Skoda Rally Blog

My previous cars haven’t had much done to them originally in terms of shell preparation – indeed, basically nothing. This has had a couple of downsides, mostly in terms of long-term shell life – although seam welding does make a shell stronger and therefore stiffer (which is desirable for any car to handle properly), I don’t think I’m really in a position to tell the difference. Certainly, though, the Felicia was a lot less rattly than the Favorit was, partly because of the shell’s original strength and partly because the cage was fitted with tie-ins to the B-pillars, making it a bit stronger.

But this Felicia is going to be a bit different – firstly, the weld-in cage will brace to the front and rear suspension, which will make a BIG difference to the car’s strength, and for safety too. But it’s still worth seam-welding the shell, as it will make it stronger again, and also make it last a lot longer. And given my current financial situation, I can’t see selling the car in a hurry. Or ever.

So, today I started on the seam welding. If you’re not familiar with the construction of modern cars, most are made with spot welds – usually every 2-3 inches there is a single spot (about 4mm in diameter) which has been welded. Seam welding augments this with a stitch of weld – basically about 50% of the seam’s edge gets welded. But before you can do that, you need to remove the seam sealer, which is a total cow of a job. It’s taken a lot longer removing the seam sealer (using an angle grinder with twistknot wire brush) than the welding has taken, but as you can see from the pictures, a fair old bit has been done. Firstly, in the engine bay, the strut towers and chassis rails plus engine/gearbox mount plates – before.

And after….

The floor needs plenty too – it’s remarkably thin and flexible, and strengthening this will also make the seat mounting more rigid.

Basically, everything between the front and rear suspension will be done. This is the inner bit of the car, and the roof, etc., and then tomorrow the underside should be done too – this is mostly just the suspension mounts, etc., – the rest of the seams are accessible from above, thankfully!

Yesterday, two nice things arrived. My harnesses (Sabelt 6-pt ones with aluminium release – an ebay £200 bargain!), and the rollcage.

Well, I say “Roll Cage”, I mean “collection of bits of tube, some of which are obvious as to exactly where they go and should fit, while others are less obvious, and some are a complete mystery”. This situation is exacerbated by the lack of any picture/diagram/photo of an installed cage, which would make it mostly a piece of cake. The packing list specifically mentioned a set of photos. I looked for said photos. For 20 minutes, checking all the packaging, etc, three times. Phoned CC today, to find that there aren’t any photos of the Felicia cage as “not that many were installed”.

Oh good.

So, I’ve put some of the tubes in place in the car, but some of it is a mystery. There are varying levels of mystery:

Easy – height of the dash bar, direction of the door bars (one is full length, the other is two pieces)

Medium – the rear of the cage – the diagonal section and the rear stays – there are upper and lowers, and the lengths given are a little bit confusing

Hard – the tunnel brace and front triangulation – no idea at all about these at the moment.

I’m waiting for a phone call back from CC so I can pick their chief fabricator’s brain about this. Perhaps I was being a bit naeve, but I did expect a CAD diagram of the assembled cage – that would probably have answered most if not all of my questions….

I’ve got all of next week off, so hopefully by the end of next week there will be some progress. In the meantime, it’s just a pile of tubes and a confused / disheartened owner. You can picture that yourself!

I’ve broken a couple of Felicias (well, 3) in the last year or so, but one thing has always eluded me – being able to remove the dashboard. But to fit a Rollcentre cage, it needs to come out as the front legs are right at the front of the passenger area, and the dash then bolts back in after having large slots cut out. So I needed to get it out, and thankfully, Pete Hopkins (who built N158 UBD) mailed me to let me know that there are 6 bolts that come through from the engine bay, something I’d never thought of. Once they were out (and another 4 screws), it came out without too much swearing. In addition, quite a few kg of soundproofing (really heavy foam stuff) came out, so that went in the bin.

On top of that, the lights, wiring loom, etc have been removed from the engine bay, along with the bumpers and so on. There’s nothing else to do the car now until I get the cage and a new welding set so I can make a nice job of the seam welding and fitting the cage and the seat rails.

So, as I was a happy bunny, I thought it would be a good idea to get on with the new car. Until I definitely decide to sell N158 UBD, and then do so, I don’t have any money spare for the new car. But one thing is cheap, and that’s taking stuff off! So I spent an hour removing all the interior and so on from the new car. Two trips to the dump later…

… I had a nice clean interior. And a much lighter car!

I guess this could turn into a blog of the build of the new car, so a new section is in order.

Well, after some fouled plugs, some hoo-ha with very stale petrol (that had turned the colour of marmalade), and a few other minor issues, it starts! It’s a bit hit-or-miss at the moment because I’ve done very little to it, and there’s no ‘fast idle’ facility yet (I’m not sure it’s going to work as it’s about 1/4 throttle the way I have it set up at the moment). But, it runs, and I’m waiting for my wideband O2 sensor to arrive, ‘cos then I’ll be able to tune the engine precisely, hopefully. However, I did manage to get it idling nicely when it was warm, with what looks and sounds to be a decent mixture. The new firmware for the Megasquirt (Megasquirt & Spark Extra) is a lot more involved than the original MS code I was running before – it has a lot more options, and as a result there’s a lot more to set. However, I think that it will work a lot more smoothly when it’s done this time – partly because it’s an untuned engine, partly because I’m using a single throttle body with a manifold that is a production item, and partly because I’ll have the wideband so I can get the mixture 100% correct.

However, the car is still kinda in bits – I’ve just picked up the other wishbone, so I can get that on…. except I’m going to Newquay for a long weekend, so that’s kinda put paid to 3 days of work. Fortunately I’m quite slack at the moment, so at least I can get on with it next week – I need to as Castle Combe is only 9 days away!

Well, after some traumas with the MegaSquirt (mainly from running low on fuel and filling the fuel rail with air as a result, and then ‘breaking down’ because of it), it’s running, and reasonably well. Got a few other minor issues sorted, like getting an accurate sense of the engine load, and getting various sensors more correctly calibrated. And then took a fair bit of time getting the mapping right, which involved going for a drive with a mate, who set the map on the laptop as we went. It’s not perfect, as the O2 sensor reading is a little erratic, so I’ve set it up rich just about everywhere to make sure that it doesn’t blow up – better safe than sorry, and fine-tuning can take place later on. But the problem with running low on fuel was worrying me, so I’m going to fit a small surge tank under the bonnet, that will be fed by the original solid-state pump, and move the high-pressure pump to under the bonnet, so it always feeds from this surge tank and as a result will not feed from air unless we’re actually out of fuel. This has been made up from an airline regulator/filter unit from Machine Mart (?30) which has been adapted to suit by removing the regulator and gauge and getting hose tails to suit the hoses I’m using. Everything else seems OK, and the car drives reasonably well – it’s more driveable than it was with the carb on, and pulls 5th gear, which it didn’t before, so I can only assume that it’s making more power than before. However, my back is playing up again today, so that’s stopped play – we’ll have to get everything back together in one piece on Wednesday, which will be a reasonable amount to do, but it is do-able.

Hopefully.

Well, today has been rather productive. I got up early, and got everything bolted together – this was actually the first time all the manifold had been in one piece, so that was progress in itself. It all fitted reasonably well, so I thought I should crack on with the rest of the installation in the car. First up, the new fuel pump needed to go on, which wasn’t too bad – getting the old one off was a bit tricky, but the new one came with new connectors and even dirt shields, so it was a quick job to fit, albeit temporarily cable-tied to the bracket – a new one will take about an hour to make, so I’ll do that on Monday.

And then off with the carb, and on with the bodies. They bolted up reasonably easily (just minor filing needed to get everything to line up properly), and then connect everything up – fuel feed and return, and all the connections to the injectors, etc. As ever, parts were needed, so I had to dive off to the MDC, so it took a bit longer than I’d have liked, but finally, I turned the fuel pump on, and realised I needed to do the return up tighter – the jubilee clip had jammed, making me think it was tight. A pool of petrol on the floor said otherwise. After that, it was all good, so I ran the pump for 10 minutes just to make sure, and then left it for a bit so the spill could evaporate completely.

And then, it was time to plug everything in, and start it. And amazingly enough, it didn’t start first time….. but only because I’d forgotten to turn the fuel pump on! Moron! Once I’d done that, it fired up pretty much straight away, although running rich as hell. And then it was just a case of tuning it with the laptop. There were a few gotchas (like loose connections and so on), and the throttle bodies worked loose because the clamps won’t be here until Monday (cable ties are sufficing for the moment), but it was amazing just driving the thing at all – drive for a bit, change settings, drive again, and so on. Just like they said in the manual. I’ve got it reasonably driveable already today – it’s quite weird because the throttle doesn’t need to be ‘blipped’ like it did with the carb – it’ll just drive! There’s still plenty of work to be done, as it’s nowhere near done – it’s running rich pretty much everywhere, except a lean patch at light throttle, low revs, which is like it was before. But I think that can all be dialled out. We will see. And if not, the carb can go back on in about 10 minutes flat, which is reassuring!

Well, since the last entry, I’ve done a fair bit of work on the car. Firstly, getting it MOTd, but that’s by the by. The important work has been getting the MegaSquirt ECU up and running. Doing any project like this it’s always easy to under-estimate the amount of time it takes to do little things to get it all running, and I think I’ve done that in this case. Anyway, first up was getting a fuel return plumbed it – the system that I’m using requires a constant flow, meaning that a new return had to be plumbed back to the tank. As it’s a rally car, this needs to go inside the car (not under the floor) to stop it being ripped off on a forest stage. And that means that it needs to either be braided hose (with no joints inside the cabin) or solid metal. As the other pipes were like that, that’s what I’ve done. A high-pressure pump is waiting to replace the current one – thanks to German, Swedish and French it only cost ?55 for a new pump, rather than the ?200+ I had been quoted by some people…..

The wiring of the MS is made easier by having a relay/connection box under the bonnet, and then an umbilical connection to the MS ECU, which is better off inside the cabin of the car. This is now done, although soldering 18 connections to the ECU end while inside the car (this has to be done as the plug is too big to cut a hole for in the bulkhead to be safe for fireproofing) took a while. And again, getting the power and ground wires connected took more time than I thought they would. But getting it in the car (at the bottom of the dash), and having it power up was a great moment. And then connecting the sensors that I could do – a signal from the coil so the ECU knows how fast the engine is running, the coolant temperature sensor, and the tube for the Manifold Absolute Pressure (MAP) sensor meant that I could run the car and see the MS react as it would do if it were running the car! At last, it all seems to be making sense.

Next up was making the wiring loom, which I wanted to make as professionally as possible, so I made sure it had all the right colour coding on it, and was taped up, etc. It looks pretty good – not home-made, for sure.

And then, fitting the oxygen sensor to the exhaust. I had planned to use a new downpipe, but this turned out to be short (because of having a catalytic converter on it), so instead I’ve got a suitable bit of threaded pipe, and welded that into the downpipe. The wiring for all this uses the standard Skoda/Bosch connectors, so that looks good.

And once again, running the MS with that on and seeing just how off my mixture currently is (with the carb on it) was a good feeling – it finally feels as if it’s going to work. I’ve painted the manifold today (quite a few coats of satin black Hammerite), and hopefully should get it bolted up tomorrow. I’ve made up the bracket for the throttle cable (although I’ll have to make a stop for the pedal as it’s short travel at the moment, and making it longer travel would be too much work at the moment), and I think that it’s all ready to go – obviously there may be a few minor ‘gotchas’, but it should all be running (at least ticking over) by Sunday morning. Which admittedly only leaves me three days to get the mapping right, but given how wrong the carburettor is, I’m hoping that even that will be better than what I have!

I’ve been putting the injection off for a long time. This was mainly because of the most difficult part to manufacture. There are some manifolds available for the Skoda engine to fit side-draught carbs (like DCOE and DHLAs), but they won’t help if you’re fitting motorbike throttle bodies. So the only option there is to make your own. Which sounds simple in principle – after all, it’s just four tubes, isn’t it? Unfortunately there’s a fair bit more to it than that – firstly there’s the issue of spacing – the ports on the head are in two pairs, which need to go to four reasonably evenly spaced ones. This makes things difficult. On top of that, the throttle bodies I have are 34mm I.D. and the ports are more like 25mm, so the pipes aren’t just off-the-shelf items, they’re tapered. Which is a pain to make. And finally (for now), the manifold needs to be strong; the mating faces need to be 5mm or so thick, and the tubes need to be thick enough to be strong, and also be capable of being welded to the thicker plates without problems.

Cutting the plates out proved to take a LOT longer than I initially thought – the two plates that fit to the head took about 3 hours. I gave up using the jigsaw in the end, and used a cutting disc on an angle grinder, and then grinding these rough shapes down to size. And then there was the big plate for the throttle bodies to bolt to (via GSXR1100 carb rubbers), and so on. I won’t go into the details, but it took a lot longer than I thought, and it’s definitely a “Mk.1” version; I think it will be functional, but there are some allignment issues which isn’t fantastic, but it’s the best I can do considering I was cutting the holes out with a hole-cutter mounted to a drill. Next time I’ll find someone sympathetic to make the plates up with a mill – perfect accuracy.

But it’s finished – there’s no vacuum tapping for the servo done yet, or inlet temperature sensor mounting, but that aside it’s just in need of some dressing and a paint. So here it is!

So, now I guess I need to get the fuel return sorted out, and get wiring! Can I really get this done in time?