>>>Part of the problem was RF leaking back into the power supply
>>>compartment on the B+ line. No amount of bypassing at the base of the RF
>>>choke would cure it, because the RF is being induced onto the B+ lead.
>>Ian -- How much RF potential did you measure across the 0.01uF
>>HV-bypass without a glitch resistor?
>I didn't even try. If you think about it, such a measurement is
>well-nigh impossible in that location, which is very hard to get to and
>bathed in an intense RF field.
If the bypass cap has a low XC at the frequency of operation, there
should be a low V across the capacitor.
>You've also mentioned several times that the bypass cap at the bottom of
>the RF choke in that amp is rather low in value. That may be so, but
>once I had fixed the glitch diode problem, the amp seemed to behave
>perfectly well on Top Band in its original configuration.
The 1nF bypass cap has 88.4-ohms of reactance at 1.8MHz. There is
roughly 1A of RF current in the choke.
>wirewound glitch resistor in the B+ feed, any RF current at the bottom
>of the RFC has to go through the bypass cap, so the problem is over.
A typical WW glitch resistor has c. 8uH.
>>>What else? Oh yes, a few things to speed up the TX/RX changeover, and
>>>save the antenna relay from arcing on semi-BK or VOX. With modern
>>>transceivers, when you hit the key in semi-BK you have maybe 10-15
>>>milliseconds between the PTT line going low and the RF appearing. It's
>>>simply a race against time, to switch the antenna relay in the amp
>>>before the RF arrives.
>>Some radios R/T switch in 5mS.
>Unfortunately, yes. In that case, either change to a fast vacuum relay
>or improve the radio... preferably both!
>Some older radios do their internal R/T switching using a slow
>multi-pole relay that also switches the PTT line to an external amp. In
>a lot of those cases you can win back valuable milliseconds by taking
>the amp PTT signal from the line that *drives* that slow relay.
>Generally the relay is driven by a switching transistor, so simply add
>another transistor sharing the input switching signal, and wire that
>directly to the PTT output.
>More modern radios have a built-in TX delay, but you can lose a lot of
>that time by using the optional relay output for the PTT. *Always* use
>the solid-state PTT output - and if it isn't capable of switching your
>amp. Modify the map so that it is... but not by using a relay! Remember
>that it's all about *time*.
>>>First, install a transistor switch in the SB-1K/AL-80A (powered from the
>>>+12V DC line) so that you can use the direct solid-state PTT option in
>>>your transceiver without blowing it up. That gets the PTT signal into
>>>the amp several milliseconds faster than going via the relay inside the
>>>transceiver. Second, install the K1KP/W6XX speed-up circuit for the
>>>antenna changeover relay in the amp itself.
>>What is the measured make-speed of the stock relay with this circuit?
>I didn't test that particular one, but I have done measurements with a
>selection of other relays, and the K1KP circuit will roughly halve the
>switching time. With a 24V line, 24V coax relays come down from 15-20ms
>to 7-10, and Kilovac HC-1s come down to about 1.5ms.
** With 120VDC-OC limited to 80mA, I measured the HC-1 make time at c.
>With an open-frame relay as used in the SB-1000 you might guess at
>20-30ms before speed-up and 10-15ms after. These open-frame relays are
>not precision mechanisms, so you can expect quite a lot of variation
>from one shot to the next.
>Bottom line with the SB-1000 is that after adding the speed-up circuit,
>the relay only arcs occasionally on first pressing the key. Without the
>modification, the relay would arc every single time. That's with a radio
>that delays the RF by about 12ms (IC-746), and a solid-state PTT driver
>to get the PTT signal to the relay ASAP.
>I haven't made detailed measurements on this amp, but am pretty sure
>that even with speed-up there is not much reserve of time between the
>relay switching and the RF arriving. The occasional arcing must be due
>to variations in switching speed (this is not a relay in good
>condition). So what we have in this particular case is not a complete
>cure but a considerable improvement.
>With the VHF amps using coax relays, the speed-up circuit has made all
>the difference between regular arcing and absolutely none. It really is
>well worth trying!
>73 from Ian
- R. L. Measures, a.k.a. Rich..., 805.386.3734,AG6K,