|
Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members. |
|
Thread Tools |
28th Dec 2016, 7:23 pm | #21 |
Dekatron
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
|
Re: Simple RF probe for DMM
I think the 2N3819 can be used as an amplifier/oscillator up into the UHF region but I think it isn't really suited to that VM circuit. An initial crude analysis would guesstimate the Gm of the 2N3819 at about -1V Vgs to be about 2.5mmho and therefore the small signal output Z of the buffer would be 1/0.0025 = 400 ohm. If you then consider a pair of (1N5711?) Schottky diodes in a doubler detector with maybe 2uA forward bias then you could easily be looking at a shunt capacitance of 4pF just in the detector.
So even before you look at other strays the RC rolloff of 400R and 4pF here will be significant up at 150MHz. So I think that a typical 2N3819 isn't going to perform well here if you are expecting flat performance to 150MHz for small signals of a few mV. Ideally you also need to factor in the 2000R resistor and the effective Rp of the detector vs frequency and other strays but (based on just this simple analysis) I think you could looking at maybe 3dB rolloff by 150MHz if you built this circuit and tested it? You might get better/worse across a spread of 2N3819 devices but it's not a good place to start in my opinion.
__________________
Regards, Jeremy G0HZU Last edited by G0HZU_JMR; 28th Dec 2016 at 7:35 pm. |
28th Dec 2016, 9:05 pm | #22 |
Dekatron
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 13,951
|
Re: Simple RF probe for DMM
In times-past I've built various RF probes. From experience, if you're using germanium diodes then it's a good idea to use a really-low-value capacitor [50pf or so] between the probe and the diode(s).
Why? Well, if you're dealing with any valve gear and probe anything that's got a few hundred volts of standing DC on it, a large-value coupling capacitor can cause enough of a current-spike to kill (or at least traumatize) the diode. A small-value coupling capacitor will at least act as a potential-divider with the diode's intrinsic capacitance and so smooth-out these sorts of spikes. My favourite probe - not made by me, I must admit: the Royal Navy should take the credit - used an EA50 diode. I'd happily poke this into driver/multiplier 'cages' where there was 500V DC on the valve anodes and a hundred volts or so of RF whizzing around. |
28th Dec 2016, 11:40 pm | #24 |
Dekatron
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
|
Re: Simple RF probe for DMM
I don't know if this helps but see the image below for the input Rparallel and Cparallel of a doubler detector that uses 2 1N5711 schottky diodes and has a forward bias of 2uA. This was built and measured on my VNA after a careful calibration that puts the reference plane directly at the doubler input.
Theory states that each diode will have a small signal resistance that is approx. n*Vt/Ib where n = 1.08 for Schottky diodes and Vt = 0.026V and Ib = 2e-6A (where Ib = 2e-6A =2uA). So each diode will have an Rp of 14.04k ohm at low frequencies because of the 2uA bias current. Because the pair of diodes look to be in parallel from the point of view of the driving signal the Rp for the detector will be halved to about 7k ohm at LF. I did a quick measurement on the VNA using a very low drive level from the VNA of -35dBm. This keeps the signal in the square law region and the input result for Rp and Cp is as the image below. Note that this typically only applies to tiny signals in the mV range but this is a mV meter after all. Bigger signals will see a different load but you can see that for small signals Rp seems to agree with theory at lowish frequencies as it shows 7k ohm and then it gradually reduces towards 200MHz. The parallel capacitance seems to be fairly consistent at about 4pF. The traces look noisy because I had to do the calibration and measurement down at a VNA power level of -35dBm. A slightly more complicated model of each diode would involve adding a series resistor to each diode model. This resistor (sometimes referred to as a spreading resistance within the diode) would be just a few ohms and the diode model would become this series resistor followed by a resistor of about 14k ohm (for that 2uA bias) in parallel with just over 2pF. If you then arrange two of them as per the detector then you get a similar graceful rolloff in overall Rp vs frequency that you see in the plot of the real detector Rp and Cp below.
__________________
Regards, Jeremy G0HZU Last edited by G0HZU_JMR; 28th Dec 2016 at 11:53 pm. |
29th Dec 2016, 12:21 am | #25 |
Dekatron
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,923
|
Re: Simple RF probe for DMM
Thanks for that Jeremy. Do you know how Ge point contact diodes would compare with the Schottkys?
I recently found that I have a number of gold-bonded Ge point contact diodes, which have been sitting in the drawer for these past...~37 years . I believe that this type of diode was used on the Marconi probe that we discussed a while ago, and that probe and some others had "heating elements" inside to ensure that the diodes were kept at a stable temperature, presumably slightly above body temperature? I'm just setting up a test to see how the signal response of those diodes change between room temperature and ~40'C. B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. |
29th Dec 2016, 4:26 pm | #26 | |
Dekatron
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
|
Re: Simple RF probe for DMM
Sadly, I don't have much experience with Ge diodes apart from playing with crystal sets and the AM detectors in AM CB/ham radio gear.
However, I've got some here that are quite old. I've got some classic (genuine) 1N34A diodes plus some OA47 diodes here. Plus a few 1N60 diodes salvaged from old CBs etc. I'd expect the 1N34A to work well in the basic diode probe shown at the start of this thread. It should have low capacitance and fairly good performance up into VHF. Not sure about the OA47. I'd expect this to be markedly worse than the 1N34A. Note that there's a minor typo in my post #21. I typed in the wrong voltage for Vgs by mistake. It should have read: Quote:
__________________
Regards, Jeremy G0HZU |
|
29th Dec 2016, 4:37 pm | #27 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: Simple RF probe for DMM
Much to my surprise, yesterday & this morning I managed to get a 'clear run' at the G0WCA circuit. The results were not encouraging. For a start, the linearity was poor (more than I would have expected) and not just on the most sensitive range, despite carefully juggling the variable resistors as provided in the design and following the set-up procedure - which in my opinion is poorly written. My earlier remark about the vaguely stated set-up for the bias diode is also relevant in that regard. Plus, I'm not exactly impressed with the arrangement of the configuration of the CA3140, either. Pin 4 should really be returned to a -ve supply rail. Amongst other things, that would enable the set zero control to be adjusted to either side of meter zero - which is done in nearly all circuits of this type. And I remain a bit dubious about the circuitry for the meter: for a start, the cct. diag. does not state the resistance of the 200 uA meter, nor the resistance in series with it. Yes, I am aware that the author states that different meters can be used, but even half-decent documentation would have had those missing parameters stated on the drawing. Finally, the many posts by Jeremy (G0HZU_JMR) have highlighted many short-comings in this circuit, some of which I would not have recognised at a casual glance - and even now, I comprehend only parts of most of those described deficiencies
So in conclusion, I do not intend to investigate the circuit further; I am sure a better arrangement can be found / made. The one in the '73 magazine' will probably the next circuit I will investigate - provided that time permits! In the meantime, my thanks to everyone who has taken the time and trouble to contribute to this thread at this point; your various inputs are valued. And a prosperous New Year to you all! Al. |
29th Dec 2016, 6:51 pm | #28 |
Dekatron
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
|
Re: Simple RF probe for DMM
Yes the CWA circuit looks very dodgy and it has even dodgier roots as it was derived from that circuit exchange website here:
http://www.zen22142.zen.co.uk/Circui...tgear/rfmv.htm When I first saw it I didn't think it was worth analysing or building. It was a design that was worth even less than the sum of its parts because of the poor way each sub circuit was integrated together. Also the original designer Rodney Byne made some wacky claims about the small signal linearity and accuracy which appeared to be based on wishful thinking rather than any measurement based reality. If I get a chance I'll dig out a 1N34A diode and do some tests on it with it configured as per the original probe shown in this thread. This is a classic design. I've built it in the past as a large signal RF probe using an HP2800 Schottky diode.
__________________
Regards, Jeremy G0HZU |
29th Dec 2016, 6:58 pm | #29 |
Dekatron
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
|
Re: Simple RF probe for DMM
Interested to see the results. I've got a Marconi 2019A here which has a calibrated attenuator on it so I'm going to develop a voltage calibration and frequency response curve for it when I get a few minutes. I built some software (python based) a while ago that works with my UT61E DMM when plugged into the PC that will apply a calibration curve to the voltage reading and give you a real voltage so it should be possible to calibrate the original design and see what the response is like. Unfortunately my 2019A doesn't have GBIP so it's going to require a lot of button plugging I don't fancy yet! I don't expect much below about 1 dBm.
|
29th Dec 2016, 9:16 pm | #30 | |
Dekatron
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,923
|
Re: Simple RF probe for DMM
Quote:
I do think the CA3140 could be usefully employed in a RF Voltmeter - in a suitably improved circuit? We seem to have jumped threads, having started on the "RF Voltmeter" thread in Test gear and then coming over to this one; might be best to commit to one or the other. Not sure if anyone saw my post earlier today about the results I got looking at the response of a point contact diode at 20 and 40'C ( posted in the RF Voltmeter thread.) I think it shows that in a professional probe, having the probe work at a slightly elevated temperature is a good design feature, but for an amateur project for use in the temperate UK, it can probably be designed out by suitable construction of the probe; depends on the accuracy required. B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. |
|
30th Dec 2016, 12:55 am | #31 |
Dekatron
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,923
|
Re: Simple RF probe for DMM
Searching for other designs for the CA3140, I came across this Practical Wireless article, for a MOSFET voltmeter, which with a diode probe goes to VHF frequencies, but sensitivity is not so good https://billingtonrepairs.files.word.../voltmeter.pdf
B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. |
30th Dec 2016, 1:45 am | #32 |
Dekatron
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
|
Re: Simple RF probe for DMM
I did consider building something similar to the PW a while back from Experimental Methods in RF Design.
|
30th Dec 2016, 1:48 am | #33 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: Simple RF probe for DMM
Barrie: I've just come onto this Forum to add a post describing a FET-based voltmeter which I found in my 'Circuits misc' collection. I got the article all-prepared and ready to post it - only to find that you've already sent in the very same article! (ref. post #31)
Must be a classic example of great minds think alike! Oh and yes - I did read your post about diode performance: interesting. There was nothing in it that I could comment on. Anyway, I'd be very interested to hear what other members here think about this latest FET voltmeter design (by GW3JGA) . . . Al. |
30th Dec 2016, 3:18 am | #34 |
Dekatron
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,923
|
Re: Simple RF probe for DMM
Spooky! I wasn't reading PW in the mid-80's so haven't previously encountered the GW3JGA designs, however...
Especially with his inclusion of his PCB layout, this isn't going to take great effort to try it out. However, feeding the diode directly in to the op-amp looks like the limit of RF detection is ~25mV(RMS) and I would need to go back and check what was being claimed for the 73 design to see if that was better. [Just checked that and his lowest range is 0-30mV with first indication of 5mV, however, his probe design is more elaborate] I've placed an "ad" in the Wanted thread to see if anyone can offer a copy of the RF design which appeared in RadCom in 2008 which describes the original version of http://www.mw0llo.com/milliwatt.aspx, which may or may not be interesting. B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. Last edited by Bazz4CQJ; 30th Dec 2016 at 3:38 am. Reason: Update |
30th Dec 2016, 4:26 am | #35 |
Dekatron
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,923
|
Re: Simple RF probe for DMM
Some info re GW3JGA; he has a website http://www.gw3jga.net/jl-aboutme and was (probably retired) a senior technician at the Electronic Engineering Department of Bangor University and is otherwise deeply involved with BATC (television club).
There appears to be at least one very happy user of his VM design https://billingtonrepairs.wordpress....r-replacement/. Could he be a member of the Forum? B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. |
30th Dec 2016, 5:25 pm | #36 | |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: Simple RF probe for DMM
Quote:
Al. |
|
30th Dec 2016, 5:42 pm | #37 | |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: Simple RF probe for DMM
Quote:
Al. Last edited by Skywave; 30th Dec 2016 at 5:55 pm. |
|
30th Dec 2016, 6:29 pm | #38 |
Dekatron
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,923
|
Re: Simple RF probe for DMM
I hadn't spotted Latin . However, it is in the Art Section. It says that he left school during WWII, so he may now be at an age where his sense of fun may be different to ours. Also, the website looks like it is being created not by him, but by a relative (see copyright declaration)?
Anyhow, it looks as if in 1986, when he produced his VM, he was at Bangor and had access to some good equipment which gave him the test results he published. That does sound encouraging . B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. |
30th Dec 2016, 6:38 pm | #39 |
Nonode
Join Date: Jun 2005
Location: Western Lake District, Cumbria (CA20) - UK
Posts: 2,136
|
Re: Simple RF probe for DMM
The Latin text id the standard "Lorem ipsum....." placeholder text generated when you set up a website.
This thread has now drifted rather far from the original subject.
__________________
Brian |
30th Dec 2016, 6:49 pm | #40 |
Dekatron
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,923
|
Re: Simple RF probe for DMM
Thanks for that info Brian.
I suggest we continue back on the Skywave thread at https://www.vintage-radio.net/forum/...d.php?t=131879 and I will now post there. Hope that solves the problem? B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. |