Re: Spurs/Harmonics fixes for v4 based on v5 board design?
Mainly based on this review:
Seems like for $60 more you are getting a more rugged device with shielding.? The difference between the Pro and the standard is on the SDRPlay product page: just a steel case vs a shielded plastic case.
After looking at the specs of the RSP1A, you may be right.? Made the wrong choice.
Will see if I can exchange it.
Thanks,
Evan?
|
Re: Spurs/Harmonics fixes for v4 based on v5 board design?
I'm curious why you went for the RSP2.
It does switch between multiple inputs nicely.
The 12bit ADC should be plenty for what we are doing.
The RSP1A is newer and cheaper, and has the 14 bit ADC.
Apparently a good set of input filters.
Only one input port.
Jerry
toggle quoted message
Show quoted text
On Wed, Jan 23, 2019 at 05:36 PM, Evan Hand wrote:
I went ahead and bought the RSP2 for the $170.? Will share results after I get it and have a chance to learn more about it.? I want to see if it can find the spurs on 10 meters.? I can also use the RF Explorer to see how it fares when changing spans for harmonic tests.? Hopefully that will be only for a short term, and the SDR software is expanded to allow for larger band scanning.
BTW, the documentation that I have read states that the current bandwidth is 10MHz.
|
Both open source based on JackYack keyer code
Keyer:
Touch Key/er:
Both are very small and perfect for QRP travel ops!
ND6T has given permission for another release of his AGC. We have added a few goodies like S-meter tap and will be available very soon! This version will not require any drilling and should (if all goes as planned) solder right in and sit flat on the ubitx PCB.
--
David
?N8DAH
|
Re: Spurs/Harmonics fixes for v4 based on v5 board design?
I went ahead and bought the RSP2 for the $170.? Will share results after I get it and have a chance to learn more about it.? I want to see if it can find the spurs on 10 meters.? I can also use the RF Explorer to see how it fares when changing spans for harmonic tests.? Hopefully that will be only for a short term, and the SDR software is expanded to allow for larger band scanning.
BTW, the documentation that I have read states that the current bandwidth is 10MHz.
73
Evan
AC9TU
|
Re: Spurs/Harmonics fixes for v4 based on v5 board design?
Hi Jerry,
Good points. Yes, using an si5351 as a tracking calibration source should work well.
Tom, wb6b
|
Re: Spurs/Harmonics fixes for v4 based on v5 board design?
What was “the mod suggested for the v5 board “?
Anybody out there built a Sweeperino or Specan? I have gathered some of the parts...
In November Ashhar wrote “where the 12 mhz from the bfo leaks into the local oscillator. A 5uh in series with a 47 pf across C202 of the LPF will suppress it. 35 turns on T30-6 gives you 5uh.” Is this still advised?
Or did it not pan out/have bad effects/prove too hard to get right/get replaced or superseded by other mods?
Jer aa1of
|
Re: Spurs/Harmonics fixes for v4 based on v5 board design?
Calibration is what I was driving at with my previous post, but should be easy enough to take care of.
We need at tracking generator, the si5351 might do.
It gives a square wave, but an SDRplay can ignore all but the fundamental.
And that fundamental should be of constant amplitude up through 30mhz and beyond, within a dB.?
So automate using the SDRplay to measure the output of an si5351 (through a suitable attenuator)
across 1 to 30 mhz, and store those readings.? Use those readings to adjust any future measurements.
The alpha release of the spectrum analyzer software only scans one 6mhz grab (or whatever)
of the SDRplay.? ?I would not have thought stepping through all of HF would have been much more difficult,
but there may be complications I am not aware of.
?
>? if it is possible to lower the sampling frequency, that may be better
I think the trick here is to have the sampling freq low enough so the ADC works well, but not much lower.
Then restrict the incoming bandwidth to maybe 1/16'th of the Nyquist max, so we are "oversampling".
We then average each run of 16 samples into one sample, or "decimate", that average will have a few
extra bits of ADC data (more than the 14 bits the ADC can normally give) by averaging over what would
otherwise be noise.? I believe the SDRplay may be able to do this sort of thing out of the box.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Wed, Jan 23, 2019 at 04:29 PM, Tom, wb6b wrote:
I'd looked at the SDRplay receivers some time ago, with the upgrade to the ADC this looks like a promising SDR. On the comment about the sampling rate and span that can be seen at one time. With software controlling the SDR to make a Spectrum analyzer, the native span of the SDR is not likely an issue. As Jerry pointed out about the span being divided into buckets, if it is possible to lower the sampling frequency, that may be better, and let the control software tune across the wider frequency span to measure the harmonics. Continuous sweeping is no longer a requirement in software controlled receivers.?
One issue may be how to measure and calibrate the receiver across a large frequency band, so you know if the signal you measure at any particular frequency is not affected by possibly varying sensitivity of the receiver with frequency. I wonder if it may be possible to use a square wave generated by a high speed logic family and then compare the harmonics with the theoretical spectrum amplitudes of the test signal square wave.
|
Re: Spurs/Harmonics fixes for v4 based on v5 board design?
I too have been thinking on how to make this modification, though after looking at some pictures of the RF Gain modes where he did drill through the board to get a ground was looking like a good way to go.? I believe there is an open area where a hole drilled through to the ground plane on the other side could then be used as an anchor point for one side of a through hole cap that then is used to tie in the two inductors. R27 is stood on end "tombstone" stile on the C22 connection side, and the other end of one inductor is soldered to the stood up end of the resistor.? The other inductor is soldered to the vacated tab of R27.
Don't forget to slab on top of R26 and R46 approx 470 ohm SMD resistors onto each to bring them down to Aprox 220 ohms (470 ohm would bring it down to 235).
Above are my thoughts, would be interested in others.
73
Evan
AC9TU
|
Hi Woody,
?
Thanks for your interest in the smart filter controller. As I answered your question via email, I'll include the core of the reply here for others that may have the same question.
?
--------
I used a generic Arduino Nano to prototype the circuit. So the chip is a ATMega328p and the pinout it that of the Nano pins.
?
The pinout is in the source code, here:
?
I should add the schematic to the github repository, however here is a post showing the generic schematic. It may be better to use gates/filp-flops that have some hysteresis to make the circuit more immune to possible clocking on noise around the switching threshold.
?
?
One additional note, I was looking at the ATtiny85 data sheet last night. To my surprise this little chip has a PLL to multiply the clock by 8. It can provide a 64Mhz clock to some of the I/O logic. It might be able to count the approaching 30Mhz frequencies of the HF bands without a pre-scalier (although I'll need to look more carefully at the data sheet to be sure). Now if there is another AVR chip that has the same high speed clock and more pins that would be ideal for this project.
?
I took advantage of the sale Microchip is having at the moment on the PICkit 4 programmer/debugger (does both PIC and AVR chips), so I'm no longer limited to the chips supported by the Arduino boards and IDE, so I may look for other processors that can handle higher counter/timer clock frequencies than the ATmega328p used on the Arduino boards.
?
The Arduino Nano with the pre-scaler is still the path of least resistance, and a good way to go.
?
Tom, wb6b
?
|
Re: Spurs/Harmonics fixes for v4 based on v5 board design?
Dave
I am thinking avoid drilling hole.? one must disconnect at R27 somehow to insert this new filter.? I think grounding C205 to some location near where you new filter will sit is most important - find a part connected to ground (use your DVM as continuity checker) and solder one end of a leaded cap there.? now find some way to place your 2 inductors together - could even be on a tiny piece of circuit board hacked into 3 islands, or a trace attached to TP13.? I am looking to try this mod myself but still thinking how I may do it.? After installing I need to verify it does not affect CW (like the xtal filter I tried) and then see if the mixer spurs are sufficiently reduced - I need about 10 dB per measurements we made at club building session.?
Curt
|
Re: Spurs/Harmonics fixes for v4 based on v5 board design?
I'd looked at the SDRplay receivers some time ago, with the upgrade to the ADC this looks like a promising SDR. On the comment about the sampling rate and span that can be seen at one time. With software controlling the SDR to make a Spectrum analyzer, the native span of the SDR is not likely an issue. As Jerry pointed out about the span being divided into buckets, if it is possible to lower the sampling frequency, that may be better, and let the control software tune across the wider frequency span to measure the harmonics. Continuous sweeping is no longer a requirement in software controlled receivers.?
One issue may be how to measure and calibrate the receiver across a large frequency band, so you know if the signal you measure at any particular frequency is not affected by possibly varying sensitivity of the receiver with frequency. I wonder if it may be possible to use a square wave generated by a high speed logic family and then compare the harmonics with the theoretical spectrum amplitudes of the test signal square wave.
Tom, wb6b
|
Can see your point Wayne
I thought I had attached images of them
Will try again.
I have several cases that are big enough X and Y But too shallow in the Z would need to lay down the display and raduino.
|
?Uhm, and?
?LOL ?I have 2, and an unassembled BITX20.
?the exciter for the 20 says 40, but it'll be 20 when I get it done. ?I spend a lot of time not feeling well, and get delayed. need to start clearing up some projects. ?Get at least one of the 40's encased.
?One I sorta did, but that case is so small, a raduino won't fit as I'd like.. ?Oh well. ?73 Y'all
?Waye WA2YNE ? West Texas.
Have 4 of these critters in our hamshack.
|
Have 4 of these critters in our hamshack.
|
Perhaps Bill's point was that pushing a stereo plug into a stereo jack
can somehow momentarily short one of the signals to ground.
A 4 ohm resistor between uBitx main board and the headphone jack would solve this
if that's of concern.
But I think all cases of momentary shorts blowing a TDA2822 involved a WX brand TDA2822.
We'll see if Raj corrects me on this one too.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Wed, Jan 23, 2019 at 12:41 PM, Jerry Gaffke wrote:
Maybe best to just make sure there isn't anything with a mono 3.5mm plug on it within easy reach.
|
Jerry
How I get around the mono/stereo problems is to wire the headphone jack onto the two end posts only. I let the base float. Stereo headphones are in essence wired in series. monaural works the same with the end and second making the necessary contacts. This means I have to isolate the base from ground but it works well.
73
Dave
|
If you happen to have only stuff with mono plugs and sockets lying about, that's a good solution.
But all new headphones these days are stereo, so you might need an adapter to go from mono to stereo somewhere.
And if you then plug one of them mono cords into the adapter, we're back to where we started with a short to ground.
Maybe best to just make sure there isn't anything with a mono 3.5mm plug on it within easy reach.
Easy for me, anything I've bought in the last 40 years has been stereo;
?
If you do want to hot-plug headphones into a uBitx and are concerned about a possible
momentary short for some reason, could add a 4 ohm resistor between the uBitx main board and the headphone jack.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Wed, Jan 23, 2019 at 10:20 AM, Bill Cromwell wrote:
I do not have a momentary short on my audio jack as the ring (the middle connector is not wired to anything. I do not expect to ever listen to stereo on the frequencies this covers and so I just left it unused. I did not install aitcraft landing lights on my van, either <evil grin>.
Meanwhile I have long been in the habit of not plugging-unplugging my radio gear while it is powered. Even some legacy gear balks at that. I have the V3 uBitx with the non-WX chip and everything is happy.
|
Re: spurs & harmonics, whew!
I am going with Gordan Gibby's daughter board that just hosts 3 relays, mostly because my local club kitted this solution - and other than cutting traces there is little rework to the main board.? I need about 10 dB reduction of harmonics on 40 and 30 meters, and less on 80 and 20 meters to be in compliance.? (As there is reported variation in the Axicom relay approach, and different rigs require a different boost in rejection, it seems to not be the ideal approach to me - and I sense in quantity our approach costs less).?
I tried using a 45 MHz xtal filter between the bidirectional amplifier and mixer - every different way I could think of.? I even left both the series resistor and cap in place on the mixer side - no matter what I do I get a reduction of 40m CW transmit down to about 2 watts -- while no other band is affected.? At least one other rig in our batch is acting like this.? So when the final mixer is biased for CW 'direct PLL making RF' - it does not cooperate with changes on the IF port.? Have not yet tried LC filtering - don't know what it will do.? I need about 10 dB here also to filter the unwanted mixer spur on some of the higher bands.?
No other issues with my v4 transceiver.? SSB works nicely using an old electret microphone originally used with a 2m HT.?
Curt
|
I do not have a momentary short on my audio jack as the ring (the middle connector is not wired to anything. I do not expect to ever listen to stereo on the frequencies this covers and so I just left it unused. I did not install aitcraft landing lights on my van, either <evil grin>.
Meanwhile I have long been in the habit of not plugging-unplugging my radio gear while it is powered. Even some legacy gear balks at that. I have the V3 uBitx with the non-WX chip and everything is happy.
73,
Bill KU8H
toggle quoted message
Show quoted text
On 1/23/19 11:16 AM, Jerry Gaffke via Groups.Io wrote: OK, thanks to Raj for clarifying this.
I remember him reporting the non-WX TDA2822 could get remarkably hot
when driving a low impedance speaker with lots of audio,
did not recall it blowing.
Several solutions for this, any should work on non-WX TDA2822's:
? Put your finger on the TDA2822 when trying high volume levels, back off if it gets uncomfortably hot (over 60C).
? Add a resistor of 10 or 20 ohms in series with your low impedance (4 ohm?) speaker.
? Add a voltage regulator to supply less than 12v to the TDA2822
When the TDA2822's first started blowing, we were all scratching our heads as to why.
Were several reports of having it blow the moment headphones or speaker got plugged into the rig
Eventually we figured out that the parts that were blowing were marked "WX" not "FCI",
and that the momentary short put them over the edge.? ?In many cases the "WX" parts
blew spontaneously and not at high volume, these clones simply can't deal with 12v.
There are web reports of other TDA2822 clones that fail miserably at anything much over 6v.
The original TDA2822 from ST (and NJR's NJM2073D) is?rated for an absolute max voltage of 15v,
a max working voltage of 12v, though being a linear class AB amp will dissipate as much power in heat
as it gives in audio to the speaker.? Operating at 12v, it is possible to exceed the max
dissipation rating if you crank up the volume, and the chip will slowly heat up.
A momentary short at the headphone jack should not blow a quality TDA2822,
it would?take time for the heat to build up enough to destroy it.
Though to be safe, not a bad idea to avoid shorts.
I have an FCI marked TDA2822 on my v3 uBitx, I have not bothered with any fixes.
Though I'm not trying to fill the house with the music of shortwave radio,
as not everyone here finds it quite so musical.
Jerry, KE7ER
On Tue, Jan 22, 2019 at 09:54 PM, Raj vu2zap wrote:
I have had some non-WX chips fail at loud levels. On Tue, Jan 22, 2019 at 10:19 PM, MVS Sarma wrote:
The TDA2822 should not have failed but for two reasons explained by
many. I am just recapitulating them.
1. Stereo out jack with mono earphone or external speaker plugged.
This situation created short at the amplifier output.
2. using the device at its extreme working voltage. See anywhere,
you find that this device is used only at 3 or? 6V.
?Solutions where implemented , the devices should be working even now.
--
bark less - wag more
|
Oops? Not solve all problems......
Glue on a heatsink, although this will solve all problems,? it helps.
OK, thanks to Raj for clarifying this.
I remember him reporting the non-WX TDA2822 could get remarkably hot
when driving a low impedance speaker with lots of audio,
did not recall it blowing.
Several solutions for this, any should work on non-WX TDA2822's:??
? Put your finger on the TDA2822 when trying high volume levels, back off if it gets uncomfortably hot (over 60C).
? Add a resistor of 10 or 20 ohms in series with your low impedance (4 ohm?) speaker.
? Add a voltage regulator to supply less than 12v to the TDA2822
When the TDA2822's first started blowing, we were all scratching our heads as to why.
Were several reports of having it blow the moment headphones or speaker got plugged into the rig
Eventually we figured out that the parts that were blowing were marked "WX" not "FCI",
and that the momentary short put them over the edge.? ?In many cases the "WX" parts
blew spontaneously and not at high volume, these clones simply can't deal with 12v.
There are web reports of other TDA2822 clones that fail miserably at anything much over 6v.
The original TDA2822 from ST (and NJR's?NJM2073D) is?rated for an absolute max voltage of 15v,
a max working voltage of 12v, though being a linear class AB amp will dissipate as much power in heat
as it gives in audio to the speaker.? Operating at 12v, it is possible to exceed the max
dissipation rating if you crank up the volume, and the chip will slowly heat up.
A momentary short at the headphone jack should not blow a quality TDA2822,
it would?take time for the heat to build up enough to destroy it.
Though to be safe, not a bad idea to avoid shorts.
I have an FCI marked TDA2822 on my v3 uBitx, I have not bothered with any fixes.
Though I'm not trying to fill the house with the music of shortwave radio,
as not everyone here finds it quite so musical.
Jerry, KE7ER
On Tue, Jan 22, 2019 at 09:54 PM, Raj vu2zap wrote:
I have had some non-WX chips fail at loud levels.?
On Tue, Jan 22, 2019 at 10:19 PM, MVS Sarma wrote:
The TDA2822 should not have failed but for two reasons explained by many. I am just recapitulating them.
1. Stereo out jack with mono earphone or external speaker plugged. This situation created short at the amplifier output.
2. using the device at its extreme working voltage. See anywhere, you find that this device is used only at 3 or? 6V.
?Solutions where implemented , the devices should be working even now.
?
--
Adrian
VK2ALF
Cooma, Australia & Otep??, Estonia
--
Adrian
VK2ALF
Cooma, Australia & Otep??, Estonia
|
N8DAH
_Dave_ AD0B
Adrian