|
Hello Erik
I am building a tracking generator mixing the lo high output +10dBm with an external lo 433,8 MHz.
Is it possible to vary the lo level/step? by software with an additional menu?
|
On Mon, Nov 9, 2020 at 10:34 PM, <Sandro.montana@...> wrote:
Is it possible to vary the lo level/step? by software with an additional menu?
Not sure I understand. Please explain a bit more. There is a menu for mixer drive. This controls also the LO output level ? --
------------------------------------------
For more info on the tinySA go to https://tinysa.org/wiki/
|
Yes, but the level settings are only 4 values.
As I use an external lo at 433 MHz + 10dBm , the tinysa lo level for mixer drive may be in range -50 to 0 dBm in 5 dB or 1 dB step, or with a direct level input.
The output(less - 8dB of the diode mixer) of the tg board will be according the lo level.
|
There is a problem if you connect the LO output directly to a tracking mixer input other then overloading the tracking mixer input The reverse leakage of the external and internal mixers will cause swamping of the tinySA IF and severely reduce dynamic range. The best you can do is apply a good amount of attenuation between the LO output and the mixer input. Even better is to use a combination of attenuation and an amplifier with very good reverse isolation. (low S12) There is no possibility to control the power of the LO independent of the mixer drive. If you want less power from the tinySA LO you can choose NOT to enable the LO output, By doing this you still get LO output from the high port but about 30dB less power. (as the internal switch is leaking in off state)
------------------------------------------
For more info on the tinySA go to https://tinysa.org/wiki/
|
Yes I see there is an updated ic that fits si4432 from silicon lab.Will study the datasheet and try with rf filters.
Meanwhile I enjoy tinysa using in lab applications.
Thank you Erik, let you know.
|
On Sat, Dec 19, 2020 at 02:00 PM, <wanderlg@...> wrote:
I did a quick test using an ADE-1-24 mixer and a 433.9MHz TX module. I don't know if I ran the risk of burning tinySA, but the result was interesting, considering that I didn't use any filtering on the 433.9MHz TX.
? ?I performed a similar tracking generator test using an external mixer, and my dynamic range was limited to around 30dB. Quoting Erik ... ? ? The 30dB limit comes from leakage of the external LO through the external mixer from its LO port to its RF port to the tinySA high input port into the tinySA mixer. As your external LO has probably something like +10dBm and the LO-RF isolation of both mixers is maybe 20dB you get -30dBm on the RF port of the tinySA mixer. To block this leakage path you best use a 30dB attenuator and a 25dB amplifier connected between the tinySA LO output and the RF port of the external mixer. This provides substantial extra isolation and thus extra dynamic range.? I get at least 80dB dynamic range.? I get equal or better dynamic range from my NanoVNA-H4 using S21 two port measurements, so I never got around to actually testing the isolation amplifier/attenuator set-up. ?- Herb
|
Hello everyone.
I just found this thread and I see that it is very close to what I was thinking of doing. I will be using an old mixer in a metal capsule and a 433.9MHz module like the one from wanderlg. I will put everything in a metal box with some internal shielding and absorbent foam. The LO output of the TinySA will be the high level input to the mixer and I will filter the powerful 2nd harmonic of the module and attenuate its signal to about -10dBm. I hope that with this drop in level and the "inverse" attenuation of the mixers (which I trust is something over 20dB in each), I get a good dynamic range. Thanks to the "normalization" feature of TinySA, no automatic gain control will be required here.
Many thanks to Erik for his great work and kind support and to everyone else for sharing your experiences
- EA1AWY Javier Muriedas (Gijón, Asturias, Espa?a)
|
I did something similar today, just a preliminary mock up on a pad board. The frequency of the oscillator can be adjusted about 150kc with the pot on the modulation input.I have a lot of attenuation in the sweep path to get the isolation so the output overall is only about -20dbm but that can be fixed with small adjustable amp to adjust the trace between -10 and -40dbm for small amplifier testing. So far for a first go it looks nice. This high pass filter looks identical on   the tracking generator on my HP analyzer.
|
On Sun, Dec 20, 2020 at 11:42 PM, John Cunliffe W7ZQ (x-N2NEP) wrote:
I did something similar today, just a preliminary mock up on a pad board. The frequency of the oscillator can be adjusted about 150kc with the pot on the modulation input.I have a lot of attenuation in the sweep path to get the isolation so the output overall is only about -20dbm but that can be fixed with small adjustable amp to adjust the trace between -10 and -40dbm for small amplifier testing. So far for a first go it looks nice. This high pass filter looks identical on the tracking generator on my HP analyzer.
Looks good John.? Just in case you are not aware, you can normalize your through measurement to 0dB using Display->Normalize to correct for your mock up configuration. - Herb
|
For those that are building? tracking generator extensions keep in mind to allow for a frequency offset to enable measurement of mixer conversion loss as described in the tinySA wiki measurement example --
------------------------------------------
For more info on the tinySA go to https://tinysa.org/wiki/
|
Herb,
I am aware of that, the trough signal is pretty it flat only has a small amount of ripple < 2db over full span. I still want to get the output signal as close to 0 dbm as possible. am thinking around -5dbm. If I understand the normalization process right,..and I confess I might be wrong... if I would normalize a -20dbm signal it would set that as 0dbm reference but also take 20db of the dynamic range
Eric, thank you for the reminder of different frequency offset. For the start, this is just something to play with, I was bored last night so I build this for fun to see how it would compare to my HP.?
|
On Mon, Dec 21, 2020 at 09:06 AM, John Cunliffe W7ZQ (x-N2NEP) wrote:
If I understand the normalization process right,..and I confess I might be wrong... if I would normalize a -20dbm signal it would set that as 0dbm reference but also take 20db of the dynamic range
? ?Normalization doesn't affect your dynamic range, it just normalizes your set-up to 0dB so that you can measure the actual gain or attenuation of a device under test (DUT) without having to do any mental math.? Its like using the Level->External Amp setting, but it compensates for for the actual gain or attenuation of your set-up using a through calibration instead of assuming a flat numeric value. ? So with the output of your external mixer/amp/attenuator connected directly to the LOW input, after you select Display->Normalize you can then insert your DUT and the actual gain/attenuation will be displayed. ? - Herb
|
Herb, You are correct. I did a few more minor changes and swept a 144Mhz band pass filter with a span of 200Mhz and a RBW of 11khz.? As you can see from the picture, the dynamic range of this setup reaches close to 90db  Not too shabby for a 50 dollar analyzer and a handful of parts from the junk box.
|
I wonder why the text gets screwed up this way when I post a picture
|
John,
because the image as posted to the group is a mere 488 by 366
pixels in size! If one really wants to post a pic of a screen, it
helps to have the camera fully square-on to the screen.
Even your iPhone8 can't work miracles. It may well have been an
email client that reduced the image to that small a size.
Meanwhile, we can probably both be grateful that no Polaroid
film, plus the squeegee, was necessary for this operation!
[exeunt, shaking at the 尘别尘辞谤测…闭
73, Stay Safe,
Robin, G8DQX
On 21/12/2020 23:08, John Cunliffe W7ZQ
(x-N2NEP) wrote:
toggle quoted message
Show quoted text
I wonder why the text gets screwed up this way when I post a
picture
|
You could also use TinySA-app or NanoVNA-app to grab a screen cap directly from the device. 73 -Jim NU0C On Mon, 21 Dec 2020 23:43:58 +0000 "G8DQX list" <list@...> wrote: John,
because the image as posted to the group is a mere 488 by 366 pixels in
size! If one really wants to post a pic of a screen, it helps to have
the camera *fully* square-on to the screen. Even your iPhone8 can't work
miracles. It may well have been an email client that reduced the image
to that small a size.
Meanwhile, we can probably both be grateful that no Polaroid film, plus
the squeegee, was necessary for this operation! [exeunt, shaking at the
尘别尘辞谤测…闭
73, Stay Safe,
Robin, G8DQX
On 21/12/2020 23:08, John Cunliffe W7ZQ (x-N2NEP) wrote:
I wonder why the text gets screwed up this way when I post a picture
|
Hi again, everybody.
I have built a tracking generator using a 433.9MHz transmitter as LO. The out of the 1st LO of the Tiny not enabled (30dB lower approximately) is amplified by a MAR-8A and is used as the high level input for the DBM. This is to reduce the 433.9MHz leakage back through the external and internal mixers. The supply comes from the Tiny and is a little low but it seems to work. Oscillator and amplifier draw 20mA.
I get a dynamic of 60dB. The photos of the measurements are from a 6.5MHz notch filter on the NanoVNA demo card and I don't know why the "artifacts" that appear in the curve are due. The deep and down are without normalizing and the light ones up after normalizing the signal.
If anyone has an explanation for them, I would be grateful.
Cheers
- EA1AWY Javier Muriedas (Gijón, Asturias, Espa?a)
|
IF you do a measurement with tracing generator you better set the RBW to manual and maximum Can you check if this removes the artifacts?
------------------------------------------
For more info on the tinySA go to https://tinysa.org/wiki/
|
Does anyone have a reliable source for inexpensive couplers? I’m a disabled vet and I have to watch my pennies
|
?
Thanks for the suggestion Erik. I will check and report.
What I did observe was that changing the RBW varied the level delivered by the tracking generator. It was caused because the frequency of the 433.9MHz SAW oscillator did not match the IF of the TinySA. Measured on three different modules it was 433.822MHz, 433.85MHz and 433.9MHz. Introducing that offset in the IF of the TinySA, the measurements are correct with any RBW.
Biggs99: I don't know which couplers you mean
-- EA1AWY? Javier Muriedas? (Gijón, Asturias, Spain)
|
John Cunliffe W7ZQ
Jim Shorney
