RAS LNA question

[jan Lustrup-LA3EQ]

Hi group,

Anyone know the transistor type used for T1 and T2 in the R.A.S.  LNA?

I am having some 0.5dB signal drops from time to time. I don’t see any oscillation going on with my spectrum analyzer at least up to 1.7GHz..

I made sure all the grounding connections are good.

I changes the N-connectors, And the variable resistor for T1 bias. Those thing did not help.

Measuring voltages around T2 showed that its current changes during these drops.

T1 current looks good and stable but I wildly unstable with the lid off.

So I suspect T2 and want to change that one. But what type is it?

 

Here is a drawing I tried to make for this LNA.

Jan

 

 

 

[Alex P]

Hello Jan,

It was probably  built by Tommy Henderson WD5AGO

email       <wd5...@hotmail.com>

Regards,

Alex P  KK4VB

[duncan campbell-wilson]

Hi

Does this LNA have surface mounted parts? The symptoms to me indicate an intermittent connection.

I have seen capacitors fail on FR 4 boards at the electrodes and become problematic with temperature and humidity.

PCB flexing due to  connector induced board twist has also been a source of failed capacitors.

I have attached a plot of  the Vela pulsar dual polarisation SNR on a 6m. The middle part of the  top trace shows a period where a  bad  joint was failing whereas the other polarisation (lower trace ) was well behaved.

The Vela pulsar is strongly  polarised.

Spray  freezing individual parts in the circuit may help to identify the culprit.

I hope this helps.

Rgds Duncan

--
--
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/54399971-98b1-4361-a4fb-7bd31ea402ddn%40googlegroups.com.

[Marcus D. Leech]

Signal drops are often NOT a loose connection, but rather "de-sense" caused by strong out of band signals presenting
  themselves to the front-end.    The transistor is driven into non-linearity, and is spending most of its capacity
  amplifying the out-of-band signal.

I've seen this on every radio telescope I've ever built.   There are a couple of solutions:

    o Use an LNA with a higher linearity (OIP3 and p1dB)
    o Put a filter *in front* of the LNA  (gack!!!)

Even in an LNA with a narrow-band matching network on the input, out-of-band signals are still going to "get in", albeit
  at somewhat (3dB?) lower levels.  If they're loud enough, that can cause compression.

The conventional "tension" in the low-noise-transistor biz has been between linearity and sensitivity--the lower the
  noise figure, in general, the poorer the linearity.  This is starting to change--driven by commercial requirements that
  need *both*.  It wouldn't surprise me if WD5AGO is using transistors that are fairly "olde skool", and while they have
  stunning noise figure at L-band, they might not have stunning linearity properties.

[jan Lustrup-LA3EQ]

Hi,

Nice plot….

That was my first thought too, so I first  used freeze spray and heat but that did not bring up the fault. T2 has cap + resistor on each of the two source leads, so I thought maybe if one failed intermittent the change would not be so large and maybe bring about the small 0.5dB problem.

So far nada….so next is the swap the transistors and see.

Jan

To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/CAFfhL0mj1Te%3D50k%2BWHo%2B5nyfApP7yikAh0fhED5W0Js-xr5FjQ%40mail.gmail.com.

[Marcus D. Leech]

On 2022-08-25 10:19, jan Lustrup-LA3EQ wrote:

Hi,

Nice plot….

That was my first thought too, so I first  used freeze spray and heat but that did not bring up the fault. T2 has cap + resistor on each of the two source leads, so I thought maybe if one failed intermittent the change would not be so large and maybe bring about the small 0.5dB problem.

So far nada….so next is the swap the transistors and see.

Jan

 

Do you have total-power plots of what your drop-outs look like?  What time-scales do they happen over?

To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/002b01d8b88d%24ba436c70%242eca4550%24%40no.

[jan Lustrup-LA3EQ]

Hi Marcus,

Here is a total power solar transit done with the RAS LNA.

You can clearly see the drop outs as dips of around 0.5dB. Also second image is a close supplied.

Jan

 

 

 

 

To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/77156b5a-ff0f-dbfb-bf6a-93ce36e0e3ee%40gmail.com.

[b alex pettit jr]

Jan,

Send this plot to Tommy Henderson WD5AGO and ask His recommendations ..

 email       <wd5...@hotmail.com>

 

Regards,

Alex P  KK4VB

Alex

[duncan campbell-wilson]

Concerned about the shape of the noise in the plots. Please what was the weather conditions like during the observation period?  

--
--
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/1475934502.76840.1661506845383%40mail.yahoo.com.

[duncan campbell-wilson]

Looks like connection changing to hi reactance.  Cable stress on a connector. 

[Alex P]

Perhaps  an Intermittent caused by weak/corroded center pin contact between cable connectors?

( I had exactly this issue last night with an F-Type connector on the TV input )

Alex P

[b alex pettit jr]

Disconnect/Reinstall/Properly_Tighten  ALL Coaxial Connections ... and retest.

Alex P

--

[duncan campbell-wilson]

Concur. I see odd jumps in my data when my dish is getting rattled by high winds.  Rgds Duncan

--
--
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/280255419.76043.1661508671119%40mail.yahoo.com.

[fasleitung3]

Hi Jan,

Of course it is always difficult to do a "remote diagnosis". Looking at your plots I would say that RFI plus some non-linearity in the chain may be a cause. In particular the close-up seems to indicate a mobile network as a possible offender.

Wolfgang

[jan Lustrup-LA3EQ]

I tried that. I put in new N-connectors. Changes F-connectors…no help….. At the end I moved the LNA inside next to the bias “T” and connected a good 50 Ohm dummy load at the input thus eliminating all outside cable & connectors…. Still the same! I changed Bias “T”. no help I changed power supply….no help….….. I measured the current in T1…stable over time…..I then measured the current in T2…the problems dip shows up as current changes at the same time as the dips!!  So T2 of it’s components are suspect. I do not know the type of transistor for T2 so I tried a spz8950z I had extra, no signal out. Replaced T2 with a mmic MAR 8, nice output. Both source resistors and caps still in place as before connected to ground pins of MAR 8.  Drane supply coil /resistor  as before. So far so good. Now running a new Sun transit with both antennas to compare…

We see how it goes.

Jan

 

From: 'b alex pettit jr' via Society of Amateur Radio Astronomers [mailto:sara...@googlegroups.com]
Sent: fredag 26. august 2022 12:11
To: 'Alex P' via Society of Amateur Radio Astronomers
Subject: Re: [SARA] Re: RAS LNA question

 

Disconnect/Reinstall/Properly_Tighten  ALL Coaxial Connections ... and retest.

Alex P

 

--

--
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/280255419.76043.1661508671119%40mail.yahoo.com.

[b alex pettit jr]

Hello Jan,

For Future Reference :  F-Connectors :

The problem on my  TV was

The male F-Type wire was Perfectly Centered and not reliably engaging the mating socket

Solution : I Bent the Wire = Problem Solved ..

Alex

[jan Lustrup-LA3EQ]

Hi all, Thanks everyone for all the good tips….

 

After changing T2 with a MAR8 it seems like the 0.5dB drops are gone.  Fingers crossed.

 

Today I did a dual meridian sun transit (blue curve) recording with the RAS  LNA outdoors on the feed probe, 27m with 70 Ohm TV cable /w F-connectors and N- adapters.

The same sun  transit was recorded with a different antenna and different LNA (red curve). Both transits are recorded and saved on “Radioskypipe” channels ch1 and ch2-

Seems what’s left looks more or less like normal noise variations. (weather-wise we are having thunder storms and heavy extreme rain showers today)

 

 

Jan Lustrup, LA3EQ

Norway

 

 

 

From: 'b alex pettit jr' via Society of Amateur Radio Astronomers [mailto:sara...@googlegroups.com]
Sent: fredag 26. august 2022 12:59
To: sara...@googlegroups.com
Subject: Re: [SARA] Re: RAS LNA question

 

Hello Jan,

--

--
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/2094195858.81428.1661511524150%40mail.yahoo.com.

[duncan campbell-wilson]

Hi,

Nice result.

Please what is the gain of the first stage in the circuit .

Parasitic oscillations:

As you mentioned instability when the lid was off the box this reminded me that I had seen this behaviour in low noise amplifier units (400).  The 12 K room temperature low noise amplifier would exhibit jumps in current consumption. Some devices were unstable while other parts  from  the same reel were behaved well.  After much stuffing around we found a parasitic oscillation well above the design frequency (840 MHz) somewhere between 3-5 GHz. Our spectrum analyser range stopped at 1.6 GHz.

The fix was to place  absorbers near the device input  and above the output stages (on the underside of the RF enclosure lid)

Several design approaches or combinations may be taken in the design and post production stage to combat parasitic oscillations.

(1) Make the shielding enclosure less than  quarter at the design frequency

(2) Isolate the first and second stages by a shielding  wall.

(3) Place an absorber in the box. I have seen cardboard used but my preference is a ferrite load rubber material fixed to the wall near the input. Not too close to the input connection.

(4) On the casing lid more ferrite material is fixed by non acetic silicone above both  first and second stages above the matching output stage.

The aim was to  have an LNA that was unconditionally stable at all phase angles as the scanning array was ill behaved in terms of impedance at some scan angles .

I have added some images below. Questions welcome.

Rgds Duncan

 

[Marcus D. Leech]

On 2022-08-26 04:09, jan Lustrup-LA3EQ wrote:

Hi Marcus,

Here is a total power solar transit done with the RAS LNA.

You can clearly see the drop outs as dips of around 0.5dB. Also second image is a close supplied.

Jan

There are a lot of "brief" dips in the response.  I've seen this before from de-sense of the LNA--non-linearity created by
  a fairly strong out-of-band signal.  If they're brief, they're more of an annoyance, and low-pass or other filtering in
  post-processing can get rid of them.

The harder-to-deal with are the longer-duration drops and increases in the baseline.  Obviously, if this is a connection
  issue, or a component issue, it's easier to deal with in many ways, than coming up with a post-processing
  heuristic that provides adequate results all the time.

You're located in rural Norway as I recall, but perhaps not far enough away from a larger town to avoid things like
  cell transmitters and the like.    This will increasingly be a struggle not just for amateur radio astronomers, but
  also for professionals.

The "usual architecture" for a radio telescope has a low-noise amplifier "up front", and usually completely unprotected
  by any filters in the first stage.   You're relying on the increasingly-excellent non-linearity properties of your 1st-stage
  LNA to tolerate strong out-of-band signals prior to inserting a narrower-band filter in behind the first LNA stage.  Even
  with a larger dish, signals can (and will!!) get into the side-lobes, and the usual doctrine for evaluating interferers
  in the side-lobes is to ascribe a gain of 0dBi to them when evaluating path loss to potential interferers.  Remember
  that such interferers don't have to be anywhere close to in-band to be a problem--modern LNAs have useful gain over
  a very wide frequency range.

If someone were to come up with a relatively-cheap band-specific filter that had low-enough loss to be considered for
  use IN FRONT of the first LNA, they'd have rather a pile of customers...

The second, and dramatically-less-pleasant, problem is that of side-bands from adjacent services.   For 21cm, in some
  parts of the world, there is navigation radar right below the radio astronomy band, and DAB right above it.  The
  regulatory spectral templates usually only specific fairly "sloppy" out-of-band requirements for these services.
  Which means that they can easily "spill over", at low levels, into the adjacent radio astronomy band.  No amount
  of filtering in your equipment can fix that problem.

THEN, there's the problem of unintended emissions from the superposition of thousands of digital electronics devices
  within a "meaningful" radius of your antenna.   Even when they are in regulatory compliance, they can still emit
  at levels that are harmful to radio astronomy....

Our radio telescope site at Carp, Ontario, is located in an area that I would describe as "quasi-rural".    We're dealing
  with low-level instability that is causing us to miss our theoretical sensitivity limit by a large factor.  We have
  a narrow-band (patch) feed, an LNA with inter-stage filtering, a "last chance" filter in front of the receiver, and the
  receiver itself has band-select filtering, IF filtering, and the usual layers of DSP filtering.     I did a 24-hour
  100MHz-wide spectral assessment in peak-hold mode the other day.   No unexpected narrow-band "surprises",
  and the overall 24-hour noise-floor variability was what you'd expect (a dB or so).  Our perturbations are subtle--
  about 0.5-2K in amplitude.   Not a problem for sources like 3C400, which produce a roughly 12K rise in our output.
  But for sources much smaller than that, it's a struggle.  No apparent hardware issues--the issues are seen at the
  same amplitude in both polarizations at 21cm, and the other two feeds (611MHz and 4150MHz) don't have the
  same problems--all are using the same master power supply.     Radio Astronomy is hard :) :)

[jan Lustrup-LA3EQ]

Thanks for  replies Marcus and Ducan and all.

I just noticed the antenna wire going from the N-connector to Cap at T1 was broken lose and ripped some metal off the cap and just barely touching. So I guess this bad connection might be the cause of the dips….So I  will order a new one and see how it goes.

Jan LA3EQ

 

From: sara...@googlegroups.com [mailto:sara...@googlegroups.com] On Behalf Of Marcus D. Leech
Sent: søndag 28. august 2022 16:46
To: sara...@googlegroups.com
Subject: Re: [SARA] Re: RAS LNA question

 

On 2022-08-26 04:09, jan Lustrup-LA3EQ wrote:

--
--
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/20a62966-2418-49b5-a11d-3c52d01d722b%40gmail.com.