Frequency standard, radio 4 or msf?

Sussed it in the end.
The problem was mainly too high a kv, the bb212 varicap has a high capacitance, just 1v correction voltage shifted the o/p a few khz, so the system was trying to make too small a change in voltage to track the frequency.

Lowering the gain a little by changing the bb212 with a pair of bb105's in parallel gets only around + and - 50 hz shift of the osc frequency for the entire 5v control range, the oscillator is now nice and stable.
I will need to replace the xtal with a 10ppm varsion, or use one of my crystal ovens.

Turns out my freq counter is off 65hz in 10m/c's, not bad seeing as all the house moves its suffered.

Well after observing the standard for a while there is a 1hz or so fluctuation at 10 mc's, much more like it.
I reckon this can be reduced also.
Heres a design mod I came up with for the receiver, I'm allready using this circuit, the stuff bottom right I think might be usefull for removing some of the am from the carrier which might be causing my drift, I'm thinking apply the received audio to a dual gate fet on the fron end, instead of the single gate fet I'm using, and implement a sort of high speed agc, the capacitive coupling is going to create a time shift so total cancellation isnt going to happen but I'm expecting a good reduction in modulation.

Well initially I was discussing whether to use radio 4 or Msf as a source for my frequency standard, hence the title.
Looks like I now have both, I recently finished my Msf standard.
It has a crystal filter on the Msf front end, 3 phase locked loops and a propriety crystal oven, o/p's are 1mc and 10mc switchable, the meter shows the loop voltages of the Pll's.
Accuracy as far as I can tell is as good as the Msf transmission.