Colin Mill

That's partly why I'm suspicious. If they have patents in place they can be more open than they are being (the patent disclosures would already have made it public anyway). If they have not got the patents in place than they have already said far too much.

Peter For me their website is far too vague to judge which makes me suspicious. It seems to spend quite a lot of time qualitatively rubbishing other transmissions and almost no time telling you what it does or, more importantly, how it does it (which makes me even more suspicious). It could be great for all one can tell. My bet is you will wait quite a long time to see it in volume production but I would like to be proved wrong. Colin

Peter The reason why traditional auto boxes generally increase fuel consumption is that they are operating with a torque converter that has significant losses compared to a conventional clutch. Epicyclic gearboxes are almost universally used in situations where gearbox weight is a primary consideration (main reduction boxes on helicopters, turboprop gearboxes etc.). However in racing cars you require a box where a high degree of flexibility in ratios is readily available. This is much less straightforward in an epicyclic than in a spur-mesh system. Also, in racing longevity is not a consideration. Colin

Peter The typical auto box is a fairly hefty device but mainly because they were designed to take the torque of 7 litre yank motors. The planetary core of these boxes is remarkably small and light as the loads are shared between three teeth in each planetary set rather than just one. Colin

Peter The use of one way clutches in this way is used in auto boxes (the 1st to 2nd shift on a BW35 for instance) so the technique in that respect does not seem new. Auto boxes also give power delivery through the changes. Perhaps modern auto boxes link to the engine management system to offload during the shifts but in the good old days they just had to cope. The 2nd to 3rd shift on a BW35 required a phased application of a clutch and release of a band and the system generally worked fine. Colin

Hi Steve From what I gather from the web page I mentioned the standard K series system has one of these valves internal to the manifold and in line with the small hose that goes up to the expansion bottle. If this gets blocked or stuck it looks as if you can get an air-lock in the head. With an all alloy engine this has to be near instant bad news. (Similar problems in the Hillman Imp virtually killed the car off). I just wonder if malfunction of this little device accounts for the repeated HGFs that some people seem to get. Colin

I reckon it's got to be BS. If the change time is zero then the relative speeds of parts of the box have to undergo instantaneous changes which implies infinite torque. Maybe this guy has also designed a perpetual motion machine. Colin

Since Lambda sensors need to be at 250C or above to work I guess you need to look for something capable of working well over 250C. Silicone works well on Methanol fuel engines but of course they run much lower exhaust temperatures than petrol. One other consideration - silicones poison Lambda sensors so I guess you need to be careful about silicone near the sensor. Colin

Since I'm new to this forum I may be bringing old news but I wonder if the head gasket failures some are getting are due to this little device getting stuck. See the following site:- http://www.lame-delegation.de/mgfcar.de/bleed/ Would very much like to know what more experienced K owners think. BTW are the 1.4 engines less prone to HGFs than the bigger bore motors? Thanks Colin

Since I heard that this seal can be a problem (and mine needed a brick hammer to get it in) I seated the top-hat on some silicone sealant. I have put a seal washer between the top hat and the sensor in the hope that any leak past the seal will simply get trapped in the sensor. You could do a better job with a modified top hat with a shallow O ring groove. Colin

Is the oil seal fully seated? The seal on mine was a swine to get in and the top hat was under a little outward pressure while fitting the circlip. Because of all the problems people have with this seal I have seated the top hat on silicone rubber (using a bit of tube over the speedo gear shaft to stop it getting on the lips of the seal. In your situation I would be tempted to try putting a little loctite 290 (not studlock!) on the circlip as I think the edge of the groove must be getting a bit rounded by now. Colin

Hi Oliver Thanks - I've done as you say and I can't see the gearbox is going to hit the inside of the tunnel however hard you corner. Adam I think your right. If I had the job to do again I'd fit the steering column after the engine as well. Colin

Just finished dropping the engine and gearbox in. Doing the job single handed called for a lot of trying to be in three places at once. I was surprised just how critical it was to get the engine at the right angle in roll. The first time I got the thing a bit stuck with the right hand lug on the bell housing fouling the steering column, the alternator hard up in the front left hand corner of the bay and the gearbox hard over to the left of the tunnel. By rolling the engine to the right the lug on the bell housing cleared the steering column before the alternator started to enter the bay and it was OK but not exactly leaving much room for error. Slackening the alternator belt bought a valuable extra bit of room. Is there a trick that makes the job easier? When the manual talks about centring the box in the tunnel do they just mean getting the gearbox mounting bolts centred in the slots? The type 9 box has a big bulge on the left side which seems the closest thing to the tunnel sides. Should I offset the mount to the right a bit or not? Any thoughts please! Thanks Colin

Oh yes, the manual is out of date I think. There is no riveting to do in the pedal box. I think they now just cut the holes accurately in the right place so no adjustment is needed. Colin

Hi In my kit the grommets were in a poly bag that was in the passenger footwell. I don't recall it being labeled in any way. Colin

Hi Chris Yes, you would need seperate feeds to the two poles of the switch and have the relay coil as the only load connected to the ignition switch. I agree, the relay would be a rather critical component and I don't fancy it at all. These days semiconductors are used a lot in mundane tasks in cars from switching interior lights upwards so the diode solution is not that off the wall. Come the 42volt revolution and CANBus there will be almost nothing on a car that is not semiconductor switched. Nice simple wiring looms but a software nightmare! Colin

While you could do it with a relay it would be a lot more work than fitting the diode. You would have to remove all the load from the ignition switch and use it to switch the relay only. The relay would have to be a two pole type so that you could switch the ECU with one pole and all electric motors off the other. That way the relay would drop out because you had open-circuited the coil drive on turning off the ignition switch. The two poles would ensure that the current from the fan(s) would not get to the ECU. However it would be a hell of a lot of trouble to go to. The diode way is a two minute job with a soldering iron or some Lucar connectors. Alternatively just leave the headlights on all the time Volvo style - at least that will reduce the run-on time. Colin

Hi Chris The Schottky I mentioned is one I found in a last years RS catalogue but I guess they still do it. The RS order code is 357-4100 and they rate it at Vf= 0.38v drop at If=20A. It has a Ifmax of 40A. Its good for a reverse of 30v so assuming none to many inductive nasties from the fan motor I would give it a good chance of surviving. There are others in the series with the usual trade-off of Vrrm and Vf performance. (I just realised I cocked up on the part number in my previous message - sorry. Its actually 42CTQ030 ) Best regards Colin

Hi Chris Yes, I'm not likely to do anything to 'cure' it on mine - some of the old air-cooled VWs would still be running when you left the car park so its not likely to get on my nerves. For those that want to try, a Schottky barrier rectifier would be good as with something like an IR 47CTQ030 you would only be loosing 0.3v across the device at 6amps. The dissipation would be a bit lower too (under 2W) so the heat sink would be easier to organise. They come out at only about 2 quid even from RS 8-) Best regards Colin

If it is the fan that's responsible we could cure it by putting a suitable size diode in series with the fan. Does anyone know what the start-up current for the fan is as we would need to spec the diode to cope with that initial load. Colin

John I'm just glad I bought the engine hoist rather than just hireing it! 8-)= Colin

Hi John Yes, its the carrier I don't like much. I used a thermostatic soldering iron on the plastic to try to get an idea of the melting point and it seems the plastic is softening at about 190C. The air temperature in the bell housing could easily be up around 100C so it would not take a lot of dissipation in the bearing for it to reach the melting point of the plastic. The bearing weighs 167g so the thermal capacity must be no more than 70J/C. I did a quick sum on the heat that could be generated in fault conditions and found that a 10% slip on the front face at a preload of 30N and 6000rpm could give 43watts of heating. 5 minutes of this would warm the bearing up by about 180C and the carrier would be history. From what you were saying about the post-mortem on one of the bearings the carrier is the bit that has really had it because they reckoned the bearing itself was still turning smoothly. Once the lateral friction between the bearing and the carrier has reduced I would expect the bearing can whirl in the carrier causing more heating. Best regards Colin

Many thanks to you all for the info I'm coming to the conclusion I'm going to have to go with the preload. I did a few measurements on my car and came up with the dimensions mentioned earlier. The spring gives a cable tension of 1.1kgf (10.8N) so I reckon that gives a preload of only 27.5N at the bearing. Since the self centering slide takes (by my crude measurement with a spring balance) 5kgf (49N) of side load to move it I can see why a preload of only 27N is going to be ineffective in keeping the bearing centred - the friction with the fingers would not be enough to create a big enough side force. However I don't see why it should need constant centering . It should go right initially and stay right. Perhaps the preload is needed to create additional friction between the bearing and its carrier to prevent it rotating on the carrier. (Since the bearing is grease filled the torque transmitted by the bearing at 6k must be quite big just from pushing the grease about that fast.) It seems that we are paying a high price for having +-1mm of self alignment in the bearing. It sounds as if the thing would last for ever if the thing could be accurately lined up with the clutch without the nasty plastic bit! Colin

Hi Tony Even I'm not old enough to remember carbon CRBs first hand! 8-) I can see that they have several competing problems to deal with. With a low pre-load the release fingers will just skitter over the face of the bearing and the face will get badly worn. Also the forces will be to low to overcome the friction of the self-aligning slide. Increase the preload and at least its the bearing is doing the turning and it should self align. But can a sealed grease-packed race do engine rpm for 100k miles? It would really help to know what the failed bearings actually look like. Is it the race itself, the front face, or the plastic carrier part that has suffered? Best regards Colin

Thanks Nick It all sounds really odd to me. Having had my ear clipped as a learner for resting my foot on the clutch and always adjusting cable clutches for free travel at the top I would really rather have the bearing come completely clear of the fingers in the released state. As I'm just putting my car together I have the option of playing about with this. I've just fitted a light compression spring over the bell housing end of the cable to act as a return spring with (hopefully!) just enough tension to to pull the cable back and push the bearing clear. At the top end I'm going to either fit an adjustable pedal stop or clamp a stop to the cable. I'm also going to pinch the idea of putting the top end cable nuts either side of the tube in the pedal-box and remove the spring existing spring from the pedal. This has got to be the only car I have seen without a stop to set the top of the pedal travel! Hopefully I can get enough use out of the car to find out if the method works or not. Best regards Colin