Tech Slot

Reported by G4DDK

These are HTML versions of the original articles in the VHF/UHF DX newsletter. Some small details have been changed to take account of time and cuurent availability.

The DX'er 50

Have you noticed all those 6m sporadic E spots on the DX Cluster this summer? I have, and because poor tropospheric conditions have led to very little dx on the higher bands, the attraction of 6m has been quite strong. I have to admit to not having been active on 6m for a number of years, so I have not experienced the "big" signals that have apparently marred operation for those amateurs of more modest means. However, the reports of the need for exceptional receiver performance have been heeded in the design that I will be describing in the SLOT.

My good friend Dave, G4HUP, described a transverter for 6m operation in the VHF/UHF Dx Book [1]. This design features high dynamic range, low cost and easy construction. As with the other designs in the book, PCB construction is used. The publisher of the book has promised that PCBs will be made available, but due to several problems that I wont go into here, they are not available yet.

Being impatient, and also fed up with buying or producing PCBs for everything I build, I thought it would be therapeutic to get back to construction using my favourite method of building using just the top surface of a sheet of copper clad PCB. Our American cousins like to call this "ugly" or "3D" construction. I prefer to call it "ground plane" construction. Basically all the components that make up the design are soldered direct to the ground plane or to each other as required to form the circuit. Short wire links bridge the gaps that the component leads alone wont otherwise bridge. This form of construction is fast, durable and yet easy to make changes to, as required. An experimenters dream! Professionally, I have used this form of bread boarding for many years, preferring it to PCB prototype board for first-pass designs.

The transverter, which I have taken the liberty of calling the DX'er 50, is a high dynamic range receive converter from 50/52MHz to 28/30MHz together with a transmit converter from 28/30 to 50/52MHz. I propose to describe the receive converter in this months SLOT and the transmit converter next month. Changes to this schedule may occur according to how the design evolves. I say this because I am not presenting a tried and tested design, guaranteed to work first time and able to blow the socks off every other design so far published. Rather the DX'er 50 is presented as an idea that you may wish to copy or use as the basis of a more advanced or simpler design. I have no plans to produce a PCB, as already explained.

My design aims were to build a receive converter with high dynamic range, be unconditionally stable and use (reasonably) easily available components. I did not consider it important that the design should be "no tune".

Figure 1 shows the receive converter block diagram.


The rf stages and the post mixer amplifier use Minicircuit MSA1104 Modamps with a TAK1H mixer. This mixer is specified at +17dBm local oscillator level and is called low distortion by the manufacturer. The infinite IP3 values for the filter and diplexer are optimistic, but are likely to be significantly higher than the active components as to make their inter modulation contribution negligible.

Using the values in the block diagram and assuming the IF has infinite IP3 and 0dB nf, then TCALC5 [1] gives noise figure as 3.8dB and IP3 as 0dBm. Obviously the IF contribution cannot be ignored (as some transverter manufacturers have chosen to do!). Assuming the IF noise figure is 10dB with IP3 of +20dBm, which is not too unrealistic for todays HF transceivers, then the overall DX'er 50 + transceiver noise figure will be 3.88dB with IP3 of -4.6dBm. Please note these are the calculated figures. At this time they have not been confirmed by measurement. The lack of broad band matching at the mixer rf port will inevitably lead to some degradation in IP3. The other mixer ports are adequately matched to ensure minimal degradation., I hope.

Turning now to the description of the circuit. The input arrangement is a single tuned circuit with the antenna connection and rf stage inputs tapped low. This ensures a reasonable degree of selectivity before the rf stages with minimal input loss. Without the input tuned circuit my 10Watt 70MHz DX Cluster link blocked the converter. With it there is no sign of blocking at all.

The two MSA1104 devices provide reasonably low noise performance with high input intercept. The combined gain here is 24dB. In the transverter the rf stages are followed by a changeover relay contact. For the purposes of this description, the loss of the contact is ignored.

The filter is an unusual inductor coupled design copied from that used by G4HUP in his 50MHz transverter [1]. The filter has a 3dB bandwidth of 3MHz, insertion loss of 5dB and a return loss of 10dB across the band. Out of band rejection is superb at image, LO and harmonic frequencies. It is also easy to align.

A Minicircuits Labs TAK1H mixer was used not only to ensure adequate receive converter performance but also it allows much greater IF drive input, leading to higher rf output with the consequent need for fewer stages of linear amplification in the transmit converter. Local oscillator requirements are +17dBm (50mw). A lower power mixer can be used, but the local oscillator level must be reduced in line with the requirements of the mixer chosen. Dynamic range will reduce if a +7 or +10dBm mixer is used. This may still be acceptable for some purposes.

The local oscillator uses a J310 FET in the classic 3rd overtone crystal oscillator design. A capacitive tap is used to match to the following MSA0504 amplifier. The J310 output was measured at +11dBm. The MSA0504 provides 7dB gain to bring the local oscillator drive to +18dBm. A three stage low pass filter reduces the second harmonic to -35dBm (-53dBc) at the expense of approximately 0.5dB insertion loss. There is no margin for using resistive matching between the LPF and mixer without introducing an additional amplifier stage, which I was reluctant to do. A 78L09 regulator provides a regulated 9 volts to the FET oscillator.

Following the mixer a further relay contact routes the IF to a third MSA1104 used as a post mixer amplifier. As explained in the VHF/UHF DX book, this provides a non-critical IF termination. A simple diplexer them follows to ensure that the post mixer amplifier is at least approximately properly terminated, since transceiver inputs are notoriously bad matches at 28MHz.

The circuit of the receive converter is shown in Hands Electronics page

How does it perform? Quietly, until a resonant 6m antenna is plugged in! Although no indication of performance, it was gratifying to hear CT1BUN on Sporadic E as the first signal within 30 seconds of switching the converter on for the first time.

Next month I shall describe the transmit converter.

Finally for this month, Practical Wireless readers will appreciate that I nearly called the design the "Neon", but then I thought someone else would probably copy the name anyway!

[1] The VHF/UHF DX BOOK. Published by RSGB ISBN 0-9520468-0-6

Part 2

73 de Sam