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 Test Gear and Workshop Equipment

Notices

Vintage Test Gear and Workshop Equipment For discussions about vintage test gear and workshop equipment such as coil winders.

Closed Thread
 
Thread Tools
Old 1st Jan 2017, 7:52 pm   #81
Bazz4CQJ
Dekatron
 
Bazz4CQJ's Avatar
 
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,923
Default Re: R.F. voltmeter: your thoughts, please

Hi Al

the Radcom design is at post 2 in https://www.vintage-radio.net/forum/...d.php?t=132584.

I think it is a very interesting read, because the author highlights that fact he sensitivity of the diodes is very much dependent on the load resistance of the amp they feed in to, and he's pushed that to 40Mohm using Darlington pairs (lower than the CA3140, but still not bad). I seem to recal that he's claiming a lower limit of detection than GW3GJA. The input of the probe, however, is still 50 ohm. I have it about 70% complete now.

We both went with the 0CWA design because of its active front end, which sadly proved to be too optimistic. The alternative hi-impedance front ends appear to be;
  1. The "Poor Man's 1 GHz Probe"
  2. Possibly that chip used in the DAB pre-amp whose designation currently escapes me
  3. A homebrew amplifier using desicrete components.
I know that in recent weeks various circuits from commercial designs (e.g. Marconi) have been presented in this discusssion, but I don't recall a commercial design with an active front end appearing? Possibly, the IPR on such an instrument would still be under wraps at this point in time? I think I'm inclined to think that having a good (working) low-impdance RFVM on the bench is probably a useful stepping-stone to going on from that to a high-impedance instrument .

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; 1st Jan 2017 at 8:00 pm.
Bazz4CQJ is offline  
Old 1st Jan 2017, 8:39 pm   #82
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: R.F. voltmeter: your thoughts, please

I'm getting dizzy with all of the recent VM circuits that have been posted

The Radcom (50R power) meter has a very suspicious claim that it's possible to get a linear response down into the classic square law region just by using a high load resistance on the output. eg the figure 1 circuit was tested giving the top curve on figure 2 that looks linear. For an untuned 50R detector this seems to defy physics. Every diode detector I've ever played with or seen equations for will show that the diode will be in square law if the voltage across the junction is less than the thermal voltage Vt. So for a 50R detector that means that square law starts below -20dBm as this is just below 0.026Vrms.

Also the +10dBm calibration procedure is flawed because the procedure doesn't allow for the Vdrop in the diode. So I'd expect the procedure to give an error in the ballpark of a dB or so. But a lot depends on the diode choice.

The overall circuit looks like it might have stability issues over temperature but I guess you could experiment with retweaking the zero as required. It might be enough to get reasonable results from it.

One of the other meters uses a BAT85 diode which seems a curious choice because it will typically have 12pF self capacitance when detecting small signals. The big resonant peak in its response graph up near 150MHz is presumably caused by the earth croc lead resonating with the 12pF input capacitance. The location of the resonance will probably change a bit depending on drive level to the detector as I'd expect the 12pF self capacitance to go down a bit when detecting higher signal levels. The '73' meter uses a doubler arrangement using regular silicon diodes and I'd expect it to have a lowish input impedance for small signals because of the bias current through the diodes and I'd question if it can really demonstrate a 150-200MHz bandwidth for small signals on the lowest (mV) range when using this type of diode. I'd expect the BW to be OK for bigger signals but I'm not so sure about small signals that are near/at square law. I don't have any vintage 1N914 diodes here and modern versions of the 1N914 are often rumoured to be rebadged 1N4148 diodes. But there may be subtle differences from some suppliers I guess.
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 1st Jan 2017 at 9:06 pm.
G0HZU_JMR is offline  
Old 1st Jan 2017, 9:45 pm   #83
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: R.F. voltmeter: your thoughts, please

I think I've worked out why the graph in the Radcom article has odd sloping on the top (no load) curve.

I've highlighted the top (no load) curve in green in the curve key on the right of the graph and also on the graph itself. It looks to me like his step attenuator might only step in 10dB step sizes so each graph only has 5 data points with linear interpolation in between? So you only need a small amplitude error on one data point to get a strange slope on the graph either side of that data point. However, even with this taken into account the slope for the no load curve looks suspicious. I'd expect the square law conformance to be quite impressive from -22dBm down to -32dBm for a wideband (untuned) detector with a 50R input load like this one. Normally the conformance error is less than 0.5dB over a much wider range than this. But his results look quite different to this.
Attached Thumbnails
Click image for larger version

Name:	stepATT.jpg
Views:	124
Size:	46.2 KB
ID:	135227  
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 1st Jan 2017, 10:29 pm   #84
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: R.F. voltmeter: your thoughts, please

I made/measured a classic 1N34A probe using a shunt diode a couple of days ago across 400kHz to 150MHz and logged the results to post up on here. I used a DVM as the 10Meg load. This is the classic RF probe as seen many times in the ARRL handbook over the years. I was going to post up the results on Mr Bungle's thread (that uses the same circuit but with a different diode) but that thread has been closed.

See below for a snippet of the results for an unbiased 1N34A diode over the square law region in a 50R system. If I'd known I was going to be demonstrating square law conformity I would have used more digits on the DVM but even with 3 digits you can see that the results for 5MHz, 50MHz and 150MHz show very good square law response below -20dBm. It is just showing signs of coming out of square law by -20dBm at all three frequencies.
The best columns to look at are the dB Step and the dB Cumulative columns.

I used a very accurate Agilent signal generator in terms of flatness vs frequency and accuracy of the 1dB steps and cross checked the flatness and amplitude flatness on a decent lab power meter. I also checked the frequency response of the sig gen on a decent Agilent 26GHz detector and it is very flat. I suspect that the max error in the sig gen would be <0.05dB over this amplitude range and much less than 0.1dB across frequency.

In square law, the detector output voltage (DC) should double for every +3dB (doubling) of the input power and you can see it does this quite well. It starts to go a bit flat (wrt square law) by -20dBm but not by much.
Attached Thumbnails
Click image for larger version

Name:	1N34A_SQL.jpg
Views:	106
Size:	63.0 KB
ID:	135241  
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 1st Jan 2017 at 10:39 pm.
G0HZU_JMR is offline  
Old 2nd Jan 2017, 1:40 am   #85
Skywave
Rest in Peace
 
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
Arrow Re: R.F. voltmeter: your thoughts, please

Quote:
Originally Posted by Argus25 View Post
Skywave,
I have attached a circuit that might help on a wide dynamic range RF probe . . . . etc.
Yes, thanks for that. It does look worthy of an in-depth investigation.

Al.
Skywave is offline  
Old 2nd Jan 2017, 3:30 am   #86
Bazz4CQJ
Dekatron
 
Bazz4CQJ's Avatar
 
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,923
Default Re: R.F. voltmeter: your thoughts, please

Yes, the numerous circuits appearing are hard to keep up with!

Jeremy; for those of who are hard of thinking with 'dB' , could you include a column with RF input in mV in your table? Attached is the comparable data I gathered a few days ago.

I've just completed construction of the Radcom design; it sort of works "out of the box", but haven't spent any time yet trying to debug it. The meter hits the FSD end-stop with a mighty whack at switch on, then slowly settles back, but won't quite zero, so looks like it's not 100% right. I did basic matching of the two Ge diodes and four transistors. At 5MHz, the meter shows initial movement at 30mV RF (pk-pk) as shown on the scope and goes to FSD (100uA) with 250mV (pk-pk) input. Comparing it with my 50 ohm dummy load, it does have ~50 ohm input impedance, as expected.

B
Attached Thumbnails
Click image for larger version

Name:	Ge diode f(T).jpg
Views:	109
Size:	32.7 KB
ID:	135248  
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch.
Bazz4CQJ is offline  
Old 2nd Jan 2017, 4:03 am   #87
Bazz4CQJ
Dekatron
 
Bazz4CQJ's Avatar
 
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,923
Default Re: R.F. voltmeter: your thoughts, please

Re the Radcom design, which I constructed exactly as per original, two questions are bugging me;

why is C2 so large (1uF), is the 'meter whacking' due to that needing to charge?

Why is the 10k resistor alongside the 4.7k in Q2's emitter, when the zero pot is only 200R?

B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch.
Bazz4CQJ is offline  
Old 2nd Jan 2017, 4:49 am   #88
Bazz4CQJ
Dekatron
 
Bazz4CQJ's Avatar
 
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,923
Default Re: R.F. voltmeter: your thoughts, please

The attached plot shows the response of the first build of the Radcom. Might be marginally better at low end if I could have set zero at zero!

B
Attached Thumbnails
Click image for larger version

Name:	Radcom V1.jpg
Views:	87
Size:	29.2 KB
ID:	135249  
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch.
Bazz4CQJ is offline  
Old 2nd Jan 2017, 12:18 pm   #89
Skywave
Rest in Peace
 
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
Default Re: R.F. voltmeter: your thoughts, please

Quote:
Originally Posted by Bazz4CQJ View Post
Yes, the numerous circuits appearing are hard to keep up with!
Agreed. For me, "keeping up with" means trying to filter out the ones that fail to meet my overall concept of what I am looking for. For some members here, that filtering process is easier than it is for me. For example, some have access to, and knowledge in the use of, circuit simulation software (I don't); many have a greater extent of knowledge and experience with these types of circuits than I have; some may have more time available to them than I have. My approach is (a) to be guided by what others have found and reported; (b) use my limited knowledge & experience to eliminate the obvious 'that won't work' ones (which are few in number), and (c) ultimately to construct and evaluate the remaining most promising ones. But with patience, determination, time availability and with the continuing aid of members here (for which "thank you"), I feel fairly confident that I will eventually meet with success.

And with that remark, today, hopefully, I can find some time back at the bench!

Al.
Skywave is offline  
Old 2nd Jan 2017, 1:10 pm   #90
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: R.F. voltmeter: your thoughts, please

Quote:
why is C2 so large (1uF), is the 'meter whacking' due to that needing to charge?
C2 provides the AC ground connection for the 50R input termination R1 and so it needs to have a low reactance at the lowest operating frequency in order to keep the input match close to 50R. With 1uF at C2 the input VSWR will be <1.5:1 down to about 10kHz or so.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 2nd Jan 2017, 1:28 pm   #91
Skywave
Rest in Peace
 
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
Arrow Re: R.F. voltmeter: your thoughts, please

Quote:
Originally Posted by Bazz4CQJ View Post
Jeremy: for those of who are hard of thinking with 'dB' . . .
I don't have too much trouble 'thinking in terms of dB'. But on the odd occasion when I do get stuck, the pocket calculator gets used! However, where I do have a problem in the discussions of circuits, such as in this thread - an R.F. voltmeter - is the use of dBm, which is a power measurement. Yes, agreed, when a dBm figure is stated and the related resistance figure is known or implied, the corresponding voltage can be calculated. But if the measurement had been stated in voltage terms, that tedious calculation would then not be necessary.

A similar viewpoint can be taken about VSWR, reflection coefficient (RC) and return loss (RL): they all relate to the same concept & measurement, but expressed in different ways. For me, I prefer return loss; others may prefer VSWR.

To summarize: in general, I am O.K. with dB. However in this thread, I do have a problem with the use of dBm; others may not. 'Horses for courses', etc.

Al.

Last edited by Skywave; 2nd Jan 2017 at 1:33 pm.
Skywave is offline  
Old 2nd Jan 2017, 1:59 pm   #92
Skywave
Rest in Peace
 
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
Arrow Re: R.F. voltmeter: your thoughts, please

Quote:
Originally Posted by Bazz4CQJ View Post
Re the Radcom design, which I constructed exactly as per original, two questions are bugging me;

Why is C2 so large (1uF) ? Is the 'meter whacking' due to that C2 needing to charge?

Why is the 10k resistor alongside the 4.7k in Q2's emitter, when the zero pot is only 200R?
The meter hitting its end-stop when item powered up. I think the reason is as follows (with the meter polarity switch in the position shown). At switch-on, with no input signal, R3, R4, R1, D1 and C3 cause Q1 and Q2 to turn hard off. Thus the meter has a large +ve pulse and one of its terminals. OTOH, at that time, Q3 and Q4 are turned on, thus completing the path for the current through the meter. The fact that the meter is subjected to such severe mistreatment is not a sign of a good design.

As for the 10k and the 4k7 in parallel, I have no idea. The equivalent value of that pair is close enough to 3k that I would have thought that that value would have been chosen - or even 3k3. Again, I'm not impressed.

Al.
Skywave is offline  
Old 2nd Jan 2017, 2:24 pm   #93
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: R.F. voltmeter: your thoughts, please

Quote:
Jeremy; for those of who are hard of thinking with 'dB' , could you include a column with RF input in mV in your table?
Here you go. I've also formatted the numbers to 2 decimal places in excel so it looks tidier.

I've got results all the way up to +22dBm input (2.82Vrms drive) at all three frequencies if anyone is interested? The plot below is just for square law.

Your test results show a fair bit of change over temperature.I didn't try any temperature tests here. However, I'd expect a single uncompensated Ge diode detector to be quite poor over temperature down at very low drive levels eg 20mVrms drive.
Attached Thumbnails
Click image for larger version

Name:	1N34A_SQL_mV.jpg
Views:	80
Size:	50.9 KB
ID:	135256  
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 3rd Jan 2017, 12:51 am   #94
Skywave
Rest in Peace
 
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
Arrow Re: R.F. voltmeter: your thoughts, please

And here's my Excel look-up chart for conversions, 50 Ω systems: dBm, dBuV, voltage and wattage.

Al.
Attached Files
File Type: xls dBm, dBuV, voltage & power conversion, 50 ohms.xls (198.5 KB, 97 views)
Skywave is offline  
Old 3rd Jan 2017, 1:37 am   #95
Bazz4CQJ
Dekatron
 
Bazz4CQJ's Avatar
 
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,923
Default Re: R.F. voltmeter: your thoughts, please

Thanks for the various responses. I've not done anything new today, but the thought crossed my mind that the results that I got from the Radcom meter were pretty much the same as I had obtained a few days ago using the "ultra-basic probe" connected to an inexpensive DMM (see below), which is disappointing!

Al, do you have the specs of any of the commercial RFVM's, like the Marconi, and can you say what sort of limit of detection they achieved with probes with no amplification in front of the diodes?

B
Attached Thumbnails
Click image for larger version

Name:	basic probe a.png
Views:	180
Size:	16.2 KB
ID:	135278  
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch.
Bazz4CQJ is offline  
Old 3rd Jan 2017, 7:32 am   #96
Bazz4CQJ
Dekatron
 
Bazz4CQJ's Avatar
 
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,923
Default Re: R.F. voltmeter: your thoughts, please

Marconi TF2603;abstract from Electronic Engineer's reference book; This instrument is a fully transistorized millivoltmeter, designed for A.C. voltage measurement from 300 microV to 3V through the frequency range 50 kc/s-1,500 Mc/s.

The 0.5in. diameter probe uses a pair of fast germanium diodes in a full-wave
circuit whose response is close to true r.m.s. with inputs of less than 40 mV and peak-to-peak in the 0.5 to 3V region.


How much less than 40mV?

B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch.
Bazz4CQJ is offline  
Old 4th Jan 2017, 3:02 pm   #97
Skywave
Rest in Peace
 
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
Arrow Re: R.F. voltmeter: your thoughts, please

Barrie - re: your post #25. Sorry about my delay is responding. At this time of year I'm always very busy with household matters.

However, here is a brief summary of specs. for commercial VTVMs, obtained from the 'Net.

Farnell TM6. Transistor & integrated ccts.
50 kHz - 1500 MHz.
1mV - 3V.
Zin: 150 kΩ @ 1 Mc.; 100 kΩ @ 300 MHz; 2.5 pF.

Heathkit AV-3U.. Valved.
10 Hz - 400 kHz.
10 mV - 300 V.
Zin: 1 MΩ @ 1 kHz.

Hewlett-Packard 400. Valved.
10 Hz - 4 MHz.
100 uV - 300 V.
Zin: 10 MΩ; 25 pF

Marconi TF1041C.. Valved.
20 Hz - 1500 MHz.
25 mV - 300 V.
Zin: 500 kΩ @ 1 MHz; 150 kΩ @ 100 MHz; 1.5 pF.

Those instruments whose freq. range extends to UHF use separate external probes for that purpose.

Al.
Skywave is offline  
Old 4th Jan 2017, 4:10 pm   #98
jimmc101
Heptode
 
jimmc101's Avatar
 
Join Date: Mar 2009
Location: Portsmouth, Hampshire, UK.
Posts: 674
Default Re: R.F. voltmeter: your thoughts, please

Interesting to not that the TM6 probe chops the RF signal before the detector rather than use a chopper stabilised DC amplifier after rectification.
The circuit is on the forum here.

Does anyone have details of the devices used?
(I've got a TM6B without a probe.)

Jim
jimmc101 is offline  
Old 4th Jan 2017, 8:24 pm   #99
mhennessy
Dekatron
 
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,241
Default Re: R.F. voltmeter: your thoughts, please

Could the Marconi TF2603 be added to the list?

50kHz - 1.5GHz
1mV - 3V FSD
Zin: 150k @ 1MHz, <2.5pF

It's transistorised. The diodes are in the probe head, and the input is balanced to reject noise. The probe is heated. The DC amplifiers are chopper-stabilised to reduce drift.

It's quite similar to the Farnell TM6 in architecture, though there are many implementation differences.

Note that the Farnell TM6 is very different to the Levell TM6.

My TF2603 is up in the attic, but could be brought down if further information is needed.
mhennessy is offline  
Old 4th Jan 2017, 11:29 pm   #100
Bazz4CQJ
Dekatron
 
Bazz4CQJ's Avatar
 
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,923
Default Re: R.F. voltmeter: your thoughts, please

Thanks for those responses; I'm working on the Radcom RF Voltmeter, referred to in https://www.vintage-radio.net/forum/...d.php?t=132584

The designer claimed great things for it but the best I can do is see a “first indication” of a response from it at about 10mV(rms) RF (see attached for the latest, slightly improved preformance), which is disappointing really. I think that to be useful, it really needs to be at least x10 better, or even more

Of course, in these probes, the perfomance of the diodes is all-important, and perhaps the vintage gold-bonded Ge ones I'm using (unmarked) are just not good enough.

I have some GEX66, maybe I should try those?

Thanks
B
Attached Thumbnails
Click image for larger version

Name:	radcom 3.png
Views:	87
Size:	9.5 KB
ID:	135379  
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch.

Last edited by Bazz4CQJ; 4th Jan 2017 at 11:37 pm.
Bazz4CQJ is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 11:37 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.