Ignition timing oddity on chassis dyno
383 with GM HEI distributor (modded with aftermarket module, curve kit, coil and shimmed for end play). Before going to the chassis dyno, I'd set my timing up (with a good timing light) for 34* total. Checked it up to about 4,000rpm and I could see the advance curve adding more timing until around 3,000rpm then it remained static. As I said, I only went up to about 3,500-4,000rpm with the timing light... just enough to see that the advance curve had finished adding timing. On the chassis dyno, we were messing around with the timing (just adding/subtracting a couple of degrees to do some fine tuning). I asked the dyno operator to take it up to 6,000rpm and keep the timing light on the balancer. He said total timing went up to something like 50* by 5,500rpm! We checked with another light and same thing. Couldn't hear any pinging (although the car is v.loud). What I don't understand is how could this possibly be happening? I still don't believe everything is at it seems, since taking timing out resulted in less power. If is was really running up around 50* total timing, there's no way the motor would be making power. I've previously tried experimenting locking the mech advance out and didn't see over 36* (with about 14* initial). Since the holes in the base plate physically limit max advance, what's going on with this wacky timing reading? It had occurred to me that perhaps the ProStreet balancer I'm using could be slipping, but it's pretty new and looks perfectly fine to the naked eye.
Anyone got any other ideas or explanations? If not, my next step will be to try a brand new MSD HEI just to eliminate the distributor from the equation. What about spark scatter or some other kind of effect?