2010 I bought a used skywatcher HEQ5Pro. It was practically new and used very little. During the first use I realized that something was not working well during the guider (sometimes frames showed the stars elongated already with focal 560mm court). I then did some tests to try to understand and solve the problem. I used to test the SC C8 reflector reduced to 1650mm focal length and MagZero MZ5m.
Well looking at the AR graph with autoguiding enabled and disabled I realized that the problem was not software but was due to the mount.
In fact, with autoguiding disabled graph presented oscillations in AR with maximum-minimum continuous + 3-4 / -3-4 pixels that occurred every 13-14 sec. Any autoguiding software enabled tried to compensate for these errors but the correction is often added up the oscillations causing even worse errors. Setting correction times same or less than 1 sec, the result was a little better.
Someone advised me to revise the frame: dismantling cleaning, re-greasing etc. After that laborious effort the problem remained.
Then I noticed in watching the moving gears that the error is repeated each time a tooth the crankshaft gear entered between two teeth of the central floating gear.
I tried in every possible way to reduce these errors by adjusting the mutual distances of the three gears, but the result was almost equal.
The possible solution could be to replace the gears with belts; in the meantime I used main telescope with focal less than 500mm in order not to blur in long-exposure.
Several companies realize the modification kit with belts, but often the modification entailed the loss the gearing ratio of the original setup making unusable the handset in the HEQ5. The only way to handle the mount would be to use the EQMOD.
A few months ago I learned that the company Rowan Astronomy (UK) realized the HEQ5 Pro Belt Modification Kit (www.rowanastronomy.com/productsa1.htm) that maintains the gearing ratio of the original setup .
So I bought the kit. The change was very simple, and in a short time.
I refer you to the album Facebook www.facebook.com/media/set/?set=a.1199861523390411.107374... where you will find more details and tricks.
The first thing you notice is the drastic reduction of the metal noise during the go-to.
The graphs before-after are eloquent. Now the shape of the curve AR is much more linear and software fixes have definitely improved.
The DEC's graph also improved, because virtually disappeared driving errors caused by backlash.