Waking up after 12 years in the barn – needs more racing DNA

[Benutzername]

A brief recap: 

Sixteen years ago, I promised a very good friend that I would buy his Lupo GTI if he ever wanted to sell it. In 2017, I sold my Lupo after 14 years and invested time in other projects. But a Lupo GTI is something special, after all right? So I reminded my friend of the promise I had made back then.

A few words about this Lupo GTI: He acquired the vehicle in 2008/2009. We also became friends at that time. Numerous modifications were made until around 2012, resulting in a road-legal vehicle that was almost a Cup Lupo (incl. original Cup Lupo suspension). But see for yourself the pictures from before 2012, including performance charts. 

He had put his Lupo GTI in a barn for the winter sometime in 2012... but he never woke it up again.

 

In March 2024, we visited the Lupo in the barn to assess its condition and discuss selling it. It was in worse shape than expected. A squirrel had made the engine compartment its home at some point, but I was determined to get this gem back on the road nevertheless. 

 

April 2024 - Vehicle recovery 

The tyres were barely inflated and manoeuvring in the barn was a feat of strength. In the end, the Lupo was on the trailer and we drove off. 
The next day, the GTI got its first wash in 12 years after I vacuumed tons of nuts, seeds, grass and straw out of the engine compartment. 

 

In the days and weeks that followed, many parts and operating materials were replaced so that the engine could be started. This involved draining all the oil and replacing it, replacing the coolant, new battery and installing a new original VW timing belt kit. The latter is particularly important because we have further plans for the engine.  

After the engine had built up oil pressure with a few turns, and the tank and all fuel lines had been flushed with fresh petrol, the engine started up fairly quickly. The next step was to deal with the exhaust. It was very badly rusted. The rear silencer had to be retained (due to German MOT requirements) and was therefore cut off, the remaining pipe was welded lengthways and fitted with a V-band clamp. The remaining exhaust pipe was replaced with stainless steel and additional V-band clamps were installed for easy installation. For further modifications to the engine, the exhaust gas temperature sensors have already been welded/built into the stainless steel manifold.  

Then new brakes (EBC turbo groove + EBC yellow), new wheel bearings, the brake calipers were repaired, new brake fluid was added and brand new tyres. In addition, a new Helix clutch with sintered pads, including a lightweight flywheel, was quickly installed. Since all the suspension bushings were either hard or defective anyway, the wishbones were upgraded directly to the technical standard of the BTC Lupo. This means, among other things, stiff Unibal and brass bearings. The rear axle was also fitted with Lupo BTC Unibal bearings.

September 2024 - MOT appointment

After various test drives and many minor repairs, the GTI received its MOT certificate. The next day, it was registered and could now be driven legally on the roads. 

Resuscitation successful! 🦾

Winter was approaching and I still had a lot to do! Let's see what comes next...

 

Edited September 19, 2025 by Benutzername

[mk2]

Nice! 👍🏻

 

[Benutzername]

End of the 2024 season – initial modifications for the coming season

A relatively simple but truly worthwhile modification is a short shifter. Together with the reinforced clutch and lightweight flywheel, shifting gears becomes pure joy! Like a 5-finger dual-clutch transmission. 😅 

If the engine needs to rev a little higher, the gearbox ratio must also be shortened accordingly. I shortened the axle ratio to 4.53:1 and also modified the gearbox housing slightly. Firstly, I milled small slots so that the oil circulates better through the chamber with 5th and 6th gear. I also installed a temperature sensor, a differential lock and, of course, replaced all bearings and synchroniser rings. 

 

Winter 2024 - shortly before Christmas, the post arrived  

 

 

 

 

[Benutzername]

What was the plan for 2025?

1.    Build a Simos->Haltech ECU adapter for plug-and-play change, including integration of a lambda controller with fast Bosch LSU4.9 sensor via Haltech CAN and a thermocouple amplifier for 4x exhaust gas temperature close to the exhaust ports, also from Haltech. All engine and vehicle peripherals should continue to be usable (except for lambda sensors), including fully functional VVT

2.    Set up the ECU, including additional sensors and basic data for the initial start-up

3.    Calibrate all necessary characteristic maps for everyday use with the series engine

4.    Hardware updates to the engine: machining/porting the cylinder head, new camshafts and valve springs, larger throttle body (keep drive-by-wire), larger injectors, 4 bar fuel pressure regulator to achieve approx. 165 hp.

5.    Update ECU mapping

 

 

We start with the Haltech ECU and an adapter instead of the SIMOS:

The Simos housing is too flat to accommodate the many cables. So I took a larger Bosch housing and only used the Simos plug. In the upper part, I built the socket for a relay, which I needed in addition to the vehicle's own relays to supply power to the ECU. After drawing up a detailed plan in Excel, I connected the corresponding pins to all the necessary sensors, throttle valve, pedal, ignition, injectors, etc. with the Haltech wiring harness and then stuffed them into the housing. There are also a bunch of other cables for additional sensors, such as oil pressure, oil temperature, transmission oil temperature and other funny gadgets. 

Now all that was missing were the additional amplifiers for the lambda sensor and the thermocouples. I installed the appropriate connectors for the external additional sensors. The adapter then looked like this, and I carried out initial tests on my desk and was basically ready to install it in the vehicle. That took care of point 1. Unfortunately, it was still the depths of winter...

Edited September 19, 2025 by Benutzername

[Benutzername]

2. Initial set-up with the Haltech ECU

Before anything can rotate, the ECU must know how the crankshaft is positioned in relation to TDC. In the second step, a ‘home signal’ is set up so that the ECU knows which TDC it is in, because we want to inject fully sequentially. Quickly adjusted with an ignition gun, then set up the remaining calibration values for the sensors, and we were ready to go with the initial start-up. It took a while for it to finally start, but the engine ran without any issues... for the moment....

It wasn't the first AVY I had mapped. This one behaved a little strangely. It wanted a lot of fuel injected, was very agile... which is ultimately a good sign, and old performance measurements already showed just under 150 hp. Later on, it would become clear why... 

During the tuining, I also tried a 60 mm throttle body instead of the original 52 mm one. At first, you would think that there would be advantages purely from a geometric point of view. Unfortunately, I was disappointed. It just makes loud noise and there was no noticeable difference at wide open throttle. So the intake manifold restricts the flow so much that increasing the cross-section of the throttle body to 60 mm has no effect. So out it goes again!

 

 

 

 

[Benutzername]

3. Improve drivability, adjust suitability for everyday use and increase performance

Once the start-up process works smoothly under all environmental conditions (in spring/summer) and all adaptations and controllers are functioning perfectly, there is initially no noticeable difference compared to the standard application in terms of drivability. However, as everything can be freely controlled, the engine's response can be significantly improved compared to the SIMOS ECU. The engine then feels like a different engine. Even the much-maligned drive-by-wire suddenly becomes your friend, because you can try out many different settings and pedal-to-throttle ratios.

The strange behaviour mentioned above became even more apparent when revving above 7000 rpm. The engine remained fairly agile up to 7700 rpm, but the injectors were maxed out already. So I wanted to exchange directly to a 4 bar regulator, but there was one installed already. That was odd and my friend didn’t know something about that too! So I measured the injectors on a injector test bench, cleaned them and measured them again. Before cleaning, they were approx. 235 cm³/min at 3 bar and after cleaning 258 cm³/min. Not bad! That's actually enough for 165 hp with this engine. But it didn't help in the end. So I set the rev limiter to 7250 rpm and completed the basic set-up.

There are, of course, lots of pictures from that time, but let's try videos. 2nd gear and pedal to the metal. 😎

 

 

[mk2]

Nice. I have lots of questions/comments... When I mapped engines, I always ran out of injection time, which sounds like you're also experiencing. It's tempting to go bigger or up the pressure, but what I've found is then it idles really badly, because short injector pulses are difficult to meter. Curious to know which option will work best for you...!

When you rebuilt the box, what were the needle roller bearings like? How tight was the magnesium cover for 5&6? I've heard that it is difficult to get the cover tight with the bearings so,there is no movement (which causes the box to break).

I like the way you used an old ECU enclosure to make a quick swap ECU setup, and so the OEM wiring isn't disturbed. It will be tricky to make it waterproof. Cool though 👍🏻

When you have to get next year's TÜV, will they not ask lots of questions? Here it is easy, but I understand in Germany the examiners are very fussy!

[Benutzername]

47 minutes ago, mk2 said:

Nice. I have lots of questions/comments... When I mapped engines, I always ran out of injection time, which sounds like you're also experiencing. It's tempting to go bigger or up the pressure, but what I've found is then it idles really badly, because short injector pulses are difficult to meter. Curious to know which option will work best for you...!

When you rebuilt the box, what were the needle roller bearings like? How tight was the magnesium cover for 5&6? I've heard that it is difficult to get the cover tight with the bearings so,there is no movement (which causes the box to break).

I like the way you used an old ECU enclosure to make a quick swap ECU setup, and so the OEM wiring isn't disturbed. It will be tricky to make it waterproof. Cool though 👍🏻

When you have to get next year's TÜV, will they not ask lots of questions? Here it is easy, but I understand in Germany the examiners are very fussy!

Thanks!

Counter question: semi-sequential or fully sequential injection? Which ECU are you using?
And what size injectors are you considering? Fuel pressure regulator with or without manifold pressure reference? What kind of fuel map configuration do you use? By MAP or throttle angle? VE or injection time? Do you use SOI or EOI?

380cc with 4 bar is no problem for idling, even with semi-sequential injection. Later, when I have my story complete, you will see the current config.  

And yes, TÜV is a problem, but I can very quickly revert the condition to an acceptable set-up, lets say.  

The only problems I ever had with these gearboxes were that the 5th and 6th gears got too hot and then these cheap plastic rings melted. 

Edited September 19, 2025 by Benutzername

[Benutzername]

3.1.  Other nice Haltech gadgets

I still had my old speedometer with shift light and rev counter up to 10k, but I was also interested in lots of other values that I wanted to see while driving.
So I quickly opened up Haltech again and ordered some new toys. 

Certainly not the cheapest way, but definitely the coolest. So I started building a cable adapter for the Haltech Dash as well, so that I could swap the original part back and forth with the Haltech Dash. The adapter basically only included the basic functions: lights, indicators, speed, handbrake, etc. The rest can be freely programmed via the Haltech CAN. 

Of course, the Haltech Dash doesn't fit well into the cockpit opening, so I took an old dash, covered it with carbon plates, painted it black and installed it in its place. The Haltech Dash now sits in front of it. 

When I first got the Dash and installed it, Haltech had just migrated the software for the display to their NSP software. At first, it was still very buggy and you could hardly change the design. But a few weeks later, after an update to the Haltech software, it finally worked. I like the old VW Motorsport logo, and I initially used a photo of the radiator grille as the start screen.

 Later, I swapped the image back for a logo and added further features such as all the indicator lights. 

 Another useful feature was an 8-button keypad, which is also connected to the ECU and display via Haltech-Can. There, I can switch VVT on and off, select other mappings, start/stop data logging, switch on the radiator fan or scroll through the display... lots of possibillities!

[mk2]

Really nice! I have never swapped an instrument cluster. Is it difficult? How do you setup all the sensor input signals- is every input software definable, like pulse, PWM, voltage, current, difference voltage, serial input like I2C...? That must take a lot of time to get right!

And I guess if you have a spare cluster PCB, you could use it to make a plug harness to keep the loom original and use the original instruments if you sell the car one day...

Edited September 20, 2025 by mk2
Just noticed the answer is yes! That is why you have the old dash...

[Benutzername]

2 hours ago, mk2 said:

Really nice! I have never swapped an instrument cluster. Is it difficult? How do you setup all the sensor input signals- is every input software definable, like pulse, PWM, voltage, current, difference voltage, serial input like I2C...? That must take a lot of time to get right!

Let's put it this way: I've more or less turned my hobby into my profession. I work as a development and measurement engineer in the automotive industry for an independent service provider. We assist with engine development at AMG, Porsche, etc.. That's why this kind of things is certainly easier for me maybe. But it's no harder than getting a third-party ECU up and running. The same software interface is used as for the control unit. The configuration is very similar. Communication between Dash and ECU only happens via CAN. Saves a lot of cables and is super convenient. 

And in the end, it's just a matter of repetitive patterns when it comes to the sensors to be used, etc. 

2 hours ago, mk2 said:

And I guess if you have a spare cluster PCB, you could use it to make a plug harness to keep the loom original and use the original instruments if you sell the car one day...

Yes, that's the point, correct! Stay flexible...

[Benutzername]

19 hours ago, mk2 said:

How do you setup all the sensor input signals- is every input software definable, like pulse, PWM, voltage, current, difference voltage, serial input like I2C...? That must take a lot of time to get right!

I will try to answer your question more precisely:

When using the Haltech Dash as a standalone system, you must first create and configure all sensors or signals that you tap into. Yes, this requires some effort. However, for a third-party provider, it is extremely flexible and self-explanatory. If not, there are small question marks everywhere, and the help file is also invaluable. 

That's an overview of what the Dash IC7 can do as a standalone device. In addition, there are vehicle functions and then the ‘generics’. This means you can integrate everything you want, at the latest. 

Since I also use the Haltech engine control unit and almost all sensors are already connected there, I can easily retrieve the values via the CAN signal. The configuration is then super simple!
Find the location where it should be displayed, select the channel from the long list of possible channels... that's it! After that, you can customise the appearance and configure thresholds and alarms. It's impressive what you can do with it. I think it ended up costing €1,100, and it's definitely worth it. 

Here, for example, you can configure the alarms. Also self-explanatory... (however, I currently only use it for oil pressure)
In operation, it looks like this: 

What also helped me personally with mapping was a page with control unit values, so that you don't have to constantly look at your notebook. Unfortunately, there are many German words, but I think it is still understandable. 

The software can be downloaded free of charge from Haltech. So if you want to play around with it...

Edited September 21, 2025 by Benutzername

[Benutzername]

4. new engine hardware

I didn't want to overdo it this time, so I bought a second cylinder head and valve cover online to work on. I also bought a set of 268°/260° Catcams and other valve springs. 

So I started with the cylinder head. When it came to porting, I really only wanted to do what was absolutely necessary. Just smooth out the intake ports and make the transitions to the valve seat rings smooth.
The exhaust ports needed to be made a little larger.
You can see the result here:

In the meantime, I've killed the ignition coil. As is well known, none of the known aftermarket replacements have any power. So I bought a ignition coils from the 1.4TSI as a ‘tuning variant’. Supposedly more ignition energy with the same dwell time level. The combustion seemed a little smoother... at least I found a technically sensible replacement. And they're red... which I thought was pretty cool.  

More or less by chance, I came across an extremely interesting component from the rallying days of the early 2000s. Who recognises it? 

Correct! The Holy Grail! The plenum from the Polo Super1600 intake manifold. That changed my plans for the engine somewhat!

Here is an example of an S1600 engine from 2005:

[Benutzername]

4.1 new hardware - new cylinderhead with new springs/retainers and 268/260° Catcams

I will show the project surrounding the S1600 intake manifold separately. So for now, let's continue with the new cylinder head, which has already been adapted to the S1600 intake manifold, , as you can see here.

New valve guides and valves were also added. The lower seat were mashined for the stronger valve springs and the valves were resealed. Furthermore, the cylinder head was milled flat once with a cutter head on the milling machine. The new cylinder head was thus ready for installation. 

Replacing the cylinder head is not a difficult task, of course, but I noticed something interesting when I looked at the old parts. 

Someone installed Schrick camshafts there. According to the numbers on the shafts, they were 256° on the intake and 252° on the exhaust. The exhaust shaft also had a 9.9 mm valve lift, which is surprisingly high on a standard cylinder head. Funnily enough, my mate, the previous owner, didn't know anything about it either. 

But then it's no wonder why the GTI already had almost 150 hp and I quickly ran into problems with the size of the injectors. But regardless of that, I had already purchased matched 315cc injectors. Red ones too... these now also run with a 4 bar fuel pressure regulator.  

After minor adjustments to the ECU, the engine ran smoothly at idle speed.

Subsequent test drives showed that the selected injector size is significantly better. At 4 bar, the injectors are 356 cc, which is 19% larger than the standard injectors at 4 bar. The fuel maps show that the torque appears to be slightly lower at low to medium speeds.  However, a significant increase can be seen from 5000 rpm onwards. The torque peak was previously at approx. 5500 rpm, but is now just under 7000 rpm. The engine runs much more powerfully at high revs and very smoothly up to 7800 rpm (current rev limiter). The duty of the new injectors is 72%. Therefore, any further changes that would consistently lead to increased performance would in turn necessitate even larger injectors. 

 

[Benutzername]

4.X Installation of the S1600 intake manifold and initial tests

Of course, it wasn't necessary, but when you get the opportunity, you just take it! With the current exhaust manifold, the S1600 intake manifold will certainly not work properly. So I gave ChatGPT the camshaft data sheet and a few other basic input parameters, and the result and explanations were astonishing. I will give the AI the S1600 dimensions to recalculate, in order to refine the exhaust manifold design. I'll definitely give it a try. Winter will be long again!

 

But now to the installation of the S1600 intake manifold:

The engine of the Rally Polo sits much lower in the engine compartment than the slanted AVY engine in the Lupo GTI. I knew beforehand that it wouldn't be that easy. That's why I did a test fit while building the intake manifold. 

With the throttle body fitted, it becomes particularly clear how crookedly the engine hangs in the Lupo. That's not really a problem. Lowering the engine a bit also makes sense in terms of driving dynamics. 

At least it was clear what needed to be done. I cut off the abutment for the Bowden cable. We had lowered engines before in the GTI, but before building a new bracket, I lowered the engine bit by bit on the right-hand side (in the direction of travel) using long screws and thick washers. This meant that the engine was still not horizontal, but there was significantly more space between the throttle valve and the bonnet. The bonnet could then be closed without any problems and overall it looked a bit better. 

At the same time, I connected everything roughly and adjusted the software slightly to accommodate the 380 Bosch injectors and the Bosch pressure/temperature sensor in the plenum.

After a few attempts, it started up. The cold start was relatively rough at first, but it quickly improved. Once it had warmed up a bit and the idle speed was stable, I took a first video. It sounds different through the huge plenum. Funnily enough, the exhaust fumes smell completely different too. 

 

I then took it for a quick spin to get a feel for it. It has become very thirsty, which is a good sign. However, there is still a bracket missing under the plenum to support the heavy weight and dampen vibrations. So now I have to build this bracket before I can seriously adapt the software on the road. The individual parts are already finished and the intake manifold has already been removed. I think we will be able to weld the bracket on Saturday. Then things will get serious!  

 

[Benutzername]

5. Update Haltech mapping – For a more reliable and powerful engine

Unfortunately, bad weather and limited time due to work meant that I couldn't experiment much with the S1600 intake manifold. The season was coming to the end and I really wanted to get the most out of the Rally intake manifold before winter set in. 

On the one hand, I still had the problem with very high exhaust gas temperatures (close to 1000°C). In the end, this was solved very quickly. They were simply pushed too far into the manifold pipe. I already knew that the measurement must be incorrect because the spark plugs did not look like they had been exposed to excessive temperatures and I had also endoscoped all four cylinders several times. Everything looked perfect, but I was still unsure.

I have already invested a lot of time in lowering the temperatures by the application. Even though it wasn't the solution in the end, it helped to significantly improve the mapping and get to understand the intake manifold. Seeing and hearing is always better, so here are two videos from the ECU tuning session uphill in the Harz Mountains. From around 6300 rpm, the sound changes noticeably. At 6350 rpm, the intake manifold resonance kicks in. From this speed onwards, the engine becomes really aggressive.  But at that time, there was still the issue of exhaust temperatures that were supposedly too high. Therefore, we were still driving very cautiously. 

 

 

On 18 October, the problem with the thermocouples was solved and I was able to concentrate more on increasing performance again. The intake sound is really wonderful above 6500 rpm. It sounds slightly aggressive, but not like with ITB, much brighter and hissier. 

The resonance in the intake manifold is really intense. At such resonance points, the engine often requires significantly more fuel. The mapping was still slightly too rich, but at this point it was always too lean. I have now adjusted the support cells of the characteristic map so that I have a higher resolution around this speed point. Unfortunately, I have run out of time to test this.

What you can also see clearly in the log file is that when you suddenly open the huge throttle valve, it initially runs lean. Fortunately, this can be fixed relatively quickly. But it looks like it won't be possible this year. 

There's still a small chance. The weather forecast predicts dry weather for Friday. Let's see... if it works out, I'll take the GoPro with me.  

 

[mk2]

Add fuel!

[Benutzername]

55 minutes ago, mk2 said:

Add fuel!

Yes and no! The goal must be for the ECU to be able to adapt itself in the long term. Therefore, I am only creating the possibility for the ECU to adjust the fuel quantity itself at precisely this speed point, as well as shortly before and after it. 

And as you can see here, the engine is already consuming a lot of fuel and then less again afterwards. So it could also be a timing problem (EOI) within the resonance. 

Edited October 26, 2025 by Benutzername
update

[mk2]

Yeah, looks good. But EGTs @ 1000c?!

[Benutzername]

21 hours ago, mk2 said:

Yeah, looks good. But EGTs @ 1000c?!

No - as I explained, it was due to not correct installed thermocouples. After installing a set of new ones, the temps became more realistic. At part load  they are around 700°C and @ FL around 770°C. This is with my EGT optimised mapping.  

[Benutzername]

Before winter arrived sooner than many had expected, I was able to take a few videos with my GoPro. So here are a few impressions:

 

 

 

[tchouki85]

Impressionnant 🥰

[Benutzername]

The first stage of the Rally intake manifold's further development does not change the fact that a suitable air filter box still needs to be installed. Well-known airboxes do not fit properly, or rather, their cross-sections are simply too small. The Lupo BTC already managed with a very small water cooler in terms of surface area, and I have already had experience in this area in the past. 

That's why I installed a cooler with a width of only 430 mm, which suddenly left plenty of space for an air filter box that is open at the front and offers a very large surface area, most of which is exposed to the airflow.

 

The air filter itself is actually from a Golf MK5 R32. The dimensions fit perfectly next to it to make every square millimetre usable. The pipe connection is 89 mm. There will also be two pipes that extend to the throttle valve. 

 

This is what it looks like at the moment. Next week, I will make some minor adjustments in the engine compartment, as I have also removed the air conditioning system in the meantime. After that, the front will be reinstalled on the vehicle. I am really excited about it!

 

 

 

 

 

 

 

Edited February 7 by Benutzername

[Benutzername]

The first fitting showed that the filter box was a little too deep and the positioning of the 89 mm connection was not ideal. So I made some minor adjustments and now it fits exactly as I had imagined. To connect to the throttle, I printed a curved pipe that tapers from 89 mm to 76 mm. The connection to the throttle is then made with a 45° reducing elbow. 

Edited February 16 by Benutzername

[Benutzername]

Pictures say more than words...

This could be really cool!