|
Vintage Test Gear and Workshop Equipment For discussions about vintage test gear and workshop equipment such as coil winders. |
|
Thread Tools |
8th Jan 2017, 1:51 pm | #121 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: R.F. voltmeter: your thoughts, please
Following David's post above - and possibly those of others - it seems to me to be a good idea to briefly sketch out what it is I am aiming to construct.
Essential features. 1. Probe input impedance of at least 1 MΩ and a capacitance low enough to produce reasonable measurement accuracy (e.g. ± 3% FSD) up to 100 MHz. 2. An indicated range on a moving coil meter from 30 mV. to 10 V. RMS (30 V. would be nice) and with a dynamic range commensurate with that. Range switching that is 'sensible': e.g. a 1-2-5 or a 1-3 sequence. 3. Perfect linearity of scale is not essential, provided that it is consistent. 4. Adequate temperature stability in the range 15 → 30 deg. C. 5. Power source: battery or mains-powered. Other features are optional to a degree: i.e. physical size, weight, circuit complexity and cost, etc. I do not believe that I am being naive in that spec. If anyone disagrees with me on this, then please let me know! Then we can discuss the matter; thanks. Looking back at my reports here to date, there have been a few designs that have promised to meet most of those specs. - and where they fell short, I feel that I could have made a few necessary changes. However, several of those simply failed to meet their claimed specs., so further experimentation was, therefore, pointless. Al. |
8th Jan 2017, 2:03 pm | #122 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: R.F. voltmeter: your thoughts, please
David - I would not go so far as to say that the 'Art of Electronics' is an outstanding book. It is my opinion that the subject it attempts to deal with - as implied by its name - is simply too small to cover 'Electronics'. It does deal with both analogue and digital electronics and in my opinion, the analogue / R.F. part is too scant. I feel it would have been better to have split it into at least two volumes - say analogue and digital (yes, with the necessary overlap when required).
There are also few places in that book where I either disagree with what has been stated; feel that what has been written has been badly stated; or possibly I have misunderstood what has been written therein. Nevertheless, I would not wish to be without my copy and do indeed value it. However, the discussion of the merits of this book is off-topic, so if this topic needs further discussions, perhaps it should be in a new thread. Al. |
8th Jan 2017, 2:34 pm | #123 | |
Dekatron
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,924
|
Re: R.F. voltmeter: your thoughts, please
Quote:
I think it may be the case that an RFVM is somewhat like a GDO; not something that finds universal favour among professional engineers, but potentially quite useful for a hobbyist's instrument, and to that end, that's why we've been able to find a good number of designs published in ham radio mags over the decades (including Radcom and 73, which I would have some confidence in....previously.... ). But unlike any ham radio project I've ever worked on before, it seems like all of these various published designs claimed performance levels which they did not actually meet, and by some margin. Time and effort has been dissipated working on some of these, but I think some lessons have been learned and those provide pointers for the way forward. I think Al said he needed to take some time to do other things, not that he had completely lost interest, and I certainly intend to do some more work on this. B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. Last edited by Brian R Pateman; 8th Jan 2017 at 7:06 pm. Reason: Quote format corrected. |
|
8th Jan 2017, 3:21 pm | #124 | |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: R.F. voltmeter: your thoughts, please
Quote:
Indeed: I have not "lost interest". However, it can be somewhat de-incentivizing when you have seriously attempted various published designs (and not all from the Amateur Radio fraternity; some decidedly professional) to subsequently discover that the result was not as promised. As for 'time availability', as I have previously stated in another section of this forum, certain changes in my domestic arrangements have substantially reduced the amount of time that I used to have to pursue this hobby of mine, since retirement. But my interest in 'electronics' has always been strong (it provided employment for all of my working life). But to keep on-topic, it has long been my experience that nothing worthwhile (in Life) comes easy. I will persist with this project, but will also strive to retain a 'hobby - home responsibility' balance. Thank you all for your patience & understanding. Al. |
|
8th Jan 2017, 4:21 pm | #125 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,803
|
Re: R.F. voltmeter: your thoughts, please
100MHz bandwidth, 1 meg input Z and low capacitance 30mV to 10V, 1,2,5,10 sequence... why not just nick the input stage and attenuator of a 100MHz scope?
If you're not too insistent on the moving coil meter, just nick the whole scope. I think the AoE is outstanding because it has dominated the educational market for so long, because there are a few areas I can pick nits with (most other books see to have more) True, it's light on RF and the authors do explain this... I've actually had RF questions from one of the authors when he was developing a pulsed HF source. To get people going, to show that relatively easy calculations will design things and to give people the background on transistors, opamps and gates, it does a good job. There is no equivalent in the RF world. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
8th Jan 2017, 6:25 pm | #126 | |
Dekatron
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
|
Re: R.F. voltmeter: your thoughts, please
Quote:
Yes, that does look to be an interesting circuit. Back in post #62 in this thread I mentioned that HP used a similar concept in one of their old noise meters. They also used a current source to drive the doubler/detector diodes. I recall that they used an NPN transistor with an inductor in the collector to help optimise the current source. It produced a highly linear detector. I suspect that the AOE circuit will have limited signal range and it won't be a true rms detector but I'd expect it to be quite linear even for very small signals. So it could be a contender as an RF mV meter for sine wave signals and it might be possible to get it to work up into the VHF region but some experimentation would be required.. All the 'magazine' circuits I've seen on here have been very suspicious in terms of the performance claims made and I did try and warn about this. The one in the 73 magazine that used a pair of jellybean silicon diodes looked to be a dead cert for failure in terms of small signal bandwidth alone. That's why I wondered if there was something special/magical about the 1N914 diodes produced back in those days?
__________________
Regards, Jeremy G0HZU |
|
8th Jan 2017, 6:31 pm | #127 | |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: R.F. voltmeter: your thoughts, please
Quote:
Al. |
|
8th Jan 2017, 8:36 pm | #128 | |
Dekatron
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
|
Re: R.F. voltmeter: your thoughts, please
Quote:
I don't use an RF voltmeter here very often but I do have an old Racal 9300 true rms meter. This works across 5Hz to 20MHz and measures from 30uV up to many volts rms. It is an old friend and I used to use one of these a lot when measuring receiver sensitivity as it could measure S+N/N at the AF output of the receiver quite effectively. It can also cope with various waveforms as it has a decent spec for crest factor. So it can measure noise quite accurately for signals within its BW. Racal did produce the 9301 variant that works up to 1500MHz but it has a limited voltage range of 100uV to 3V and it only works down to about 10kHz. I've often thought about buying a 9301 but I really don't think I would use it very much. The input impedance of the 9301 is 100k ohm // 3pF but you can fit a x10 attenuator ahead of this to boost the input impedance up a lot and this also allows measurement up to 30V rms. The accuracy will suffer slightly with the x10 head but it is a true rms meter so it is a class above some of the other commercial meters mentioned so far.
__________________
Regards, Jeremy G0HZU |
|
8th Jan 2017, 8:55 pm | #129 |
Dekatron
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
|
Re: R.F. voltmeter: your thoughts, please
Probably the greatest praise I can offer the little Racal 9300 meter is that we still have a couple of them at my place of work. This meter has a DVM port at the back (that converts the AC voltage to a DC voltage) and you can improve the accuracy/uncertainty of your measurements if you use this feature and connect it to a DVM (set to read DC volts) and you exploit the digital display and extra features of the DVM.
This is really useful if you want to make lots of relative measurements on a dB scale. You can connect up a DVM and set it to dB/relative to get a 0.00dB digital display as your starting reference. The 9300 + DVM provides a very stable display like this with remarkable linearity over each of its ranges. If you terminate the input with a 50R load then you effectively have a versatile (true rms) LF power meter with a digital display. The lowest frequency any of my lab (thermocouple) power meter heads will go down to is 100kHz. The 9300 goes down to 5Hz and it is way more sensitive than the thermocouple head. For operation up to 15-20MHz it will take some beating as a sensitive RF voltmeter especially if you used it with a Marconi TK2374 RF probe with a x10 attenuator head fitted. You then get true rms performance from 300uV to about 3V rms across LF to maybe 20MHz with very high input impedance from the RF probe. If you used the Racal 9301 with its probe and a x10 attenuator you would get similar performance over a much wider frequency range but only down to about 10kHz.
__________________
Regards, Jeremy G0HZU Last edited by G0HZU_JMR; 8th Jan 2017 at 9:25 pm. |
8th Jan 2017, 9:21 pm | #130 |
Dekatron
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,241
|
Re: R.F. voltmeter: your thoughts, please
That looks like an interesting unit. I like the Racal stuff from that era - well made, and good performance.
I'd nominate the Fluke 8920A as a worthy contender. It's also 20MHz (apart from the most sensitive range - 2mV FSD - where it falls to 2MHz). It uses a thermal converter, so as near to "True RMS" as it gets, and combines a 1999-count digital readout with a basic moving-coil meter that allows you to watch for trends. And the "killer feature" for me, as an audio person, is the dBm mode. You choose the impedance - which for me is 600 ohms - and it does the math! Relative mode is included too. Mine came with the frequency counter output module, which is a post-attenuator feed of the signal. It's nice. Could I also add the Marconi 2610? We still have one at work. Very similar to the Fluke (25MHz though), but not quite as nice to drive. Again, a thermal converter. Newer design, so GPIB is an option. |
8th Jan 2017, 11:40 pm | #131 |
Dekatron
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
|
Re: R.F. voltmeter: your thoughts, please
I've not seen the Fluke 8920A before. The specs do look good on it. It's a shame it loses performance on the lower range but apart from that it looks to be a high performance meter. I'm impressed. It looks a bit strange in terms of styling but I guess this doesn't matter. I had a look on ebay and they seem to be quite expensive. £200 typical? That seems very high but then it does look to be a decent meter. By contrast the little Racal 9300 can often be bought for next to nothing. I think I paid £40 for mine from Stewart of Reading and that was quite a few years ago.
Purchases of the Racal 9301A are a little riskier/costlier and the remote RF probe often tends to suffer in terms of damage and wear and tear.
__________________
Regards, Jeremy G0HZU |
9th Jan 2017, 1:35 am | #132 |
Dekatron
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,241
|
Re: R.F. voltmeter: your thoughts, please
Mine was £31 from eBay - that was perhaps lucky, but you shouldn't have to pay much more than that if you're patient... Ignore the "optimistic" sellers
I agree the case looks a bit odd. Fluke used that style for a while, but moved away from it because you couldn't put anything in there that ran warm*. The idea was multiple units stacked up could be physically locked together (which works well in practice). However, if you put it on a shelf such that you only see the front panel, it looks great *Dave Taylor - an ex-Fluke guy - gave away a lot of this sort of info on the EEVBlog forum and during an Amp Hour podcast. The Marconi also loses bandwidth on its most sensitive range - I'm guessing that, like with the Fluke, an amplifier is having its gain increased by 20dB. From a quick look at the Racal, it looks like the gain stages are all fixed, but differing attenuation is switched in as required. They've done well to make that work down to 100uV FSD The Fluke has a button that prevents it auto-ranging down into the 2mV range if desired. I've often thought about building an external amplifier to provide the extra gain, but haven't ran into a situation where it would be useful as yet. IIRC, it becomes inaccurate below 180 counts, so 180uV would be the lowest signal that could be measured. I've used it a lot for audio, and rarely need to get down there (around -70dBu). |
14th Jan 2017, 11:12 pm | #133 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: R.F. voltmeter: your thoughts, please
For those who would like to know. It's been some time since I made a post here, but that doesn't mean that I've given up on this project. Things are happening behind the scenes and I will report accordingly when I have something worthwhile to say.
Al. / Jan. 14, 2017 // |
15th Jan 2017, 2:43 pm | #134 |
Dekatron
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,924
|
Re: R.F. voltmeter: your thoughts, please
Good; me too .
B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. |
23rd Jan 2017, 2:55 pm | #135 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: R.F. voltmeter: your thoughts, please
Just a 'quickie'. No progress at all last week: heavy dose of 'flu. Now slowly recovering.
Al. / Jan. 23 // |
4th Feb 2017, 8:32 pm | #136 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: R.F. voltmeter: your thoughts, please
What was that remark about "It never rains but pours?" Quite so. Whilst almost recovered from that awful dose of 'flu, one morning I awoke with a blinding headache. I recognised the 'type' of ache immediately: very characteristic, having had it before a few years ago.
For those interested, this is continued in Member Introductions and News (to avoid taking this thread too far OT!) Al. Last edited by Skywave; 4th Feb 2017 at 8:46 pm. |
8th Feb 2017, 11:44 pm | #137 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: R.F. voltmeter: your thoughts, please
Apart from pulling this Thread back on track, just a couple of rather fundamental aspects I'd like a bit of clarification on, please.
The 'front-end' of an R.F. voltmeter: there appear to be two approaches that are popular . . . (a) a series capacitance feeding a shunt diode acting as a half-wave rectifier, thus converting the to-be-measured R.F. to d.c. and (b) a resistive probe feeding a high-input impedance R.F. buffer, possibly followed by R.F. amplification, and the R.F. then rectified to d.c. In both cases, further d.c. amplification my take place in order to drive a display or meter. My concern here is the input impedance presented to the source by the R.F. voltmeter. Obviously, we would like that Z to be as high as possible and, ideally, constant over the freq. range of interest. A few moments thought reveals that that is not possible (unavoidable input reactance is a function of frequency). So, it seems that the next best target is an input Z that remains suitably 'high' at the highest freq. of interest. In that regard, to me, it seems that option (b) above is superior to option (a). And that is because the half-wave diode will present a low impedance when it is conducting, whereas the resistive probe will present a more uniform input Z and will present a higher Z at the same freq. So, where are flaws in my thinking there, please? What other factors have I failed to take into account . . . etc. Before I get any more deeply involved in this, I just want to check that I have a few basic facts clear and correct in my mind. Thanks guys. Al. |
9th Feb 2017, 12:00 am | #138 |
Dekatron
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,924
|
Re: R.F. voltmeter: your thoughts, please
Hi Al; I think that (a) or (b) is very much determined by the required sensitivity. If a limit of detection of ~10mV rms is OK, then happily go with (a) but for significantly <10mV (Marconi's did 300uV) I think you have to go with RF amplification in front of the diodes, so it's some version of (b). However, some companies (Boonton) seem to have access to some fantastical diodes, which don't need pre-amps and still go very low.
B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. |
12th Feb 2017, 11:59 am | #139 |
Octode
Join Date: Jun 2013
Location: Huddersfield, West Yorkshire, UK.
Posts: 1,783
|
Re: R.F. voltmeter: your thoughts, please
I have the Marconi TF2603 RF Millivolt meter, which has or HAD the mechanical chopper after the Diode probe. I thought I could substitute the amplifiers etc with more modern devices, eg opamps, either chopper types or not...... I successfully replaced the mechanical chopper with Fets, but after hours of work, I decided that I could not better the TF2603, thus all components were reclaimed and went onto the next idea.
The great thing about the TF2603, is that the amplifiers and final detector scale the ranges with the need to only print 3 scales. Who ever designed that animal, deserves a pat on the back.
__________________
Should get out more. Regards Wendy G8BZY |
12th Feb 2017, 12:51 pm | #140 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,803
|
Re: R.F. voltmeter: your thoughts, please
There have been more than two types, and many of the crudest, a valve or point contact diode in a little probe housing as part of or as an accessory to a multimeter have been shipped in very large numbers over the years. This makes them popular even if they aren't very good.
None of the methods is without some serious drawbacks, so i suppose you just have to pick one and live with it. A FET buffer with appropriate bootstrapping can give fairly low residual capacitance, but then the mechanical bits of a probe tip (or worse; input cable) puts the C up enough to disturb many everyday circuits. Maybe the best that can be done is what's found in active probe units intended for scopes and spectrum analysers. These are still in the onesy-twosy picofarad league. Their biggest problem is the low voltage range leading to high risk of destruction. Just throwing loss at the problem with a resistive probe still leaves the stray C of the probe tip. One thing that is vexing me is the issue of using a crystal filter with 1.8k//1.5pF load spec. The FETs I have available are over 2pF. SMT pads are about 0.3pf each So attempts to make a low C termination are ruined from the start. Do a resonant transformation and the tolerances will make that C value wander and may even go negative.... Grrr! So I think the issue comes down to choosing not what your RF voltmeter will do , but rather what it will Not do. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |