Configure DX Spider for use with custom server

[w8be]

I have been trying to get the DX Spider to work with :
dxc.w3lpl.net on socket 7373

So far I think the dx spider is connecting to the default server.

Can you please advise on the proper way to set this up. I am running the service on a rpi5 (trixie).

[accius]

Hey — your screenshots look correct, the setup should be working. Here's how to verify:

1. Make sure the DX Cluster Source dropdown is set to "Custom Telnet Server" — this is the dropdown above the host/port fields in Settings → Station. If it's still set to "DX Spider (Proxy)" or another option, the custom host/port fields are ignored even though they're visible.

2. Check your server logs — in the terminal where you run node server.js, you should see a line like:

   [DX Cluster] DX Spider: connected to dxc.w3lpl.net:7373 as W8BE-56
   

   If you see it connecting to dxspider.co.uk:7300 or one of the other default nodes instead, the source dropdown isn't set to custom.

3. Callsign — make sure your callsign is set in Settings → Station (or via the CALLSIGN env var). W3LPL's cluster requires a valid callsign login. OpenHamClock will use YOURCALL-56 automatically, or GUEST if no callsign is configured (some nodes reject GUEST logins).

Since you're on an RPi5 running as a service, you can check the logs with:

journalctl -u openhamclock --no-pager -n 50 | grep "DX Cluster"

(Replace openhamclock with whatever your service name is.)

Let me know what the logs show and we can go from there.

[w8be]

I went back and looked at .env. I changed...

Your callsign for DX cluster login (default: CALLSIGN-56)

DX_CLUSTER_CALLSIGN=W8BE-56

to..

Your callsign for DX cluster login (default: CALLSIGN-56)

DX_CLUSTER_CALLSIGN=W8BE

---

Now I see this using journalctl ...

---

I don't think I am getting a response. However I see spots.

# ===========================================
# OPTIONAL - External Services
# ===========================================

# ITURHFProp service URL (for advanced propagation predictions)
# Only uncomment if you have your own ITURHFProp service running
# ITURHFPROP_URL=https://proppy-production.up.railway.app  # Default — uses public OpenHamClock P.533 service. Override only if self-hosting.

# DX Spider Proxy URL (for DX cluster spots)
# Only uncomment if you have your own proxy running
# DXSPIDER_PROXY_URL=https://your-dxspider-proxy.com

# DX Cluster source: auto | proxy | hamqth | dxspider
# auto = tries proxy first, then HamQTH, then direct telnet
# proxy = use DX Spider Proxy (set DXSPIDER_PROXY_URL above)
# hamqth = HamQTH CSV feed (HTTP, works everywhere)
# dxspider = direct telnet to DX Spider nodes (works locally/Pi)
# DX_CLUSTER_SOURCE=auto

# ===========================================
# DX CLUSTER SETTINGS
# ===========================================

# Your callsign for DX cluster login (default: CALLSIGN-56)
DX_CLUSTER_CALLSIGN=W8BE

# Spot retention time in minutes (5-30)
SPOT_RETENTION_MINUTES=30

[w8be]

I went back and logged into dxc.w3lpl.net on port 7373 with Putty. No issues getting in. It did ask for my callsign but I could not enter sh/dx. It did not lock up but it did not seem to respond to that command. fyi

[accius]

Good news — your logs confirm it's working! The line:

[DX Cluster] Sending command: sh/dx 25 to dxc.w3lpl.net:7373 as W8BE

means OpenHamClock successfully connected to W3LPL, logged in as W8BE, and requested spots. And the 21 spots in your DX Cluster panel confirm data is flowing.

The reason sh/dx doesn't respond (both from OpenHamClock and from your PuTTY test) is that W3LPL runs AR-Cluster, not DX Spider. AR-Cluster uses a different command set — it ignores sh/dx but that's totally fine because AR-Cluster automatically pushes spots to all connected clients after login. So you won't get an initial batch of 25 historical spots, but the live stream works perfectly — which is what you're seeing.

TL;DR: everything is working as expected. Your spots are coming from W3LPL's cluster. No changes needed on your end.

(I also notice your port changed to 7000 in one of the screenshots — both 7000 and 7373 should work for W3LPL, just make sure the UI and your actual config match.)

[w8be]

Very good. A label indicating the cluster would be a nice touch. tu

[w8be]

I did on additional test . I logged on the the VA Qso Party K4T cluster and inserted a test message and it showed up in the dashboard. well done...