On Mon, Oct 23, 2017 at 12:11 pm, Gary O'Neil wrote:
I considered that... at least in the context of the PTT input from the mic switch, and that may well be the source of the intermittent bursts in sideband mode. I’ve shunted the PTT input haven’t attempted to verify this though. 75 ms is a very long time for a pair of 7 ms relays to settle however, and the 6 ms burst in CW mode is consistent and reliable, as is the dropout that follows and then proceeded by the start up of the initial CW element.v1.26 greatly reduced the burst but, as Gary mentioned, a short burst still remains.
I've done some more measurements and I finally found the cause (and a fix!).
In v1.26, we temporarily turn off the VFO as soon as the PTTsense line goes HIGH, then wait for 75ms before turning it on again.
This basically works well, however there is one problem: The raduino program is running in a loop, and one cycle takes about 11 ms. So depending on the timing that the PTT is keyed, it may take up to 11 ms before this is sensed by the PTTsense line. Until that moment a burst may still be emitted.
The solution is to use an interrupt service routine which is executed IMMEDIATELY as soon as the PTTsense line goes HIGH, rather than 'polling' the PTTsense at 11 ms intervals.
By using an interrupt we can reduce the reaction time to only 300 us which is negligable.
I have implemented this approach in v1.27 and it completely eliminates the RX=>TX burst in SSB as well as in CW!
(one downside however is that when we use CW in semiQSK mode, the initial dit/dah will be canceled due to the delay).
I will release v1.27 shortly.
Note that the above fix only eliminates the RX=>TX burst (the 'speaker pop' is not affected).
Also, while experimenting I noticed that occasionally a TX=>RX burst is still emitted.
For some reason the TX=>RX burst never occurs in semiQSK mode (when the relays are activated by the bypass transistor across the PTT switch). It only occurs sometimes when the PTT is keyed manually.
I therefore believe the TX=>RX burst may be caused by contact bounce of the mechanical PTT switch. Perhaps this can be fixed by placing a small capacitor across the PTT switch (haven't tried that yet).
73 Allard PE1NWL