hi
thankyou for your in-depth reply
firstly, yes i had to remove the whole oil pump assembly in order for the machining to take place. as far as i am aware the machinist then lay the pump on two parallels either side of the drive tangs/ pump gear and bored it out square in relation to the mating surface of the pump and then made a sleeve. i may of not had to remove the sump if the gasket was not disturbed in the process of removing the pump. however as i was in there i thought just go the whole way and do the sump gasket plus crank seal other applicable gaskets.
yes you are correct the crank gear does have a rear flange to prevent the belt walking and the front will meet the pulley. however the width of the belt is significantly less than the gear so it can walk back quite a way before it hits the flange. by this time the belt is a few mm off the back the the water pump idler and cam gear.
now, before your most recent reply i noticed while looking for a flanged pump that the older style pumps (flanged) have 20teeth. i thought this may be an issue with my setup as the pitch may differ slightly and could be part of the reason for the older belts to have 1 less tooth or this may be due to the adjustment that can be made on the water pump mounts.because of this i had doubts about fitting the older gear to my newer pump. anyway this is where it got interesting as i noticed that although my water pump was not flanged, it had 20teeth much like the older flanged style gears. i had not noticed anyone else's later style engine with a water pump that had 20 teeth and a fixed position. so i began looking for a 18 tooth pump with a fixed mounting position much like all the others for sale and that can be seen pictured on the internet. this time i bagged a hepu pump with 18teeth and no flanges. when comparing the hepu pump to the fai pump installed the fai pump gear was larger and with more teeth. i roughly worked out the pitch to be around 2.8 on the hepu gear and 2.9 on the fai gear. so although the gear had a bigger size it seemed similar in pitch.
when testing the new hepu pump the cambelt seemed well where it should be however when the engine was running it tracked back towards the engine just like before. i made tweaks to the tensioner just to see what would happen i even contemplated and then acted upon the fact that the angle of the engine was so sever due to being on an incline and with the front on axle stands that the belt may be wandering back ever so slighlty. i corrected the engines angle. if i gave the belt more tension the process of belt walk slowed somewhat. the process was taking about 20 seconds of normal cold engine tickover. it was noticed that the belt was wandering slightly left to right on the idler on the way up to the tensioner while running. this was why i thought to increase the tensioner preload for testing purposes.
the idler was checked for squareness in relation to the water pump as the belt was wandering most visibly in this location. by eye the idler seemed to lead slighlty inwards away from the turbo side of the engine ever so slightly. in other words if you put a straight edge on the idler and waterpump the gap on the right hand side of the waterpump gear seemed more on the right than the left.however after loosening and turning the idler it really looked dead square to the water pump. this could have been caused if the boring process of the pump or sleeve was not quite correct although the idler seems dead on square by eye now.
so i started the engine again and the belt remained in the centre of all gears apart from the idler as for some reason it is 2mm proud of the water pump. in other words much like before if you use a straight edge from the idler to the water pump a gap of 2mm is between the rule and the water pump. I'm not sure if this is normal??. the belt remained centred for at least a minute which seemed to be an advancement. however i began to rev the engine quite high every second or so and watched the cambelts movements. it then tracked back quickly to the back of the engine. this test was repeated and the same results occurred each time.
when the belt runs correctly it is always near the edge of the idler. perhaps when i rev it up and its wandering increases on the idler as previously mentioned, once the belt meets the edge of the idler it pulls it the rest of the way and the belt then subsequently walks back on all gears and tensioners. this is why i would like to know if the idler should be flush with the waterpump or perfectly inline.
it seems as though any tweaks of adjustments i make alter the initial characteristics of the walking however the outcome always remains the same. i am looking at getting 2mm turned off the idler so that it sits inline with the waterpump. worth a try i think.
it seems that unfortunately the car had this problem when i purchased it however the position of the belt was not so extreme as it is now. the belt has always been on the back of the cam gear.
what you have mentioned about the cylinder head makes sense i hope this is not my issue but it could well be.
image of the idler and waterpump with a straight edge
https://ibb.co/NxFhFSB