UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Amateur and Military Radio

Notices

Vintage Amateur and Military Radio Amateur/military receivers and transmitters, morse, and any other related vintage comms equipment.

Closed Thread
 
Thread Tools
Old 1st Feb 2012, 3:18 pm   #21
Chris Parry
Retired Dormant Member
 
Join Date: Jun 2011
Location: Waterlooville, Hampshire, UK.
Posts: 237
Default Re: Receiver aerial input impedance: questions.

Yep, lossless feedback works. I've used it many times at VHF. Never seen it done on a valved HF radio though. Over the years, I have worked on valved radios with push-pull front ends and even one that had a complicated quadrature front end. None of them worked any better than the ordinary type as regards noise figure. (Which is the key indicator of SNR recovery in a given bandwidth for an ordinary AM/SSB radio.) The performance benchmarks here are as follows. Prewar technology is won by the DST100 with 8dB NF on ten metres. The postwar category is won by the Eddystone 940 which manages an astonishing 5.5dB, despite being a scummy design with tiny RF coilformers and a paxolin bandchange switch. Worth recording here, is that of the many "souped-up" HROs etc that have come my way with EF183s and the like fitted in the front end, not one of them has worked appreciably better than a healthy example of the standard article as regards noise figure. Where significant improvments >can< be made is in the area of strong signal handling. Here, the Americans led the field decisively, as shown by designs from Collins & Squires-Sanders in particular. But that discussion falls outside the scope of this thread. Chris.
Chris Parry is offline  
Old 1st Feb 2012, 3:27 pm   #22
MichaelR
Retired Dormant Member
 
MichaelR's Avatar
 
Join Date: Mar 2006
Location: Sheffield, South Yorkshire, UK.
Posts: 1,587
Default Re: Receiver aerial input impedance: questions.

Hi Dave,

Just to clarify this statement that no filter can reduce noise. I am assuming that you are referring to thermal noise. Often the term noise is used in a less specific way for instance noise generated by plasma screens etc. This type of noise I think you are referring to as interference because obviously filtering can remove this.

Mike
MichaelR is offline  
Old 1st Feb 2012, 3:53 pm   #23
jimmc101
Heptode
 
jimmc101's Avatar
 
Join Date: Mar 2009
Location: Portsmouth, Hampshire, UK.
Posts: 674
Default Re: Receiver aerial input impedance: questions.

Apologies for the duff link in my earlier post, this one should work.
http://www.ece.vt.edu/swe/lwa/memo/lwa0071.pdf

Jim
jimmc101 is offline  
Old 1st Feb 2012, 6:03 pm   #24
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Receiver aerial input impedance: questions.

Thanks. I've seen that type of amp before, but I didn't realise it could go so low in noise figure. One thing puzzles me: in thermodynamic terms, how can you have something which looks like a resistor (the input port of the circuit) yet does not generate thermal noise? If there is dissipation there ought to be fluctuation too. I suspect that part of the story is that the input power mostly ends up at the output, and augmented by the transistor gain, so it acts a bit like an anti-attenuator.
G8HQP Dave is offline  
Old 2nd Feb 2012, 4:16 pm   #25
emeritus
Dekatron
 
Join Date: Jan 2012
Location: Brentwood, Essex, UK.
Posts: 5,316
Default Re: Receiver aerial input impedance: questions.

My recollection is that, to a good approximation, as the centre conductor of a 75 ohm transmission line has a smaller diameter than a 50 ohm line, for a given outer diameter of coaxial line, 75 ohms gives the optimum voltage rating and 50 Ohms the optimum power rating. I used to be able to do the sums as a student, but that was long ago now!

Re reflections, the engineer I trained under had worked at Belling-Lee in the 1950s shortly after Band III transmissions had started. To illustrate the concept that there is usually no "best" thing in engineering, but only the best thing for a particular situation, he recounted how the BL engineers had developed a low-loss cable for the new services and were singing its praises to one of the directors. "So how does it compare with bell wire?" he asked, and they had to rig up a bell wire feed and do a side by side comparison. To their embarrassment, the bell wire gave a far sharper picture! They worked out that the reason was that, with the new cable, the signal suffered multiple reflections at both the aerial and the set, and because it was so low loss, several of the reflections were strong enough to be detected, causing horzontal smearing. Conversely the bell wire was vey lossy and not a good impedance match, but still delivered enough direct signal to produce a good picture. By the time the first reflection had travelled back to the aerial and been reflected back to the set, it was too weak to affect the image. Of course the bell wire was only superior because they were in a strong reception area, and at the fringes the new cable would have been far superior.
emeritus is offline  
Old 2nd Feb 2012, 4:43 pm   #26
ppppenguin
Retired Dormant Member
 
ppppenguin's Avatar
 
Join Date: Dec 2003
Location: North London, UK.
Posts: 6,168
Default Re: Receiver aerial input impedance: questions.

Quote:
Originally Posted by emeritus View Post
....They worked out that the reason was that, with the new cable, the signal suffered multiple reflections at both the aerial and the set, and because it was so low loss, several of the reflections were strong enough to be detected, causing horzontal smearing......
I've come across that problem in a different context. Serial digital video is 270Mb/s, 1.5Gb/s or 3Gb/s, normally on 75R co-ax. The return loss is specified as minimum 15dB up to the bit rate. In practice the links work with lousy return loss but a SHORT cable is more likely to fail than a long one because of repeated reflections. Cable losses deal with the reflections quite nicely for longer runs.
ppppenguin is offline  
Old 2nd Feb 2012, 6:26 pm   #27
kalee20
Dekatron
 
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,061
Default Re: Receiver aerial input impedance: questions.

That's where I was coming from in my earlier post.

If the cable is properly matched at one end, there can't be multiple reflections bouncing back and forth.

Since it's not going to be possible to ensure an arbitrary aerial will match the 75Ω cable, over the whole tuning band, the receiver itself should provide a 75Ω termination at the aerial socket. (And then, as the direction of energy flow is from the aerial to the receiver, there won't even be one reflection).
kalee20 is offline  
Old 2nd Feb 2012, 10:38 pm   #28
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Receiver aerial input impedance: questions.

You don't need a particularly good match to quickly damp down reflections. Whether reflections matter depends on the modulation rate, which is why TV is fussy and audio is not.

Assume that half the power is reflected at each end. Then just 5 back and forth journeys will drop the signal by a factor of 1000 (-30dB).
G8HQP Dave is offline  
Old 7th Feb 2012, 1:59 am   #29
Skywave
Rest in Peace
 
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
Arrow Re: Receiver aerial input impedance: questions.

There are a few errors in my thinking in post 10: I shan't dwell on them: no-one else noticed anyway. (If they did, they didn't report them! ) What I shall expand on here is a quick 'conceptual' measurement that I have recently made.

I used an HP 8640B sig. gen., my Eddystone 888A receiver and an oscilloscope. The sig. gen. was set to 3.7 MHz, 30% A.M. @ 1kHz, connected direct to the receiver's aerial input terminals and tuned in on the receiver (A.M detector in), AGC switched off, 'scope monitoring the receiver's A.F. output (which was terminated in a 3 ohm resistor). The sig. gen. level was set at 3 mV; the receiver's gain controls and the 'scope's vert. atten. settings were adjusted for 4v. p/p on the 'scope. [That's about 670 mW output]. I then found that when I inserted a 120 ohm resistor between the sig. gen. and the receiver, the level on the 'scope dropped to 2v. p/p: i.e. half of what it was prior to inserting that resistor.
I repeated that procedure for sig. gen. levels of 300 uV and 30 uV, adjusting the receiver's gain controls and 'scope atten. controls as necessary for a suitable level at the 'scope. In both cases, the insertion of the 120 ohm resistor reduced the output by half - just as before.

An application of some basic theory indicates that the input resistance of the receiver is that value of inserted resistor minus the output resistance of the sig. gen., i.e. 50 ohms. So, 120 minus 50 = 70 ohms: the receiver's aerial input resistance @ 3.7 MHz and at the input levels chosen.

Yes, there are a few assumptions and a few compromises in that set-up, but I cannot think of any that totally invalidate the procedure nor seriously jeopardise its returned value. As I said, it was the practical application of a conceptual notion to see if it returned a result that looked 'sensible': I think it does. What I need to do now is to expand it to include reactance at the input and 'clean up' the various R.F. interconnections, i.e. refine the procedure. And then repeat at other frequencies and possibly other settings of the R.F. gain control and also to see if the setting of the AGC switch has any bearing on the measurement.

I can see I could be busy for some time! Nevertheless, fascinating stuff!

Al.
Skywave is offline  
Old 7th Feb 2012, 12:55 pm   #30
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Receiver aerial input impedance: questions.

Yes, I think you have shown that the input impedance is 70 ohms. Maybe the designers had a particular reason to ensure this. Over most of the HF band low noise is not necessary, so the tradeoff is between front-end selectivity and good signal handling. The first needs a big step-up (low input impedance) so the antenna doesn't load the first tuned circuit. The second needs a small step-up so the signal voltage at the grid is small - this gives a high input impedance. Or you can just go for a power match, and accept whatever selectivity and signal handling this happens to provide; this is what they appear to have done. An advantage of this is that you can add attenuators and they do what it says on the tin.
G8HQP Dave is offline  
Old 7th Feb 2012, 2:30 pm   #31
Skywave
Rest in Peace
 
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
Arrow Re: Receiver aerial input impedance: questions.

Quote, G8HQP: "Maybe the designers had a particular reason to ensure this". (Input impedance = 70 ohms.)

As is the case with many Eddystone communications receivers (and other makes) there are three aerial input terminals, (one being earth / chassis), thus catering for unbalanced or balanced transmission lines from an aerial to the set. With such an arrangement, is it common to find that either input is designed for 70 ~ 75 ohms, (although, IIRC, 400 ohms can be found in some), thus attempting to provide a good impedance match between aerial feeder and receiver. Of course, 70 ~ 75 ohms is chosen since this will provide a good match between the remote end of that feeder and a simple half-wave dipole. It follows, therefore, that when an aerial other than a half-wave dipole is in use, a device to perform an impedance transformation is necessary between the incoming feeder and the aerial input terminals. This latter matching aspect has been discussed recently at this forum, so we need not discuss that further here and thus cause this thread to go off-topic.

As I stated in my last post, I intend to take my investigations and measurements on this further. At present, it is primarily the cold weather that is hampering progress in that direction! When conditions improve, I shall return to this thread.

Al.
Skywave is offline  
Old 7th Feb 2012, 3:26 pm   #32
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Receiver aerial input impedance: questions.

Assuming the receiver does not start with a permanently engaged atttenuator, the input impedance depends on the input tuned circuit: Q, dynamic impedance, coupling, turns ratio. Some these will vary with frequency, unless they have deliberately designed this out (e.g. by resistive slugging of the Q). It will be interesting to see how it changes across a range. Other things being equal (but they are probably not), dynamic impedance of a capacitor-tuned circuit should rise with frequency (roughly as square root) unless skin effect counteracts this, so the input impedance should show the same rise across an individual range.
G8HQP Dave is offline  
Old 7th Feb 2012, 6:15 pm   #33
Skywave
Rest in Peace
 
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
Question Re: Receiver aerial input impedance: questions.

Quote:
Originally Posted by G8HQP Dave View Post
Other things being equal (but they are probably not), dynamic impedance of a capacitor-tuned circuit should rise with frequency (roughly as square root) unless skin effect counteracts this . . .
You sure about that? Dynamic resistance, Zd = L / (C.R), no?
So for fixed L and R, Zd is inversely proportional to C.

Apart from that, everything else you've written I agree with.

Al.
Skywave is offline  
Old 7th Feb 2012, 7:02 pm   #34
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Receiver aerial input impedance: questions.

Yes, provided that the only resistance is in series with the coil and fixed in value. If the coil has a roughly constant Q then the effective series resistance rises with frequency, either through skin effect or core/former losses. Constant Q gives a square-root rise in dynamic impedance, as the inductive reactance rises as a square-root. Permeability tuning would be different, but that seems to have been rarely used. I guess the answer is: it all depends!
G8HQP Dave is offline  
Old 7th Feb 2012, 10:50 pm   #35
BobGreen
Retired Dormant Member
 
Join Date: Jan 2011
Location: Egham, Surrey, UK.
Posts: 219
Default Re: Receiver aerial input impedance: questions.

I majored in line and radio comms a very long time ago. I am very impressed by the knowledge exposed by you guys here.

Bob
BobGreen is offline  
Old 8th Feb 2012, 6:50 pm   #36
kalee20
Dekatron
 
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,061
Default Re: Receiver aerial input impedance: questions.

Quote:
Originally Posted by G8HQP Dave View Post
One thing puzzles me: in thermodynamic terms, how can you have something which looks like a resistor (the input port of the circuit) yet does not generate thermal noise? If there is dissipation there ought to be fluctuation too.
My thoughts are as follows:

The amplifier's 75Ω input resistance doesn't have to involve dissipation. Even if it did, it doesn't have to generate noise. (A thought experiment could consist of a valve amplifier - infinite Zin - shunted by a 75Ω resistor which is cooled to cryogenic temperatures. Perfect match with no thermal noise. Only the resistor is cooled - the amplifier isn't).

As for how a 'resistive' condition can arise with no thermal noise, if the resistive condition exists with no thermal dissipation, then there may well be no thermal noise. Clearly, energy must be absorbed, and if it's converted to heat then thermal noise comes into play, but if it's converted to another form of energy then I'm inclined to think not. (How about a motor raising a weight by a winch?)

The circuit in question is a common-base amplifier (very low input impedance), driving its load. But a transformer feeds some of the output voltage back, in series with the input. Thus, the source now sees a significant voltage at the input terminals. Correct choice of turns ratio allows the effective input impedance to be whatever is desired. At the same time, as transformers don't introduce noise, the noise of the 'input resistance' is only whatever noise is generated by the amplifier's load, coupled back by the transformer.

As I don't regard myself as a noise guru - please comment!
kalee20 is offline  
Old 8th Feb 2012, 7:29 pm   #37
Skywave
Rest in Peace
 
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
Question Re: Receiver aerial input impedance: questions.

Quote:
Originally Posted by G8HQP Dave View Post
Constant Q gives a square-root rise in dynamic impedance, as the inductive reactance rises as a square-root.
I'm sorry but I don't understand that. Can you expand on that with a bit of basic algebra, please, so that I can fully understand the reasoning? And when you say "rises as the square-root" . . . . square-root of what?

Thank you.

Al.
Skywave is offline  
Old 8th Feb 2012, 7:51 pm   #38
kalee20
Dekatron
 
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,061
Default Re: Receiver aerial input impedance: questions.

I had a quick think about this, Al.

Zd = L / (C.R) at resonance, approximately provided that Q is large enough.

Case 1: Suppose effective coil resistance is constant, a reasonable assumption at lowish frequencies, before skin effect and coil-former dielectric losses kick in.

Since f = 1/[2.π√(L.C)], then for fixed L, C = 1/(4.π².f².L)

Substitute in the formula for Zd:

Zd = 4.π².f².L²/R, ie dynamic impedance is proportional to the square of the frequency, with constant inductance and variable-capacitor tuning.

Case 2: Suppose coil exhibits constant Q as frequency varies. This tends to be the case at higher frequencies, when frequency-dependent skin-effect losses and insulation losses dominate.

Since Q = X/R = 2.π.fL/R, then R = 2.π.f.L/Q. Substitute in the formula for Zd just found:

Zd = 2.π.f.L.Q, ie dynamic impedance is proportional to frequency.

Thus, over a tuning range, it's likely that as you tune up in frequency, that at first Zd will rise very quickly (case 1), and then the more slowly (case 2)
kalee20 is offline  
Old 8th Feb 2012, 8:08 pm   #39
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Receiver aerial input impedance: questions.

kalee20 (post 36): A resistance will dissipate if there is any signal. The possibility of dissipation means that it will generate thermal noise. Of course, if the resistor is cold then it won't produce noise.

I think your last paragraph may be getting close to the answer. You can't 'manufacture' resistance from reactance, but you can transform resistance. The very low emitter input resistance is boosted by the feedback, because the emitter only sees a fraction of the input voltage. I'm still puzzled, though.

Skywave: I think I'm getting too old for mental algebra. You are right. I was getting confused by the frequency varying like 1/sqrt(LC). Constant coil Q means that dynamic impedance increases linearly with frequency. Constant series resistance means it changes with square of frequency (I hope! - more mental algebra).

PS Just seen post 38.

Last edited by G8HQP Dave; 8th Feb 2012 at 8:10 pm. Reason: clarify
G8HQP Dave is offline  
Old 8th Feb 2012, 9:03 pm   #40
Skywave
Rest in Peace
 
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
Thumbs up Re: Receiver aerial input impedance: questions.

Thank you Kalee20 - that's the type of explanation I wanted. All is quite clear now.

Al.
Skywave is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 1:46 am.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.