Tempo's Complex Impedance Analyser for HF
An Evaluation
This article is an operational review of the Tempo Research Complex Impedance Analyzer, CIA-HF model 5012-5000, S/N 0255. The CIA-HF can be used to characterize 50 ohm systems between 400 kHz and 54 MHz, such as transmission-lines, baluns, antennas etc..
I have exchanged data and several emails with the manufacturer before releasing this article and we both expected I would have a new unit for review before this article was released for publication. The unit I did review was an early unit and the factory has now gone through an improvement in process control that ensures maximum accuracy for the critical bridge circuits.
When another unit is available I will re-review its accuracy because, by reports, the new process control has improved the measurements. The manufacturer is also forwarding the new data acquisition software for evaluation, so I will keep you posted.
Introduction
In my October 1999 article for antenneX, I showed you how to construct your own resonant normal-mode helical antennas and in another article before that, I showed how you may employ spiral end-loading to bring a short element to resonance. The CIA-HF was mentioned in both these articles because it provided the experimental measurements that made these articles possible.
Several inquiries later, and prompted by my desire to know the operational accuracy of the CIA-HF, I finally sat down and measured the performance of my unit. I did this without employing an accurate bridge, network analyzer or other expensive equipment and, no, you don’t have to find expensive reference components and plug them in one-by-one!
A terminated transmission line will reflect a complex impedance at its input. For example, any length of ideal coaxial cable with a characteristic impedance of 50 ohms terminated by a perfect 25 ohm resistor will exhibit a Standing Wave Ratio (SWR) of 2:1 at all input frequencies. These input impedances should fall on the 2:1 SWR circle when plotted on the Smith Chart. When the terminated cable is shorter than a quarter wave it will exhibit inductive reactance and when it is between a quarter wave and half wave long it will exhibit capacitive reactance. This behaviour alternates through each successive quarter wavelength and will exhibit resonance when the length is a multiple of a quarter wavelength.
The value of the termination and the electrical length of the line determines how much resistance and reactance is presented at the input and these impedances can be determined from the transmission line equations. I refer you to a very well presented article written by Wilfred N. Caron [ref 1] which uses this technique to characterize the Tennatest Admittance Bridge, the Palomar Engineers R-X noise bridge, a Twin-T Admittance bridge and his own excellent Hybrid Junction Admittance Bridge.

The CIA-HF contains a microprocessor controlled DDS synthesizer (400kHz to 54MHz accurate to 200Hz) with low power impedance bridge (1000 to 0 ohms with 2.5 digits which may display ±0.1 ohms) and a lower power SWR bridge, an LCD graphical display, a keypad and an RS232 serial interface.
The unit is powered by 8 AA batteries (not included) and is connected to the r.f. load via an SO-239. A short length of 50 ohm coax is run between this connector and the Printed Circuit Board. This length must be accounted for when performing accuracy comparisons. It is also advisable to connect an external power supply, as the unit consumes 150mA and the display contrast tends to fade and other functions tend to falter with low battery voltage.
The unit is housed in a robust grey plastic case which is provided with a handy flexible adjustable stand.
The unit is supplied with instructions which describe all operations, how to construct simple r.f. accessories, how to perform measurements such as check coax for shorts, open-circuits and power-loss, and how to wire the external computer interface, which is provided by a 3.5mm stereo audio socket . The photo above shows the CIA-HF with its RG58C/U coax test harness.
Transmission-Line Equations
The transmission-line can be characterised by the general equations that follow. If the line is lossy then equation 4 must be employed, however, if the line is reasonably good you may use equation 5 without introducing significant error [reference 2].
The reflection coefficient at the input due to the load is described by:

The consequential standing wave ratio will be:

The transmission-line propagation constant is:



Where f is frequency in MHz, A100 is attenuation per 100 ft, Vf is the velocity factor and l is the wavelength. A100 can be derived from the manufacturer’s attenuation data at 1 MHz and s is derived from the attenuation slope; A100 = 0.37 dB / 100 ft and s = 0.58 for RG58A/U [reference 3.0].
The input impedance for a general transmission-line terminated by Zl is:

While the input impedance for an ideal line, or nearly so, will be:

Where l is the physical length of the line, Zo is the characteristic impedance and Zl is the load or terminating impedance g and b and are defined in equation 3.0 and 3.2 respectively.
The Tests
Several test harnesses were constructed to determine how the unit performs.
The first harness was a 2.71m length of coax (this length includes the length of the small piece of coax present between the SO239 and the PCB inside the unit) which was terminated by a T-connector and two 50 ohm Ethernet terminators.
The next test harness was constructed from 6.59 m (including the internal coax) terminated by two 47 ohm resistors in parallel, which measured 25.7 ohms at d.c.
Appropriate frequencies were selected and the SWR, Absolute Impedance (Z), Resistance (R) and Reactance (X) of the coax input were obtained from the data display. Care must be taken determining X because the unit only displays the absolute value of X and determines the sign of the phase angle by the behavior of the input as the frequency is swept.
The R, X, Z and SWR values can be compared against the theoretical values derived from the transmission-line equations.
The Results
In Figure 1 you can see the variation in the magnitude |Z| or absolute value of the input impedance with frequency. The initial rise is probably caused by some capacitive coupling inside the unit; this effect was not present above 400 kHz
Figure 1 – Variation of |Z| with frequency.

The |Z| behaviour is quite reliable and indicative of the transmission-line behaviour. You can even see the slight decrease in amplitude caused by increased line attenuation at higher frequencies.
The Figure 2 – SWR below illustrates the SWR variation with frequency. This is a measurement of a value, which according to the ideal transmission-line equation, does not vary for the same load. In fact for RG58C/U, this value does not vary much over the measured range, the variation is almost certainly an artifact of the CIA-HF measuring technique.
Figure 2 – SWR vs Frequency

The next figure illustrates the variation of the measured phase angle with the frequency:
Figure 3 – Phase Angle vs Frequency

The phase angle is described by the arctangent( X / R ). The phase angle is zero when the impedance has 0 reactance which means the circuit is resonant.
The coaxial line phase variation has been measured and plotted above in Figure 3 – Phase Angle vs Frequency. Aha, something strange is happening, the unit is exhibiting phase discontinuities at subsequent resonances. The unit indicates zero phase across a spread of frequencies because the unit is erroneously measuring zero reactance at these points. That is, the zero phase point will actually occur somewhere inside this spread and the slope of the phase changes will be incorrect when the reactance values are small.
Figure 4 – Theoretical Line Impedances at 2:1 SWR

Figure 5 – Measured Line Impedances at 2:1 SWR

The overall performance can be judged by looking at a plot of the resistance and reactance values at various frequencies. The input reactance should initially be positive, or inductive, and then subsequently negative, or capacitive, as the line passes through the first quarter-wave resonance and then positive after the half-wave resonance and so on through all resonances. Any line or other measurement losses will cause the trace to tend towards the transmission-line characteristic impedance, i.e., very long lines look like their characteristic impedance, irrespective of the terminating load. The attenuation trend will appear more pronounced at the parallel or half-wave resonances. Notice the phase discontinuities where the reactance is zero for several spot frequencies.
Figure 6 – Relative error

In the relative error graph, Figure 6, the SWR error has been calculated against the theoretical value and the difference between the theoretical impedance values and the indicated values have been compared as a percentage of the magnitude of the theoretical impedance Z.
The CIA-HF indicates the magnitude of the complex impedance as Z and indicates the absolute value of imaginary part as X and the real part as R i.e., Z = sqrt( R^2 + X^2) and determines the phase angle from arctan( X / R ) and the sign of the angle from the behaviour.
You will notice that the SWR error is quite small, while the Z error is a bit larger and the X error is the most significant.
At this point there was meant to be a Spreadsheet that was not backed up and thus is not available, sorry about this – MD0MDI
These test results can be found in the supplied spread-sheet (which is virus free ~ Ed.). If you look at the results for the shorter coax, which only exhibits one wave between 1 and 35 MHz, you will notice that the phase discontinuity occurs for small Z only. Thus the unit appears to be functioning reasonably satisfactorily, however, the unit obviously has periodic errors depending on circumstances.
Conclusion
The CIA-HF was characterised with a test harness exhibiting a relatively high SWR (2:1) input to the unit as a means to compare it against other bridges in accordance with the work of Wilfred N. Caron [ref 1]. The CIA-HF measured these impedances better than all bridges except the “Hybrid Admittance Bridge” especially designed by Caron.
The CIA-HF exhibited the largest errors in the calculated reactance values when those reactances were near zero – i.e., near resonance.
Since the unit employs an SWR and Z bridge, these values (SWR and Z) remain reasonably accurate and are a good indication of the connected antenna system performance. However, the X and sometimes R values are sometimes questionable. These later problems may have been addressed in the later units as a significant process control correction has been made. You will need to wait for the results of the review I intend to perform when a new unit arrives at my home QTH.
The CIA-HF is a very handy piece of equipment. It contains an accurate frequency source and the ability to measure SWR, Z, R and X in 50-ohm systems. It has diverse applications which I intend to present in another article.
References:
- Wilfred N. Caron, “The Hybrid Junction Admittance Bridge” The Antenna Compendium Volume 3, published by the ARRL in 1993 (ISBN
0-87259-401-7), pp 223 – 230.
- Ramo, Whinnery, Van Duzer, “Fields and Waves in Communication Electronics”, Wiley Internation, 1965.
- Losses in Transmission Lines, ARRL Handbook, 1999 page 19.5.
- CIA–HF Complex Impedance Analyzer, Operating Manual, Tempo Research.
Table 1 – Specifications:
Item | Specifications |
---|---|
Frequency range | 0.4 to 54MHz |
Resolution | Increments of 1KHz |
Accuracy | ± 200 Hz |
Display width | 0 to 10MHz |
Harmonics and spurious output | < -30 dB |
SWR Impedance | 50 ohms |
SWR range | 20:1 |
Impedance and Resistance ranges | 0 to 100, 0 to 250, 0 to 1000 ohms |
Return Loss range | -1 to -40 dB |
Phase angle | -90 to +90 degrees |
Q Factor range | 1 to 1000 (defined as 2:1 BW/Fc) |
Measurement speed | 1.2 seconds per sweep |
Antenna connector | SO-239 |
Output power | <5mW into 50 ohms |
DC voltmeter | 2.5 digits ±10%, 25 volts max. |
Power requirements | 8 AA cells, 12 to 16 VDC @ < 150mA |
Size | 4.3″W x 2.25″H x 8.5″L (including connector) |
Weight | 1 lb 10 oz (including batteries) |
Table 2 – includes the various scales where provided by F4
Item | Scale 1 | Scale 2 | Scale 3 |
---|---|---|---|
SWR | 18:1 to 1: | 6:1 to 1:1 | 2.8:1 to 1:1 |
Z (Ohms) | 1000 to 0 | 250 to 0 | 100 to 0 |
X (absolute ohms) | 1000 to 0 | 250 to 0 | 100 to 0 |
R (ohms) | 1000 to 0 | 250 to 0 | 100 to 0 |
Originally posted on the AntennaX Online Magazine by Ralph Holland, VK1BRH
Last Updated : 16th May 2024