| Electronics > Projects, Designs, and Technical Stuff |
| Do all GPSDO's radiate 10MHz from Coax? |
| (1/2) > >> |
| NY2KW:
I use a Thunderbolt GPSDO plus TADD-1 distribution amp as a reference for my spectrum analyzer, signal generator, scope and counter. These all share a common earth chassis ground. I usually work on boatanchors but recently wanted to get into SDR and I bought the UBitx HF SDR transceiver kit. Calibration is needed and in firmware is done by zero-beating to WWV or Signal Gen on 10.000000 MHz carrier. I couldn't do it because of a loud birdie around 10.000.000 Mhz. The UBitx has 12MHz and 25MHz crystals so I didn't think the radio was the source. I quickly found that with UBitx powered off, and just using a broadband antenna into my SA, there is a -60 dBm signal at 10.000.000 MHz. I got rid of the TADD and focused on just one coax Freq Ref between Thunderbolt and any device on my bench other than the SA. If coax is disconnected from Thunderbolt, signal disappears (< -130 dBm) so coax is acting as antenna. I thought the Thunderbolt must be poorly shielded but I also have a Lucent GPS XO/Rb pair shielded like a tank and also a Leo Bodnar GPSDO reference and I see the same 10 MHz -60 dBm spur on my SA with these also. I changed to RG-402 semi-rigid shielded coax and same thing. I have a FLEX radio about 10-15' away from my workbench and it's noise floor goes down to -120 dBm and no spur there. Is this expected for any GPSDO as frequency references on the workbench? Jerry |
| ArthurDent:
WWV at 10Mhz doesn't exist in my little lab. I have so many 10Mhz cables running around, any radio is dead quiet at 10Mhz because of all the separate sources adding together. I have 3 GPSDOs, multiple 10Mhz xtal/rubidium standards, and a 10Mhz distribution amplifier with about 20 outputs and the unused outputs aren't terminated. I wouldn't even try to eliminate the radiated 10Mhz. I suspect even ground loops and the center/shield of the coax acting as a poor transformer may add to the problem as well. You shouldn't need to get WWV at 10Mhz when you have a far more accurate source available. You might have to run a long length of coax to another part of your house where the 10Mhz is near zero and calibrate the receiver there. I assume this is something that has to be done once or rarely so this may be the simplest way out. |
| NY2KW:
Thanks for replying to my post. WWV doesn't exist on my bench either :-) It was the recommendation of the UBitx user manual but I always turn to my Rigol signal generator. Have you ever looked on your SA for any buried signal around 10MHz ?? Yes, for the UBitx calibration is infrequent (hopefully) but not sure how moving location will help.... I would need to move my UBitx AND Signal Generator so I still would need a long coax from one of the GPSDO to the SigGen and I dont see why that won't act as an antenna much the same as on my workbench. Jerry |
| NY2KW:
I was thinking to try to add common mode chokes on both ends of the coax coming from GPSDO. Jerry |
| thinkfat:
If the output of the gpsdo and the inputs are proper 50 ohms impedance, the cable will not radiate. Since it does, you have an impedance mismatch, most likely at the inputs of your test equipment. You can try common mode chokes, like, winding the coax around a ring core. But I doubt it will be effective. Gesendet von meinem Nokia 6.1 mit Tapatalk |
| Navigation |
| Message Index |
| Next page |