[time-nuts] strange PPS TI holdover curve on manual holdover initiate of Z3801A

[mike cook]

Hi,
 I have recently been suffering form excessive jitter in the PPS TI values as seen in Z38xx ( thanks for a great tool Ulrich) and am trying to find the origin.
I don't have a full view of the sky and  get forced into holdover due to satallite loss once ore twice a day. Checking with another 2 Oncore VPs , using the same antenna/position as I use for the Z3801A did show that the onboard receiver was always picking up at least one less bird, but swapping the receivers didn't change the problem status.

So I decided to force the Z3801 into manual holdover and see if taking the locking out of the loop would change the  status.
As you can see from the screenshot the jitter dropped to what I was seeing when all was ok.

So it looks like I have an issue with locking..still could be my bad antenna position I guess.

Anyway, that isn't the subjet of my question.

After putting the receiver in manual holdover ( :sync:hold:init ) I saw as expected that the PPS TI values fell off with respect to the EFC which was following a steady slope ,  which from the doc I interpret as being that calculated to counter aging alone. However after about 8hrs in this state I see the PPS TI curve moving back up again. I would have thought the delta to be constantly increasing. 

My screenshot exceeds the 128k limit on attachments here size, so I have put it at 
http://gluon.stratum1.d2g.com/timenuts/strange-holdover-TI-profile.PNG
Hope you can see it.
Has anyone seen this before? Is it normal? 

Thx.
Mike