CIS Flow mods, who can help/enlighten
 

First of all if your here to post to tell me it can't be done, don't waste your breath. I am not so full of myself that I don't believe there may be a remote possibility that we can't do it. However if you are here to help/share knowledge so that everyone on this board can benefit, please do help and post constructive criticism. I have spent my career doing things that were never done before, many times after being told it wasn't possible.

I'm asking people to please be polite, respectful and help me help everyone on this board that has come to this crossroad only to be told its too expensive or not possible.

I'm looking for a couple metering spools besides the one i have. The first thing we are going to try and do is make our own spool. That requires skilled journeymen machinists which I am not. But I do know where to find them. I do not have any 930 spools so if anyone has one they can donate for the purpose of blueprinting it, we wont modify it unless you ask.

I will be taking this metering valve to the lab and looking at it under the scope. I will measure the profile and see if its possible for us to machine it. We can grind a new electrode and re-machine the slots on one of our turbine blade machines. I may scrap a spool or two so if anyone has any they want to sacrifice in an attempt to learn something for everyone, that's what will happen. I have no interest in keeping secrets. My interest is to find out how this thing works and report back. I want a CIS head to flow 400 HP and maintain idle characteristics up to mid range. If I can do that the sky may be the limit on modifying these heads. Using a modified fuel head and Reanimotions Digital WUR I think this can work for a lot of cars.

This machine is not the backyard machine. So obviously this isn't something that anyone can do. Maybe we cannot. However given 32 years in the marketplace producing extremely high accuracy EDM machines for the Jet and Land based turbine industry as well as the fuel injector market I do believe its possible. It may be machined from scratch (metering spool) and then burned. Pictures of it I have seen lead me to believe that may be possible.

When I read posts about modifications made by working with Universities in Germany I believe this can be done with one of these machines.

We currently have machines that use round tungsten electrodes with a .010 diameter and a hole in the center of them for water to flow thru. We don't buy these machines, we design/build them. For all the nay-sayers this is old technology. These electrodes can burn thru almost any alloy known to man. We can move the part in slow increments to create shaped holes with this machine. The result is hundreds of cooling holes blasted in hundreds of blades and veins, all flow tested. The highest grade of components for commercial and Military jet engines. Certainly more accurate than a CIS metering spool. We can grind/shape electrodes to make a one cut pass. If Bosch did it 35 years ago, the right guy with the right machine can do it now.

If anyone on here has more information they would like to share please speak out. I will admit I'm new to how this head works, but I think its time for the "black magic" of CIS to be exposed.

The simple goal is to be able to modify something and keep relative stock appearance for the older turbos and the engines with aftermarket turbos and superchargers as well as the robust CIS injection. Its almost bulletproof. If all a guy has to do is modify a fuel head and and a digital WUR to get 500-700 HP of fuel thru these heads EFI is much less desirable for many.

I am an Electrical Engineer so going full EFI is not something I am afraid of. Tuning is something I am learning with the help of others.

Below is a picture of an airbox, made by the machinists at the shop. It has been detailed and built on a fixture, not with clamps and visegrips. Everything done here is precision and repeatable.

http://forums.pelicanparts.com/uploa...1461782847.jpg

Any pics of what a metering spool looks like, where it can be found, etc? Most of us haven't been deep into the guts of our CIS systems, so any pointers would likely help get the ball rolling.

I'm guessing he is referring to the fuel head metering slits?
Basically the fuel head consists of the balance/flow chambers and the individual slits fed by those.
I'm guessing he is talking of modifying the slits? as per Imagine FDs.
Alan

subscribed and fascinated

For starters, buy this book:

http://www.amazon.com/Automotive-electric-electronic-systems-Robert/dp/3184191109

It covers a lot of CIS principles in detail. Don't buy a newer edition, all of the CIS stuff is left out (ancient history).

Bosch made their own electronic CIS in the mid - '80's:

http://forums.pelicanparts.com/uploa...1461807386.jpg

Notice that they have added a potentiometer to give airflow feedback. This allows the CP adjustment to be coupled to the air meter position.

Rather than burn through a bunch of barrels (they are fitted by class to specific head bore diameters BTW), you should get an engineer to calculate the flow produced by different slit widths. I think you have all the data required. I fear you will burn through a lot of barrels otherwise.

Bought the cheapest copy of that edition available. Looks good!

What vehicles were these fitted to? Not the 930?
Regards
Alan

Mercedes mostly. Merc V8 is a great option for a 930 BTW.

I will locate the book. I may already have the earlier edition as i purchased it in the mid 90's. We have a fluid engineer and as luck would have it, he is a car enthusiast. All of this is good information. I am really looking for a 0 438 100 031 right now to compare with the 0 438 100 097. We are going to look at both of these heads. Even a head for a larger displacement V8 from Mercedes. We want to understand all aspects of the upper and lower chambers and their volumes with relationship to the size of the metering slits. if there is anyone in Michigan that has experience with rebuilding these heads please chime in.

I will tell you this, why the hell are you the 1st person actually wanting to do this? I've been pondering this since day 1 on the limitations of CIS questioning why hasn't anyone made one flow up to 1000hp, or 1500hp, while improving known weak areas?

I used to be an F15 engine mechanic and rebuilt 1000's of them. I know the blades and their cooling holes quite well.

I have a fairly thorough CIS Primer on our home page that is free for all to use. Go to Porche Enthusiasts Page and you will see it. Jim Williams put this together years ago for the N/A systems, I consulted on the turbo versions.

Larry Fletcher and I worked on metering slit modification a couple of years ago in an effort to reproduce the 934 metering valve. The machine used was EDM, the tolerance needed is 0.001". The resulting piece will support 700HP. I don't know what you could get out of a 911 fuel distributor as our starting point was double the capabilities of the -097 FD.

You can see in this photo the metering slits on the modified version are twice that of the stock 930.


http://forums.pelicanparts.com/uploa...1461814611.jpg

Did you mean 0.0001" tolerance?

Tapered metering slits that are stock width at the bottom and get wider towards the top might work well too.

There's no way the tolerance is .0001 inches. That's not achievable with any EDM machine nor is it necessary. The blades, veins and injectors in every vehicle on the road and jet in the sky are not manufactured to that tolerance. Its simply a slot, with exacting flows. EDM is very precise and overburn is very controllable and repeatable. The small amount of flow differences in a EDM produced slot would easily be adjusted on the flow head. I would venture to say the the remaining variables in these engines, compression, flow of intake runners, airbox construction, cylinder leakage etc etc etc would easily outweigh any machined differences in this EDM produced slot. It simply wont show up in any liquid flow test. It is that precise. After running my engine on a dyno and recording AFR's on each cylinder and EGT's on each cylinder it appears the only way to make this this best it can be is adjusting the head under full load at full boost. Over and over again. Different RPM, different Air temps, adjustable boost all contributed to different EGT's. Then map it. That wont be possible with CIS. We have to target a flow and AFR that keeps all 6 cylinders happy at best. We have all been running these engines boosted and NA without really knowing whats going on in each cylinder. At best an AFR gauge in each pipe. I would say if we can increase the flow then it will be exact enough by the methods everyone is currently using. A 3-5 minute test capturing fuel from every injector and weighing it on a gram scale. Lifting the sensor plate different amounts and recording weights. It shouldn't matter if we lift it, lower it, during a test as long as the result is 3% across the board on all six injectors. I have done this at least 100 times, all with the same result, 2-3% with the injectors installed. The only test I have not done it to place the injectors in a pressurized environment to see how it affects flow under boost. This would be a static test so not so definitive, but it may shed a bit of light in the affects of boost. The engineers I spoke to that design fuel injectors say a higher pressure opening injector reduces the affect of boost. I don't think this is possible with CIS.

That's what we are trying to produce. Trying to maintain idle characteristics yet give us a point where we can make it fat. If the Digital WUR is working we may be able to produce an area in the slot that the metering valve moves to only under heavy boost. Its the secondary chamber volumes and the springs (possibly) that also affect flow. I will leave that for the fluid engineer. Somebody has got to be able to sort this out. We don't want to burn anything until we are convinced what the target is. We know we can change the slot. What we don't know is how much to change it and how it affects the other components in the head. Where is that German guy that designed this today..... He could surely save us some time.

I'm certainly not the first person, I may be a bit more determined and open to share any results that we come up with. Secrets on something this old is just absurd. Given the complexity of something able to burn these slots is why its not commonly done. We make a small machine that fits on a desk. It burns injector nozzles. Has a mechanical indexer, rotates the injector and the the EDM burns the hole. One of these machines can probably do the job with a larger EDM generator. Once tooled up relatively easy to produce. Locate the first slot, burn and index. We will shortly know how accurate the original slots are. (Positional accuracy of the slot in relation to the rest of the spool). Our guess is that during this time a mechanical rotation device was used. These were easily accurate to .001 inches. Rested on a mechanical hard stop ant each index point. 6 slots at 60 degree intervals. My job was easy, sequence the device. The skill was the journeyman machinists that built these indexers. A trade that is sadly shrinking. My idea was to find some old scrapper parts, locate a 60 degree indexer, and boom... we are modifying things. A machinist makes the electrode, puts it in a collet, picks up the first slot, and I press a button. 60 minutes later its done. All for the rock bottom price to pelicans of $10,000.000.00 per spool. Quantity discounts available. Tell all your friends.....

Hasn't this already been done - tapered slits - Imagine FD?
Alan

The Imagine Auto modified fuel heads from 8 years ago were rebuilt and modified by Larry Fletcher of CIS Flowtech CIS Flowtech - Rebuilding Bosch Fuel Distributors & Warm-Up Regulators
Larry's a good guy and he does nice work.

The default ones flowed ~20% more fuel than stock and you can trim that with an adjustable control pressure regulator if it's too rich during cruise and boost but then the air flow sensor plate is a little higher in it's venturi cone than with a stock fuel head restricting intake air flow during boost.
I have one, it works well, and I don't think the metering slits were tapered. I don't think thats been done yet.

This is great stuff. Looks like I left the party too early, but I wanted to get my car up and running asap.

I'm watching this. I like the old cis system and anyone who will help keep it alive is alright with me. Yes I know there is much to be gained by going efi but with my simplified system if anything goes wrong on the road I can diagnose the problem and fix it right there and be back on my way.

I'm not sure what the question is, this has been done before, we've done it, and you have the equipment so you can do it. You are telling everyone you can and will do it. If you need metering pistons and barrels just buy some junk fuel distributors. There is no black magic, plenty of manuals and schematics out there.

I tried to get an answer from Cisflotech, and its my understanding Gavin did as well. 20% isn't enough for what I need, probably not enough for most people. I didn't get a reply to my email, Gavin can say what happened with his request as I wasn't a part of that discussion. 930 heads flow more than Sc heads.

I have read the manuals, and posts indicating that changing the slot width is only a part of the solution. In that post no explanation on that statement.

So making a statement like that just creates more questions. In the end it's not helpful at all. Seems a bit secretive to me.

I'm simply asking for help and answers. The answers I get here everyone can see. the answers I am forced to find, once again everyone will see.

This was posted by Edelweiss


" The slitz
We have done this already - with the help of a German university, what we found out is that the slits are just a part of the story and it was the simple one, later it gets a little more complicated.

However you are right - CIS can do much more: on a 8,5 : 1, 3,4 L 934 we see 720 hp using a modern cam profile..

And still it is not running lean - so the limit is not CIS !

Best reg."

So whats the rest of the story? It would be nice to know.

Do the lines limit the flow?
If so what is the limit thru the lines?
Do injectors need to be changed?
If the slits are opened what else needs modification?
What affect does changing the spring pressure have?
How much fuel can a stock 911 pump flow?
do I need two pumps like a 930, or will a single 200l/h 8 bar motorsports pump be enough for 400HP?
If i have a 930 head and I want 500 HP what do I need to modify?

What we want to do is find the limits. I don't see that information anywhere on this site. Plenty of people running out of fuel and lots of threads about attempted workarounds.

I'm trying to get more from my head, for my car. If I learn something everyone here will learn it too.

I imagine that it would be very difficult (and rightfully so) to get free, detailed explanations and instructions on how to extensively modify a fuel head from those that have successfully done so, via vast experimentation, great personal expense, years of learned knowledge and so forth, and who do this type of stuff for a living (Larry, Brian, Edelweiss, et al.).

OK these are questions the collective can help you with. We do very basic modifications to the SC fuel distributor to get it to feed race cars putting out 250HP. No idea what it would take to double that figure. I'm pretty sure what would be left when done would not be recognizable as an SC unit. You couldn't use the fuel lines or the injectors, or the pump, or the WUR. Anything can be done but boy is this going to get involved. It is true that not everyone has access to an EDM machine (I live in Rocket City so there are several) but the time and effort for research is tremendous. Protecting your business is critical. The internet has destroyed many a business by giving to the masses for free the toils of the few.

These heads are almost 40 years old. I am not trying to break into anyones business. I sent someone in that business an email. They didnt reply. A fellow pelican in the same situation did as I did and he ended up going full EFI. After 40 years I would think this would be a freely discussed topic.

Back to the problem at hand. Edelweiss posted modifying the slits was not the only solution. Doing flow tests on volume the stock SC pump will not be enough. I don't want two pumps like the turbo so I am trying to buy a Bosch 200 l/h 8 bar pump.

This is the part number B 261 205 413-01

So ABEC-9 or 11 ball bearings aren't held to that spec?

I have an interest in purchasing (2) of these pumps. Please let me know if you find a Bosch Motorsports supplier that can get the pumps at a reasonable price.

Thanks!

You are correct again, Tippy. Higher-end EDM machines can split thousandths.

That's what I was told, hence buying EDM

This isn't a discussion about whether and EDM machine can split thousandths. For a fuel head it quite simply isn't necessary. We don't supply bearing manufacturers, we supply companies that burn holes for flow. I don't know what bearing companies do but holding a dimensional tolerance vs. a flow tolerance it quite different. since the heads are somewhat adjustable .001 or .0001 isn't an issue. There are so many more variables in these engines we cannot change or control.

So I will add this story..

These machines can cost upwards of 1 Million USD. I don't know what higher end means but it would be interesting to see.

Wire EDM can split thousandths. Sinking machines are different. Electrodes wear a tapered amount while cutting. Unless someone can invent an electrode that does not get smaller during the spark/arc process taper is always an issue. When a slot is cut, if there is enough depth that the electrode can be longer, taper can be minimized and tolerances tightened and the hole is "finish" cut on a unworn portion of the shaped electrode. Sometimes a second electrode is sent thru the hole for finish size.

I don't have one in my hand (a spool), but I will soon. It may be that these were wire cut, but getting tooling inside a small diameter is difficult. Wire EDM is like a band saw.

Wire EDM uses a spool of wire, exacting tolerances. Because as the wire is constantly moving and it constantly replaced with new wire finer accuracy can be achieved. Taper/Wear is not an issue.

An example is a fuel mixture plate for a Delta Rocket. We made the machine that makes them. The entire theory is based upon two elements, fuel and oxidizer. They are directed at each other thru thousands of holes in patterns called quadlets. the size of the hole and the ability to direct these two at one another is critical. They also must me in relation to the machined parabolic dish surface. Some extremely talented (probably German) engineer came up with this idea in the 50's. We produced a machine that probed the entire parabolic dish and recorded a series of measurements (thousands) of the surface of the dish around each quadlet pattern. We then manipulated a 7 axis machine and electrode to blast the quadlets in relative to the surface that we had probed. Every hole, every pattern has multiple offsets. Some finer than .00005 inches.

The result, the first rocket engine exploded on the pad. Why?

The answer, some 45 years after the theory was proven and placed into production, machining tolerances and EDM was not capable of making every pattern of fuel and oxidizer produce the theorized contact spray. So to compensate the manufacturing process increased the sizes of the fuel and oxidizer holes and after repeated testing, came up with enough thrust. The engineer was told he was close, but his theory was slightly flawed.

Prior to this a very skilled journeyman layed out the entire dish with high spot blueing. I'm not sure how but they then scribed lines on this dish where every pattern went, every hole. They placed it in a machine which the went to the calculated position. this machinist then looked thru an optical scope, moved the machine to the crosshairs, then burned the hole. this took hundreds of hours. He didnt just move in x-y plane. He using years of learning manipulated at least 5 axis to hit the lines. then recorded each set of adjustments to try and maintain the angular relationship to the other holes in that pattern. On a parabolic shaped dish.

Our machine probed and burned in 1 week.

The reality, we produced a plate to blueprint tolerances with as close as you can get to perfect considering flow and angles. With the increased electrode sizes due to manufacturing inconsistencies we produced more heat and more thrust. the increased heat caused the plate to exceed its operating temperature and burned thru. Reaching the source of the fuel and oxidizer above the plate and exploding.

The solution. Go back to the original engineers calculations on hole sizes. This included bring the 80+ year old "rocket scientist" back to the plant to help us find the cause.
He came in a handicapped accessible van with his wife. I became the coffee runner to keep him and our mathematician busy calculating. Called in occasionally to explain my formulas for directing the 7 axis machine to the probed locations.

He left and I will never forget it. A man told for years he was close, but not quite there. He left with a chip on his shoulder and a wife that thanked me for helping to prove he was right all along. I didn't solve the problem, he did. In reality we owed him thanks, he saved us from a 12 Million Dollar loss. He passed away a couple years later, his wife said he left that day a different man.

When you think your smart, someone always comes along to humble you. I have made it a part of my career to learn from those people. I despise people unwilling to share their knowledge. If this was today's technology and a formula 1 win was at stake, I would understand.

Back to fuel injection. A day you don't learn something, is a day wasted. He told me that on the way back to his van.....

This needs more than a forum post to tell it. Bloody amazing story.

Just reread it - gave me chills.

The age of the technology is irrelevant, if someone has invested time and money into research why should they be despised for not wanting to give that away? Would you give away your toil for free that is your livelihood? I seriously doubt it. There is still a market for CIS mods, these folks make their living on it.

I expect to change the injectors, a new better pump is sourced. It appears the lines will flow enough, The fluid engineer is working out flows thru the lines. He wants to better understand the pressure system. Pressure affects velocity which affects volume.

Taking the pumps, filter and accum. Will do these tests as it would be in the car including the length of lines from tank to engine.

He will have the head tomorrow. I'm taking it off a car i have. I was unable to find a Bosch head on ebay for 6 cylinder porsche. Every place I called to buy one wants exchange. If somebody has one that needs rebuilding we will rebuild it for free if we can use it to learn.

The steel pump diaphrams have been traced and can be cut on the wire EDM. If we start it at the end of the day, we will wake up to a new one in the morning. stacked on top of each other we can cut several at a time. Waiting to hear what affect those play in the system. (I got new ones in a rebuild kit).

I can tell you this. If these folks had answered my questions or Gavin's, I wouldn't have this thread. I really wanted to send my head out and get it back in a few days with more flow. That didn't happen. Gavin was a few months ahead of me, and in those months came no answers. So now I am attempting to learn, and I am surrounded by engineers that are also learning. I don't know everything, and I am surely not asking CISflowtech or Eidleweiss. I don't know CISFlowtech, or Larry. He may be a great guy. He may have a great product and business. I do know they didn't help me.

I've got a cast iron fuel head off of a '79 930 engine if you need another one.

https://farm4.staticflickr.com/3705/...2d2f39a3_c.jpgIMG_0890 by willtel, on Flickr

I have this one that you can play withhttp://uploads.tapatalk-cdn.com/2016...9ed7fc8ca6.jpg

Sent from my SM-N910T using Tapatalk

It's off to the lab. Trying first with a Lamda head.

http://forums.pelicanparts.com/uploa...1461957352.jpg

Just let me no where to send it..Kevin

Sent from my SM-N910T using Tapatalk

That head is from a 930? Are you selling it? I would have to look up the number....

Send it to

American Oak Resort
112 Jefferson
Houghton Lake MI 48629

The manager is always there when UPS comes.

I can prepay shipping if you like. If you need any kind of receipt I can fax/email you one...