesting Thom. I have always run the .7 spring in mine but I don't have anything scientific to offer as I have never logged knock. Good to hear that this has helped your long standing problem and the theory above backs up something I heard a long time ago. This was that you should always use the weaker spring as it gave you more control. This was in relation to the Motec boost control PID setup for closed loop. I cant remember the exact explanation but I think the idea is that the boost controller is doing the work not the spring and hence the control is better and more accurate.
This is called overshoot : http://en.wikipedia.org/wiki/Overshoot_%28signal%29 If you cant generate a control action before the signal (boost) reaches its steady state value the signal will probably overshoot its desired value. Ideally you want to open the wastegate to slow the approach of the boost pressure as it reaches the required value. If you have a 1 bar spring the EBC is effectively vitoed until the boost reaches 1 bar, as the pressure used to overcome the spring is manifold pressure which can't be greater than 1 bar until ite reaches the set value (1 bar). Therefore the EBC cannot control the turbo until the 1 bar boost is reached. Effectively I think TTM was relying on the spring to control boost pressure. I suspect if you had d/c the boost controller you might not have seen much difference.
TTM
Well-known member
Thanks for your input. That was a short joy though - knock has been back today, stronger than ever at now only 0.7 bar. Same tank of fuel, same weather conditions. Within the next two weeks the car will be sporting a modern ECU and a new wiring loom. Can't wait to get rid of this headache.
George Elliott
New member
Thom, sorry to hear that, hope you get to the root cause of it soon. Keep us posted. good luck George
Thom, I hope this resolves your issue. Hopefully you have a good map to start from as in my experience it takes a long time to get a fully functional map with a standalone that works as well as a factory map. i.e. all temperature conditions, cold start, hot start, idle quality etc etc.
Nice! Sounds like hours of fun![ORIGINAL: TTM The ECU upgrade is complete and the engine runs, well at least it idles, crystal clear [
] Need more details & pics!! Also where did you get the intake pipe from the filter cone?
DivineE
New member
Superb! Let us know how it goes. I will be particularly interested to know how the 3.0 SPS turbo equipped engine runs with stand alone ECU as they all seemed to have a few small character floors with the standard set up. You mention the idle. The biggest thing I've noticed with stand alone, compared with all the other 944s I've owned and driven is the smoothness in the way the engine runs. At idle and when holding at high rpm say 5000rpm the engine is perfectly silent and smooth like a new car where it just felt a bit more lumpy and organic in traditional 944's. No idea why an ECU should change what I thought was a physical characteristic of the big four pot but it's definitely, noticably different. Regards, Ben
George Elliott
New member
very interesting, standalone ECU systems seem to be making some progress, I always regarded them as Nick describes them. Good to hear you have tickover Thom. Barks, I have another request for your ECU[], pls over ride the feature where the engine wont fire a spark until it reaches 285rpm. I feel sure the 944 engine would churn for less time without this. It sounds awful compared to a 911. George
], pls over ride the feature where the engine wont fire a spark until it reaches 285rpm. I feel sure the 944 engine would churn for less time without this. It sounds awful compared to a 911. GeorgeThom, as ever, it looks like a very neat and tidy installation. What ECU did you go with? Do you have any concerns with heat where the coils are located? Have you put in any heat shielding? I only as as mine are in a similar position. Hope this clears up all the minor but irritating issues you've had. []
]
sawood12
New member
Isn't that a safety feature to allow the engine to build up some inertia against the springiness of the compression before ignition to prevent it firing and trying to turn it in the wrong direction and/or over-stressing the con rods?ORIGINAL: George Elliott very interesting, standalone ECU systems seem to be making some progress, I always regarded them as Nick describes them. Good to hear you have tickover Thom. Barks, I have another request for your ECU[
That sounds very plausible. On a lot of older / more interesting aeroplanes you have separate starter and ignition switches for the same reason. The drill is to get the engine turning before you switch on the ignition, to prevent it trying to start backwards and possibly breaking something.
TTM
Well-known member
The ECU is part of a complete "plug & play" kit supplied by Vic at Pauertuning. The kit revolves around a Vi-PEC ECU and includes : - a custom, ready-to-fit wiring harness that connects to all original sensors - a J&S Safeguard Interceptor that triggers down ignition timing/boost through the ECU when detecting knock - Bosch EV14 high impedance injectors, 1200cc units in my case as I plan to run E85 occasionally and this is a 3.0L engine - 4 GM LS2 coils & a hall effect camshaft position sensor for sequential mode, with CNC-machined brackets - a base map to start the car and work from there. - at last but not least, technical support! I added several options such as a PLX DM-100 52mm gauge that allows me to monitor in real time up to four datas, namely AFR, MAP, EGT and oil temp. I'm yet to modify the oil drain plug though. Using the ECU only for real time monitoring would have meant using the laptop constantly, not the best for ergonomics, and the PLX gauge allows for monitoring up to 4 variables at the same time. Boxes and wiring harness :
The 5 PLX boxes are Wideband/Boost/EGT/Oil Temp/52mm gauge. The ECU has its own internal MAP sensor.
Placing the new harness in the engine bay :
It's been years since there has been so little fuss in the passenger footwell. The plate holding all the boxes actually attaches to the original black plate holding the original DME/KLR boxes :
"Sly, will that be enough for the weekend?" :
Sly doing wonders again as he rearranges the layout of the oil catch can :
I chose to have the brackets anodised in gold as the camshaft housing is black and the usual plated parts such as the fuel damper (etc) are still quite shiny to this day. I was hoping to give a black/gold Lotus JPS effect to the engine bay, I hope the end effect isn't too garish, what do you think? Even though I have been playing the Vi-PEC software since I ordered the kit several months ago there is a fair bit of "homework" to do. Good thing when going for a kit like this is that all parameters such as dwell settings, injector dead time values and a myriad of others I have still little idea about are all set and basically I shouldn't have a lot more to do than having the car on a dyno to get the timing right to the edge of the knock threshold. I would never have moved to a standalone ECU if I had to start with a blank map, just too much hassle for a non-professional like me, and it does feel good to have an engine idling superbly just after loading to the ECU a calibration file just received by e-mail []
]
Posts made and opinions expressed are those of the individual forum members
Use of the Forum is subject to the Terms and Conditions
