Welcome to m0rjd.co.uk

M0RJD Calling CQ & Standing By...
Bob, M0RJD, is a British Amateur Radio Full Licensee, mainly interested in constructing and designing Amateur Radio equipment.


This picture of me "calling CQ" was taken some years ago by a work colleague. See if you can guess where the cable goes...
I wasn't licensed then, so there is no RF, which may be why I didn't hear any replies?


This website is a Work In Progress, please return from time to time.

Currently in progress is a PICASTAR (STAR) H.F. transceiver, which uses a DSP system for everything beyond the 2nd IF amplifier (Receive) or up to the same point (Transmit). The RF front-end is modern yet conventional, consisting of Band Filters, DDS LO, H-Mode mixer, crystal roofing filter and switching second mixer. The 2nd IF is at 15kHz for processing by the DSP unit, which produces audio output for SSB and CW.


I am now commissioning my STAR.

During this commissioning process I found that the LO signal to the H-mode mixer was too low and set about increasing it to the recommended +7dBm. Whilst doing this it became obvious that many other constructors had suffered the same problem, using a variety of methods to achieve their goal. Although I managed to reach the magic target by 'changing things', this did not satisfy a desire to understand why it was a problem and to devise a universal solution in place of the ad-hoc methods previously used.  See more of this

I have concluded that, although the DSP methods are excellent, some of the analogue/RF circuitry has errors. This shows how much attention was put on getting the DSP working through many code iterations, sometimes at the expense of the remaining circuit, which has largely been put together from chip-manufacturers' application notes or circuits copied from others. This has sometimes resulted in poor interfacing between these elements, and in some cases the perpetuation of originally flawed circuitry.

This is a great shame, because it means that the excellent DSP is not being presented with the best available signal.
No-one would expect that there would be gratuitous impedance mismatch, incorrect operating points, matching circuits that don't match, and just plain design errors, but all these exist in the design.
Over the years some improvements were introduced to PICASTAR, such as H-mode mixer and switching 2nd mixer to replace expensive diode mixer modules; better & more modern DDS, and so on. Some introduce new problems, such as lack of LO amplitude, and one has actually introduced a mismatch.

For many of us, the attraction of PICASTAR was the use of inbuilt DSP for the 2nd IF and subsequent processing, the DDS-based 1st LO, and the stable state of the design. Whilst the design is stable, in that it has had no official modifications for several years, it is not free from problems. It will work well anyway; you will probably not perceive any trouble whatsoever in operation. Many units have been constructed to the published design and are in regular use.
By far the most common problem encountered by recent constructors has been the difficulty of obtaining the requisite LO amplitude. This was my own experience.

Whilst in the process of commissioning my own ComboSTAR, as a newcomer to the scene, I have uncovered several areas that are in need of change, either to remove design errors or to avoid pitfalls. Other people have found problems in some areas also.
If you are put off by this, then please read on with an open mind. PICASTAR has first class capability, with superb results from the DSP. The main design problems can be rectified quite easily - I have a personal vested interest in ensuring that, wherever possible, changes do not require daughter boards or PCB modifications other than component changes. Without these changes it will work properly; many have built it as-is with satisfactory results.

For now I have dedicated these web pages to identifying and correcting the problems on PICASTAR, solely with a view to getting the best performance possible by avoiding marginality. It transpires that there is possibly a 1dB  performance improvement as measured by receiver sensitivity, but in any case the correction of design problems can only result in a better STAR.  There is neither need nor excuse for perpetuating design errors of the past. Look at any small MDS improvement as a bonus!

The changes I personally propose here are all reversible by reinstating the original components, should you wish. The notes are all aimed particularly at users of the ComboBoard, a large PCB from Glenn, VK3PE  which gathers the original sub-module collection together on a single PCB, but are equally applicable to the original modular build method.

Read on and see what you think.

If you know of an area of the design or a part of this site that is in need of tidying or correction, please let me know and I will consider either adding your changes to this website, properly crediting you, or add a link to your own information.  Naturally it is best if not only do you identify a problem or quirk, but also propose a cure!

The links below address problems and solutions:

Squarer part of First Mixer (recently added and rather rambling)
Low Amplitude from DDS / First LO,
Poor design of 2nd IF,
Mismatches in 1st IF bidirectional amplifier and surrounding circuit,
Bias point of Bus Switches used as analogue multiplexers,
Ferrite Transformers page to be added maybe.

In the meantime, please don't be put off by these things. Many constructors have built PICASTARs and are getting very good results from them.  But that isn't a good reason for leaving wrongs wrong when they can be made right.

I have put these items on this web-site to make them accessible to anyone who may be interested or just curious.
Please be aware that any alterations that you make to the published Picastar design will make it non-standard and could hinder fault-finding.  I suggest you don't make any changes to the published design without first getting the unit functioning; change only small sections at a time and verify that the changes don't upset things for you.

None of these items will ever be incorporated in the standard design, which has long been frozen.
On the Picaprojects Yahoo! group Peter says:
"Having given due notice that my time is going to be devoted entirely to PICaYAGI, please do not expect any ability to devote my time to any of your issues with the closed/sealed/finished projects.
They all work without reservation as published here.
This is not from any desire to not support published projects - but several years on each has to come to an end as they must if we are ever to move on. So if you have 'found' these projects several years after they were finished - or have taken several excess years to finish - please be aware that I am now not prepared to swap my present commitments to the present work for the old stuff. This last para is written because I am being continuously and unreasonably pestered.
Peter G3XJP"

The stuff here is a way of interchanging 'new' information on the design whilst avoiding Peter G3XJP being 'pestered continuously and unreasonably'. As he says, he is completely and unreservedly content with his published design.