FrankenCIS installation on a 930 (from mechanical to digital WUR ) ??

[carrpet]

Replacing the AAV with a Bosch Idle Valve

• I measured the ID of the AAV inlet and worked out the cross sectional area (CSA)
• I took a video with a timeline of the AAV in operation, from fridge temperature - 5C, to fully closed (5 minutes +) on the bench then played it back and estimated the % of the total area which was open for air to pass through every 10 seconds
• My start point was to copy the AAVs opening action and replicate this using the MS controlled IV based on variable aperture area over time.
  I measured the dimensions and CSA of the idle valve (IV) aperture then calibrated the %CSA against the Microsquirt PWM (pulse width modulation at 92Hz) duty output required to operate the valve. 30% duty corresponded with a fully closed IV and 70% corresponded with fully open.
• I produced a spreadsheet so that I could correlate the AAV's open CSA versus time versus Microsquirt duty required to give the same open CSA for the IV
• My plan was/is to use the IV in open loop control so that it would vary the size of the aperture with engine temperature. MS also offers the capability to have a pre-determined (by me) cranking temperature v IV duty map and then a warm-up v IV duty map. This means that I can program a much wider aperture (more air bypass) until the engine fires, then 5 seconds later (adjustable), it moves onto the warm-up curve with a smaller starting aperture transitioning to fully closed at a rate which is directly related to engine temperature.
• So I installed the MS controlled IV started the car and the idle was jumping around all over the place with irregular misfiring. Checked everything over, hose connections , electrical connections, MS settings etc. Tried again. Same result !!!
• I reinstalled the AAV both electrically and pneumatically but with a view to connecting up the IV electrically only so I could watch it's opening action over time.
• I uninstalled the IV in MS firmware and started the car. Ran perfectly.
• I then re-installed the IV in MS and started the car again. The same misfiring problem resulted.
• I ran the logging facility in MS to try and see what was happening
• I started the engine with the IV installed in MS firmware but with the electrical plug out of the IV. Car ran perfectly until I plugged in the power and then the misfiring started.
• I went through the MS log and shared the log with James Murray (who is an MS consultant wrote the MS hardware manual which includes a section on Idle Valve installation (Microsquirt_Hardware-3.3.pdf (megasquirt.co.uk)).  James had never experienced a problem like this before suggesting it could be a 911 related issue. The AFR was jumping around coincident with the bad running, but it became clear that the AFR spikes were driven by misfiring which appeared to be ignition driven. Hmmm ! I had assumed that the CDI+ being installed 2cm away from MS could provide a source of interference to MS but no evidence of this up until now.
• I then, as best as I could, separated the IV wiring away from the rest of the loom and away from CDI+. There was a reognisable improvement but some way from being acceptable, however the clue as to what the problem might be, was clearly there.
• I decided to leave it for the evening and give it some thought. I remembered going back to the time when I couldn't get MS to register rpm. Initially I was using the Black-Violet tachometer driver wire to drive the rpm input for MS (which is square wave) but having discussed the issue with Jonny Hart we came to the conclusion I should try the analogue sinusoidal input from the distributor, which is the often referred to "green wire", which is screened. The clue is in the screening !! I had connected MS directly, soldered to the CDI+ socket female spade connector in the CDI+ plug, under the rubber boot. However in doing so I had unwittingly created an aerial for receiving PWM interference. So the theory was, it wasn't CDI+ interfering with MS it was the other way round !
• The next morning I reinstated the CDI+ tacho driver output connection (Black/Violet) and disconnected the ~25cm green wire spur , which I temporarily tucked out of the way behind the CDI+ box. Started the car, problem solved ! I then cut back the spur to the green wire. So the learning is that the grounded screening of the green wire is essential to prevent picking up interference which screws with the CDI+ box.
• I then reinstalled the IV replacing the AAV and it works like a dream. Much better starting and no unnecessarily long periods of higher than needed rpm during tickover and warm up, and it is very easy to adjust using Tuner Studio to program MS.

Edited January 20, 2021 by carrpet

[carrpet]

[carrpet]

Remember its Covid lockdown so I have too much time to think about this stuff !

One of the issues I have is this;

Porsche use mechanical thermo/time devices in their CIS cars. The WUR, AAV and thermo-time valve for switching off the CSV. All of these devices have pros and cons. All these thermo-time devices are power heated, all of the time the engine is running. However the action of all of these devices deteriorates with time and its is not necessarily easy to diagnose the issues that arise.

The WUR with it's coil wound heated bi-metallic strip does however warm up much faster than the car. 2 minutes from cold to hot and full control pressure. Our cars do not go from cold to hot in 2 minutes. Therefore to use a thermo (non time related) map in MS is a challenge because CHT does not rise at the same rate as the WUR heating element. Also our 911/930 engines like to start at relatively low control pressures and hence low (ie possibly sub 8 AFRs) eg at 3C my WUR generates 10 psi of cold pressure but 2 minutes later it is a 52psi and the engine is very happy with that. My engine CHT rises by 12C in 2 minutes from starting. 6C to 18C and reaches full CP (warm pressure) after 3.25minutes. nb. that is not a fully warmed up engine, just warm enough to get to CP max. Therefore MS using CHT as the reference temperature  produces a much richer mixture for longer, unnecessarily, during warm up. If I could get a faster thermo response this would mitigate the extended warm-up period.

So thoughts on potential solutions are:

1. Use a thermocouple rather than a thermistor sensor which have a much reduced thermal mass and therefore lower inertia
2. Locate the sensor closer to the source of heat
3. Question to self - Is a sensor probe with oil splash or oil immersion going to respond faster than a dry location such as the outside of the cylinder head cooling fin ? If so, go for an oil splash probe ?
    

All thoughts greatly received and answers on a postcard are satisfactory !!

Edited January 21, 2021 by carrpet

[carrpet]

AAV v Idle Valve
Now that I have had a chance to play with the Idle Valve settings I thought it is worth sharing what I have learnt.
The first thing I would say is that the 930 seems to like a lot of bypass air during cranking to promote firing ie. more than my AAV was delivering.
It also seems that once the engine has fired up, it immediately likes less air by pass, A fast idle of 1400 rpm is more than sufficient and that can be dropped rapidly as the engine warms up. The AAV cannot deliver this because the AAV does not know the difference between cranking and idling.
Some owners are reporting 9 minutes to fully close the AAV from ignition on. Mine was 5 minutes. I cannot see the justification for that. My engine is happy to have no additional air bypass after about 3.5 minutes and a good idle starting from cold at about 5C.
Now that the AAV is removed I am now getting about 5kpa extra vacuum suggesting that the AAV was continually passing air even when it was fully "closed"
The learning continues.

[Jonny Hart]

Hi Peter,  I was having a read of your post....

• Quote

   

  • I decided to leave it for the evening and give it some thought. I remembered going back to the time when I couldn't get MS to register rpm. Initially I was using the Black-Violet tachometer driver wire to drive the rpm input for MS (which is square wave) but having discussed the issue with Jonny Hart we came to the conclusion I should try the analogue sinusoidal input from the distributor, which is the often referred to "green wire", which is screened. The clue is in the screening !! I had connected MS directly, soldered to the CDI+ socket female spade connector in the CDI+ plug, under the rubber boot. However in doing so I had unwittingly created an aerial for receiving PWM interference. So the theory was, it wasn't CDI+ interfering with MS it was the other way round !
  • The next morning I reinstated the CDI+ tacho driver output connection (Black/Violet) and disconnected the ~25cm green wire spur , which I temporarily tucked out of the way behind the CDI+ box. Started the car, problem solved ! I then cut back the spur to the green wire. So the learning is that the grounded screening of the green wire is essential to prevent picking up interference which screws with the CDI+ box.

   

   

Unwittingly creating 'antennas' is something we all have to watch out for but there could also be another reason.   Essentially, you're sharing the distributor signal between two input circuits (CDI+ and MS).   I haven't looked at the input circuit to MS but its impedance may degrade or reshape the signal from the distributor.    I first noticed this effect working on twin plug 911 systems that run two CDI units.  Effectively you have two input circuits in parallel, so any input resistance for example is halved so component selection is critical.   In one case, this had the effect of changing the shape of the distributor waveform in such a way that it now had two zero crossing parts per cycle, meaning the coils were firing unexpectedly - all sorts of trouble.   You can only really debug this with a scope obviously.

I did some further thinking on running from our black/purple tacho output, my initial concern being the smoothing that we apply to the tacho drive, introducing a lag on the true RPM.  With hindsight, the bob weights/springs in the original mechanical setup would ALSO have a lag as they cannot move instantaneously.  Who knows what this mechanical delay would be would be less that the electronic software damping which is around 10ms.

 

[Jonny Hart]

On 21/01/2021 at 18:14, carrpet said:

Remember its Covid lockdown so I have too much time to think about this stuff !

One of the issues I have is this;

Porsche use mechanical thermo/time devices in their CIS cars. The WUR, AAV and thermo-time valve for switching off the CSV. All of these devices have pros and cons. All these thermo-time devices are power heated, all of the time the engine is running. However the action of all of these devices deteriorates with time and its is not necessarily easy to diagnose the issues that arise.

The WUR with it's coil wound heated bi-metallic strip does however warm up much faster than the car. 2 minutes from cold to hot and full control pressure. Our cars do not go from cold to hot in 2 minutes. Therefore to use a thermo (non time related) map in MS is a challenge because CHT does not rise at the same rate as the WUR heating element. Also our 911/930 engines like to start at relatively low control pressures and hence low (ie possibly sub 8 AFRs) eg at 3C my WUR generates 10 psi of cold pressure but 2 minutes later it is a 52psi and the engine is very happy with that. My engine CHT rises by 12C in 2 minutes from starting. 6C to 18C and reaches full CP (warm pressure) after 3.25minutes. nb. that is not a fully warmed up engine, just warm enough to get to CP max. Therefore MS using CHT as the reference temperature  produces a much richer mixture for longer, unnecessarily, during warm up. If I could get a faster thermo response this would mitigate the extended warm-up period.

So thoughts on potential solutions are:

1. Use a thermocouple rather than a thermistor sensor which have a much reduced thermal mass and therefore lower inertia
2. Locate the sensor closer to the source of heat
3. Question to self - Is a sensor probe with oil splash or oil immersion going to respond faster than a dry location such as the outside of the cylinder head cooling fin ? If so, go for an oil splash probe ?
    

All thoughts greatly received and answers on a postcard are satisfactory !!

So I take it that there is no time based map in MS?  How about using an external timer of some kind?  Solid state relay timer or similar?

 

https://www.ebay.co.uk/itm/Trigger-Cycle-Timer-Delay-Switch-Circuit-Board-MOS-Tube-Control-Module-DC-12-24V/193580168145?hash=item2d1246f3d1:g:fcwAAOSwgAFfEg92

 

 

 

 

[carrpet]

4 hours ago, Jonny Hart said:

So I take it that there is no time based map in MS?  How about using an external timer of some kind?  Solid state relay timer or similar?

https://www.ebay.co.uk/itm/Trigger-Cycle-Timer-Delay-Switch-Circuit-Board-MOS-Tube-Control-Module-DC-12-24V/193580168145?hash=item2d1246f3d1:g:fcwAAOSwgAFfEg92

Thanks and good input Jonny. Yes that is an option. I have a few other options to optimise first I think

Meanwhile what I am learning is that the Porsche thermo-time (WUR and AAV) devices provide conditions which are a compromise between start-up and warm-up requirements. Microsquirt is able to separate these two states out using rpm. Anything above (this is adjustable) 400 rpm MS sees as warm-up and everything below it is cranking. My engine cranks at 270rpm in the winter with a fully charged battery.

The WUR calibration based on my growing learning is based more on starting requirements (than warm-up) so once the engine is running I can lean out much faster to get to warm pressure.

So there is a MAP in MS called warm-up enrichment which enables adjustment between -100% and 500%. (see a couple of posts below)

There is another MAP (see also a couple of posts below) which is Control pressure v Engine Temperature which is the bit of the code which just emulates the WUR. In the Porsche manual there is a calibration chart for each WUR with a tolerance band. My WUR is for a 930.66. I have now moved to the very top of this band and at higher temperatures ie above 30C even above specification.

The second thing I have done is programmed in a negative warm-up curve, so that instead adding in more fuel (the WUR curve has already over delivered on this) I can immediately take fuel out of the mixture 5 seconds (adjustable)  after the engine has fired. Using the data logging facility I am able to check that this is happening in reality, and it is. So I have a tapered curve of negative enrichment which goes to zero negative enrichment at 35C CHT. I have gradually been backing more fuel out using this curve but you only get one cold start opportunity a day to test /optimise it.

The second area of opportunity is to use a much lower thermal mass thermistor probe (I have since clarified that you cant use thermocouples with MS). These devices are used in the air cooled motorcycle MS installations which seem to be growing in popularity and I have one on order. Any improvement in thermal responsiveness would be useful so I will give that a go. They are threaded (3-4mm diameter) so will provide some new opportunities for location. Let's see. They are also relatively cheap.

From a 2.5C start-up this morning I got to maximum control pressure (52psi or 366 kpa in MS language) in about 3 minutes. I have made some more adjustments and will try those tomorrow. Let's see !!

 

 

 

Edited January 24, 2021 by carrpet

[carrpet]

4 hours ago, Jonny Hart said:

 

Hi Peter,  I was having a read of your post....

•  

Unwittingly creating 'antennas' is something we all have to watch out for but there could also be another reason.   Essentially, you're sharing the distributor signal between two input circuits (CDI+ and MS).   I haven't looked at the input circuit to MS but its impedance may degrade or reshape the signal from the distributor.    I first noticed this effect working on twin plug 911 systems that run two CDI units.  Effectively you have two input circuits in parallel, so any input resistance for example is halved so component selection is critical.   In one case, this had the effect of changing the shape of the distributor waveform in such a way that it now had two zero crossing parts per cycle, meaning the coils were firing unexpectedly - all sorts of trouble.   You can only really debug this with a scope obviously.

I did some further thinking on running from our black/purple tacho output, my initial concern being the smoothing that we apply to the tacho drive, introducing a lag on the true RPM.  With hindsight, the bob weights/springs in the original mechanical setup would ALSO have a lag as they cannot move instantaneously.  Who knows what this mechanical delay would be would be less that the electronic software damping which is around 10ms.

 

Thanks Jonny, that is very reassuring about use of the Tacho Driver square wave output. It seems to be working very nicely incidentally !

[carrpet]

The top table is the WUR MAP of control pressure versus temperature for starting

The bottom table removes fuel from the mixture once the engine fires and is in the warm-up phase. I have arrived at this with a combination of intuition and trial and error !

Edited January 24, 2021 by carrpet

[PorscheApostle]

As admirable as your endeavours have been Peter, I cant help but think that this will never be a cost effective method of adapting a CIS setup to run more efficiently. 1). Porsche (Bosch) developed this system to run effectively as a homogeneous setup. 2) these cars are worth a significant amount and most folk wish to keep them original as designed from factory where the system does work for the tech available at the time. 3) EFI alternatives have been developed many years ago for this application and are readily available should you wish to adapt a 930/66 engine for this application.  4) with respect I can't see why this setup would appeal to anyone versus an efi setup that would be more cost and comprehensively more effective. 

This ground is well trodden and as apposed to JH product that deals with deficits in ignition and maintains originality  this solution cobbles together later tech (ICVs) CHTs  etc to deal with a problem that for the majority of 930 owners doesn't exist or present a tangible issue. 

I like the idea and as a standalone project for yourself I see the appeal but as you've discovered establishing a stable cht without using say a Carrera setup where the cht was machined into the head the dissipation of heat from the heads does not give a fast reaction to the control pressure that you would otherwise see in the oe platform from a live feed. 

An EFI system wiyh with larger injectors on Carrera manifold would realise more bhp and mapable fueling(Megasqurt , Motec etc). The stock system has its limitations and they are not exclusive to the AAV or the or the TTS . The CIS system can support BHps up to 400 mark and then the CIS system cannot deliver the fuel even with a 3rd injector such as Kremer developed. 

I bolt on removable solution is usually the appeal for most but the question you need to ask is why would people do this? I've refirbishe 3 original wastegates for a 930 last year at serious cost purely because owners want the original systems. A turbo smart wastegate is by every measure a better product. But people want the original setup 🤷

Not hating just my experience.

 

DBS

 

 

 

[carrpet]

17 hours ago, PorscheApostle said:

As admirable as your endeavours have been Peter, I cant help but think that this will never be a cost effective method of adapting a CIS setup to run more efficiently. 1). Porsche (Bosch) developed this system to run effectively as a homogeneous setup. 2) these cars are worth a significant amount and most folk wish to keep them original as designed from factory where the system does work for the tech available at the time. 3) EFI alternatives have been developed many years ago for this application and are readily available should you wish to adapt a 930/66 engine for this application.  4) with respect I can't see why this setup would appeal to anyone versus an efi setup that would be more cost and comprehensively more effective. 

This ground is well trodden and as apposed to JH product that deals with deficits in ignition and maintains originality  this solution cobbles together later tech (ICVs) CHTs  etc to deal with a problem that for the majority of 930 owners doesn't exist or present a tangible issue. 

I like the idea and as a standalone project for yourself I see the appeal but as you've discovered establishing a stable cht without using say a Carrera setup where the cht was machined into the head the dissipation of heat from the heads does not give a fast reaction to the control pressure that you would otherwise see in the oe platform from a live feed. 

An EFI system wiyh with larger injectors on Carrera manifold would realise more bhp and mapable fueling(Megasqurt , Motec etc). The stock system has its limitations and they are not exclusive to the AAV or the or the TTS . The CIS system can support BHps up to 400 mark and then the CIS system cannot deliver the fuel even with a 3rd injector such as Kremer developed. 

I bolt on removable solution is usually the appeal for most but the question you need to ask is why would people do this? I've refurbished 3 original wastegates for a 930 last year at serious cost purely because owners want the original systems. A turbo smart wastegate is by every measure a better product. But people want the original setup 🤷

Not hating just my experience.

DBS

 

Quote

Hi DBS,

Thanks for taking the time for what is the longest response to this thread.

So to distil out your key points;

1. This is not a cost effective replacement for the fuel delivery side of CIS.
2. EFI is available off the shelf for those enthusiasts who want a significant performance upgrade (>400bhp)
3. Most Porsche owners, based on your experience, want to retain an original set-up to retain the value in their cars
4. Most Porsche owners are happy with the performance of their "homogenous CIS installation"
5. Classic retrofit's CDI+ box is OK because there is a deficiency in the Bosch CD!+ box and therefore Jonny's Box has validity because it fixes the deficiency and it looks identical to the original

There are no right and wrong answers to this but there are alternative ways of looking at the subject which may lead Porsche owners to different conclusions, depending on the assumptions that are made.
I don't know, lets hear what other Impact Bumpers have to say ?

Here are a few thoughts;

1. The car has not yet been out of the garage onto the road yet ! This project has not yet reached the stage where the benefits can be evaluated so it's impossible to talk out cost effectiveness without the data on the benefits which will only be realised realistically, post Covid.
2. The stock 930.66 generates 300BHP. So there is quite a lot of opportunity up to 400BHP which appears to be of interest to many 930 owners
3. 930 owners are sitting on 30-50 year old CIS components which may no longer perform as originally specified and are faced with diagnostics, replacement or refurbishment. Replacement is certainly not cheap and refurbishment is highly dependent on the person carrying out the work, and that is not cheap either.
4. MicroSquirt (MS) is a fantastic diagnostic tool which makes identification of defective components much easier as the system generates real data for fault finding. MS costs £350 and with the relevant sensors can probably be purchased for £100 with the necessary knowledge. That's not bad value even if MS is used just as a monitoring / diagnostic tool on an ageing CIS system, in a "valuable car".
5. Trouble shooting and Fault finding on CIS systems can be extremely challenging and often very frustrating. It often requires a significant level of patience and knowledge, which Porsche owners may need to acquire, in order to solve their CIS problems. This can be quite hard work because the most difficult issues to solve are often multifactorial, where more than one component is partially working or has failed. My next door neighbour has a 911 - his WUR failed in 2020 after changing his fuel hoses. My WUR was outside specification contributing to a cold start issue, and needed to be adjusted. The diagnostics on these failures were both long and drawn out. This might not have been the case if, for example, MS had been installed. The same applies to my partially working AAV and CSV which was also contributing to the cold start problem.
6. It is a well known phenomenon with the 930 that the standard CIS K-Jetronic system is at, or beyond its limit when driven hard on full boost which leads to higher than optimal  AFRs. The FrankenCIS WUR with a Microsquirt map offers the possibility of addressing this issue. Let's see.
7. Some Porsche owners want to retain the ability to return their cars to full authenticity, if they wish, as I do, without necessarily sticking to this during their tenureship. The FrankenCIS WUR and microsquirt could be removed from my 930 in 2 hrs work. The fuel mixture and tickover settings are identical as both systems generate a Warm Pressure of 52psi. I would need to acquire a 3/8" BSP blank to block off the IAT sensor aperture.
8. The CHT solution described in this thread is taken straight from the TurboKraft EFI installation manual. So the "deficiency" referred to applies to many EFI installations as well.
9. EFI is a significant and expensive change to make. I bet there aren't many people that would revert an EFI installation back to standard once they have "burnt the bridges". That is a big call for any Porsche owner to make and one that may be a "Bridge too Far" for many, despite the performance improvement.

So when the project is complete, and we have a better view of the benefits, we can carry out a review and see whether or not this makes sense to some Porsche owners (or not!).

The judgement on the cost effectiveness of the benefits of what will be a £1000 investment is going to vary from owner to owner. Lets see what assumptions people make ! They may not be the same as yours and mine ?

The cost of a "new" WUR from Porsche is £1300 and an AAV is £800. Whilst they are not the assumptions I would use, they are facts !

PCC

 

Quote

 

 

Edited January 25, 2021 by carrpet

[flatsix777]

There is probably a lot of truth in DBS’s critique if the object is to develop and somehow market improved CIS for IBs but I didn’t see it as that.  It’s an interesting project and if you have the skills it can be good to take on these technical challenges. CIS is much maligned but back in the day was a good system used by Porsche, Mercedes, Ferrari, Ford and no doubt others. It does have limitations, particular with turbo charging, and the kit is 30 or more years old in our cars. Trying to use current tech to improve it, whether dealing with old parts that don’t work or overcoming things that were constraints back in the 1970’s but need not be now, has a place.

Personally I have no wish to EFI my 930, although I do see the benefit over CIS and would happily do so in other circumstances. I am quite interested in making CIS work reliably and, if possible better. I also like doing some mods just because I can. When I do this I’m only trying to make myself happy, nobody else. 

Some of the things that Peter is doing might have benefit, time will tell, but we’ll definitely all know a bit more about what is possible with CIS. 

David
 

[carrpet]

56 minutes ago, flatsix777 said:

There is probably a lot of truth in DBS’s critique if the object is to develop and somehow market improved CIS for IBs but I didn’t see it as that.  It’s an interesting project and if you have the skills it can be good to take on these technical challenges. CIS is much maligned but back in the day was a good system used by Porsche, Mercedes, Ferrari, Ford and no doubt others. It does have limitations, particular with turbo charging, and the kit is 30 or more years old in our cars. Trying to use current tech to improve it, whether dealing with old parts that don’t work or overcoming things that were constraints back in the 1970’s but need not be now, has a place.

Personally I have no wish to EFI my 930, although I do see the benefit over CIS and would happily do so in other circumstances. I am quite interested in making CIS work reliably and, if possible better. I also like doing some mods just because I can. When I do this I’m only trying to make myself happy, nobody else. 

Some of the things that Peter is doing might have benefit, time will tell, but we’ll definitely all know a bit more about what is possible with CIS. 

David
 

There is certainly a lot of interest in this development in a 911/930 on both sides of the Atlantic. I have a similar thread on Pelican under the 911/930 Turbo section of the Pelican Parts forum and it has >6000 views and a 3 star rating so the interest is there. Every person I talk to about this has a different perspective on what would make them think about installing FrankenCIS / Microsquirt  or bits of it in their cars; I had a cold starting issue on my car which has been there for at least a decade which is now resolved as result of this project identifying deficiencies in 3 pieces of 30 years old CIS hardware, which individually did not appear to be a problem.

Ian Comerford is very keen to get the car on a rolling road, and do the before and after. I have a plan, and the parts, to use two, 3 way ball valves to enable hot switching, between the two WURs, to get the base line.  

As David pointed out many of the cars designed in the 80's with K-Jetronic injection systems were also designed to operate within the constraints of hardware like thermo-time devices which are a clever but pragmatic compromise to give a manageable engine performance over a range of conditions from start-up, warm-up, normal running and full load. The engines use an ingenious combination of centrifugal force, engine vacuum, engine temperature, boost pressure and thermo-time devices to control the engine ie. there is a lot which needs to come together for the engine to run well under all of these conditions and a lot that can go wrong ! Perhaps there is a better way using modern technology without rebuilding the top of the engine to get a moderate improvement in performance and driveability with more control enabling continuous improvement, and better diagnostic capability at a realistic cost ?

This assessment is and will ever be in the eye of the beholder, but there is more evidence yet to be generated, and it is a lot of fun !

 

Edited January 25, 2021 by carrpet

[Ian Comerford]

I am interested in the e-WUR aspect of this project as it may have utility as a cost -effective upgrade to my SC’s CIS.  If it gives more accurate fuelling throughout the range from cold, through warming up to full temperature then surely it’s worth understanding or trying to quantify the benefits?  I gather from Peter that just doing the e-WUR part with the Microsquirt, e-WUR and associated parts is about £1k.  You could get close to that figure by buying new WUR /AAV from Porsche and not seriously improve the situation.

Any thoughts on this?    
 

Thank you

[carrpet]

This is a graphical representation of today's start-up log.

Colour coding

Light Blue and yellow - target control pressure and actual control pressure (v time along the bottom) - see how closely the eWUR maps the programmed in warm up curve ie. it does what it says on the tin !
Green = CHT (please note that the control pressure and idle valve duty are directly linked to the CHT rise hence the importance of rapid and accurate CHT response during warm up)
Light Pink = Idle valve duty (fully open corresponds with 70% duty and fully closed is 30% duty)
Violet = MAP (please note this is absolute pressure in kpa so the lower the absolute pressure the higher the vacuum ie. you can see engine vacuum increasing with time as the engine warms up, obviously the open idle valve reduces vacuum in the early stages )
White = rpm
Red = AFR (Please note that the sudden hike in AFR straight after start up is due to the delay while the internal heater in the Bosch UEGO sensor has to get up to temperature before it will start reading)

 

Edited January 26, 2021 by carrpet

[carrpet]

16 hours ago, Ian Comerford said:

I am interested in the e-WUR aspect of this project as it may have utility as a cost -effective upgrade to my SC’s CIS.  If it gives more accurate fuelling throughout the range from cold, through warming up to full temperature then surely it’s worth understanding or trying to quantify the benefits?  I gather from Peter that just doing the e-WUR part with the Microsquirt, e-WUR and associated parts is about £1k.  You could get close to that figure by buying new WUR /AAV from Porsche and not seriously improve the situation.

Any thoughts on this?    
 

Thank you

Hi Ian,

The cost of £1000 for the eWUR with fuel injector and damper, MicroSquirt plus sensors and wiring costs are correct. As of today The Porsche WUR cost is £1300 and the AAV is £794. A Cold Start Valve is £450.

If you look at the log graph above you will see how accurately the actual control pressure maps over the target control pressure. It doesn't get much better than that ! So the technology can deliver the fuel pressure you want, the clever bit is linking that with CHT, IAT, MAP, RPM and AFR which will come with starting auto-tune and then exposing the engine to a full range of stress conditions while driving. The final bit of tune would be on rolling road. My view is that autotune will take us to a position which is better than the existing WUR because my base AFR map table is already informed by flighlead404's 930 tune which he has passed on to me.

We can't start to objectively evaluate the benefits until the car is on the road.

Regards

Peter

 

Edited January 26, 2021 by carrpet

[PorscheApostle]

2 hours ago, carrpet said:

Hi Ian,

The cost of £1000 for the eWUR with fuel injector and damper, MicroSquirt plus sensors and wiring costs are correct. As of today The Porsche WUR cost is £1300 and the AAV is £794. A Cold Start Valve is £450.

If you look at the log graph above you will see how accurately the actual control pressure maps over the target control pressure. It doesn't get much better than that ! So the technology can deliver the fuel pressure you want, the clever bit is linking that CHT, IAT, MAP, RPM and AFR which will come with starting auto-tune and then exposing the engine to a full range of stress conditions while driving. The final bit of tune would be on rolling road. My view is that autotune will take us to a position which is better than the existing WUR because my base AFR map table is already informed by flighlead404's 930 tune which he has passed on to me.

We can't start to objectively evaluate the benefits until the car is on the road.

Regards

Peter

 

Totally get the idea, but a fully refurbished wur is £250 from kmi. Agreed an AAV is not so that cost is there. On balance as you eluded to, project isn't finished yet so once an affordable cost has been established that would serve all CIS applications the appeal would to owners of Sc's , turbos Carrera 3s and early cis 911s the market would be there for all the options. I've never had a CIS car I haven't been able to get to run well by Purley applying the CIS basics.

The system does work but if there are any deficiencies with the engine (vac leaks ignition etc ) it will not tolerate those. As I'm Johnny will attest, you can't put an upgraded spark box on a sick engine in order to solve what appears to be an ignition issue.  

On a healthy engine having a reliable electronic fuel system control would be very desirable. Much like having a modern spark control and circuitry is desirable .

Look forward to seeing the end product.

DBS

 

 

 

 

 

[Phill]

Don't forget 1k for CDI+........

[carrpet]

11 minutes ago, PorscheApostle said:

Totally get the idea, but a fully refurbished wur is £250 from kmi. Agreed an AAV is not so that cost is there. On balance as you eluded to, project isn't finished yet so once an affordable cost has been established that would serve all CIS applications the appeal would to owners of Sc's , turbos Carrera 3s and early cis 911s the market would be there for all the options. I've never had a CIS car I haven't been able to get to run well by Purley applying the CIS basics.

The system does work but if there are any deficiencies with the engine (vac leaks ignition etc ) it will not tolerate those. As I'm Johnny will attest, you can't put an upgraded spark box on a sick engine in order to solve what appears to be an ignition issue.  

On a healthy engine having a reliable electronic fuel system control would be very desirable. Much like having a modern spark control and circuitry is desirable .

Look forward to seeing the end product.

DBS

Hi DBS,

Yes aware of KMI's very good value and great service. He is definitely one of the good guys and arguably underpricing his services ! I hope he's not on Impact Bumpers !! I think KMI refurb AAVs as well.

Fair assessment.

Lets see where we end up with this. The jury is still out although the indications are good. 

Regards

PCC

 

 

 

 

 

[carrpet]

I have made a breakthrough in Cylinder Head Temperature sensor responsiveness using motorbike technology.

Kenso make a threaded Thermistor Temperature Probe/Sensor with a thread size of 4.5 * 0.75mm.

I have installed this by drilling and tapping directly into the cylinder head at a position marked by the red circle in the photograph below - you can see it's position in relation to the TurbKraft temperature sensor.

With the heat exchanger removed I was able to use a 3.7mm drill a 12.5mm deep hole (from the front surface of the tin plate) between two stub fins and used a 4.5 * 0.75mm tap. I then applied thermal conducting grease to the thread and installed the probe as in the photos below

The results are easy to assess objectively using the data logging facility in MS comparing two start up curves of the two different temperature sensors.

I compared the temperature rise over the first 45C  from start-up temperature.

Turbokraft installation bracket/adaptor plus Porsche CHT sensor (£305 delivered cost from Arizona)) took 227 seconds

Kenso sensor (£15 delivered from the UK) took 187 seconds.

That is a big difference, The Kenso is getting me to maximum Control Pressure in 111 seconds versus 168 seconds for the TurboKraft solution.

This must mean that I will get much faster responsiveness to changes in CHT under load which microsquirt will react to.

Also having a much more cost effective solution means I can now complete my costings for full installation !

 

 

Colour coding

Light Blue and yellow - target control pressure and actual control pressure (v time along the bottom) - see how closely the eWUR maps the programmed in warm up curve ie. it does what it says on the tin !
Green = CHT (please note that the control pressure and idle valve duty are directly linked to the CHT rise hence the importance of rapid and accurate CHT response during warm up)
Light Pink = Idle valve duty (fully open corresponds with 70% duty and fully closed is 30% duty)
Violet = MAP (please note this is absolute pressure in kpa so the lower the absolute pressure the higher the vacuum ie. you can see engine vacuum increasing with time as the engine warms up, obviously the open idle valve reduces vacuum in the early stages )
White = rpm
Red = AFR (Please note that the sudden hike in AFR straight after start up is due to the delay while the internal heater in the Bosch UEGO sensor has to get up to temperature before it will start reading)

Edited January 30, 2021 by carrpet

[carrpet]

Further opportunities to take cost out

1. Find a UK supplier for a Dodge Viper injector modified to 52lbs of fuel per hour
2. Use a round relay socket that accepts square relays (flightlead404 on Pelican manufactures these)
3. UK source Toyota fuel pressure damper
4. I suspect you could still get MS for £350 if you shop around and negotiate

Edited January 31, 2021 by carrpet

[carrpet]

Updated costs

[carrpet]

Now the fun begins

The first stage in tuning is to generate an AFR table from the engine performance specifications.

This is a reference table for the tune and represents AFR targets across the full range of RPM and MAP values. Remember I have a Turbocharger so my maximum MAP is going to be 1.8 bar or 180 kPa if I have an original turbo wastegate spring which opens at 1.8 Bar.

The next tool is the Fuel VE1 table which starts life as 100% across the board.

As soon as the AutoTune button is turned on, the VE1 table values are adjusted live by Tuner Studio, in the car, by either adding or subtracting fuel for each incremental engine load condition, gradually filling in the matrix. There are some pre-conditions which need to be satisfied before auto-tune starts like a minimum CHT of 100C and a minimum rpm of 1200 (both of these conditions are set by the user). This prevents auto tune tuning areas of the table you don't want it to influence. Remember I have already "tuned" start-up and warm-up and they are controlled using specific tools designed just for this purpose. Auto tune is for an engine at normal running temperature. With the laptop on the passenger seat you can see the VE1 table being adjusted as you drive "live". It would really need a video to show this process although as you will see when we get there, the new VE1 table will be very different to the 100% across the board.

Experienced MS users believe it takes 3 runs to get this close to a good state of tune.

Edited February 4, 2021 by carrpet

[PeterK]

If the laptop is in the driver's seat, are you sitting in the passenger seat 🤣

I'm enjoying your journey - thanks for the steady updates

[carrpet]

4 minutes ago, PeterK said:

If the laptop is in the driver's seat, are you sitting in the passenger seat 🤣

I'm enjoying your journey - thanks for the steady updates

Good spot !!