开云体育

Somebody posted a report recently on this group about matching a batch
of CTS crystals from Mouser. The frequency spread across the entire
batch of 100 crystals was a mere 110 Hz, which is not enough for
building lattice filters with bandwidths that people are likely to
want. Crystals from other suppliers, especially older crystals, might
have more variation.

If memory serves, the lattice design also tends to have problems with
group delay distortion; that is, differences in the phase delay of
different frequencies in the passband. (All filter designs have some
phase delay but some are worse than others. And there are differences
both in the amount of delay and in how it varies.) In voice modes
group delay makes the audio sound "bad" in a way that can be heard but
hard to describe. It's much more of a problem for data modes because
it can interfere with decoding.

I don't want to discourage people from building filters and
experimenting! Just saying that it's something that you should test
for, along with frequency response, especially if you're going to use
the filter in a radio that uses wideband digital modes.