CW - Not working with Artemis.

[Scottay1967]

Testing CW / WinKey support in Artemis with SunSDR2dx.

Hardware:

• SunSDR2dx v88.8
• Artemis SDR
• K46 CW Modem / WK3-compatible WinKeyer
• External CW keying cable connected to SunSDR KEY input
• Same hardware tested successfully in ExpertSDR

What was tested:

Initial WinKey setup:

• COM3 detected correctly in Windows
• WK3Demo 4.2 used instead of older WKTestBed
• K46 entered WinKey host mode correctly
• WK3Demo successfully connected to the K46
• “WK SMT Version 31.03” displayed correctly
• Status changed to “Port Open”
• No further COM or USB issues after correct startup sequence established

K46 operation:

• WK3Demo Tune function generates sidetone correctly from the K46
• CW generation inside WK3Demo appears normal
• K46 host mode and communication confirmed working

Physical hardware verification:

• Same K46, cable and SunSDR hardware tested in ExpertSDR

• CW keying and TX worked immediately in ExpertSDR without any hardware changes

• Confirms:

  • K46 working correctly
  • cable/pinout working correctly
  • SunSDR KEY input working correctly

Artemis testing:

• Radio placed into CWU mode
• Semi break-in enabled
• Hardware sidetone enabled
• Primary connection set to Radio
• Tested both external K46 keying and Artemis internal CW memories/keyboard

Results in Artemis:

• External K46 keying does not trigger TX
• Manual shorting of KEY input also does not trigger TX in Artemis
• Artemis internal CW keyboard/memory also does not produce TX or RF output
• No MOX/TX/RF activity occurs during any CW attempts
• Sidetone may be generated locally, but no RF transmission occurs

Important observation:

• Exact same hardware/functions work correctly in ExpertSDR
• Failure appears isolated specifically to Artemis CW TX/keying path

Possible causes:

• CW TX chain not fully implemented
• CW MOX/key routing issue
• External KEY input not being armed/handled
• CW profile/state issue
• Regression in current Artemis build

At this stage the issue appears software-side within Artemis rather than hardware, USB, COM, WinKey or cable related. I'm mot a very experianced CW user but, I'm starting to get into it & learn it. So my CW terms maybe a little off but, the end story is it worked straight away with EESDR3 but, couldn't get it to function with Armetis.

PS. Keep up the good work I absolutely love this software!
Appreciate the hard work!
Regards,
Scott
VK4SHG

[F6Bernie]

Hi Scott, I've tested the CW for the first time and this mode doesn't work with the 2.1.6 version, my key Benscher BY2 directly connected to the Sun Pro (same settings than Expert sdr 3).
Maybee the "semi" button has no effect ? The key doesn't turn the transmiter on...
73' Bernie

[Scottay1967]

Thanks for the reply Bernie, I’m just venturing into the CW world so there’s every chance it was just me and a missed setting?🤣Regards,ScottSent from my iPhoneOn 23 May 2026, at 12:01 am, F6Bernie ***@***.***> wrote:F6Bernie left a comment (kk68/ArtemisSDR#44) Hi Scott, I've tested the CW for the first time and this mode doesn't work with the 2.1.6 version, my key Benscher BY2 directly connected to the Sun Pro (same settings than Expert sdr 3). Maybee the "semi" button has no effect ? The key doesn't turn the transmiter on... 73' Bernie —Reply to this email directly, view it on GitHub, or unsubscribe.Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you authored the thread.Message ID: ***@***.***>

[F6Bernie]

"Thanks for the reply Bernie, I’m just venturing into the CW world so there’s every chance it was just me and a missed setting?🤣"
Don't worry Scott, I'm sure there is a real issue with CW. As I said previously , maybe it's the "semi" button that is not doing the job ;) About this topic, it would be useful for the tests to fix the monitor ;)
73' Bernie

[kk68]

Hi Scott, Bernie — sorry for the silence; I've been out of town traveling.

CW being broken on both DX (Scott) and PRO (Bernie) at v2.1.6 is a real bug, and Scott's diagnostic is unusually clean: external WinKey path, internal CW memories, manual KEY-input short — all three fail to trigger TX in Artemis, same hardware works immediately in EESDR3. That points clearly at the CW MOX / keying chain inside Artemis rather than anything to do with the WK3 / cable / radio.

I don't have time to dig into this properly while I'm still light on availability, so it's going to sit open as a real-known-bug for a bit longer than I'd like. To set expectations honestly: CW keying isn't part of my own daily operating, so I haven't been hitting it in regression testing, and the path likely needs end-to-end review (KEY input → CWX state → MOX gate → TX wire → radio key line). That's not a quick fix.

If anyone reading the thread is comfortable with the codebase and wants to take a look, the relevant entry points are:

• console.cs chkMOX_CheckedChanged2 for the MOX path
• console.cs CWFWKeyer / CWPitch flow for keyer mode handling
• sunsdr.c SunSDRSetPTT for the wire-side TX entry
• The Setup → CW tab options to confirm none of them are unintentionally gating

I'll get back to this when bandwidth allows, but no promises on timing. Apologies for the slow response.

73, Kosta

[Scottay1967]

No Problems at all Kosta! You’re allowed to have a life too!I’m not doing a lot of CW at present but when you get to it you get to it! Thanks again for all you’ve done already!Regards,ScottSent from my iPhoneOn 17 Jun 2026, at 4:19 am, kk68 ***@***.***> wrote:kk68 left a comment (kk68/ArtemisSDR#44) Hi Scott, Bernie — sorry for the silence; I've been out of town traveling. CW being broken on both DX (Scott) and PRO (Bernie) at v2.1.6 is a real bug, and Scott's diagnostic is unusually clean: external WinKey path, internal CW memories, manual KEY-input short — all three fail to trigger TX in Artemis, same hardware works immediately in EESDR3. That points clearly at the CW MOX / keying chain inside Artemis rather than anything to do with the WK3 / cable / radio. I don't have time to dig into this properly while I'm still light on availability, so it's going to sit open as a real-known-bug for a bit longer than I'd like. To set expectations honestly: CW keying isn't part of my own daily operating, so I haven't been hitting it in regression testing, and the path likely needs end-to-end review (KEY input → CWX state → MOX gate → TX wire → radio key line). That's not a quick fix. If anyone reading the thread is comfortable with the codebase and wants to take a look, the relevant entry points are: console.cs chkMOX_CheckedChanged2 for the MOX path console.cs CWFWKeyer / CWPitch flow for keyer mode handling sunsdr.c SunSDRSetPTT for the wire-side TX entry The Setup → CW tab options to confirm none of them are unintentionally gating I'll get back to this when bandwidth allows, but no promises on timing. Apologies for the slow response. 73, Kosta —Reply to this email directly, view it on GitHub, or unsubscribe.Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS and Android. Download it today! You are receiving this because you authored the thread.Message ID: ***@***.***>