Well, I've completely redone the design.

The 74HCT7046 is gone, replaced with a Hogge phase detector (built from=20
a HCT74 DFF, a HCT86 XOR and a HCT125 O-C buffer), and a homebrew=20
HCT04-based VCXO. A BB419 varactor provides sufficient tuning to lock=20
against an incoming signal which meets the Broadcast Teletext Standard.=20
As a bonus, I get retimed data thrown in for free. Data is -- going by=20
the scope output -- valid on the falling edges of the clock signal.

I need to fiddle around with the loop filter a bit, though -- it's=20
currently a 30k + 100pF LPF which seems to work, but probably isn't even=20
remotely optimal. I could do with finding a good reference for setting=20
up PLL loop filters...

But for now, at least, it's behaving itself...

Thanks,
--=20
Phil.
piclist@philpem.me.uk
http://www.philpem.me.uk/
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.