[Heathkit] SB-610 phase networks needed

[Peter Markavage]

The bottom line is that it really doesn't make much difference if the
output tone frequency has shifted due to component aging. It's a single
or two tone test. As long as the tones are somewhere generally between 1K
and 2K, the "test" can be done. Of course, if it's determined that the
PEC has an internal component failure, then a re-build is necessary of
the PEC.  The tones are not critical unless you have the desire to match
the original specification. I built one out of discrete components years
ago in the HO-10 when a lead broke off the PEC right at the substrate.
Because of component tolerances, I think I was off by about 150 cycles.
Life went on, and in time, I didn't even remember it wasn't exactly per
the spec.

Pete, wa2cwa


On Sat, 22 Sep 2007 10:00:42 -0700 "Kenneth G. Gordon"
<kgordon2006 at verizon.net> writes:
> On 21 Sep 2007 at 15:08, K3PID wrote:
> 
> > The phase shift
> > networks P.E.C 84-39 & 84-40 are the frequency determining 
> components
> > and my guess is that the internal capacitors have shifted enough 
> to
> > move the frequency.
> 
> Usually what happens with those PEC things is that the 
> resistors drift HIGH, although the capacitors usually don't 
> drift at all.
> 
> > Does anyone have these in NOS or if not then does anyone have the
> > detailed specifications for the networks?
> 
> In some Heathkit rigs, the schematics show the values of 
> the components inside that PEC, and then one can be 
> made very easily out of discrete components. I do this 
> regularly for the SB/HW transceivers which use a PEC as 
> the tone generator for the CW sidetone.
> 
> As I said, in every case I have examined so far, the 
> RESISTORS have drifted high, thereby raising the tone 
> frequency. One of the main culprits is the (usually) single 
> EXTERNAL resistor, which is also part of the phase-shift 
> network. In the SB/HW transceivers, it is the resistor 
> connecting the PEC to the tube.
> 
> As Sandy mentioned, there is VERY little information on 
> designing those things. The math is not trivial, I might add...
> 
> Ken Gordon W7EKB