AR7070 Progress Updates

Status of the AOR 7070 Receiver, 22-10-2008

I am pleased to say that my friend John Thorpe, the designer of the AOR 7070, has made considerable progress with the production design, so there should be a prototype with early software by the end of the year. Martein Bakker and I have both contributed in the research for the front end design. Apart from being a bit of a genius John is a very good production engineer and he has optimized most of the circuitry for surface mount components in the basic front end design without compromising the technical performance.

As things stand at the moment the receiver noise figure will be 11dB for an IP3 of 45dBm at 7MHz rising to 50dBm at 30MHz.This is at 50 KHz spacing but my experience suggests that this should be maintained to about 15KHz two tone test spacing. A critical element in the design is the first IF amplifier which gives a NF of 1.3dB for an IP3 out of 40dB and a gain of 10.5 dB. Having a low noise figure and a reasonable output IP3 is not easy to achieve. It is expected that the radio will retail for about 1300 and it will also have a good LF performance.

In the process of optimizing the design John seems to have discovered what is important in the design of the transformers for best IP3 in an H mode mixer .Martein may be able to take advantage of this to further improve the performance of his already excellent front end design particularly on the higher frequency bands.

I have already extracted a promise from John to give a talk and demonstration of the new radio at Warrington Amateur Radio Club as soon as it can be put on air. It should be worth waiting for.

Status of the AOR 7070 Receiver, 21-11-2009

At the present time there are two prototypes one of which is being used for DSP software development by the receiver designer John Thorpe. SSB and CW signals have been received with good results but there is still quite a bit of software to write before the radio can go into production. Overall noise figure for the receiver before any optimisation is 12dB. This is about 0.26 uV for 10 dB SNR in a 2KHz bandwidth for SSB reception without the need of a preamplifier before the H mode mixer.

There is an issue regarding the 45MHz fundamental mode monolithic crystal filters used in the design. The AOR 7030 receiver used filters manufactured by Hertz technology in Japan and it was intended to use these in the AOR7070 because of their excellent performance in terms of in band and out of band inter-modulation performance (IP3).

Hertz have been taken over by Kyocera and the 45MHz filters were discontinued. We do not want John to be diverted from the software development so Mark Sumner of AOR UK and I have been looking at other sources of supply. Martein has very kindly tested a number of products we have obtained for in band and out of band IP3 and also to see if their linearity with signal level obeys the third order law.

Clearly there is still some way to go before the radio is in production. However we already know that the radio will have superb IP3 performance and excellent phase noise performance for an up-conversion radio. So I think it is a winner and worth waiting for!

Status of AOR7070 Receiver, 18-02-2010

Since my last update my friend John Thorpe, the AOR7070s designer, has made considerable progress with the DSP software to the point that the receiver functions on CW, SSB, AM, FM and synchronous AM and has variable bandwidth DSP filters. There was a break in the snowy weather so I took the opportunity to travel to Matlock to listen to the prototype for about two hours.

John knows I am a CW enthusiast so we first had a listen on CW. The fidelity of the audio and the way the CW seemed to stand out above the noise floor was immediately apparent. There was a station at about S3 level and an S9 plus signal about 300 Hz away. The linearity was such that the 'ears' could still lock on to the S3 signal in the presence of the 9+ signal. Bringing in the DSP filter bandwidth removed the S9+ signal completely.

There is a band scope that works in real time covering a bandwidth of 25KHz around the on tune frequency. It can see a blip from a 0.1uV signal and has a dynamic range of 110 dB. We used the band scope to watch J38XX working splits. The bandscope appears on the RX LCD display but you can connect a lap top to have a larger display (the lap top uses an optical link to connect to the receiver). It is intended to implement another mode for the band scope so you can scan the complete band to look for signals when you are waiting for a band to open.

We left the lap top band scope connected whilst we listened to SSB signals and the other modes. Again in all modes the superb fidelity was immediately apparent and to call the subjective performance of this radio 'fabulous' would be an understatement. It would be nice if we knew for certain why it sounds so good.

All I can say is that NF is around 11dB (12dB max) and the IP3 of the radio for signals within the roofing filter bandwidth is +21dBm and +36dBm at 10 KHz spacing. Also those figures were obtained without adjusting the first H mode mixer for optimum balance

The Photo shows a view of the radio from the back. The front end analog board is on the left with the digital board on the right which also connects to the front panel board. As with the AR7030 receiver the photo shows the excellent constructional techniques used in these designs.

The AR7070 when in production will be in a class of its own for an upconversion radio. It is probably the first super-linear upconversion receiver

Roofing Filters In Up-conversion Short Wave Receivers, 13-05-2010

Martein's development work on his 9MHz crystal roofing filters using the high quality crystals made for him by the German firm Quarz Technik (QT) has implications for up-conversion radios.

Martein's measurements of IP3 on his 9MHz 500Hz and 2.5KHz bandwidth roofing filters showed conclusively that for a given quality of quartz the narrower filter had a worse inband IP3 at 100Hz test tone spacing compared to the 2.5KHz filter at 100 Hz test tone spacing. In fact it appeared to degrade at 6dB per octave of bandwidth for the narrower filter design. This had nothing to do with the close in 100Hz test tone spacing as the 2.5KHz wide filter was 12dB better for the 100 Hz test tone spacing

In view of this, there doesn't seem to be much point in high end rigs made in Japan offering 3KHz wide roofing filters as tests for linearity have shown that a 45MHz fundamental mode crystal filter of 15KHz bandwidth is likely to have superior in band and out of band IP3 compared to a 3KHz wide filter. As a result it is difficult to know what the Japanese designers are trying to achieve by using a narrower filter. If the 3 KHz bandwidth filter is used, CW signals within that filter bandwidth will have worse IP3 compared to whether a 15 KHz wide filter is used.

In the Japanese rigs the problem may be that the processing of the filter in band signal in following circuitry may compromise a good in band linearity of the wider roofing filters.

In the AOR UK AR7070 receiver the linearity of the circuitry following the 15 KHz wide roofing filters gives a receiver in band IP3 of 21 dBm at 1KHz test tone spacing. Given Martein's measurents on his 9MHz crystal filters an IP3 of 21 dBm should also be achieved at 100 Hz spacing. If that is not achieved the limitation will be in the IP3 performance of the 24Bit audio ADC.

Status of the AOR UK AR7070 HF Receiver, 21-07-2010

Those of you who have been keeping an eye on this project will know that RIMARTON Ltd, the company behind AOR UK ceased trading a few weeks ago. This left the production of the AR7070 receiver in a precarious position.

The good news is that a company prepared to provide the funds, premises and of long standing experience in the RF communications world has stepped forward. It is hoped to manufacture a pre production run around September this year, followed by the first production run in early 2011.

It is still possible that the receiver will be marketed under the AOR brand name although like the AR7030 it will be manufactured in the UK.

John Thorpe, the receiver's designer, is still working on some of the firmware for the pre-production units. However he hopes to do an extensive series of technical measurements shortly on the prototype. It is hoped to present these results on Martein's web site.

I am sure you will realise that Martein is interested in state of the art receivers particularly those based on the concepts behind the design of the CDG2000. He would not normally publish details of what will become a commercial product. He and I have both contributed to the design of the AR7070 and it's detailed architecture is unlike any other commercial up conversion radio. As a result I hope he will continue to allow the technical details of this receiver to appear on his web site until the AR7070 is in full production!

Status of the AOR UK AR7070 HF Receiver, 14-10-2010

Extensive measurements of the AR7070 2nd prototype have come available. For all the details see the link below!

Colin, G3SBI


AR7070 - Prototype2 Measurements


Back to Up Conversion Frontend by G3SBI

Back to the TOC