Hi
I'm changing from 193 swirl port heads to Vortec heads and after market camshaft on a 1993 Chevy small Block L05 factory roller engine.

How do I measure for the correct push rod length with hydraulic roller lifters?

 

Light Checking spring on the head.

 

A light spring weight checking spring. You don't need to buy a specific test spring. You can always go to places like home depot etc and see what they have in their selection. You'll also need one that won't bind when at full lift.

You only need a spring strong enough to keep the valve closed when the lifter is on the base circle of the cam. As you bar the engine over to check the wear pattern on the valve tip, the lightweight spring won't be strong enough to compress the hydraulic lifter.

 

This is how is did mine:

 

The simplest and least error prone approach is the Mid Lift Method a.k.a. the Jim Miller method.
Valve closed. Full running spring. Rocker placed loosely on the stud. No need to turn the engine over. Just measure. The only tools needed are a straight edge, a vernier caliper, an adjustable push-rod and the cam specs.
Similar to Jari's example above but more precise as there's no "fiddling" involved.

For what it's worth, the lightweight checking spring introduces an error into the
measurement as the valve train is not completely loaded. While it may not represent an issue in a street orientated build, once things get more towards the high performance, big cam, heavy spring territory, the error can become significant.

Here is an excellent tech paper on the theory and reasoning behind this approach. While the history may not interest you, the approach might.

http://www.engineprofessional.com/TB/EPQ210_20-30.pdf

 

Trouble with that is that I run triple springs in my heads. An adjustable pushrod would snap in half if it tried to compress my valve springs. That's why it's always recommended to use lightweight checking springs like the ones seen in the video.

The above video doesn't compress the spring but they're still in place just to make it easier.

 

With the method I'm proposing, it isn't necessary to load the valve train at all. The rocker arm sits on the stud and the adjusting nut isn't required. The lifter is on the base circle, measurements are taken and the pushrod adjusted to suit. No preload and no stress on the components other than the spring is fully loaded (which by the way, with a heavy valvetrain as you have is all the more reason to avoid the checking spring).
The difficulty is in the measurement. Finding a suitable reference plane with which to measurement to the trunnion centreline and the rocker pivot can be problematic. I built a jig to extend the retainer mounting surface (where is contacts the spring) and it was easy to measure from there.
Pictures available upon request...

 

I run shaft rockers so the way my measurements are done is completely different than when using rocker stud rockers. The end result is the same. Correct pushrod length has the rocker at 90* to the valve stem at 1/2 the lift. Shaft rockers also center the rocker over the valve tip better than guide plates do. That also meant clearancing a few of the pushrod holes through the head to allow more clearance for the larger diameter pushrods. It's whatever it takes to get good valve train geometry.

 

Once you get into engines of your performance level, the game changes. With the enormous spring pressures, shaft mounted system is probably the only real viable option.
The Jim Miller I referred to with regards to the mid-lift method of measurement also builds an industry leading shaft mounted arrangement.