m11. Grid-dip-meters + xtal tester


m1. LC and Q-meters
m2. RF detectors, level meters, attenuators, dummy loads, signal dividers
m3. RF signal- and power generators
m12 Power- and VSWR-meters
m21 Norwegian instruments
m22. Old measuring instruments
m23. Signal-to-noise-meter
m31. LF/Audio instruments


#11.1: Philips professional GM3121 Grid dip meter









Philips GM 3121 Grid Dip Meter.
Note the Type 4662 indicator valve, this is a very old neon indicator which was used as magic eye in some receivers
of the mid 30's, the length of the light depends on the voltage across it, but I suppose it wasn't so useful in the
30's when valves when expensive, wages low, and dual types were even more expensive, but later and in particular
when war surplus were overwhelming, prices low and wages higher, it was no reason for not choosing a 12AT7,
6BQ7A or ECC85, so the optimum ideas have really changed, but this circuit is still very rare construction.




GDM data sheet




Philips type 4662 neon indicator tube is a strange type of indicator tube, probably designed around 1935 for signal strength indication in broadcast receivers, but seem to have been used very little for this purpose, only one receiver has been found where this indicator was used. But it works ufb for the GDM..
It is easy to find data for it in Frank Philipse's tube database, but the link has changed, so I won't put it here.


#11.2)Tech TE-15 GDM


You may notice that the 2m coil is opened and modified. It also another 2m coil,
but they cover almost the same range. Also note that all coils and segments of
scale are colour-coded, using the normal colour code: brown-red-orange-yellow-green-blue




Inside view. Have removed some of the rubbish which wouldn't operate properly, and gounded the so-called "negative side" so that it will work on +12V without shortcircuiting against other chassises.




Tech TE-15 Dipmeter (Tradipper).

The instrument was purchased from Schünemann in Berlin in 1967, and have been in use regularly since. One problem was
that it uses 9V battery, and it was usually flat when you need it. So I connected it to external 12V supply, but the problem is that it has plus to
chassis, and every time the chassis touches other equipment you the wellknown SHORT-CIRCUIT between dipmeter chassis and grounded
chassis state occurs. So I decided to convert it to positve supply voltage. The new circuit diagram is shown. As the oscillator frequency isn't
particularly stable and clean, I wasn't too interested in the monitor function, it wasn't found particularly useful at all. OC45 (455kHz IF amplifier
transistor) is experienced to be one of the best DC-amplifier transistors for low input voltage, better than the audio amplifier type used in the
original griddipmeter and was therefore used to improve the meter.

One problem is on the highest range, it simply won't operate properly. The reason was found to be the coil, it is not really a coil, but a short
between two pins inside the connector. Remove the short, and wind a coil using 18SWG copper wire, 5turn with 8mm diameter - not critical -
it will cover 2m and above it, slightly lower frequency range than it was earlier, but it works as DIPMETER, and you may now dip your 2m coils.
I also have the Philips GM-3121 griddipmeter, it is somewhat better than TE15, and must be used with care since the ECC85 exites higher
RF level, but I must admit, it is other, better ways to measure coils, but GDM provides an easy way to check resonnances and signals in tuned circuits.


#11.3:



Alternative modern version to the GDM

When you already have a modest signal generator (preferably not thumbweel tuned), but miss a reliable griddipmeter, this may be something to
consider. Tom (LB8X) says that MFJ259 will work fine using the described principle, and you need not buy adapter to try (see next figure).
It is not a GDO, GDM or transistor-dipmeter or such, it is an alternative.
The notes below are based on ideas described in an article, see QST May '86 pp.14-20: W7ZOI Hayward "Simplified Scalar Network Analyzer" p.14 (Feedback Aug p.40), with the subtitle:
Beyond the dipper (using signal generator)
here is shown my version - tested up above 200MHz, frequency limits are somewhat odd to discuss, it really depends on making inductive coupling to a tuned element, and change the shape to fit for different purposes. For UHF strip-lines one side of the probe should be flat for maximum coupling, while on LF - using 88mH toroids it is not difficult to make a 1-2 turns link into the toroid, but it is worse to use with clock cores.


#11.4) Another alternative GDM





Dip-meter using signal generator, this is a simpler version of the W7ZOI type dipper, in function is similar to using MFJ259 withthe set of test coils described in the previous notes. It seems important to use a signal generator with large range variation, HP-651B has 10:1 ranges, it is not stable and accurate enough for receiver tests, but very suitable here, toggle-switch tuned generator is hopeles device, you will tune so slow that you won't be able to see the dip


#11.5) Single capacitor type GDM



LA7MI dipmeter with single capacitor and coil without tapping (AR pg305 nov 82)



Other means of GDM's to consider (DUBUS VHF UHF Technik Vol.1 1980)
UHF-Dipmeter .................................................DL7QY pg.350-355
VHF-UHF Resonance Meter by DJ2HF and DC0DA pg 355-358
Superregenerative UHF Dipmeter................. DL7HG pg. 339-341


#11.6) UHF Dipmeter (DL7QY)

The construction is 25-30 years old and some changes are evident, but still some good ideas could be
seen.



This dipper is working well in the range of 130-1400MHz. The high sensitivity enables the user to dip small strip line circuits with good reliability. This is achieved by using for different probes, each containing the whole RF circuitry expecially designed for the corresponding frequency range. The tuning of the oscillator is done with variable capacitance diodes. The tuning voltage is controlled by means of a good potmeter (fig.1), and indicated by a large panel instrument which can be calibrated for frequency directly.



Power supply for DL7QY UHF dipmeter

Two stabilized voltages are needed, +12V for oscillator and level amplifier, and +30v for varicap tuning. Here is used a 15V line transformer with voltage doubler, followed by two regulator circuits - as shown in fig.1. The tuning voltage is stabilized twice by parallel regulators; IC TAA550K and a zener diode ZF27, which is temperature compensated by two 1N4148 (1N914) type diodes in series.


Dip-meter RF circuits and level indication:

All four probes have oscillator transistors in grounded base configuration, fig 1b shows the connection schemes of the recommended semiconductors. The varicaps are of type BB109, BB105 and BB141. They have low dissipation in the UHF range. Fig 1c shows their capacity variation due to tiuning voltage changing. The oscillator RF voltage is rectified in a voltage doubling circuit in each probe, and amplified and indicated by the remote "Level Indication" meter in the power supply box (fig 1). The rectifier diodes are biased with an adjustable voltage. It is adjusted to that voltage, at which the meter needle begins to rise from zero (when transistor is not oscillating - finger across the U-shaped quarterwave strip-line inductor).
The probes are connected to the power supply unit by a shielded 5-wire cable, using stereo pick-up connectors. The probe boxes are made of tinned iron plate 1mm thick, with outlines of 100 x 20 x 20mm. It is convenient to prepare at first long sheet strips 20mm wide and then cut off the desired lengths. The parts can be soldered together easily.

BB105 and BB109 varicap tuning diode capacitance (suppose these diodes are rather old now)

Connexions for the semiconductors
AF279, BFR34, BFR91 (Isn't BF479, BF679 or BF979 some more commonly found PNP devices?)





Oscillator levels for the different ranges



GDM Probe A (130-280MHz)




GDM Probe B (270-560MHz)





GDM Probe C (530-940MHz)




GDM Probe D. 900-1420MHz (DL7QY).
At this range oscillating was achieved with an npn silicon tranistor of the type BFR34A. Ft=4GHz.
Two varicaps BB141 connected in series tunes the frequency from 1420 down to about 800MHz. Because of the low L/C ratio, good sentitivity is fround from 900MHz and up. The lower part will not be used. Some spurious resonnances with flat peaks occur around 1100MHz. In this range dipping is possible, when the level indication will be recontrolled by turning the potmeter "selectivity". Fig.4a shows the outlines of L1. It is bended to shallow bow out of the probe case to improve the coupling. The front view (fig 4b) shows the arrangement of the two variacaps beneath L1. Their proper sites, especially the nearer the varicaps will be placed to L1, the higher the upper frequency will go.


11.7. Superregenerative UHF-Dip-meter (DL7HG)



Superregenerative UHF Dipmeter (DL7HG)
DUBUS VHF UHF Technik (Handbook) Vol.1 1980
pg. 339-341

see more VHF/UHF/SHF techniques on page D.


11.9 Xtaltesters


My old xtal tester, built in 1970, and modified several times.




11.10) Simple xtal tester which works with a frequency counter. Somewhat more open possibilities for making experiments.
Connect a VOM (preferably an anague type or 100µA-meter with series resistor) to the testpoint may indicate the crystal activity.
It is almost difficult to find a transistor which won't work. It is a good idea to mount some different xtal sockets on a box.
The variable capacitor is measured to 10-30pF.Most NPN general purpose transistors will operate (suppose the only transistor type
which doesn't work is 2N3704)



The xtal tester checking if it works with 307kHz xtal - although the optimum component values are for 3-15MHz


11.11) The xtal tester as VXO (variable crystal oscillator)

It was interesting to see how much swing I could get with some representative selection of xtals for considering some VXO circuit. The table below is only meant to show the comparison between xtals on different frequencies and overtone xtals operated on fundamental frequency. It is no secret that larger tuning range could have been achieved with some xtals, but that is another story, the comparation was more interesting for my application. In the first instance I am planning a close-in frequency sweep-oscillator to analyse ultimate selectivity of receiver's IF filters.

The 10-30pF tuning capacitor is certainly not optimum, so some larger range might be possible for several xtals, and the capacitors on the transistor could be optimized, too.
If you try to make the tuning range large, the output level may vary a lot, and it may become a problem to limit the output voltage. Too hard limiting may worsen the noise sidebands,

Frequency F min max deviation notes
3686.4 3686.14 3687.76 1.62 kHz HC18
9216 9214.80 9221.43 6.63  
11981.35 11983.62 11993.29 9.67kHz  
26786 8927.97 8935.33 7.36 HC18, R/C
27005 9000.60 9007.80 7.2 HC25, C/B
37800 12594.66 12604.51 9.85 HC18, R/C
14318 14318 14328 10kc Computer
15700 15697.50 15707.80 10.3kc AP22
46.3 15417.93 15430,6 12.67kc overtone
13352.080 13350.07 13360.54 10.47 Storno CQM
10700 10746.88 10755.67 8.79 Trio, HC6U
11541.34 11538.92 11549.23 10.31 HC6U, 30pF
14 866.0 14 864.54 14 874.37 9.83 AP22
15 539.30 15 537.12 15 546.11 8.99 AP22




Improved variable xtal oscillator (VXO)




Resonance tester.
A sometimes useful instrument, when it is not possible to dip a tuned circuit, it may be placed in a screened can.
Problem is that it doesn't always work, perhaps it is room for improvement here. It is neccessary to have DC-
return via the coil to instrument ground. In some cases an RFC could do this, but often it would upset the measurement.


connections for some actual transistors



It seems problem with definite dc level on the second transistor, and excitation control for the first stage, so it seems after few tests that this could be a solution. It doesn't pull the frequency down much compared to the dip-meter frequency. A later version using BF245B, BF324 (PNP) and BF314 (NPN), the two latter semiconductors has same pin connection as for BC547/BC557, in spite of what data sheets may tell! Wished to use some definite RF devices, but if they are difficult to find, replacement for BF324 is 2N3906 (=BC557 and BC558), and for BF314 is 2N3904 (= BC547, BC238), MPS918 or BF199 types are better NPN devices.

The last version (minus one). The coil shown has very high Q, it is the tank coil from Heathkit HX-20


e/m


BACK

Last update: 2005.01.08