http://www.delphi.com/pdf/ppd/pwrtrn/gas_ignics.pdf

Didn't Delphi just declare backruptcy??? Guess the future of ion ignition didn't prove quite so lucrative ...

Too funny!! Finally a guy that can read past the post above his and use a search.

Here is where I posted that first:

http://forums.pelicanparts.com/showthread.php?s=&threadid=248643&perpage=20&pagen umber=5

This is where the poor dog died an unforgiving death.

Good thing I already got my Delphi software......free.

Muhahahahaha!

JC's above thread leads to many threads by the same author. Much of it may seem remote but in can be connected to a land engine. I started by copy&paste relevant sentences in a separate file. You can not operate a land engine by following exactly what an aircraft engine needs but you can connect the dots if the reader studies the game.

I also use JP Instruments operation manual for the EDM-700 gage. It can be downloaded at its web site.

Sky Ranch Engineering Manual 2nd ed , it's 500 pages of intense info. Engine Performance chapter is 75 pages. Troubleshooting chapter is another 75 pages. It's also inexpensive.
from Sacramento Sky Ranch Inc.
916-421-7672

How To Build Horsepower Vol 1, How To Build Horsepower Vol 2, by David Vizard. These are 2 hot rod store $18 ... 8" X 11" paper books from Summit.
Vizard also has many other technical writings. His 2 books are written around carbs so a reader will have to connect the dots again. He has great info on the Individual Runner, IR, carb.

also personal experience & text from working on boat engines, etc. Boat engines are always under load.

Meanwhile my act involved around tuning to a EGT around 1,300-1,350F measured about 1" from the head. The consistancy between each cylinder is important as we are really tuning 6 indivudal cylinders. This deviation has been tuned to a lower reading than good fuel injection deviations. Paul Gagliardi loved it.

For entertainment.
CHT has a lot to do with hp output. My CHT increased around 25degF, generally, when I changed the cams and induction from 1977 stock CIS and cam to carbs and E-cam.

Grady,

Bosch is not the only company using the term accelerometer with reference to knock sensors ... GM was using such terminology as early as 1976 in a patent application for engine timing control with knock limiting:

http://not2fast.wryday.com/patent/p4002155.pdf

http://not2fast.wryday.com/patent/p4384473.pdf

http://not2fast.wryday.com/patent/p4111035.pdf

http://not2fast.wryday.com/patent/p4106447.pdf

The concept of a retrofit system of automatic retarded ignition timing based on a knock sensor seems feasible for either 3-pin or 6-pin Bosch CDI ignition systems, though the 3-pin unit seems an easier job, and given that there are more high-compression engines using the 3-pin units [C3 and ROW SC engines being the rare exception] ... maybe there is more of a market for such a system on the early 'S' engines with 9.8:1 compression!

Here is a Texas Instruments technical report from 1995 about using DSP to process knock sensor signals:

http://not2fast.wryday.com/ignition/spra039.pdf

Bosch P/N for the knock sensor used on the earlier Saab Turbos is 0 261 231 006. Readily available, cost about $40. It's a simple thick disc, maybe 1" diameter, with a hole (maybe 10mm) in the middle to bolt it to the block, and two terminals. It is piezo BTW. If there was a stud on the 911 case where it would fit, it would be a piece of cake to bolt up.

David Vizard teaches at the U of N Carolina. I mentioned him as the author in the previous page. I'm not sure of his title but he's always doing dyno and fuel experimenting at the school. Induction seems to be his pet expertise. I have his office phone number if you want an e-mail. I really don't have enough brains to carry on an educated conversation concerning your project idea with him. He could probably become a serious part of the project as that's the impression I have about his act.

I have notes from around 50hrs of trials using various jets and idle settings and switching from 36 venturis to 34. I even did rides knowing settings weren't correct just to see EGT/CHT reactions to a bad setting. My formula was to set a good base line in 5th gear, maintaining a steady throttle for a few miles between 3 - 4.2k rpm, and then hope that the rest would fall into place. It worked. I haven't even fine tuned the accelerator yet as the improper volume became good enough till I catch up to on other 911 projects.

"I’m willing, how about you and others? "
------ I'm retired, have zero responsibilities, and have plenty of time. Tell me to do something and I'm moving before you finish the sentence.

please excuse me for hoping in on this thread, im not pretending to be anywhere near as well educated on this subject as some (if not all!) of the people here but what are those peoples views on boxes such as the j & s safeguard? are these sort of units sub standard or are they just a bandaid?

http://www.jandssafeguard.com/

Vizard literally wrote the book on the BMC A-series engine, 850cc-1275cc as used in Sprites, Midgets, etc. I built a 948cc engine for a Bugeye Solo I car following his guidlines, then later a 1340cc 12.5:1 engine for the same car. He's big on research and testing, he's tried and tested everything on those engines. This particular book delves deeply into intake tracts (including carbs), combustion chambers, exhaust systems, cam timing, ignition advance mods, etc. I spoke with him once on the phone (he was associated with a CA-based specialty parts supplier, I think he taught in CA for a while?) and he seemed like a friendly, helpful guy.

I am wanting to use knock sensing with my upcoming Megasquirt/EFI conversion. It seems a likely place for the sensor placement could be in the center of the case, under the shroud. Thoughts?

Andy, to answer your question on the J&S, it's a great unit. I have (as well as many of my /4 customers) used the unit and it delivers as promised. The only drawback is the price, but then again there are no other self contained knock sensor/retard systems available that I am aware of.

Here is the info from SafeGuard that Andy (Pelican adomakin) provided.
I slightly edited it to correct spelling, typos and punctuation without changing the content. I left some when it could affect the meaning.


”Shown is our Marine/Racing version. Features include high energy inductive discharge ignition and adjustable rev limiter. This system is ideal for turbo, supercharged, high compression, or even nitrous applications. An optional display shows the amount of knock retard.

Using a single knock sensor, the system detects the onset of detonation and retards the timing on a per cylinder basis. A mode switch lets you select a maximum of either ten degrees or twenty degrees of knock retard. In the ten degree range, the unit retards one, two, or three degrees per ping. Double that for the twenty degree range. The system is always trying to re-advance to stock timing. In the ten degree mode, it re-advances at the rate of one degree every twenty revolutions.

The system does not need a cam or crankshaft reference to determine which cylinder to retard. The unit is programmed to "know" that the knocking cylinder is the one that just fired, and that it won't fire again for two more revolutions. When the knocking cylinder comes around to fire again, software dials in the retard amount for that cylinder. It does this as each cylinder goes by, building up a different retard amount for each cylinder.

The unit has a high energy ignition with constant energy dwell controller. The system monitors the coil current, and adjusts the dwell time to achieve seven amps of coil current. Compared to a GM HEI, this is a 62% increase in energy stored in the coil. The system can also be used to trigger an MSD.

The Marine/Racing version features an adjustable Velvet Touch rev limiter. At the selected limit, the system fires two cylinders, then skips one, and repeats this sequence. For six cylinder engines, the sequence is fire three, then skip two. If the engine goes 100 RPM above the selected limit, the system will skip as many cylinders as needed. A test point is provided to adjust the limit, scaled at one volt per 1000 RPM. For example, if you want to set the limit at 6250 RPM, adjust the knob so that the RevTestPoint reads 6.25 volts.

Standard automotive units are also available for single coil foreign and domestic vehicles as well as multiple coil applications, including Honda, Miata, Subaru, Eclipse, Toyota.

For engine builders and dyno operators, check out our Knock Finder Dyno Timing Controller.


J&S Knockfinder Highlights 2001 Superflow Engine Conference
Harold Bettes of Superflow showed the J&S Knockfinder Dyno Box at the 2001 Superflow Advanced Engine Conference. Already in use by top engine builders the unit sparked a lot of interest especially in the top ranks of NASCAR where every horsepower is hard to come by and where the schedule to meet races and develop dozens of engines is relentless. Major teams said the unit would save them valuable time considering their responsibility to develop so many combinations in that it was easy to hook up and the slide pots gave them instant control over each cylinder's timing.


New: One Channel/Two Channel Ultra SafeGuard unit for imports
New 1 Ch/2Ch "combination" unit for imports with one or two coils. UltraSafeGuard versions feature boost retard as well as knock retard, adjustable MAP Limiter, TPS trick, VTEC squelch


New: Four Channel Ultra SafeGuard unit
New Ultra SafeGuard Direct Ignition System for newer vehicles running two, three, or four ignition coils are available. The four coil version can be used on the '96 thru '98 Mustang. These units feature boost retard as well as knock retard. Closed loop timing control allows you to run aggressive timing and not risk blown head gaskets or broken pistons. For full details contact J&S or go to our Products page for more information on what a J&S Ultra SafeGuard can do to maximize your vehicle's potential and protect your investment!

(continued)

(continued)

Tech – Detonation
Background
Engine Knock-that annoying rattling sound that sometimes comes from under the hood of your pride and joy, is a killer. We have all heard the stories of blown head gaskets and broken pistons. Maybe it has already happened to you, too. But just what is knock, sometimes also referred to as detonation?

The knocking sounds you hear are the cylinder walls set into oscillation by intense pressure waves, caused by abnormal combustion. Normal combustion is a controlled burn that starts from the spark plug and spreads outward, causing a pressure rise in the combustion chamber. This pressure is then converted into torque on the crankshaft. Ideally, the peak pressures will occur about ten to fifteen degrees after top dead center (TDC), as the piston is on its way down.

Detonation is a form of abnormal combustion that starts off right, but at the last millisecond, something goes wrong. The remaining air-fuel mixture, called the "end gas", explodes all at once, instead of burning in a controlled way. Resultant engine damage is caused by an instantaneous pressure rise that can exceed 1500 psi. This is more than double the normal peak combustion pressure, and will blow head gaskets, break piston ring lands and hammer the rod bearings. Another form of damage seen is that the tops of the pistons will be eroded and can even melt.

High octane fuels are resistant to detonation because they contain compounds that slow down the chemical chain reaction we call combustion. If left unchecked, these chain reactions would quickly escalate, resulting in increasing damage to the piston and other engine components. All fuels, regardless of octane, have a knock limit. This is reached when the temperature of the "end gas" reaches an autoignition point. Combustion chamber designers use high swirl inlets and large "quench areas" to fight this "autoignition" problem. There are other factors beyond these mechanical design features which influence "end gas" temperatures. Some of these are: (1) Intake charge temperature, (2) Coolant temperature, (3) Compression Ratio, (4) Boost pressure, (5) Spark timing, (6) Air-fuel ratio, and (6) Humidity.

An increase in compression ratio, boost pressure, or spark timing will increase peak cylinder pressure, which in turn raises the "end gas" temperature. Higher inlet and coolant temperatures also increase the "end gas" temperature. Richer mixtures can be used to cool the charge. At some point beyond about 10:1, however, will again increase the tendency to detonate. A decrease in humidity will also tend to increase detonation.


Solutions
Some things you can change, and some you are stuck with. Obvious things to do are get cold, fresh air to your air cleaner, use the best form of charge cooling (intercoolers) you can afford, and work on your cooling system to bring the temperatures down. Get an Air-fuel meter (O2 meter), and if necessary, install a fuel enrichment device. For turbo/supercharged engines water-alcohol injection can be very effective, but the volumes required mean constant refilling and maintenance, so in the long term it is a specialized solution that has a lot of drawbacks.

Spark retard, within limits is the most powerful means of controlling detonation. Traditionally, this has been a manual affair by simply cranking the distributor back "x" degrees, Some electronic devices can retard timing in relation to manifold or boost pressure and other devices offer simple manual knobs you can alter the timing with if an audible "ping" is heard. Many newer cars are quipped with knock sensors that retard the timing when detonation occurs. This form of closed loop spark control where the vehicle's ecu automatically retards all cylinders when detonation occurs, a situation we call "retard all". On a select few modern engine controllers (Porsche, Corvette) the microprocessor is programmed to retard individual cylinders as knock or detonation occurs. The problem with either of these approaches is twofold: First, the OEM carefully maps a complex set of spark curves and only allows a small degree of retard to take place, typically 4 degrees; Secondly, these ecus are highly specific and are not readily adaptable or programmable for other applications.


J&S SafeGuard versus the Detonation Dragon
J&S Electronics has developed the SafeGuard which uses a knock sensor to provide feedback to it's microprocessor which, in turn, controls the timing for each cylinder on an individual, optimized basis. Firms like DINAN Enginneering, Kenne Bell, and CarTech have successfully used SafeGuards in their installations. DINAN used SafeGuards in their high dollar integrated BMW turbo kits which set very high standards for performance and sophistication. Kenne Bell has used SafeGuards in Syclone and Typhoon V-6 turbos for years to complete a very successful upgrade program for these General Motors vehicles. CarTech Stage III Mustangs completed hundreds of runs at the dragstrip running 12 pounds of boost with times in the low 12's and high 11's. In the last few years supercharged Mazda Miatas have become the rage, but owners have found out that killing the stock timing was not the way to horsepower heaven. By adding a J&S SafeGuard and returning timing to factory specs the missing low end performance and extra horsepower were rediscovered.

J&S Electronics has more than a decade of experience in wrestling with the thorny issues that surround knock-sensing and detonation. It is not a simple matter of hanging a microphone on the engine and listening for the tell-tale "knock". If you have discussions with OEM powertrain engineers they can tell you of the difficulties they face in the areas of signal discrimination or "noise" as well as the issues related to how the actual control over the timing should be done. J&S's SafeGuard has sophisticated mathematical controls or algorithms than discriminate between noise and actual "knock", and has user adjustable controls that allow for sensitivity as well as amount and speed of retard. Whereas OE retard schemes are limited to say 4 degrees, the J&S SafeGurad can be set for up to 20 degrees of ignition retard. This prevents catastrophic failure than can occur should you loose engine coolant or some other unforseen event. The J&S SafeGuard will control the timing of each cylinder in proportion to the degree of detonation occuring, thus maximizing performance and preventing engine damage.

Detonation is the precursor to preignition and only J&S Electronics can provide you a real-time solution to your ignition timing that both maximizes performance and protects your investment.



Tech – Preignition
Preignition and detonation are two separate and distinct events. It was first pointed out as far back as 1906 that the two phenomena were not only quite distinct but were in fact not related to each other. In the first place, preignition in itself does not produce an audible "knock" and if it is audible at all it could be described as a "dull thud". Because preignition is frequently brought about as a result of persistent detonation, the distinct "knock or ping" of the latter came quite erroneously to be associated with it.

It is by no means uncommon for preignition, or in this case it would be more correct to describe it as autoignition, to occur at the same phase as the timed spark. In this case the ignition can be switched off, and the engine could continue to run perfectly steadily without the slightest observable change in performance, sound, or any other characteristic. The danger, however, lies in the fact that all control of timing can be lost and ignition may creep in earlier in the cycle.

The danger of preignition lies not so much in the development of high pressures but rather in the very great increase in heat flow to the piston and cylinder walls when the ignition occurs too early in the cycle. This increase in heat flow, in turn, raises still further the temperature of the hot spot or surface which is causing the preignition resulting in even earlier ignition. At some point the temperatures are elevated to the point where the incoming charge is ignited, causing backfiring in the inlet tract. The belief, still widely held, that preignition can give rise to dangerously high cylinder pressures is totally false. Under no circumstances is the peak pressure resulting from preignition appreciably higher than from a spark-initiated ignition and, in both cases, the peak is reached when the maximum pressure is attained at or just after top dead center, that is to say, about 10 degrees earlier than the normal optimum. As the time of ignition is further advanced by either advancing the time of the spark or by earlier preignition, the maximum cylinder pressure falls again due to the excessive heat loss, for the piston is then compressing gas at or about its maximum temperature, and the intensity of heat flow is increased many times. The danger lies not in the production of excessive pressures but of excessive heat fow. The intense heat flow in the affected cylinder can result in piston seizure followed by the breaking-up of the piston with catastrophic results to the whole engine.

In nine cases out of ten, preignition is initiated by overheating of the sparkplug electrodes or some sharp point or edge that has gone "critical". We are accustomed these days to focus all our attention on the subject of detonation for it is the limiting factor controlling the performance of a spark-ignition engine. We are apt to forget that the real danger is that it leads on to preignition. In itself, detonation is not dangerous... It is the preignition it gives rise to that can so easily wreck an engine. J&S Electronics SafeGuard is your best defense against detonation and the harmful effects it can lead to. “


Best,
Grady

Here's what the Saab part looks like:

http://forums.pelicanparts.com/uploa...1131902873.jpg

"On a select few modern engine controllers (Porsche, Corvette) the microprocessor is programmed to retard individual cylinders as knock or detonation occurs. The problem with either of these approaches is twofold: First, the OEM carefully maps a complex set of spark curves and only allows a small degree of retard to take place, typically 4 degrees;"

- Grady Clay -

The above is incorrect about Porsche. The 964/993/928 Motronic units can
retard the timing for each cylinder to a max of 9 degrees.

You mean on modern EMS software you can actually manage ignition events by cylinder???

Now that's a revelation!!!! I thought all I had to do was screw a "knock sensor" under the shroud!!

Muhahahaha!

The pic below I'm pulling 8 degrees on one cylinder and 4.5 on the other.....gotta fix the timing maps...or fuel???


Beat that dog!!!!

http://forums.pelicanparts.com/uploa...1131904977.jpg

"The pic below I'm pulling 8 degrees on one cylinder and 4.5 on the other.....gotta fix the timing maps...or fuel???" - tsuter -

Only 20+ years behind Porsche/Bosch, e.g. the 959 Motronic unit more advanced than
the '96 993 Motronic unit!

Some fail to realize the Porsche/Bosch technology/resources or where they've been
and continue to think that the after-market tuners have new discoveries,
e.g. 3.2 performance chip tuning is now a high school auto shop project!

Please!!!!!

Here is freeware for hacking the channels on your Bosch Motronic ME7.5 and peeps think they'll just bump channel 9 a few degrees....

But they don't know ME7.5.

Can you spell "adaptation"?


http://forums.pelicanparts.com/uploa...1131906362.jpg

Well I see TI has already done exactly what I was suggesting. Using an A to D converter to sample the incoming signal, filtering it with a DFT to lower the workload and then send it on to a DSP to be analyzed with spectral analysis for it's signature. This is the same basic process used in voice recognition.

Technically speaking even in this scenerio an analog filter is still used since all A to D converters have a low pass filter on their input to remove any frequencies above their given sample rate to prevent aliasing.

Reading their report I can see they were headed the same direction as I was suggesting. Using sampling technology to digitize the incoming signal so it could be more accurately analyzed and discriminated against.

I must say though, the solution outlined by Ion looks pretty elegant and has the opportunity to nip the problem in the bud before it actually happens where as knock "sensing" is after the fact. Of course, as has been previously outlined however, low detonation is not necessarily destructive so maybe it doesn't matter.

\ I agree that it's a great system but it is Saab still using it? I thought they weren't on one of their newer models - maybe because it wasn't really a Saab or something. Just curious were the state of this was.
-Chris

SAAB still uses Ion detection on all it's cars except on it's sneakiest version with Opel-sourced 1.8 N/A motor (which uses Simtec). They are on Trionic 8 version by now.

Trionic 8 is a really complicated piece of work involving both AMM and MAP sensor. SAAB got knock-detection part nearly perfect more than decade ago so last 10 years it was mostly development toward solving emission problems. Most of extra complexity in Trionic 8 is there just to piss off mechanics, help cold-start emissions and disscourage SW-type of chipping.

Performance-wise, Trionic 5 was the best that there is. It's a fully sequential MAP-based EFI with coil-on-plug ignition and Ion-knock detection. It can pull back timing on individual cylinders after one(!) crank revolution and has fuzzy-logic that allows it to map itself to whatever fuel quality you use. It's called "adaptation". It's also a reason why SAAB's loose lot's of power when you pour low-octane brew in them.

What followed were T7 and T8...a mixed blessing. They are overly complex (and that really means complex considering that Trionic 7 was/is more complex than anything else around) and it's LMM's are crappy. It has (badly implemented imho) FBW throttle with artificial filtering to try to cure turbo-lag.

All in all, it's really too complex for it's own good. It can detect difference between AMM-measured and MAP-calculated air-mass and log a error code in case there is a vakuum leak etc. Thottle is nothing but a "torque requester" and it always outputs same amount of torque regardless of altitude and air temperature.

With other words, it will boost 0.9 on cold day with 98 octane and 1.2 on warm day on 95 octane to deliver same amount of power etc.

What it would like to see is cheap MegaSquirt-type of EFI with Ion knock detection. It's actually not so complicated, it's just that Mecel got it so damned right so many years ago that it's hard to go around the patent.


Detecting knock by microphone and doing DSP-filtering is like measuring vehicle speed by doing optical recognition of trees passing by just beacuse somebody patented speedo-pickup on differential.

P.S.
If SAAB were as good at building cars as they are at makind engine managment systems, they would be a market leader. Unfortunately, new SAAB 9-5 engines are nothing but 9000 engines with all fun removed.