HOW ABOUT COVERING (MONITORING) FM & AERO BANDS

Alex_KM5YT

I would like to see coverage of the FM and AERO bands - even if they had to be excluded from pre-selection for whatever reason.  Both provide information and entertainment when one's preferred ham bands are dead, and in general.

In addition, emergency interoperation with aviation should probably not be excluded.

Alex/KM5YT

KY6LA_Howard

Easy to do Just add a transverter

Steve-N5AC

In a direct sampled system such as the FLEX-6000, the spectrum will be broken into a series of Nyquist zones related to the sampling frequency.  Each zone is the width of the sampling frequency, fs, divided by two.  In the FLEX-6000 the sampling frequency is 245.76MHz and so the spectrum is broken into a series of zones that are 122.88MHz wide as such:

Zone 1: 0 - 122.88MHz
Zone 2: 122.88MHz - 245.76MHz
Zone 3: 245.76MHz - 368.64MHz

and so on...

Frequencies from each zone alias into the other zones, wrapped about the Nyquist frequency if a filter has not been employed to prevent it.  For example, A signal at 98.9MHz will alias to 146.86MHz (122.88 - 98.9 = 146.86 - 146.86).  To prevent this, filters are used to select a single Nyquist zone to pass to the data converter (ADC).  The filters must reduce the signal to a level where the signal will not be present at all when aliased so that it is not visible in the other band.  So we have to filter the broadcast FM band to a level so low that it will never appear when we look at the 2m band.  To achieve this with strong FM transmitters, we use both a band-notch filter on the FM broadcast band and a bandpass filter on the 2m band.  The result is that the FM band is long gone in the FLEX-6000.

If you'd like to understand all this in more detail, here are a few references:

Analog Devices: What the Nyquist Criterion Means to Your Sampled Data System Design 
Texas Instruments: Intermediate Frequency (IF) Sampling Receiver Concepts 
Wikipedia: Nyquist-Shannon Sampling Theorem

Richard McClelland, AA5S

I wonder how far out a sampling frequency of 491.52 Mhz is?  Then our zones can be 245.76 Mhz which should be sufficient for coverage of the FM bands. No rush, I only have been listening to the FM bands in my car on the way to work.

Richard McClelland, AA5S

But why stop there?  We could go for a sampling frequency of 983.04 Mhz and then we can have our zones to be 491.52 Mhz.  10 Khz to 432 Mhz should please even this tough crowd.  Ok, I'll stop.  I promise.

Steve-N5AC

All very possible ... it just gets more expensive the higher you go.

Richard McClelland, AA5S

:-)  Go for it. What's a few dollars between friends?

Alex_KM5YT

Steve.

I am generally familiar with Nyquist.  It was beaten into me in EE, in doing real time A/D sampling in laboratories, etc., etc.  I have looked (not read 100%) at the first two references.

So - are you telling us that the A/D sampling frequency could not be moved to ~270 Mhz to handle this case (what is the published limit for the device?)?  I see the arcana of the "zones" - that's not worth the agony.  If 245.76 MSPS is the sampling limit for the device, it sounds like 2M was a serious strain.  Seems like one should strive to keep all the bands to be handled seriously w/in the primary sampling rate (~2 Fmax), no?

(No transverter - another box - wanted here - yet!)

Alex/KM5YT

Steve-N5AC

The device we are using in the FLEX-6000 (AD9467) comes in two bins: 200Msps and 250Msps so 250Msps is the maximum for the device.  As a general rule of thumb, the devices with lower sampling rates have better performance, but you are then restricted to less bandwidth.  HF is not particularly performance demanding in noise figure, but it is extremely demanding in dynamic range.  As you go up in speed, you have to have digital devices that can ingest more and more data.  For example, at 247.56Mbps each data converter produces 3.9Gbps of data.  At 983.04Mbps, a converter would spit out 15.7Gbps of data.

There are a lot of design choices that occur in the selection of the ADC: will it require a preamp on the front?  If so, can you buy a preamp that will not degrade performance?  How much data will have to be processed and what does it cost to process that data?  What interface is the data converter and what type of FPGA will be required to process that data?  How much do the converter and the FPGA cost together?  How much power and cooling will they take?  What zones and frequencies will they cover?  What is the device's SFDR, where will the spurs produced in the device fall and how will they affect you?  When you select the sampling rate, can you move the products of sampling out of the bands of interest?  Does the converter then perform well at that sampling rate?  and so on...

Selecting other zones is not that big of a deal -- you just pick a filter that isolates the frequencies of interest in that zone.  But then the spectrum is inverted for even zones (tuning up goes down in frequency) and and the converters perform successively worse as you increase the zone because of under sampling and sometimes because of response of the device in the zone of interest.  You also have to write all the software that understands and deals with the different zones.

I guess my point is that there is no "free lunch."  Every design decision has tradeoffs.  If there were no tradeoffs, everyone would gravitate to the "best" data converter.  But as you can see in the market, everyone chooses a different device for their project because there are tradeoffs and you need to make them advisedly.

Alex_KM5YT

OK - transverter now; FM & Aero in the 7000 series (new 'hybrid optical 1000 MSPS device') - and built with a VME card cage to allow field repairs and upgrades, too! Optical Ethernet link.

Ciao

Alex/KM5YT

Richard McClelland, AA5S

Is there a (potential) application on the other side that could use 10 Gbps of data?  I realize that there are 1 Gbps optical links but there wouldn't be any point in having less than 10 Gbps; otherwise, a regular copper Ethernet connection would suffice.  Maybe Flex could put a connector on  the back and let the end users choose whether to go with a 1 Gbps copper SFP or a 10 (or even 40) Gbps fiber SFP?  That would be a nice feature.

Richard McClelland, AA5S

While we're finishing up KM5YT's design, all of the various models, from the 7300 to the 7900 could be built with the same identical hardware (VME cage, power supply, modular Ethernet connections, etc.) and the final model/product number would be determined by which cards, SFPs, etc. are plugged into the chassis.  There has to be enough increase in manufacturing efficiency for the 7900 price point to be about where the 6700 is today, surely I'm not dreaming.

Steve-N5AC

We build systems today that use SFP+.  Our LBRX has four 10Gbps SFP+ connectors for a total of 40Gbps output.  But I'm not convinced that amateurs are going to want to digest this much data nor pay for it (anytime soon).  Our customers that do that today have computers with over 100GB of memory (RAM) and are very expensive.  The last SFP+ modules I bought were $350 each.  Probably not something hams want to spend money on seeing as just to get the data out of the system you'd have to have a couple of those, a router that will do 10Gbps or a card for the computer that will do it, the large computer, etc.

We also have built card-cage type of systems, but they are also very expensive.  There are gobs of little parts, rails, screws, sheet metal parts, etc. in card cage systems.  I just don't think that will ever be a good choice for hams.  In a previous life, I was on a team that built a card-cage repeater controller.  By the time we stopped production, the small card cage was $300 and was a big part of the cost of our system.  That was 20 years ago...

Richard McClelland, AA5S

It's easy for us to say, "Wouldn't it be cool?"  Making it happen is another matter.  Best wishes as you make it happen.

Alex_KM5YT

It seems that the availability of a faster front-end multiplexor - a   t r i v i a l   inevitability - will make this m i n o r    extension entirely possible.  I think we can sit back and wait for the device.  So please stand by for the device and - as they say - 'make it so!'

Alex/KM5YT 

Mike VE3CKO

Howard, can you point me in the right direction. Been googling 88 - 108 mHz transverter with nothing close.

Ken - NM9P

Using a $20 RTL-SDR dongle (has a metal case and SMA connector) and some free software, I have been able to listen to as many a 8-12 airband channels simultaneously on my computer.
addition of a suitable FM Band blocking filter would improve its performance quite a bit.

My old IC-706MK2G also makes a great Airband monitor - better than the AM detector in any of my regular VHF/UHF rigs (IC-208H, IC-207H, T90A, Standard C168, etc) because of its superior NB, AGC, and dynamic range.

While it would be "super-cool" to use the great SSDR panadapter for Airband, or PSB monitoring, I think that, unless I were a deep-pockets government or scientific agency, a redesigned 6000 series would be a cost effective approach.

As others have said, however, building a simple wide-band receive converter may be possible.

In the early 80's I once built a converter on a solderless breadboard to convert airband to an AM radio.  Then I later modified it to convert airband to a CB radio -- all just for fun and education when I was l learning about dual-gate mosfet mixer circuits.  I think the injection oscillator was a tap from my GDO, if memory serves me correctly.

Dan -- KC4GO

and for a bit more look at SDRplay.com SDR Receiver only but works very well for all the way to 2 Gigs.  Several models < $200.00 

Mike VE3CKO

I'm not going to use a dongle or purchase another SDR   I have plenty of FM radios around if I wanted just to tune in and listen. I've got a Flex 6700 and a 5000 wanted to see the FM broadcast spectrum from what I already own so perhaps DX FM stations. A cheap transverter is what I'm looking for.

Alex_KM5YT

The "Not Planned" reply is ~irrelevant and misleading.  The capability to implement this coverage will come with the use of an inevitably faster front end sampling multiplexor.  QED.

Ciao,

Alex/KM5YT

Gerald-K5SDR

Alex, the concise answer is no. Gerald

Alex_KM5YT

Gerald - I hope you understand I am not referring to modding the 6000 series for this capability, but that it would/should be ~trivial for the "7000" series.  So please indicate if this frequency range is going to be avoided for other-than-technical reasons.

Ciao,
Alex/KM5YT

Tim - W4TME

We cannot indicate anything regarding products we have not designed.  

Patrick

After reading this post, an old idea I had been thinking about came back to the surface. How about Flexradio building it’s own transverter that would cover let’s say 74 MHz to 1300 MHz. With the following features to enhance it’s integration with the 6000 series radios. Power amp to back fill what is missing on 2 meters and power for the upper vhf and uhf frequencies. On 6000 series radios provide integrated RX with sampling direct conversion like the base radio. It would integrate with the 6000 series radios via the RF I/O of the radio and communication with the transverter to/from base radio via Ethernet. Transverter would have at least 4 RF connectors for designated bands or BW. A 4 port gigabit hub would be integrated with the transverter to provide connection to the base radio, PGXL, and internet and any future devices like the HF tuner. For me this would be a WOW addition that would give the 6000 series a 50 kHz to 1300 MHz fully integrated capability. The hub would allow stand alone operation when internet com is not possible but allow compatible IP devices without additional equipment.

Just dreaming of course!

Aloha and Mahalo,

Pat

WH6HI