Well for the few of us that are all motor in here it seems as though lost motion assemblies have come up in conversation as of late. i wanted to fully explain these things to you guys so you can make an informed judgement as to what you need to do about them.
First off an explanation
lost motion assemblies have a very simple job to do. they basically provide constant pressure to the rocker arm that the vtec lobe travels on. as for the reasons to why this is necessary i have no deffinitive answer but its fairly obvious. when you are not in vtec and the rockers are not bridged together with the locking pins the rocker arm that the vtec lobe rides on basically just sits there. without the lost motion assembly the rocker would just sink or most likely fling itself around from vibration. so this item, the lost motion assembly sits underneath of the vtec rocker arm and keeps it held up so the vtec lobe on the cam can ride on it. i would also assume that without this device in there and the rocker arm being either sunk or flinging around it would be alot harder for the pins to slide and bridge the rockers for vtec operation. i would also assume from just looking at the design that by keeping the rocker suspended and allowing the lobe to ride on it would help balance centrifical stress.
these are the two versions of prelude lost motion assemblies.the spring style lma on the left IS IN FACT the newer style from the 99-01 h22a4 according to the helms. i had previously thought the cylinder style (on right) was the newer style but after getting off my ass and actually looking in the helms and talking to some people I WAS WRONG. sorry, get over it!
this is a picture of the b16a lma (left) and the type r (b18c5) lma (right)
now if you will look at the tips of the lma you will see what the difference is. the actual head of this thing is where the problem arises. its obvious by looking at the type r lma that the head has a much greater surface area. this is why the type r lma's are able to support much higher lift cams in b series motors. i will show exactly why later. see its not actually a straight up and down movement that is going on here. remember the rockers are moving in an angular movement. so the actual top to bottom compression isnt what is looked at.
now if you go back and look at the actual h22 lma's and compare them you will see the newer style has a slightly larger diameter head on it. this is why these are better for higher lift cams
next is a series of pictures borrowed from ek hatch on h-t. these pictures demonstrate the movement of the rocker arm and how it actually rides on the lma
this is a picture of the rocker arm depressed on the lost motion assembly with what would be the lift of a normal cam
this is a picture of the rocker arm depressed on the lost motion assembly with what would be the lift of a bigger cam that exceeds the amount the lma can actually handle
after seeing those pictures i hope you can now understand the whole lost motion assembly thing thats been floating around. so its quite obvious that the actual surface area of the head of the lma is what is important.
as of now b series builder that use prelude lma's need to shim the installed hieght .040 to work in thier motors. if one were to use type r lost motion assemblies they would need to shave that amount off of them for them to function correctly and sit at the correct height. if these things to do not sit at the right height you are asking for problems so do not take my word as law and do what you are supposed to do when you assemble a motor and measure everything, check clearences and rotate everything before actually using it to check for binding.
First off an explanation
lost motion assemblies have a very simple job to do. they basically provide constant pressure to the rocker arm that the vtec lobe travels on. as for the reasons to why this is necessary i have no deffinitive answer but its fairly obvious. when you are not in vtec and the rockers are not bridged together with the locking pins the rocker arm that the vtec lobe rides on basically just sits there. without the lost motion assembly the rocker would just sink or most likely fling itself around from vibration. so this item, the lost motion assembly sits underneath of the vtec rocker arm and keeps it held up so the vtec lobe on the cam can ride on it. i would also assume that without this device in there and the rocker arm being either sunk or flinging around it would be alot harder for the pins to slide and bridge the rockers for vtec operation. i would also assume from just looking at the design that by keeping the rocker suspended and allowing the lobe to ride on it would help balance centrifical stress.
these are the two versions of prelude lost motion assemblies.the spring style lma on the left IS IN FACT the newer style from the 99-01 h22a4 according to the helms. i had previously thought the cylinder style (on right) was the newer style but after getting off my ass and actually looking in the helms and talking to some people I WAS WRONG. sorry, get over it!

this is a picture of the b16a lma (left) and the type r (b18c5) lma (right)

now if you will look at the tips of the lma you will see what the difference is. the actual head of this thing is where the problem arises. its obvious by looking at the type r lma that the head has a much greater surface area. this is why the type r lma's are able to support much higher lift cams in b series motors. i will show exactly why later. see its not actually a straight up and down movement that is going on here. remember the rockers are moving in an angular movement. so the actual top to bottom compression isnt what is looked at.
now if you go back and look at the actual h22 lma's and compare them you will see the newer style has a slightly larger diameter head on it. this is why these are better for higher lift cams
next is a series of pictures borrowed from ek hatch on h-t. these pictures demonstrate the movement of the rocker arm and how it actually rides on the lma
this is a picture of the rocker arm depressed on the lost motion assembly with what would be the lift of a normal cam

this is a picture of the rocker arm depressed on the lost motion assembly with what would be the lift of a bigger cam that exceeds the amount the lma can actually handle

after seeing those pictures i hope you can now understand the whole lost motion assembly thing thats been floating around. so its quite obvious that the actual surface area of the head of the lma is what is important.
as of now b series builder that use prelude lma's need to shim the installed hieght .040 to work in thier motors. if one were to use type r lost motion assemblies they would need to shave that amount off of them for them to function correctly and sit at the correct height. if these things to do not sit at the right height you are asking for problems so do not take my word as law and do what you are supposed to do when you assemble a motor and measure everything, check clearences and rotate everything before actually using it to check for binding.