| FlexRadio 6600 |
| Kilka słów od Rob'a Sherwooda: I am on a computer now because answering your Flex question will be rather involved. Since you asked a pointed question on a public forum, I have to be completely candid about the past few months and the results of my lab testing. In the third week of February I borrowed a Flex 6600M from a long time friend, and began testing it in my Denver lab. Initially it had a PLL problem, as this circuit wouldn’t consistently lock up to the internal TCXO clock oscillator, an external 10 MHz source from my Rubidium house standard, or the optional GPSDO oscillator. After much fiddling, the PLL locked up solid, and I ran tests for the next 6 hours. I made my normal suite of measurements, dynamic range, RMDR, noise floor, sensitivity, etc. The numbers were not as expected, being inferior to a Flex 6700 in respect to dynamic range and RMDR. The other numbers were reasonable, and as long as the flaky PLL locked up on a given boot, the radio generally worked OK. I sent a preliminary report to Flex, and obviously with the PLL issue I was issued a return authorization and the rig was sent back to Austin. At that time only the PLL problem was corrected, and unfortunately the other performance issues were not addressed. The 6600M came back to Denver a few weeks later with a repaired PLL circuit. I ran the rig through the lab again, and all my measurements were within 1 dB of my original lab report. Dynamic range was still lower than expected for receiver A, and really low for receiver B. At that point I let the rig run the rest of the day and over night for a 20 hour warm-up. At that point a thermal issue was observed, as the dynamic range of receiver A dropped to 84 dB, and receiver B increased to 84 dB. RMDR was still about 10 dB below factory specifications. In the mean time, I was able to borrow a 6600 non-M to test to determine whether I simply had a bad sample or if there was a consistent problem. The 6600 tested virtually identically to the 6600M, and I reported this back to Flex. By now it was March 15th, and almost a month had gone by since my initial long form report had been emailed to Flex. Some additional time went by before Flex confirmed that they were now seeing the same issues I had reported in February. The 6600M was sent back to Flex for a second time, and was there for a few weeks while the hardware issues were being addressed. Around three weeks ago the 6600M was shipped back to me in Denver, after having undergone hardware updates for the thermal issue and the RMDR problem. The dynamic range problem was software related, and production software has not been released as of today to correct that issue. The latest release I see today on the Flex website is v. 2.1.33. I was able to test dynamic range with non-production software, and with that pre-release software, the 6600M is performing similarly to the 6700 I tested March 17, 2017. That data set for the 6700 is listed on my website as a “second sample”. There are now several radios with more than one sample listed: IC-7300, IC-7610, K3 and K3S and the R8600. There is also data on three IC-781s going way back to 2006. In general the direct sampling radios have more data scatter than legacy superhet radios. This seems to be due to sample to sample variations in individual ADC chips. Just as the League does, I give any OEM time to correct issues that come up when I review a transceiver or receiver. It has been just short of three months since I found the problems I have explained above. It is my understanding that the hardware upgrades have been in production for several weeks. I will touch on the original data and the improved results in my Contest University presentation on Thursday before Dayton/Xenia. I’ll add the 6600M to my website before I leave for the Dayton Hamvention. Hams have been asking me for months about the new Flex products, and now less than a week before Dayton I felt it was time to respond to the question you posted today. |