Differential Radiometer Experiment

224 views
Skip to first unread message

Dave Cohen

unread,
Jun 20, 2021, 10:36:43 AM6/20/21
to Society of Amateur Radio Astronomers
I am testing a possible differential radiometer setup which consists of the following components:

Two RTL-SDR (blog v3) dongles
Two Nooelec SAWBird H1 amps (powered by RTL-SDR bias tee)
One wire grid array (same as used for scope-in a box)
One 50-ohm load
NUC computer
25-ft of coax for each receiver chain

The reference LNA is connected to the 50-ohm load, and the "sky" LNA to the antenna.  LNA's are mounted close to each other in one enclosure, and SDR's in another. Observations are started with the following commands:

soapy_power -d rtl=0 -g 45 -r 2.5M -b 1024 -t 60 -c -l -O spec_rtl0_20210620.csv

soapy_power -d rtl=1 -g 45 -r 2.5M -b 1024 -t 60 -c -l -O spec_rtl1_20210620.csv

I made sure to turn on both bias tees before recording.  What I'm getting is the attached figure.  I wasn't expecting such a large difference in the height of the curves, as well as the difference in their shape.  The hydrogen signal (blue) is clearly present is the sky graph around sample 550.  I've been impressed with the decent response of this little antenna.

I'm trying to figure out why the spectral spectral responses differ so much and how to do baseline correction using the reference channel.

Any thoughts?

Dave 


combined.png

Marcus D Leech

unread,
Jun 20, 2021, 10:55:22 AM6/20/21
to sara...@googlegroups.com
Don’t forget that your 50Ohm load looks like a blackbody at about 300K. Whereas your Sky receiver is seeing perhaps 20-50K. 

Plus there will be total system gain differences of perhaps 3dB, maybe more, just due to the vagaries of RF components. 



Sent from my iPhone

On Jun 20, 2021, at 10:36 AM, Dave Cohen <weze...@gmail.com> wrote:

I am testing a possible differential radiometer setup which consists of the following components:
--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/c47c81a0-9f87-47f0-a3ef-956ad148dc4cn%40googlegroups.com.
<combined.png>

Lester Veenstra

unread,
Jun 20, 2021, 10:57:20 AM6/20/21
to sara...@googlegroups.com

Might I suggest a single LNA with a SPDT SMA Coax switch between LNA and antenna with the termination on the other switch port.

If no other for  computer to  to control switch, you might consider similar to theset:

https://www.aliexpress.com/item/1005001355318820.html?spm=a2g0o.productlist.0.0.5c8169f3T4ZkYI&algo_pvid=0902e18a-0335-4b60-b4bc-03801fe0b8d2&algo_exp_id=0902e18a-0335-4b60-b4bc-03801fe0b8d2-1

https://www.aliexpress.com/item/4000091488766.html?spm=a2g0o.productlist.0.0.5bd5502c8xuHJQ&aem_p4p_detail=202106200754443364999659624000027144097

 

Lester B Veenstra  K1YCM  MØYCM  W8YCM   6Y6Y

les...@veenstras.com

 

452 Stable Ln (HC84 RFD USPS Mail)

Keyser WV 26726

 

GPS: 39.336826 N  78.982287 W (Google)

GPS: 39.33682 N  78.9823741 W (GPSDO)

 

 

Telephones:

Home:                     +1-304-289-6057

US cell                    +1-304-790-9192

Jamaica cell:           +1-876-456-8898

--

Lester Veenstra

unread,
Jun 20, 2021, 12:58:16 PM6/20/21
to Lester Veenstra, sara...@googlegroups.com

Dave Cohen

unread,
Jun 21, 2021, 5:08:47 AM6/21/21
to Society of Amateur Radio Astronomers
Marcus,

I think you got it right on both accounts.  See the attached plot.

That begs the question if it is even worth using two receiver chains versus a coaxial relay as Lester suggested.  Can the differences in response between the receiver chains be calibrated out?
compared.png

Marcus D Leech

unread,
Jun 21, 2021, 9:34:14 AM6/21/21
to sara...@googlegroups.com
The static gain difference can be calibrated out. Just artificially multiply the weaker signal by a constant that makes them roughly equal. 

The trouble will come In the dynamic gain changes due to temperature changes. They won’t be precisely identical in each receiver chain. Sometimes, this is not a big deal particularly if relevant components strongly share a common thermal “fate”. 

Sent from my iPhone

On Jun 21, 2021, at 5:08 AM, Dave Cohen <weze...@gmail.com> wrote:

Marcus,

Marcus D Leech

unread,
Jun 21, 2021, 9:44:27 AM6/21/21
to sara...@googlegroups.com
Also, this version of the sawbird LNA has a built in switch and terminator. 


The noise figure is a bit higher than you’d like due to the switching device. 

Sent from my iPhone

On Jun 21, 2021, at 9:34 AM, Marcus D Leech <patchv...@gmail.com> wrote:

The static gain difference can be calibrated out. Just artificially multiply the weaker signal by a constant that makes them roughly equal. 

Lester Veenstra

unread,
Jun 21, 2021, 10:26:26 AM6/21/21
to sara...@googlegroups.com

Even if you calibrate out one time there gain vs temp functions may not be the same.

I am just assembling SCOPEINBOX so , so far I have yet got into the software,  to see where you might want to control the switch. 

 Time to start searching EBAY for low cost surplus SMA RF Switchews

 

 

Lester B Veenstra  K1YCM  MØYCM  W8YCM   6Y6Y

les...@veenstras.com

 

452 Stable Ln (HC84 RFD USPS Mail)

Keyser WV 26726

 

GPS: 39.336826 N  78.982287 W (Google)

GPS: 39.33682 N  78.9823741 W (GPSDO)

 

 

Telephones:

Home:                     +1-304-289-6057

US cell                    +1-304-790-9192

Jamaica cell:           +1-876-456-8898

 

 

Lester Veenstra

unread,
Jun 21, 2021, 10:30:38 AM6/21/21
to sara...@googlegroups.com

 

Lester B Veenstra  K1YCM  MØYCM  W8YCM   6Y6Y

les...@veenstras.com

 

452 Stable Ln (HC84 RFD USPS Mail)

Keyser WV 26726

 

GPS: 39.336826 N  78.982287 W (Google)

GPS: 39.33682 N  78.9823741 W (GPSDO)

 

 

Telephones:

Home:                     +1-304-289-6057

US cell                    +1-304-790-9192

Jamaica cell:           +1-876-456-8898

 

 

From: 'Lester Veenstra' via Society of Amateur Radio Astronomers [mailto:sara...@googlegroups.com]
Sent: Monday, June 21, 2021 10:26 AM
To: sara...@googlegroups.com
Subject: RE: [SARA] Differential Radiometer Experiment

 

Even if you calibrate out one time there gain vs temp functions may not be the same.

Lester Veenstra

unread,
Jun 21, 2021, 10:45:44 AM6/21/21
to sara...@googlegroups.com

Interesting   Will try and see how much degradation to the T of the antenna G/T there is with this option.

 

Lester B Veenstra  K1YCM  MØYCM  W8YCM   6Y6Y

les...@veenstras.com

 

452 Stable Ln (HC84 RFD USPS Mail)

Keyser WV 26726

 

GPS: 39.336826 N  78.982287 W (Google)

GPS: 39.33682 N  78.9823741 W (GPSDO)

 

 

Telephones:

Home:                     +1-304-289-6057

US cell                    +1-304-790-9192

Jamaica cell:           +1-876-456-8898

 

 

From: sara...@googlegroups.com [mailto:sara...@googlegroups.com] On Behalf Of Marcus D Leech
Sent: Monday, June 21, 2021 9:44 AM
To: sara...@googlegroups.com
Subject: Re: [SARA] Differential Radiometer Experiment

 

Also, this version of the sawbird LNA has a built in switch and terminator. 

Marcus D. Leech

unread,
Jun 21, 2021, 12:40:19 PM6/21/21
to sara...@googlegroups.com
On 06/21/2021 10:26 AM, 'Lester Veenstra' via Society of Amateur Radio Astronomers wrote:

Even if you calibrate out one time there gain vs temp functions may not be the same.

I am just assembling SCOPEINBOX so , so far I have yet got into the software,  to see where you might want to control the switch. 

 Time to start searching EBAY for low cost surplus SMA RF Switchews

There is much "it depends" with this approach.

If the LNAs and receivers share the same thermal and power-supply environment, then this can work very well in some situations--
  in particular where the phenomenon being measured is larger than the discrepancy in the gain/temperature function on the
  two sides.  I've used this in my riometer designs with very great success.


Lester Veenstra

unread,
Jun 21, 2021, 1:38:25 PM6/21/21
to sara...@googlegroups.com

Their

Dave Cohen

unread,
Jun 21, 2021, 11:32:37 PM6/21/21
to Society of Amateur Radio Astronomers
Lester,

There are some reasonably-priced surplus SMA coaxial switches on Ebay,  in the 20-30 dollar range.  I consider this a last resort as there would need to be a way to activate the switch.  The RTL-SDR dongles come with GPIO, but I don't think it's brought out to terminals or a solder pad.  You might have to solder directly to one of the leads on the RTL2832 chip.  Then you need a source of 28 VDC to turn on the relay, and some interface electronics to convert the GPIO signal to 28 V to drive the relay.

So I was thinking - maybe instead of a 2nd receiver chain, you characterize the single receiver at different temperatures into a 50-ohm load, and see how the curve varies with temperature.  Subtract the curve from the actual sky data, and when you get a straight line (at points other than the hydrogen signal), then you might have the best fit.  It's sort of like taking "darks" and "flats" for those that dabble in astrophotography.  I'm collecting data right now, and I'll compare the curves as temperature falls during the night.  I don't have enough data to make a final conclusion, but it looks like the response is shifting vertically with increasing temperature, but with the same overall curve shape.  At least that's what I'm hoping for.  Then it's a matter of adding a constant value, as Marcus said.  If the curve actually distorts based on temperature, then I'm going to take some days off from work and think deeper on this problem.  I can't wrap my head around that quite yet.

I'm interested on the results you get with the relays you ordered.  Please publish.  Thanks for your insight.

Dave

Hamish Barker

unread,
Jun 22, 2021, 1:03:43 AM6/22/21
to sara...@googlegroups.com
Pretty sure the slope of the gain curve varies with temperature as well as shifts vertically. At least that's my experience with sawbird h1+ and both rtlsdr and nooelec smartee ( which is the same receiver chip). 

a possible problem with the idea of Calibrating 50 ohm load at various temps is that the antenna is likely not quite perfectly 50 ohm and likely has at least a little slope in frequency dependent gain ( and reactance). Since the h1 signal is so small, this may be a problem. But go for it with attempting calibration of temp dependence, since real data is worth much more than handwaving arguments.

To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/8e86ffbb-9a66-4d7d-a2d1-80c5833c4411n%40googlegroups.com.

Jeff Kruth

unread,
Jun 22, 2021, 8:13:56 AM6/22/21
to weze...@gmail.com, sara...@googlegroups.com
One aspect that I have not seen addressed is the mechanical switch wear & tear: Most radiometers use electronic switches for temp cal/subtractions. Mechanical contact wear will change the insertion loss and give unreliable readings. Most systems used, in particular, ferrite latching switches in waveguide or coax. These are generally un-obtainium for amateurs and not many companies make them now. Of course this depends on the frequency of switching. But a new mechanical RF relay switch is only good for 1 million cycles and who knows how old the switch you get is.

Other methods work well for noise cal, including the use of directional couplers for noise -added schemes. High level noise can be PWMed to give lower level readings, since the integration smooths it out. Point at the trees, using an accurate outdoor temp reading, then set your PWM width to give the same reading. Turn noise on/off as fast as you like when needed. Lots of literature about this including some good chapters in "Serendipitous Discoveries in Radio Astronomy" (title? IIRC).(not enough coffee yet...)
 
Jeff Kruth

mike....@gmail.com

unread,
Jun 22, 2021, 8:25:56 AM6/22/21
to Society of Amateur Radio Astronomers
The Nooelec sawbird+h1 that I just got has a 4 term interface which the documentation says one terminal is a "switch" that terminates it to a 50 ohm load for calibration. I have not tried it.  

Marcus D Leech

unread,
Jun 22, 2021, 8:34:52 AM6/22/21
to sara...@googlegroups.com
Yes the “barebones” version of that LNA has a built-in switch and terminator. The packaged version in an enclosure does not. 

I helped in the design of that device. It uses a TQP3M9037 first stage and a low-loss switch. It had higher noise figure than any of us were expecting. So the follow-on versions in enclosures used a different parts lineup and eliminated the calibration feature. 

Now having said that, the barebones version with the built in switch will perform at least as well as any external-switch setup would. 

Sent from my iPhone

On Jun 22, 2021, at 8:25 AM, mike....@gmail.com <mike....@gmail.com> wrote:

The Nooelec sawbird+h1 that I just got has a 4 term interface which the documentation says one terminal is a "switch" that terminates it to a 50 ohm load for calibration. I have not tried it.  

Lester Veenstra

unread,
Jun 22, 2021, 10:54:24 AM6/22/21
to sara...@googlegroups.com, weze...@gmail.com

Good point; but we are not trying for Dicke precision or switching rate.  

 

I would advocate for mechanical vs diode switching to maintain lower insertion loss.

 

For example on the NOOK LNAs, the standard one claims a NF of    · 0.8dB noise figure at 1420MHz    and the unit with the internal terminated switch  · 1.05dB noise figure at 1420MHz.  (Actually not that bad considering the price class)_

 

I will exercise one of my SMA microwave switches with long term square wave switch drive and measure before and after.

 

Remember this is not NARO precision but a small step up from the manual single time switch to a 50 ohm term in place  of the antenna feed.

 

For the purist, do not forget to stick a thermister on the 5o ohm load, but for RTL-SDR, the simple looks fine.

Lester Veenstra

unread,
Jun 22, 2021, 10:56:29 AM6/22/21
to sara...@googlegroups.com

Does the “BareBones” and the “Switch Terminated” fit in same case?

Marcus D Leech

unread,
Jun 22, 2021, 10:57:27 AM6/22/21
to sara...@googlegroups.com
I don’t think they do. 

Sent from my iPhone

On Jun 22, 2021, at 10:56 AM, 'Lester Veenstra' via Society of Amateur Radio Astronomers <sara...@googlegroups.com> wrote:



Lester Veenstra

unread,
Jun 22, 2021, 11:09:31 AM6/22/21
to sara...@googlegroups.com

 

 

Nooelec SAWbird+ H1m Barebones  with case

Physical Specifications

Parameter

Value

Unit

Length

80 (3.15)

Millimeter (Inch)

Width

30 (1.18)

Millimeter (Inch)

Height

20 (0.79)

Millimeter (Inch)

Weight

52 (0.11)

Gram (Pound)

 

 

 

SAWbird+ H1t  barebones

Physical Specifications

Parameter

Value

Unit

Length

75 (2.95)

Millimeter (Inch)

Width

25 (0.98)

Millimeter (Inch)

Height

15 (0.59)

Millimeter (Inch)

Weight

14 (0.03)

Gram (Pound)

Paul Oxley

unread,
Jun 22, 2021, 4:41:22 PM6/22/21
to weze...@gmail.com, sara...@googlegroups.com
Jeff

An alternative is to place the switch only in the path of the calibration signal. The antenna is fed via a low loss path with the calibrator fed by a high loss path. Unequal ratio hybrids could be used. However for the amateur, the simplest would be to have the calibration signal injected by a small probe (~1 wave length) on the edge of the antenna. The only loss in the desired path would be the shadow of the calibration probe.

This type of arrangement also brings the calibration signal lower nearer the desired signal level.

Of course, one would need to establish the values of the calibration signal over ambient air temperature. The impact of dirty relay contacts would be minimal since the only thing impacted is the value of the calibrator which is much higher in level.

Paul

Marcus D. Leech

unread,
Jun 22, 2021, 4:43:29 PM6/22/21
to sara...@googlegroups.com
On 06/22/2021 04:40 PM, Paul Oxley wrote:
Jeff

An alternative is to place the switch only in the path of the calibration signal. The antenna is fed via a low loss path with the calibrator fed by a high loss path. Unequal ratio hybrids could be used. However for the amateur, the simplest would be to have the calibration signal injected by a small probe (~1 wave length) on the edge of the antenna. The only loss in the desired path would be the shadow of the calibration probe.

This type of arrangement also brings the calibration signal lower nearer the desired signal level.

Of course, one would need to establish the values of the calibration signal over ambient air temperature. The impact of dirty relay contacts would be minimal since the only thing impacted is the value of the calibrator which is much higher in level.

Paul
That's roughly equivalent to a directional coupler approach (noise adding).  I've used that also.  The direct-path loss on a good
  directional coupler is very low.




Hamish Barker

unread,
Jun 22, 2021, 4:51:18 PM6/22/21
to sara...@googlegroups.com
I have both and will check

To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/081e01d76776%24c4e542c0%244eafc840%24%40com.

Hamish Barker

unread,
Jun 22, 2021, 4:56:08 PM6/22/21
to sara...@googlegroups.com
or get a glassblower to make a neon tube to fit through the waveguide on a slant. (or fit one of the small diameter low wattage fluorescent tubes, or scavenge one of the 2mm diameter ones from an old laptop screen backlight or flatbed scanner).

When it is not energised, nonconductive gas and no noise. When energised, supposedly flat noise source. I think I read that a calibration noise source used in the holmdel horn used to discover the CMB was a gas discharge tube.

To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/150304284.3666358.1624394451106%40mail.yahoo.com.

Paul Oxley

unread,
Jun 22, 2021, 4:59:07 PM6/22/21
to sara...@googlegroups.com
Hamish

Great! Simplest solution without a switch or excessive loss in the desired signal path.

Paul

Paul Oxley

unread,
Jun 22, 2021, 5:10:59 PM6/22/21
to sara...@googlegroups.com
Hamish

An additional thought. A simple gas discharge tube is a neon light such as a NE51 bulb. 

Paul

Jeff Kruth

unread,
Jun 23, 2021, 1:31:46 AM6/23/21
to sara...@googlegroups.com
This is called a Mumford tube, invented in the mid '50's, and is not such a great idea as there is much more to the concept than stated. I have many waveguide gas tube noise sources. They need a high voltage firing pulse and a very good controlled current supply as the noise supplied is dependent on the current. The length of the tube at L band would need to be about 24 inches based on my S-band tubes. The neon pressure is critical.  Also the angle is critical for low reflection coefficient. Loss can be very significant is care is not taken to insure the VSWR is not hosed by sticking a 1/2 inch diameter (yes, thats the size for L band) in a horn system.
A 35 dB ENR solid state noise source coupled into the horn thru a small 20 to 30 dB coupling loop is MUCH MUCH better. The high ENR sources are available, or you can make one using MMIC amps, just make sure you put them in a little temp controlled oven to keep the gain stages constant. I have presented these approaches several times at SARA conferences over the last 26 years or so.

We use similar calibration systems in both our 1421 MHz feed and the 8.4 GHz DSN system here at the university, both designed by me and vetted by outside reviewers.
 
The main issue is to not contribute much noise to your system in the "off" state (dont want to spoil your system noise temp), and to have a relatively precise and STABLE noise source in the "on" state.
Do a Y factor measurement by observing the cold sky with the noise off, then turn it on, do the simple Y factor math and viola! you have your Tsys!! Great for tracking system performance and degradation. Also the absolute temperature contribution can be set and this can be used to measure source temperature in the sky, providing "cal" marks on your chart! You can figure out how many Janskys the thing you are looking at really is!
 
Jeff Kruth
 
 

Hamish Barker

unread,
Jun 23, 2021, 1:52:14 AM6/23/21
to sara...@googlegroups.com
jeff has real world experience, i defer to that, thanks!

One question: if the tube is glass and is switched off, does the glass tube alone have an effect on the horn/waveguide? I thought it would have no effect but does the glass dielectric have an effect?

To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/1379788306.3799303.1624426300439%40mail.yahoo.com.

Jeff Kruth

unread,
Jun 23, 2021, 2:20:56 AM6/23/21
to hamish...@gmail.com, sara...@googlegroups.com
Hi Hamish!

Yes, its does. The Er of glass is 4 and so has a significant reflection coefficient. Also there is loss associated with it, depending on the type of glass. Nice thin-walled quartz tubes would be best. Lowest loss, good properties.
 
All of the "big boys" have gone to solid state noise sources for reliability, ease of use, etc. That what they use at Greenbank on the GBT and others. I recently received a truckload of equipment here from Goldstone (JPL) and in it was an S-Band noise calibration system where they used 3 solid state noise diodes, temperature controlled in an oven assy, coupled into the main receive path by different value couplers so that the could "adjust" the noise injection in discrete steps without and step attenuators or complicated controllers.
 
The Mumford tube is still used at NIST (NBS) for noise source calibration because it is a fundamental physical standard and can be related back to basic measurements. IIRC, the fill gas of choice was mostly Argon. The company that made most of the noise sources was called Signalite and they sold their gas tube business in the early 2000's (IIRC) to a company called High Energy Physics, who I believe may now be out of business. When I contacted them a while ago (5 yrs?) about a millimeter wave noise standard tube for 110 GHz, they told me that the guy who made the tubes had died and the demand was so low that they just dropped the product line. So gas tube noise sources (may) have gone the way of the dinosaur. I still a bunch, esp for the bands above 18 GHz as they are a great reference to compare a solid state source to.
 
My S-band noise tube assemblies are about 3-1/2 feet long. Most are made with a pointed lossy iron waveguide termination in them (300K low VSWR temp source) that is removable by unbolting the waveguide flange, to insert the source in a standard waveguide system. They produce (are designed for) a standard ENR of 15.2 dB which (using 293K as a reference) means that the effective hot temperature of the source is 30 x 293 or about 9000 Kelvin.
 
When doing Y factor (hot/cold) measurements, you want the "hot" temp to be MUCH greater than the "cold" temp (if possible) 
 
We have been doing Y factor calibration to determine G/T for our dish in the DSN band for commissioning or station, DSS-17, into the NASA DSN network. We use the moon which JPL and others have figured has a radiometric temperature at that frequency of about  185 K and we use cold sky of about 9K (Zenith in Kentucky, once again calculated for us by JPL). While not great, we can get atmospheric contribution at low altitudes this way by tracking the moon as it rises. That way we can develop a G/T profile VS Elevation angle for our system. 
 
I have spent a lot of time with "noise"....
 
Jeff Kruth

Marcus D. Leech

unread,
Jun 23, 2021, 8:46:07 AM6/23/21
to sara...@googlegroups.com
On 06/23/2021 01:31 AM, 'Jeff Kruth' via Society of Amateur Radio Astronomers wrote:
This is called a Mumford tube, invented in the mid '50's, and is not such a great idea as there is much more to the concept than stated. I have many waveguide gas tube noise sources. They need a high voltage firing pulse and a very good controlled current supply as the noise supplied is dependent on the current. The length of the tube at L band would need to be about 24 inches based on my S-band tubes. The neon pressure is critical.  Also the angle is critical for low reflection coefficient. Loss can be very significant is care is not taken to insure the VSWR is not hosed by sticking a 1/2 inch diameter (yes, thats the size for L band) in a horn system.
A 35 dB ENR solid state noise source coupled into the horn thru a small 20 to 30 dB coupling loop is MUCH MUCH better. The high ENR sources are available, or you can make one using MMIC amps, just make sure you put them in a little temp controlled oven to keep the gain stages constant. I have presented these approaches several times at SARA conferences over the last 26 years or so.

We use similar calibration systems in both our 1421 MHz feed and the 8.4 GHz DSN system here at the university, both designed by me and vetted by outside reviewers.
 
The main issue is to not contribute much noise to your system in the "off" state (dont want to spoil your system noise temp), and to have a relatively precise and STABLE noise source in the "on" state.
Do a Y factor measurement by observing the cold sky with the noise off, then turn it on, do the simple Y factor math and viola! you have your Tsys!! Great for tracking system performance and degradation. Also the absolute temperature contribution can be set and this can be used to measure source temperature in the sky, providing "cal" marks on your chart! You can figure out how many Janskys the thing you are looking at really is!
 
Jeff Kruth
 

The noise source I use for my riometers is a 1N4744 "Zener" diode in reverse bias in series with a forward-biased 1N4001 or 1N914.
  This is then mounted onto a chunk of aluminum.  Feed the whole thing through a roughly 47K-100K series resistor, and the RF
  "take off" point is between the resistor and 1N4744.   Use about 20V (laptop power supply) to stimulate it.

Zeners and ordinary silicon diodes have temperature coefficients that are of very similar magnitude, but opposite in sign.  So the
  arrangement tends to be self-stabilizing.

The output at 40Mhz is about +37dB ENR.  I haven't measured at L-band but I expect that it's considerably less.


Jeff Kruth

unread,
Jun 23, 2021, 10:18:04 AM6/23/21
to patchv...@gmail.com, sara...@googlegroups.com
Very good! Forward bias B-E junctions of NPN bjt's have also been used for HF-VHF noise sources, sometimes followed by a lot of gain blocks to get high level noise.
You are correct that there will be an appreciable noise slope due to shunt capacitance effects, both in the junction and without.
Jeff Kruth
 

Hamish Barker

unread,
Jun 23, 2021, 4:45:34 PM6/23/21
to sara...@googlegroups.com
Cheap effective noise calibration (source and procedure) would be a good topic for a reference document to go on the Sara website. Jeff and Marcus posts have just about covered it. A schematic and couple of pictures to round it out?

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/713175305.3921516.1624457878020%40mail.yahoo.com.

Marcus D. Leech

unread,
Jun 23, 2021, 4:58:27 PM6/23/21
to sara...@googlegroups.com
On 06/23/2021 04:45 PM, Hamish Barker wrote:
> Cheap effective noise calibration (source and procedure) would be a
> good topic for a reference document to go on the Sara website. Jeff
> and Marcus posts have just about covered it. A schematic and couple of
> pictures to round it out?
>
Attached a .pdf of a schematic.

YMMV. Void where prohibited, etc.


untitled.pdf

Paul Oxley

unread,
Jun 23, 2021, 5:06:58 PM6/23/21
to sara...@googlegroups.com
Marcus

Looks good as a noise generator. However, some of the concerns expressed were on how to switch from the calibration source to the antenna. 

Maybe the schematic should show the full arrangement.

The level of the noise signal may also be of concern. 

Paul


--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+unsub...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/60D3A06F.3060901%40gmail.com.

Marcus D Leech

unread,
Jun 23, 2021, 5:09:10 PM6/23/21
to sara...@googlegroups.com
This was a follow up to the sub-thread about noise adding rather than switching. Something that you yourself suggested. 



Sent from my iPhone

On Jun 23, 2021, at 5:06 PM, Paul Oxley <oxl...@att.net> wrote:


To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/578218532.4101760.1624482414016%40mail.yahoo.com.

Paul Oxley

unread,
Jun 23, 2021, 5:22:53 PM6/23/21
to sara...@googlegroups.com
Marcus

It would be more useful if the unequal ratio combiner or directional coupler was shown on the schematic. Maybe someone with a working RA receiver can verify what works best.

The level of the calibration signal should also be more specific. Do you need to add "gain blocks"? If so, how much gain is needed.

Paul 



Marcus D Leech

unread,
Jun 23, 2021, 6:07:09 PM6/23/21
to sara...@googlegroups.com
Well maybe someone else wants to do that. My schematic capture program doesn’t do generic drawing. There are dozens of directional couplers “out there”. My parts library doesn’t have any. 

Jeff has also discussed putting a small coupling loop inside the waveguide red. Y the source. I have no way of showing that in my schematic capture tooling.



Sent from my iPhone

B & MR Randall

unread,
Jun 24, 2021, 4:47:36 AM6/24/21
to sara...@googlegroups.com

Hello Jeff and All,

Back in 2005 I did a SARA Journal paper on noise calibration.  (Attached)  Checked links & they still work.  From later work in lab at 408MHz,  I discovered that mismatch at antenna or receiver input alter the calibration a bit.  There is a reason Greenbank receivers use a lot of ferrite isolators.  LNA's of that vintage always had a good bit of mismatch at the input.

 

I drew the coupler as I did because that corresponds to the physical position of the connectors on most coax line or strip line couplers.  Many couplers will have an internal termination on one port.  I have seen surplus couplers with internal termination blown out.  Those are not usable!  If there is an internal termination, it normally will check out as 50 ohms on an ohmmeter through one of the other ports.

 

Jeff Kruth supplied one of my directional couplers, a 30 dB coupler for 200 to 400MHz.  It has loss in the main line below my measurement ability.

 

With the help of the mini VNA's available now, you could characterize a surplus coupler.   My characterization of couplers in 2005 used a URM26 signal generator with a diode detector ant an HP 415.  A couple of super good 50 ohm terminator are needed.  A Mini VNA can give answers for each measurement in seconds, instead of 5 minutes. 

                Terminate all ports except one.  Measure return loss for each port this way.

                Terminate all ports except two.  Measure insertion loss between each pair.

You can now establish the personality of the coupler.

 

Years ago Charles Osborne fixed me up with a couple of super good terminations, with data from his lab.  I keep them in a padded box & measuer DC resistance to assure it did not change in storage before I use them.

 

Sorry about rambling on here.  Waiting for comments from Jeff & Charles who both understand RF measurements better then me.

 

Bruce Randall

 

 

From: sara...@googlegroups.com [mailto:sara...@googlegroups.com] On Behalf Of Paul Oxley
Sent: Wednesday, June 23, 2021 5:23 PM
To: sara...@googlegroups.com
Subject: Re: [SARA] Differential Radiometer Experiment

 

Marcus

Noise_Cal.pdf

Lamar Owen

unread,
Jun 24, 2021, 11:25:05 AM6/24/21
to sara...@googlegroups.com
On 6/22/21 8:12 AM, 'Jeff Kruth' via Society of Amateur Radio
Astronomers wrote:
> One aspect that I have not seen addressed is the mechanical switch
> wear & tear: Most radiometers use electronic switches for temp
> cal/subtractions. Mechanical contact wear will change the insertion
> loss and give unreliable readings....


Contact resistance can change from one closure to the next; seen that a
few times in a few areas.  Mechanical switching is really OK for very
light duty only.


Directional couplers have the disadvantage of increasing Tsys, but with
a low enough LNA temp that can be dealt with.  The RAS LNA is specified
for 20K (0.29dB NF).  You then 'key' the noise source at the coupler on
or off as needed with the DC power driving the source.  The noise source
needs to be a precision noise diode, or again you get uncertainty from
one 'key' to the next; the DC power may need to be tightly regulated,
too, and you wouldn't want a mechanical relay doing that keying (a simple


> Turn noise on/off as fast as you like when needed. Lots of literature
> about this including some good chapters in "Serendipitous Discoveries
> in Radio Astronomy" (title? IIRC).(not enough coffee yet...)

"Serendiptitous Discoveries in Radio Astronomy: Proceedings of a
Workshop held at the National Radio Astronomy Observatory Green Bank,
West Virginia on May 4, 5, 6, 1983"  Published by NRAO.
https://www.amazon.com/Serendipitous-Discoveries-Radio-Astronomy-Kellermann/dp/9995393859

Lester Veenstra

unread,
Jun 24, 2021, 11:26:58 AM6/24/21
to sara...@googlegroups.com

For a low cost, uncalibrated, noise source, one of the simplest is a series , or even one, of amplifiers with the input terminated.  For L-Band (i.e. 1450 MHz), the ubiquitous TVRO line amplifiers are good, as is, or modified for external (non coax) power feed.

 

 

https://images-na.ssl-images-amazon.com/images/I/31d0pPjIz8L._AC_.jpg


ASKA 20 dB In-Line Amplifier Satellite DBS LNB 2.5 GHz 950-2400 MHz Signal Amp

 

 

https://www.amazon.com/Line-Amplifier-Satellite-950-2400-Signal/dp/B01KKAS042/ref=sr_1_2?dchild=1&keywords=tvro+line+amplifier&qid=1624546964&sr=8-2

https://www.ebay.com/itm/224459605410?hash=item3442d5c5a2:g:qn4AAOSw47VgnFVg

 

 

in SMA

BIAS TEE

https://www.ebay.com/itm/353031317783?hash=item52324e6917:g:e7sAAOSwA9xehF4q

 

https://www.aliexpress.com/item/4000354917121.html?spm=a2g0o.productlist.0.0.6c553d19keJVuC&algo_pvid=7505f2bc-a224-4c64-86b8-887f9e465a94&algo_exp_id=7505f2bc-a224-4c64-86b8-887f9e465a94-3

 

 

20 dB Amp

https://www.ebay.com/itm/174791411880?hash=item28b26148a8:g:Kw0AAOSwwg1guZff

 

https://www.aliexpress.com/item/1005002064326648.html?spm=a2g0o.productlist.0.0.6c553d19keJVuC&algo_pvid=7505f2bc-a224-4c64-86b8-887f9e465a94&algo_exp_id=7505f2bc-a224-4c64-86b8-887f9e465a94-42

image001.jpg

Marcus D Leech

unread,
Jun 24, 2021, 11:30:20 AM6/24/21
to sara...@googlegroups.com
The problem with that approach is that they’re much more sensitive to supply voltage variability and temperature changes than a simple diode noise source. 

But yes, this will work. 

Sent from my iPhone

On Jun 24, 2021, at 11:26 AM, 'Lester Veenstra' via Society of Amateur Radio Astronomers <sara...@googlegroups.com> wrote:



For a low cost, uncalibrated, noise source, one of the simplest is a series , or even one, of amplifiers with the input terminated.  For L-Band (i.e. 1450 MHz), the ubiquitous TVRO line amplifiers are good, as is, or modified for external (non coax) power feed.

 

 

<image001.jpg>


ASKA 20 dB In-Line Amplifier Satellite DBS LNB 2.5 GHz 950-2400 MHz Signal Amp

 

 

https://www.amazon.com/Line-Amplifier-Satellite-950-2400-Signal/dp/B01KKAS042/ref=sr_1_2?dchild=1&keywords=tvro+line+amplifier&qid=1624546964&sr=8-2

https://www.ebay.com/itm/224459605410?hash=item3442d5c5a2:g:qn4AAOSw47VgnFVg

 

 

in SMA

BIAS TEE

https://www.ebay.com/itm/353031317783?hash=item52324e6917:g:e7sAAOSwA9xehF4q

 

https://www.aliexpress.com/item/4000354917121.html?spm=a2g0o.productlist.0.0.6c553d19keJVuC&algo_pvid=7505f2bc-a224-4c64-86b8-887f9e465a94&algo_exp_id=7505f2bc-a224-4c64-86b8-887f9e465a94-3

 

 

20 dB Amp

https://www.ebay.com/itm/174791411880?hash=item28b26148a8:g:Kw0AAOSwwg1guZff

 

https://www.aliexpress.com/item/1005002064326648.html?spm=a2g0o.productlist.0.0.6c553d19keJVuC&algo_pvid=7505f2bc-a224-4c64-86b8-887f9e465a94&algo_exp_id=7505f2bc-a224-4c64-86b8-887f9e465a94-42

 

 

 

Lester B Veenstra  K1YCM  MØYCM  W8YCM   6Y6Y

les...@veenstras.com

 

452 Stable Ln (HC84 RFD USPS Mail)

Keyser WV 26726

 

GPS: 39.336826 N  78.982287 W (Google)

GPS: 39.33682 N  78.9823741 W (GPSDO)

 

 

Telephones:

Home:                     +1-304-289-6057

US cell                    +1-304-790-9192

Jamaica cell:           +1-876-456-8898

 

 

 

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.

Lester Veenstra

unread,
Jun 24, 2021, 11:45:47 AM6/24/21
to sara...@googlegroups.com

Agree:  Series amplifiers:  Cheeep and simple but, yes, not as stable as (for a real astronomy application) a real noise source

 

However there are a number of moderate priced, Chinese,  real diode noise sources (I ordered some to play with)

 

https://www.ebay.com/itm/174747976989?hash=item28afca851d:g:rgYAAOSw7AFgh6cy

 

 

https://www.ebay.com/itm/265206649801?_trkparms=ispr%3D1&hash=item3dbf8c53c9:g:geMAAOSwiOBg1KQw&amdata=enc%3AAQAGAAACkPYe5NmHp%252B2JMhMi7yxGiTJkPrKr5t53CooMSQt2orsStdebXPz4ZTXCT8FI9kPBhwAMGfNRpXW7i4Y927WY%252FFyj7K%252FnN%252BDzOvPloAADPHyNHqb8nJFhG3dE%252BooimN8A0mEKbpAeWOINUzuCiQqGtaDr4GPWbm5ANiKX4wrU8heYrCRM%252FA2Jp9WJomm2bu3nb2zWnjgi%252F0RE2JvLWCVqtU%252BK6WsxIxTBeNBZnJua0FnP8CyOEvirwBwL8FBYOfFrA4qD%252FiAxEqdY%252FiozQfBkw5Cc%252Bry21e1RzKT6TsY1UwsKVB4UL9yU4L9IWyLdaLKQVfOawB8zeuhMWkkHq4Jt2ZBFLDoem47mfG0yl9Jsxo09lDjOiLwQLxiH40eH2xd1irM4STtxbLn8TlR6Drws7aTrbFnwWumdFHkTqNSTJeNeib8idv9XLS5j5kZgF%252FZHhG8oGeWfj1Wk74H5dKoiPLeI%252FlArYdh3HJMaf89r06fR%252BnufN8cz%252B7NRrkABpjedq6I4Fjs4iDf3l9NzkcHWPUHtFwwyWgmN%252B6gCrKK69kdjJ5wyPgn79U9AN0CwmiHJwwA18OYJCEMpYZbJj8uUeSFihRi9WahiKz21Ak9v1Q%252FBiLtFC%252FjZvZ9uKf0%252BPmLg%252BwROIO5XFr8u2ZbVxtB%252Bv4ESEtwzWmgWLiV6Y0ES8Q0FyIaaTCcb72N45rrkNsJZlXN6CSREy7JC2bxwkvVMbUICvQw8zprC0efRYg%252Bx0iEVZlcqhvwrkqmTHHXIDDLTOfqHqQaUdJzIFP0wf6gM5KREvxD3Zp%252B%252BK8TSXr%252BR2SMbB%252BbWCFg1hiQaKaJilkU%252FkTFaMQB2YX%252FW4Er6n1t3q8KVVGjbEQAXk5rkdRHMm7Qd%7Campid%3APL_CLK%7Cclp%3A2334524

 

 

 

 

https://www.ebay.com/itm/221866344023?_trkparms=ispr%3D1&hash=item33a843be57:g:nX8AAOSwNRdX4fKV&amdata=enc%3AAQAGAAACkPYe5NmHp%252B2JMhMi7yxGiTJkPrKr5t53CooMSQt2orsSvtkx670Z0mbyfWqmxLFLYedKTr05Vf3oa373pOOgxaQLeuDcQG4%252BRivndNLu3QavSmCJwH0DeIWLqAC6D%252F%252B%252FIauhQt9ATRkRiRpeWsQiGnx8891XZ5xY%252FuGAsrFki4GMeLYY94UpJ51k9dlY%252BhL9xVLkwQen6geIMs7OmUQipfBcfu91FfnxtrIBRZsjaZK6VqmxH9WqYkc6GgXeKJUPlCKhii%252BkCeAfhEZtN2s19r9fsPbQQXQpp47K%252B9uU%252B9XnNwuUnPx%252BKPHH97DKoPKUE3CsPF%252F2EKGmAYJz4RqakvWn6aT8LNKluLCyDPtald00ZM%252BkgsZ8DNv4Ka0D5J3yPpguwqwd7bTMe6PK3IHxpxSkDYRvtXw7x70MzhIuHuBFKKO3EzGWIJgxvTuRlyiF390MPecOA%252B8m3tykzY50pUl%252FQtCHFoqqfHfVHDbxrDjEgtz6YHLyMI7kduWHAaku3EoWnaCEXB2tYJc1%252F1sBBH65UpMtwP6fMeuEuvGwxrw71NFDWXARkZ%252BeKbFbQlfqDIxgjttk4cp8kY4gFjSg3YWDkT4b%252FLRR4csQabWFSRIlrpAp3pihpXHHBs8UHPTKq9uG2rv6na%252FJJJYYlR0MqFK3YQfa5HhmJi2yGT7Vl4g879VTaGKTo8Enk8pBlN9JHW2Bkvs8CDS68smUzDEGWDIdLijYFvSQv4LQCiUb3nqxjw7VNVgRkamFcV1fSicTFEgUlsQXOFz7x%252BSW180HHCYvfmi2%252FMmMdc%252B0RqMoaXcsf%252B%252F07uNHmjW4aO8MP%252FWb8%252FxuiWVgLJI16r7sfn5pMDQkqZmRXSuJmME9VEXxKMoF%7Campid%3APL_CLK%7Cclp%3A2334524

 

 

Or if you want to go hole hog:  (This is quite a nice attractive price vs the real HP device)

 

Noise Source Generator Calibrated (10 -1600 MHz; ~ 15 dB ENR, SMA) Noise Figure

 

Noise Source Generator Calibrated (10 -1600 MHz; ~ 15 dB ENR, SMA) Noise Figure 
US $269.00

 

https://www.ebay.com/itm/162785457428?hash=item25e6c4f514:g:oakAAOSwVORZ0uaH

 

 

 

Lester B Veenstra  K1YCM  MØYCM  W8YCM   6Y6Y

les...@veenstras.com

 

452 Stable Ln (HC84 RFD USPS Mail)

Keyser WV 26726

 

GPS: 39.336826 N  78.982287 W (Google)

GPS: 39.33682 N  78.9823741 W (GPSDO)

 

 

Telephones:

Home:                     +1-304-289-6057

US cell                    +1-304-790-9192

Jamaica cell:           +1-876-456-8898

 

 

From: sara...@googlegroups.com [mailto:sara...@googlegroups.com] On Behalf Of Marcus D Leech


Sent: Thursday, June 24, 2021 11:30 AM
To: sara...@googlegroups.com

Subject: Re: [SARA] Noise Source

 

The problem with that approach is that they’re much more sensitive to supply voltage variability and temperature changes than a simple diode noise source. 

 

But yes, this will work. 

Sent from my iPhone

 

.

image001.jpg

Lester Veenstra

unread,
Jun 24, 2021, 11:49:59 AM6/24/21
to sara...@googlegroups.com

--

--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.

image002.jpg
image003.jpg

Shef Robotham

unread,
Jun 24, 2021, 10:05:26 PM6/24/21
to 'fasleitung3' via Society of Amateur Radio Astronomers
I bought this high level noise source from Amazon,, was cheap , ~ $20 w lots of ENR… I did need to put a pad as apparently the output is affected by the load impedance,,, no surprise.


This hole thread seems to be a lot of noise over noise….. hmmmm !!!

Shef

On Jun 24, 2021, at 11:26, 'Lester Veenstra' via Society of Amateur Radio Astronomers <sara...@googlegroups.com> wrote:

For a low cost, uncalibrated, noise source, one of the simplest is a series , or even one, of amplifiers with the input terminated.  For L-Band (i.e. 1450 MHz), the ubiquitous TVRO line amplifiers are good, as is, or modified for external (non coax) power feed.
 
 
<image001.jpg>

ASKA 20 dB In-Line Amplifier Satellite DBS LNB 2.5 GHz 950-2400 MHz Signal Amp
 
 
 
 
in SMA
BIAS TEE
 
 
 
20 dB Amp
 
 
 
 
Lester B Veenstra  K1YCM  MØYCM  W8YCM   6Y6Y
 
452 Stable Ln (HC84 RFD USPS Mail)
Keyser WV 26726
 
GPS: 39.336826 N  78.982287 W (Google)
GPS: 39.33682 N  78.9823741 W (GPSDO)
 
 
Telephones:
Home:                     +1-304-289-6057
US cell                    +1-304-790-9192
Jamaica cell:           +1-876-456-8898
 
 
 
--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.

Lamar Owen

unread,
Jun 25, 2021, 9:35:47 AM6/25/21
to sara...@googlegroups.com
On 6/24/21 10:05 PM, Shef Robotham wrote:
> ...
> This hole thread seems to be a lot of noise over noise….. hmmmm !!!
...
But isn't the whole or radio astronomy all about noise?  At least it
started that way.....

Lester Veenstra

unread,
Jun 25, 2021, 10:41:51 AM6/25/21
to sara...@googlegroups.com

Well when I get my HP 8970B / 346C working I might be able to calibrate some of these padded sources.  Threre are others out there that should be able to do the same.

 

Lester B Veenstra  K1YCM  MØYCM  W8YCM   6Y6Y

les...@veenstras.com

 

452 Stable Ln (HC84 RFD USPS Mail)

Keyser WV 26726

 

GPS: 39.336826 N  78.982287 W (Google)

GPS: 39.33682 N  78.9823741 W (GPSDO)

 

 

Telephones:

Home:                     +1-304-289-6057

US cell                    +1-304-790-9192

Jamaica cell:           +1-876-456-8898

 

 

From: sara...@googlegroups.com [mailto:sara...@googlegroups.com] On Behalf Of Shef Robotham
Sent: Thursday, June 24, 2021 10:05 PM
To: 'fasleitung3' via Society of Amateur Radio Astronomers
Subject: Re: [SARA] Noise Source

 

I bought this high level noise source from Amazon,, was cheap , ~ $20 w lots of ENR… I did need to put a pad as apparently the output is affected by the load impedance,,, no surprise.

Reply all
Reply to author
Forward
0 new messages