Keyboard Shortcuts
Likes
Search
Arrgh! What happened?
|
I was enjoying listening to my radio & transmitting to see if it could be picked up on some SDR radios, and suddenly I have very annoying tuning"clicks" (although it sounds more like a helicopter when tuning quickly) and the birdie at 7.199 is VERY loud, whereas before this it was just noticeable. What just changed? Just for fun, I added some bypass caps to the 5V regulator on the Raduino board, but that made no difference. I'm stumped... |
|
check if the ground wire has come off somewhere. - f On Fri, Apr 14, 2017 at 9:13 AM, davetelling . <davetelling@...> wrote:
|
|
Good & bad news.. After trying a variety of ways to figure out where the tuning noise was originating, I finally tried powering the Raduino board from a separate 12V supply. That totally eliminated the tuning clicks, but did nothing for the 7.199MHZ birdie. However... when attempting to add additional power supply bypassing on the Raduino, I accidentally hit the si5351 with 12V. That did not go well.. Now, I have 10 pcs. on order, so I can blow up a few more if necessary. At any rate, it may be worth investigating the clicks in other radios. See if yo ucan split the Raduino and mani board power source & see if it helps. |
|
I added 1000pf cap and .1uf cap parallel on the 5volt regulator and a 50 ohm resistor in series with the audrino power and I don't have any clicks anymore but still have the birdie but I can deal with that for the time being. On Fri, Apr 14, 2017 at 6:55 PM, davetelling . <davetelling@...> wrote:
|
|
People have reported in the past that a large electrolytic cap across the Raduino 12v can clean up the clicking, which agrees with your finding. ?I'd go with a cap plus resistor as Farhan is using to filter the 12v to the audio amp at U1 rather than go to an entirely separate supply. ?Having the library shut down the PLL when changing frequency is something else. ?If I was serious about eliminating the 7.2mhz birdie, I'd either go to a 19mhz vfo or shield the heck out of everything from the BFO up to the crystal filter. On Fri, Apr 14, 2017 at 03:55 pm, davetelling . wrote:
I finally tried powering the Raduino board from a separate 12V supply. That totally eliminated the tuning clicks, but did nothing for the 7.199MHZ birdie. ? |
|
Good call.. what really puzzles me the most, though, is that until yesterday, I had no tuning clicks and the 7.199 birdie was almost inaudible. Suddenly, both were very noticeable and annoying. I tried several different Raduino sketches to see if any made any difference, and none did. I wonder what changed? Grounds are all good.
|
|
I think what changed was your voltage regulation. These chips can degrade under some conditions,
especially the cheap clones. You are getting noise feedback through your power supplies. As for the clicks, the data sheet does not indicate a discontinuity in coverage similar to that of the Si570. I think what happens is that there is a small discontinuity during the I2C transfer; what is being heard is actually the I2C frequency feeding back through the power supply and maybe a small pause durng reading of the registers to change the frequency. Some voltage regulators are good at suppressing this noise; some are not. Anything that changes the VR characteristics leads to a vulnerability for feed-through. Since there is variability in the VRs themselves, that would probably explain why some experience clicking noise from the beginning, and some never do. The solution is two-fold: make sure you have good voltage regulators from reputable places to use as parts; and make sure there is good filtering of the power supply (or supplies) to suppress transient noise. This includes both the Raduino and BITX power supplies. Also use a common ground point to eliminate or minimize ground loops. These solutions have been hinted at throughout the messages. Another solution is to turn off frequency generation during I2C transfer -- but I think this is masking the problem and not attacking it at the source. I could be wrong (Heaven knows I have been many times before). But this is what makes the most sense from experience and study of the data sheet. Asher opines that it is a "glitch" in the chip while changing frequency, but the data sheet does not support this. However, it seems to make sense. Jason, NT7S, is currently working on a library solution to the problem. He is the author of the now very popular Si5351 library and DDS board. The birdies problem is entirely separate and is related to the given IF frequency and the 40m mixer. The true solution to these is shielding and changing the VFO frequency to a higher range. Note that this 5 MHz./12 MHz combination will work perfectly well on other bands besides 40m. I would certainly avoid 17M with this combination, however. john AD5YE |
|
OK, great news! I got 10 pcs of si5351 from Digikey today, and was able to successfully solder in a new one on my Raduino board. I also added a 10 ohm series resistor to the 5V regulator input, and a 100 uF cap from the regulator input to ground, and I have a working radio again, with no tuning "clicks". Woo hoo!
|