CrazyDave

That's a bit more like it. 3" is a far more sensible size for the pipework. The water in / out tappings could do with being higher up though? I have a bit of trouble with air locking in the top of the CC, but I guess a couple of tapped holes higher up the end plate and a blanking plug is easy enough. I've always seen these mounted upright, water in at the bottom and out at the top, so it doesn't airlock. Not sure if you'll have space for that though? Sounds like a trial fitting job to me :) Have you got Aquamist? Mmm doin turbo time eh... not good.

4" Inlet and outlet pipes seems a little over the top! Maybe if you contact them they could make them a little smaller. The water feed points would need to be on top to prevent airlocking so the air inlet / outlet connections are on the wrong side. They would be better 90deg either way. And better still the hose tails at 45deg to the cooling core. I've seen these things used on a few engine test beds and they work very well, but they are fed with 10deg chilled plant water. I ran 440cc injectors with Megasquirt and 8psi and they are a bit big for idle fuelling (running at about 1.7ms~). 330cc would be a better bet as Kev said already. Get some pics on Karl, £1500 turbo kit is a bit of a bargain.

Nice work Rob, coming along well! Have you sorted out a wideband AFR meter? Can't stress enough how important this is, and don't want to be a doom monger or anything. But you will destroy your engine very quickly if it runs lean on boost. Your fuel system is going to be loaded up a lot more than it has ever been before, so fuel pump, blocked filter, low voltage at the pump, all or any of this could cause the fueling to go lean. Innovate LM-1 or LC-1 and gauge would be my choice. Mr Beige got his from Bill Shruvington (a megasquirt guy in the UK who is an Innovate agent). The LM-1 is good as you can also datalog whilst driving, do some gentle runs up the road and see how it's running. Then you can build it up a bit at a time, checking the AFR after each run and making sure its nice and rich 12-11:1 ish. Could save you the cost of melted pistons?

Now that sounds like the fuel pump. It may have come loose in the tank? Check it out ASAP as this can cause a very lean running condition and you may do some damage if under heavy load at the time, even though you've got knock sensors.

Yes it has two analog outputs which can be scaled. My idea for the lambda control was to use the wideband probe as feedback to the piggy box, then have my own map for target o2 and adjust the switching point of the narrowband signal to Motronic. This all depends on how big the Motronic authority is and what adjustment I can get away with. Should be good if it will allow enough to richen the part throttle just a bit. I must admit, having driven an OBDII Golf, the general refinement is far better than OBDI. Maybe a future plan :)

Oh and I forgot that my 4 year old daughter always wants to go out in the Corrado, she wants one when shes old enough, as long as we can paint it pink :lol:

Nice thread! Speaking for myself (being closer to the 40 mark than 30!), the Corrado is a really special car. Right from the first time I drove one back in 1995 I was in love. Mind you it was a special day organised my the Grantham VW dealer with all the demonstrators up from Milton Keynes for test drives. Looking down the list of cars to try was exciting enough, but seeing a Corrado VR6 on the list was the best! Give them your driving license and they gave you the keys. Anyway, my friend and I took the C out around the 'Pre-set' route and trashed the living daylights out of it. The grinning didn't stop for days! Then if you get bored of just having a standard Corrado you could always put a turbo on it and scare yourself silly every day like I do :lol: Keep it, enjoy it while you can, it's a rare and unique machine.

Sure would. The general idea is that you can filter out anything that upsets the balance for Motronic. Trouble is if you go that route you'll be wanting your BEGI back! Guess I'll have to make you an offer quick then?

Yep, display will be available in the car. Anything wired in can be displayed (boost, MAF, O2, speed), and it has a pair of serial ports if I can figure out the VAG ECU login I can get all the ECU channels (just like vag-com). Also has an SD card interface for some serious logging potential. Not certain, but I guess load is derived for engine speed, TPS, and MAF. Air flow is perhaps the biggest input, as it would be an indication of work being done. Just been out in the VR, I reckon its much better than it was on squirt. Need some more timed runs now!

Always wondered how the pistons went back in on a VR! Shouldn't be long now then Chris, a man of your talents will have that back together in no time. Shame the builders aren't as rapid!

Vagcom & AFR logging mainly plus a bit of scope work to see whats really going on with the MAF signal. The MAP sensor has most of the answers, as you close the throttle at anything but idle you get a very high vacuum, this and boost in the inlet is the cause of the MAF spike when the BOV dumps, so the MAF gets clamped to an idle value when this happens. The clamp has a bit of variation against engine speed just to keep Motronic happy its a valid signal, it throws a MAF implausable if you don't do this. The transfer of the signal is very fast (dsPIC microcontroller running at 40MHz!) as the sampling is running at 1kHz. More than enough for the old Motronic ECU :) . A filter kicks in at low engine speeds to stop the turbine disturbance I saw on the scope you also don't need full MAF rate at idle as it's mainly TPS that kicks the engine into life. The sampling could go much faster as the dsPIC has hardware digital signal processing, but haven't needed that yet. It's funny, but I needed the understanding from running Megasquirt to get to the bottom of the Motronic problems. In fact one of the Megasquirt problems is that it's entirely MAP sensor based, it needs a TPS - MAP blend to get best results. MAP reacts as a result of the throttle changing, but TPS changes first, so best response needs TPS as the primary input not MAP. MAP should be used as compensation for boost conditions. Wow thats how DTA works! Thanks Andy, it's even quicker now! The guy I buy tyres off sees me as the cornerstone of his business. Sorry, it's already been earmarked for a Mk2 Golf Turbo project that a friend of mine is doing 8)

That's that panic over then :lol: What you need is a nice slow noisy smelly diesel to really make your VR feel great extra special!

If its a standard car then no need for this stuff. 'When' you start to mod thats when the delicate balance VW set it all up for gets disturbed and needs fixing. Now you could have it mapped and sort it that way, but I don't like the idea of MAFless remaps and £££. Standalone is the very best way to go for serious changes like turbos and supers, but ££££. When I get the lambda trickery sorted we'll see how it feels, may be a good mod for a standard VR. I know mine drove really well with a bit more fuel at part throttle on Megasquirt. Gets rid of that very slight hesitation on part throttle that all OBDI cars have. OBDII is way better as it has wideband O2 (I think) and lots of other nice stuff like staggered injection timing. That's why Golf VR's make 190hp when RR tested with only 2.8lts (and a Corrado needs 2.9). OBDII would still need the MAF sorting for a turbo / BOV combination though. CrazyFilter mmmm catchy :lol:

Michael, car looking awesome as always! Wish mine looked that good :lol: , but I do spend to much time under the bonnet.

Bit of an update: I've been doing some experimenting with using a standard Motronic ECU to run my VRT. At this point the Megasquirt has been put to one side due to its lack of sequential injection. This is what's fitted at the moment... Standard OBDI ECU (Late VR CP model). Standard mapping. Standard Injectors. BEGI rising rate regulator. Carbon cannister and vapour recovery valve fitted. Breather connected. ISV between post CC boost pipe and shortrunner. Innovate LM-1 AFR meter. Standard narrowband O2 sensor. Inlet pipework: Cone filter behind bumper at the side - straight pipe run through standard MAF housing - 3" to 4" transition pipe - Turbo - 2.25" Pipework - Tial 50mm BOV - Charge cooler - TB - Shortrunner Manifold (CrazyDave special :) ). Exhaust: Cast manifold - Turbo - 3" SS downpipe - narrow & wideband lambda - No cat - Jetex 2.5" Setup the BEGI with a pressure gauge as per the instructions. Initial pressure set as per instructions. Slight further adjustment needed for NA to boost transition. Then a couple of turns on the gain valve whilst watching the AFR. Now set at 12:1 and pretty steady right across the boost range. Now the above runs very well. But a few problems with the BOV during gear changes causing over fueling then cough when back on the gas (only slight). So some 'elastic trickery' called for! I came up with a small box of electronics to intercept the MAF signal. This also has a MAP sensor connected to the inlet manfold. The MAP detects the condition just before MAF clipping occurs and filters it out, replacing the signal with a simulated NA engine throttle closed MAF signal. It also needed a bit of filtering to stop something I noticed with a scope attached, I think it was low speed turbine disturbance of the MAF signal. The MAF signal had that 'wab wab wab' oscillation you can hear in the exhaust note? No problem for the signal processor. TPS should say MAF in the above picture, I started with TPS but decided MAP would be better. http://www.d2designs.vispa.com/videos/M4V00629.MP4 Sorry no reving action, couldn't reach the pedal while holding the camera :lol: Anyway the result.... super smooth VRT running on standard Motronic ECU! Gearchanging is now crisp, overfueling has gone, no popping and banging, behaves itself in all driving conditions. If I was being really fussy (which I am) it could do with a bit more timing down the bottom end, or maybe fuel? Future plans: Trim the narrow band sensor from the wideband sensor. The turbo setup could do with a bit more fuel on part throttle. I plan to trick the ECU into thinking it's running lean when it actually isn't by simulating a lean narrow band signal. This should make the idle a bit steadier, as Motronic trims the idle fueling very hard for emissions.

Hope you've got a friendly MOT man, emissions are soooooo close without a cat!

Sure is a nice example, lets hope it goes to a good home. Good job my bank balance isn't very healthy at the moment! Still suffering from turbo expenditure :lol:

Bit of an update: I've been doing some experimenting with using a standard Motronic ECU to run my VRT. At this point the Megasquirt has been put to one side due to its lack of sequential injection. This is what's fitted at the moment... Standard OBDI ECU (Late VR CP model). Standard mapping. Standard Injectors. BEGI rising rate regulator. Carbon cannister and vapour recovery valve fitted. Breather connected. ISV between post CC boost pipe and shortrunner. Innovate LM-1 AFR meter. Standard narrowband O2 sensor. Running fixed 8psi boost. Inlet pipework: Cone filter behind bumper at the side - straight pipe run through standard MAF housing - 3" to 4" transition pipe - Turbo - 2.25" Pipework - Tial 50mm BOV - Charge cooler - TB - Shortrunner Manifold (CrazyDave special :) ). Exhaust: Cast manifold - Turbo - 3" SS downpipe - narrow & wideband lambda - No cat - Jetex 2.5" Setup the BEGI with a pressure gauge as per the instructions. Initial pressure set as per instructions. Slight further adjustment needed for NA to boost transition. Then a couple of turns on the gain valve whilst watching the AFR. Now set at 12:1 and pretty steady right across the boost range. Now the above runs very well. But a few problems with the BOV during gear changes causing over fueling then cough when back on the gas (only slight). So some 'elastic trickery' called for! I came up with a small box of electronics to intercept the MAF signal. This also has a MAP sensor connected to the inlet manfold. The MAP detects the condition just before MAF clipping occurs and filters it out, replacing the signal with a simulated NA engine throttle closed MAF signal. It also needed a bit of filtering to stop something I noticed with a scope attached, I think it was low speed turbine disturbance of the MAF signal. The MAF signal had that 'wab wab wab' oscillation you can hear in the exhaust note? No problem for the signal processor. The box has a little screen so you can see whats going on. The numbers are just internal digital values at the minute, but can be calibrated to give boost pressure / vacuum, MAF volts, AFR etc. TPS should say MAP in the above picture, I started with TPS but decided MAP would be better. http://www.d2designs.vispa.com/videos/M4V00629.MP4 Sorry no reving action, couldn't reach the pedal while holding the camera :lol: Anyway the result.... super smooth VRT running on standard Motronic ECU! Gearchanging is now crisp, overfueling has gone, no popping and banging, behaves itself in all driving conditions. If I was being really fussy (which I am) it could do with a bit more timing down the bottom end, or maybe fuel? Future plans: Trim the narrow band sensor from the wideband sensor. The turbo setup could do with a bit more fuel on part throttle. I plan to trick the ECU into thinking it's running lean when it actually isn't by simulating a lean narrow band signal. This should make the idle a bit steadier, as Motronic trims the idle fueling very hard for emissions.

Probably don't need the canister as it's just the extra air thats needed, but I guess a percentage of the air has fuel vapour in it so might not be quite right without. I can see ££ signs in your eyes though! Watch out if it's been mapped.Vince levels the maps so it might not have any effect. It's amazing how much Motronic has to cope with and it does a fair job most of the time. Usually takes a few changes before anything starts acting up.

Lots of people on hear complain of VR6 idling problems, cutting out etc etc. Now this is usual ISV, MAF etc but here's something I discovered today! I've had a fair bit of this on my own car just lately, and I always suspected it was the modifications I'd carried out. Well some of it is but I had a major breakthrough today that seems to have cured it! Motronic ECU's have a concept called 'calibrated air leaks' it has 4 separate MAPs to handle these situations. These are things that happen during certain conditions the engine encounters ie vapour recovery from the carbon cannister, the engine breather etc etc. Now when I fitted all my turbo stuff, I happily unpluged all these things thinking I wouldn't need them, plugged up the holes and blanked off the pipes etc. This is when my problems started. The ECU expects air to leak under certain conditions and if it doesn't it'll try to conpensate but not fast enough and by then the engine has stalled. Today I refitted my carbon canister and valve (with a slight mod to cope with having boost) and crank case breather pipe. It now idles perfect, doesn't stall when hot after driving hard etc. Great :) So if you are having problems with cutting out on a standard car. Have you removed the carbon canister? Removed the breather and dangled it down the back of the engine? These are all air leaks that the ECU expects to see, and will have a hard time adjusting if they've gone. Good eh!

Sounds like loss of fuel pressure maybe caused by a dicky fuel pressure regulator or leaking injectors It should hold fuel pressure when switched off but if its leaking past the regulator it'll take a few extra seconds to prime and run correctly. Try switching the ignition on and off 4-5 times before you start up. This will prime the fuel system ready as the pump runs for a second when you first switch on the ignition.

Mmm sounds like lambda probe? Closed loop control isn't active until engine water gets warm. Does your car have one? It should have. But some decat systems don't have a bung to screw the sensor in. Without the lambda probe it's a bit hit and miss as to if it run OK, I've had both with my car. The first time I unplugged the probe it ran great, but when I reset and tried again it ran like a bag of nails. But keep in mind that it could still be an air leak. The lambda probe measures oxygen content, so if the system is leak free it's the mixture it reads. If you have an air leak the extra oxygen is being read instead. So the fuelling will be all wrong.

Not sure if anybodys interested but most of my Corrado related effort is going into this at the minute. It's for modifiying Motronic sensor signals to allow engine mods to be used that would normally upset the ECU. Mass air flow Throttle Position Exhaust Lambda Engine Speed +4 spare inputs These will be intercepted into the above box. Then modified for certain conditions and sent on to Motronic. In addition to the above: 4 analogue outputs. Graphic LCD display for adjusting settings etc So for me it'll allow changing of the MAF signal so that I can run a standard Motronic box, with a BEGI regulator, Tial dump valve, turbo etc and get perfect driving like a standard car. The ECU is completely standard, but the signals fed to it are modified to meet the new conditions. Two of the conditions that I'm looking to adjust: Accelerate up the road and change gear. The MAF reads correctly as the engine revs climb, but when the throttle closes the dump valve opens and releases all the boost (like an air leak). The MAF reading shoots right up, this over fuels and causes the ignition timing to be out for the actual condition the engine is in. The solution is to block the high MAF reading when the throttle closes suddenly. As the throttle closes, the MAF signal will be limited by a new map contained in the above box. The ECU will think it's running a normal engine and set the correct fuel and ignition timing. Turbocharged engines like a little more part throttle fuel than NA. The ECU lambda sensor signal on early Motronic cars is narrow band, like a switch, rich or lean. With a wide band sensor fitted into the standard exhaust bung, the ecu can be tricked into richening the mixture a little at part throttle. Normally the ECU would try to get 14.8 AFR at idle and part throttle. Based on the throttle position the o2 signal can be changed so that Motronic thinks it's controlling to 14.8 but in fact its at 14.0 (or any other value). Early days at the minute but I'll keep you all posted! It could even be used to eliminate the MAF altogether, and replace it with a MAP sensor.

If you found MAF wire problems it would be worth checking the throttle position wiring as the two are closely related to determine igntion timing and fueling. I get a similar problem with my car at the moment (but caused by something very different, a dump valve!), but you'll get a similar thing for the same reason. An air leak would cause the MAF to read incorrectly during throttle closed at higher rpm. MAF is a major input for determining ignition timing so if it gives the wrong reading it'll make it rubbish to drive. Any popping from the exhaust? An air leak would also cause an incorrect O2 reading, and as a result general bad running. Check for air leaks in the manifold / downpipe area and all of the inlet system. Hope this helps? :) edit: and yes it could just be the MAF anyway, but check the other stuff first.

Nice job Stu (not forgetting mic!), doesn't take long when you get stuck in does it! My front panel was a mess, but it cleaned up after shot blasting at work, then some nice stove baked paint and it came up a treat. Very surprised though on a 95 car, a bit longer and it would have needed a new one. Keep up the good work! :)