[R-390] IF sweep alignment

[Shoppa, Tim]

> Ringing mechanical filters, IIRC.
> There was also at least one thread in the good old days of this
> list that strongly advised against trying sweep alignment of the
> IF strip, because the ringing filters distorted the wave form.

It's not that mechanical filters prevent sweep alignment.

It's that mechanical (or, indeed, any higher-order) filters can have a large group delay and there can be milliseconds of "skew" between the "in signal" and "out signal". As a result a sweep generator at a fast sweep rate (as might be appropriate for a couple of simple IF transformers) can produce misleading indications because the "out" is delayed so much from the "in".

Sweep generators above the bottom rung of the TV-radio-repair-shop-level usually have a knob/switch that sets the sweep rate, and if you slow the sweep rate they will do just fine.

Some filter theory: it's the variation of group delay across the passband that results in ringing. IMHO the Collins mechanical filters are not quite as optimal as a constant-group-delay or Gaussian shaped filter, but they aren't bad compared to some other options (including many modern quartz crystal filters). It's not that any one filter component is necessarily superior to the others, but mostly that there's been a trend in the past half century in filter specsmanship to optimize the shape factor at the expense of increased ringing. The fact that the ham radio magazine reviews publish the shape factor, but do not say anything about ringing, has led to this unfortunate situation. CW-oriented homebrewers have been singing the praises of constant group delay filters for a long time now but few of the ham manufacturers have listened (but note that many of the military radios have had constant group delay filters for digital modes for decades now.)

Tim N3QE