It's all about the Software

How do you calibrate the rx-888? Help button is broken. Do you have to be in am mode?

You can calibrate through hardware. After disassembling the metal casing, there is a potentiometer above the shielding cover of the VCXO unit, which can be fine-tuned according to the standard frequency.

The software option in SDRC is under Radio - Options in the LH panel. AM, CW, ECSS or DSB modes are all ok if you use the carrier of a standard frequency station as the reference, as they all show a line to fix over the centre of the signal.

Attached are the frequency drifts (apparently well within +/- 1 Hz) over 30 minutes that I measured for my RX888 and my KiwiSDR.

Some caveats:

1) Room temp was stable to about +/- 0.5C. (There was no room heating/cooling with its corresponding temperature fluctuations.)
2) Supply voltages were assumed constant from the various regulated supplies (including those in the PC), but they were not monitored.
3) Measurement frequency was 10 MHz and no adjustments were made to the rx's. (Absolute accuracy was not measured, just the frequency variations.)
4) An ovenised GPSDO was used as the reference, and is assumed stable (enough!).
5) It was also assumed that the PC sound card did not significantly contribute to the drift, but this was not monitored.
6) All equipment had been running for hours prior to any measurements.
7) I think I recall that the GPS constellation is not great here at about 4pm local, and I see some evidence of that in the Kiwi's satellite monitoring panel. These measurements were made around 3.30pm local.

With all that in mind, it is interesting to see how the RX888 (with an internal TCXO) wanders more slowly, but further than the Kiwi does (with its internal GPSDO).

The Kiwi seems to initiate GPS frequency corrections about every 30 seconds. (It is not in a box here; an enclosure is on order which may help with its oscillating drift behaviour.)

Lastly, the RX888 might benefit from a (10 MHz?) reference input, but it is not too bad.

Regards,
Charles.

The main reason is that the ADC LTC2208 generates too much heat, which causes the internal temperature to vary greatly. I will consider providing an external clock input option on the RX999 to facilitate access to reference clocks such as GPSDO

Please consider a seperate powersupply input too
73 PD3LK Leon

To fei666888,

Thank you for your great work; you will have to call yourself fei666888999 !

Regards,
Charles.

How easy would it be to modify an existing RX-888 to add an external reference input?

BTW my RX-888 is moving over 4 ppm at UHF during warmup, so I assume it doesn't use a TCXO, just a simple uncompensated XO.

That's about what I found. 4ppm after 2 hours. For that check I took off both end plates to cool it down. But it is probably more sensitive to draughts in the long term. I have the intention of trying a fan to blow through. I did that on my 666 blowing on to the ADC.
I've not taken the board out yet, pictures I've seen make me wonder if it has a 27MHz reference like the 666 under the screen. I've not seen any circuit notes for an 888. My reference is not a good one, the local 1296MHz beacon is wobbly. But not on my 666. Any mods would probably need a bigger box unless an external reference was used.
A quick check of performance shows fair above 60KHz through HF. VHF and above again fair, not as good as a FCD.
Was it here someone suggested tightening the SMA nuts? Not a good idea without adding washers behind the plate. The body of the socket is set back, not flush with the edge of the PCB.
Alan