Key in, TX Delay and CW Delay (holdoff)

Geoff, our apologies - this fell off my radar. I'll ask again

Here is the response:

The Transmit and Delay Logic diagram illustrates the general PTT-driven transmit flow used for voice and data modes. CW operation follows the same high-level transmit pipeline, but includes an additional timing behavior known as CW Delay (sometimes called holdoff).

CW Delay is a transmit hold time that keeps the radio in the TX state after a CW element completes, rather than immediately returning to RX. The purpose is to prevent unnecessary TX-to-RX-to-TX transitions between closely spaced CW elements, characters, or words. This is especially important at higher CW speeds and when operating with QSK enabled you would see 0 for CW Delay.

Conceptually, CW Delay occurs after RF emission has begun and before the radio would normally transition to “User Releases PTT” in the diagram. Instead of unkeying immediately when a CW element ends, the radio evaluates whether another CW element is scheduled soon enough that remaining in TX is more efficient than dropping back to RX and re-keying.

In other words:

• CW Delay does not add time before RF begins.
• It extends the transmit state after a CW element, based on CW speed and timing.
• If the delay expires with no new CW element pending, the radio proceeds through the normal ramp-down and RX transition shown in the diagram.

Because CW timing is element- and speed-dependent, this behavior is not explicitly drawn in the generic PTT timing graphic, but it fits logically just ahead of the “User Releases PTT” / ramp-down portion of the transmit sequence.

Geoff - you’re close, but the published timing chart is a generic PTT flow. CW follows that same pipeline, but “PTT” is effectively synthesized by the radio and driven by the CW element scheduler + FIFO.

Key points

• TX Delay is a PRE-key delay.
  • It happens before RF is allowed to start.
  • Purpose: give external devices (PGXL, relays, antenna switching, etc.) time to get ready before RF is emitted.
• CW Delay (holdoff) is a POST-element hold time.
  • It happens after a CW element finishes.
  • Purpose: keep the radio in TX briefly so it doesn’t bounce TX -> RX -> TX between closely spaced elements/characters.
• CW elements are not truncated by CW Delay.
  • CW Delay only decides how long we stay keyed after an element ends.
  • It does not chop an element short.

Your original example (30 WPM: dit = 40 ms)

Assume you send ONE dit only.

1. Key down event (or internal keyer schedules element start)
2. Radio enters “TX requested” state (equivalent to “User asserts PTT” in the generic diagram)
3. Apply configured TX delays (TX1/2/3 as applicable, plus TX Delay)
4. RF ramps up and RF output begins
5. Dit is transmitted for its full scheduled duration (40 ms at 30 WPM)
6. Dit ends (key up)
7. FIFO drains to empty for that element
8. CW Delay timer starts (hold TX state, no CW modulation during this idle time)
9. If NO next element arrives before CW Delay expires:
   • begin RF ramp down
   • transition back to RX
10. If a next element DOES arrive before CW Delay expires:

• stay in TX
• key the next element at its scheduled time (without dropping to RX)

Direct answers to your two numbered questions

If TX Delay = 30 ms, does RF output begin at 30 ms with the complete 40 ms dit?

Yes.

• TX Delay shifts the START of the CW element later.
• When RF begins, the dit still lasts the full 40 ms. It is not shortened.

If CW Delay = 15 ms, does the dit complete or is it truncated?

The dit completes.

• CW Delay starts after the element completes (after the FIFO drains for that element).
• CW Delay only affects how long the radio remains in TX AFTER the element ends.

Fixing your step list (single dit case)

Your items 7-11 were the uncertain part. Here’s the corrected order:

1. Key down (element scheduled)
2. TX requested (internal PTT assert)
3. Apply TX delays (TX1/2/3 as applicable + TX Delay)
4. RF ramp up -> RF output starts
5. Dit sent (40 ms)
6. Element ends -> FIFO drains to empty for the element
7. CW Delay (hold TX state)
8. CW Delay expires (no next element pending)
9. RF ramp down -> RX

Notes:

• “User releases PTT” for CW is essentially “no more elements pending” (not necessarily a physical user action).
• “FIFO Buffer Empty” corresponds to end-of-element for the waveform data.
• CW Delay occurs after FIFO empty (end of element) and before ramp-down.

Practical note (PGXL / missed first element symptoms)

Reports like “VA copied as UA” at higher WPM are consistent with front-edge timing (insufficient pre-key delay / amp not ready / envelope shaping).

• That points at TX Delay (pre-key) and overall keying path readiness.
• CW Delay (holdoff) helps avoid TX/RX thrash but won’t fix a mangled first dit/dah.

73, N7HQ