m1. LC- and Q-meters.


c. Amateur radio technical experiments
m2. RF detectors, level meters, attenuators, dummy loads, signal dividers
m3. RF signal- and power generators
m11 Grid dip meters + xtal testers
m12 Power- and VSWR-meters
m21 Norwegian instruments
m22. Old measuring instruments
m23. Signal-to-noise-meter
m31. LF/Audio instruments




Several old constructions and instruments are still a good idea to use today, the Rohde & Schwartz 50 years anniversary book contains several stories about some of their oldest, but highly appreciated instruments.


1.1)



Some German LF/HF-coax connectors from the 50...60's.
Above is the coax connector usually found on CF instruments from Wandel&Goltermann and Siemens, and below is the Fernsehen connector which is practical for video switch panels. I use them for selecting between HF antennas. (also shown on page m-12)



1.2) Analogue RCL-measuring instruments



Radiometer MM1f, an old, but still very useful multimeter to maintain!



Radiometer MM1f multimeter.
Modification in particular to achieve better reading between 20nH-2µH coil inductance. Note the series coil, it is the secret why it is possible to measure inductance values down towards zero. (Since it is a gif file it prints better than it seems on the screen). It is usually no problem to find inductance values with this instrument when the instruments fails to work. Also have a digital L-C-R-meter and some Q-meters, but the old 1950-vintage instrument seems still worth to maintain
.
A suggested compensation for home-brewing instrument, to null out the extra connection wires with alligator clips, in my case they measure to 150nH



1.3) Radiometer QM1f (50kHz-70MHz in 8 ranges)



It seems sometimes difficult to use if you don't already know the inductance or resonnance frequency. Won't modify as it probably upsets calibration, the only modification is a mains socket on the rear and a BNC skt for frequency counter. The screws for device under test is bad design and it is often difficult to fix the coil wires properly, they tend to fall out when screws are tightened. Perhaps it was invented before they started to make holes in the screws?!.
The Q-meter detector uses an EF6, see note further below


1.4) Build your own Q-meter calibration coils

Received note (not edited): I can give you some detail if you like using an Amidon core as an example. I would think the "Q" will vary some with "doping" material used to hold the turns in place, wire spacing around the core, and consistency of the core material itself (maybe other factors). I bring up the last point because these Amidon "yellow" cores appear to be operational at a lower range than their charts indicate even thought they mention that larger cores tend to peak towards to low end of the range.

Amidon T-50-6 core, Permeability 8, Freq range 10-50Mhz, color - yellow

Winding - 23 turns of #22 wire
Spacing - equally spaced around toroid
Doping - none
Inductance - 2.5uH

Results:

Frequency (MHz) Capacitance (pf) "Q"
7.9
8.5
150 
134
340
350
9.0
10.0
119.4
94 
355
355
15.0
18.0
41.5
28.5
340
296
19.2 25 284


73s  Kees K5BCQ



Boonton 260A Q-meter:
http://www.qsl.net/k5bcq/qmeter/qmeter.html


#1.5) how to build your own Q-meter


An idea how to build your own Q-meter, with the shown ferrite core transformer you easily make very low feed impedance I redesigned my Heath Kit QM-1 meter using this technique, and used a 6BA6/EF93 as grid detector, based on the principle used for Radiometer QM1f, this uses a Philips EF6, and I believe it is better than EF93 which I ended up with as only modern alternative, while EF94/6AU6 didn't work at all.QM1f uses a small resistor and requires quite heavy excitation to achieve high enough drive, and a thermocouple to measure the power. But this is much easier with the described ferrite toroid technique, since you may measure the level on 50W side.

Had some discussions with Dick Rollema PA0SE regarding detectors, and he proved that 1N4148 type detector, didn't load the circuit so much that it couldn't be used in a Q-meter, but of course it depends on the frequency and minority carrier life time


#1.6)



The poor man's Q-meter
It was constructed as a comment to a SPRAT article to demonstrate how cheap it was possible to make a coil measure meter, suppose it could be built for less than £2.50, but depends on what you have in your junkbox. American broadcast type capacitors seems to be 2x365µµF, while the European standard is 2x450pF or 405+455pF (since we have the longwave AM broadcast band, too), it is really not so important. With different capacitors you may tune to resonnance and coils be calculated when you know the capacitance. You use an 80m receiver and the S-meter deflection indicates the Q-value. It is important to use a good capacitor at C4, not such disc ceramic(!) type, the other capacitors are not critical in other ways than that they may influence the frequency stability. C9 should be small as possible, since the signal into the receiver is certainly strong enough, it consists of two insulated wire ends twisted together.





Suggested construction on a piece of copper laminate using the dead-bug method


#1.7)



Heathkit QM-1. (140kHz-18MHz in 4 ranges).We bought some in 76 when they were offered at half price.
It has reasonable accurate Q-readings, but I never managed to align the frequency properly over the band. 6AL5 detector was a problem, mine drifted, but somebody else believe the diodes could be selected. Must admit that I used QM1 for several experiments, and it is now far from the original, but possibly totally out of calibration. Ended up with a 6BA6 detector in a similar circuit to the one in Radiometer QM1f using EF6.


#1.8



Q-meter. Grid detector using 6AU6 used with Heath QM-1. The grid circuit must have dc return circuit via the coil. The original is used in Radiometer QM1, but has an obsolete type, EF6 which seems somewhat different from more modern type pentodes.

Had some discussions with Dick Rollema PA0SE regarding detectors, and he proved that 1N4148 type detector, didn't load the circuit so much that it couldn't be used in a Q-meter, but of course it depends on the frequency and minority carrier life time, so the usage may depend on operational frequency.


#1.9) Q-meter construction



Q-meter. An alternative circuit: The primary consists of a brass plates which passes through the toroid core.
It would be some improvement in coil loading if a voltage divider goes to the diodes or an
RF amplifier could be used as buffer.

Notes for building your own Q-meter.
The first problem to consider is that you need a defined RF voltage to use this technique for building your own Q-meter, the second is perhaps even worse; the detector should not load the tuned circuit. With 20mV excitation to the coil and Q=50, the voltage on the hot side is 1V. So if you attach a 1/100 voltage divider for connecting a 50 ohm amplifier, you will transform a loss resistance of 10000 x 50 = 500kW to the hot circuit and this is really too much for a good Q-meter. I would reckon that an integrated circuit detector need at least 100:1 detecting range, so it is not a good solution. Dick Rollema PA0SE mentions the diode detector, and perhaps, but it must be checked what the dynamic impedance is, and worse is the diode characteristic which should be checked properly for any detector intended to be used.
On the other hand, you may always build something which works, but you don't know the actual calibration, and you may have good use of it whether it is perfect or not.

RF detector with 2x 1N270.
LA7MI Stein Torp made some measurements of a voltage doubler with 2x 1N270 on HF:

RF (RMS) 1MHz Detected DC RF (RMS) Detected (DC)
10V
5v
2v
28V
14v
5.4V
100mV
50mV
20mV
187mV
50mV
13.7mV
1v
0.5v
0.2v
2.6
1.25
0.45
10mV
5mV
2mV
3.4mV
0.85mV
0.137mV

DC output from a voltage doubler with varying RF voltages (LA7MI, Amatör Radio 1985-11 pg 299),
note that detected voltage may decrease for higher frequencies, but not so much below 100MHz.



He constructed a quasi-compensating amplifier with CA3140E so that the reading was within 20% accuracy. It is described on one of my pages, but I am not capable of finding it now. It may also be mentioned on page
m2 as "LA7MI LF/HF/VHF/UHF mV-meter". Some other possible diode types are OA95, AA118, AA119, AAZ15.

If output voltage is not critical, you just want to see some DC change as RF voltage varies, see sensitive RF detector using silicon diodes (1N4148) on page m2




The problem with the detector is that it must have very high impedance, the characteristique must be predictable, and the performance must be checked and calibrated. A problem with germanium diodes is internal capacitance, so they should be applied with an RF-voltage-divider, a suggestion is shown here.


#1.10) Measuring small coil values



It is often difficult to measure small coil values, this construction works fine down below 20 nanoHenry. The secret in measuring such small values is to add a 0.5-1µH coil in series with the coil under test. This extra coil should have high as practical Q-value (dependent on you desires). Q-measurements may not be accurate, but you can measure coil vales down to zero.

See also the notes and modification for Radiometer MM1f which uses the principle mentioned with the extra series coil.


#1.11) Another RF transformer construction for Q-meter



Another suggestion is to use an RF-transformer described by W4ZCB for another purpose (ref. 3): To minimise leakage inductance, the single turn braid must use brass end plates so the braid can be opened up and pushed into contact with the sides of the hole in the binocular core.
However, reasonable results can be obtained with just the braided with enamel inside and no end plates.

References:
1) Zwischen-Basis amplifier with J308 2) G3SBI Radcom TT Dec 95 pp70-71
2) Zwischen-Basis cascode-amp. with J310 1) G3SBI Radcom TT May 95 pp60
3) Zwischenbasis amplifier W4ZCB and G3SBI Radcom TT Sept 96 pp70-71
4) Zwischen-Basis JFET amplifier 4) G3SBI Radcom TT Sep 98 pp 58-59

See the original notes on page Page r22
(2003.08.06)


#1.12) Edge-mounted p.c.b. tool

Some useful tool to pull out PCB's with edge-contacts. The shape varies for different manufacturers of
telecommunication equipment


page-994 List of contents




Back

2004.07.01