Key down to RF delay problem

[Bob Henderson]

I wonder whether anyone else might have faced this issue.

Ever since I commissioned my HL2 more than a year ago, I have struggled to live with a very long key down to RF delay (around 65mS).  This causes first keyed element clipping.  Attached scope screen grab illustrates.

Initially I focused my keying on the Thetis MIDI interface.  I knew the delay was a problem but didn't know whether the cause was Thetis, Thetis for HL2 or the HL2 itself.  More recently I have tried keying via the front panel jack only to find the exact same delay.

I will spare you the detail of my months of machinations but can confirm I seem to have eliminated Thetis as the cause by instead running Spark SDR and finding it behaves in the same way.

I have poked around a little in the HL2 hardware and having done so can confirm that INTTR, LDMOS bias and LDMOS supply are all present and correct in <2mS so conclude these do not contribute to the problem.  Something earlier in the chain is delaying the production of RF following key down.

At this stage it feels like my PCB may have suffered a production glitch and such a glitch may have affected a batch.

Can anyone shed any light or have any helpful ideas?

73 Bob, 5B4AGN

[James Ahlstrom]

Hello Bob,

When you key from the jack on the HL2, the HL2 generates its own CW waveform. The key state is sent to the PC so it can generate a sidetone, but the PC should not try to send a CW waveform. So when you key from the jack on the HL2 any delay is in the HL2. Using Quisk I measured a delay for a dah of 40 msec.

If you key from the PC using MIDI or its serial port there are many sources of delay, and the HL2 must wait for the PC to send it the CW waveform to transmit. To prevent clipping the first dit you must save the CW timing in a buffer in the PC and send it to the HL2 after a suitable delay. In Quisk, this is called "Start CW delay msec". Look for a similar feature in Thetis or Spark.

Jim

N2ADR

[Bob Henderson]

Hello Jim

Thank you for your note.  I am keying via the front panel jack with a Winkey and monitoring key down at the jack and RF at the dummy load.  The screen grab attached to this note better illustrates the key down to RF delay I see which stands at approximately 60mS.  In the earlier grab the first element was lost in its entirety as the code speed was a little faster (around 20wpm)..

I have not tried Quisk but the delay looks to be the same with Thetis and Spark.  As you say, keying in this way has the CW waveform generated directly within the HL2.  What I would like to understand is why it should take such a long time from key down for the keying envelope to commence.  And whether the 60mS I see is an unusually long time or typical.

Bob, 5B4AGN

[Bob Henderson]

It appears I have a lazy/sticky T/R relay.   RF sampled at a random point near to but upstream of the T/R relay reveals key down to RF to be 3.6mS.  I will need to live with the issue until I can source and fit a replacement.

73 Bob, 5B4AGN

["Christoph v. Wüllen"]

Looking at the rising edge of the first pulse it seems the delay
does not entirely come from the "digital side" (Thetis + FPGA),
but that e.g. the PA powers up slowly.

> --
> You received this message because you are subscribed to the Google Groups "Hermes-Lite" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/e034334c-1898-4647-a6fe-5af953934376n%40googlegroups.com.
> <Letter I at 12wpm.jpg>