Speed variance study with a Phoenix Roadrunner

I picked up a Roadrunner the other day so I could better measure the speed of my tables. Using a strobe was not good with my eyes and my app only went to .0. Here is what I noticed. If I set my speed at the beginning of a record to around 33.310, by the end the speed will increase by .07. So, depending on where the arm is on the record will depend on what the speed reads. Outer groove slower, inner groove faster. Not by much mind you, but arm placement has an impact.

I’m not using an Eagle because I run an Origin Live motor system. I'm thinking that I may increase the tension of the belt slightly to see if that reduces the variance. I modded out a Thoren TD 850. The motor is fabricated in a manner that allows me to customize the tension and use different size belts. However, .07 drift is really not bad but if I can improve upon that it would be a bonus.

I have 2 other tables that I’m going to check out. Both are modded and use a Music Hall CC2. They are a VPI HW19mkiii and a totally tweaked out Thorens TD 150, that is not yet complete.

This proven to be an interesting tool so far. What have others experienced when only using the Roadrunner as a tach?
Post removed 
Once things are stable, say after a couple of minutes, are you getting reading variations as large as .07???  That 33.403 with a 33.333 reference.  Also once warmed up how stable is the reading.  Once stabalized with a heavy platter and a 3 phase controller, I am getting only .001 variation with tape drive.  Having a controller with a rpm adjustment is really nice.  Belt tension, bearings warming up etc will change rpm. So you get up to change the lp, move the knob and get real close to 33.333.  The road runner is a great tool to have.

Enjoy the ride
Thanks Ray. Are you referring to U.S. Audiomart? I don't see a Roadrunner there, though there is a "Wanted" ad for one.

I have been playing around with my Roadrunner as well. Its quite handy for seeing the effects of environment and handling has on the speed. I never liked swapping a belt from 33 to 45 and back on the pulley. Seeing how much it can vary after switching back, it justified, to me anyway, going the added cost of having a power supply made that allowed switching voltage not the belt and having a pulley made to run 2 belts. Variance of .003 is pretty solid IMO.