Can a Chronograph give false high readings?

I am admittedly a newbie with both PCP rifle (25 Marauder) and chronographs, so please bear with me.

I borrowed a friend's Pro Chrony (a few years old) and set it up on a bench in my unheated garage for the winter.

The rifle was set up on a bench rest tripod and sandbag and returned to approximately the same position for each string. More on this later.

Having read about the potential issues with fluorescent lights, a desk light with incandescent bulb was positioned above to shine down on the unit. The diffusers were installed - I could not get readings without them. While shooting the overhead fluorescent near to the chronograph was turned off. 

Readings were IMHO relatively consistent for quite a while and trends made sense.

I had just installed a new Hill kit + Huma regulator and was running short strings every few days to assess the tune and consistency over time. It is noteworthy that the temperature was dropping as winter set in, but I kept the rifle, pellets, and chronograph in heated area until shooting. A five shot string only took a few minutes.

Velocities were running ~ 885 fps with ES < 1%. Then during one string I began to get flyers down to the low 800's - and this was on an unseasonably warm day. I was aware of the number of consistent shots per fill so this wasn't an issue. I took rifle back into heated area and tried later. The subsequent string was consistently in the low 800's. I was planning to do some outside shooting at distance, so cranked spring back up to again achieve 880 fps. 

Quite a few days later I shot another 5 shot string and velocities were 906, 1036 (!), 906,905, 814, 905, 814, 809. WTF? I recalled that I had 1048 fps previously, but ignored it at that time.

So now there are both low and high flyers and doubt that both are possible. I can understand the lows, but not the highs.

Subsequent investigation and conclusions:

The alignment of the chronograph to the rifle was not precise enough nor repeatable enough. This was found to contribute to low flyers.

On a whim I placed a white poster board under the chronograph and shot a string with it upside down on the posts (like a Caldwell) without diffusers. Readings were consistent, but 80 fps lower than the last string. The chronograph was then flipped over (still on white poster board), diffusers reinstalled, and an additional light was added - one above each diffuser rather than one in the middle. With more precise positioning, reading are now again very consistent with those taken with unit upside down and trends again are making sense. The reflected light from the poster board may be beneficial. We shall see.

The high flyers are still a puzzle.


 
Had similiar issues many years back with ny first chrono. Graduated to a better chrono and built a chrono table after reading about issues with lights. I built a 4 ft long bench (portable) with a trap at the far end, chrono in a stationary position in the middle and a station for the muzzle at the beginning. A light was fixed over each screen with an incadesent bulb andvthe duffusers were used. By using this bench the muzzle is always the same height over the chrono, the same distance from the chrono and always the same light source. I experimented with this setup outside in sunlight and yes the readings were different!! On cloudy days the readings were different! On extreme cold the readings were different! Bottom line is....to get consistant readings the chrono position, lighting, position of muzzle in relation to the chrono needs to be the same. Since using my chrono table my readings have been extremely consistant as long as it doesnt get used in extreme cold, in this case I am guessing that fog might collect on the windows causing erratic readings
 
I saw a youtube video that looks like what AirShot describes. This may be my next step. And I thought I could tune once and forget it ......

How far should the muzzle be from the first sensor to avoid what Motorhead describes?

Shooting a foot to 2 feet behind the first sensor should give the most consistent readings ... IMO