Premium Rcvrs, FlexRadio, HPSDR, SoftRock, SDR, DSP, PCs, OSs, etc.!

[WA1GFZ]

If I was to start over I would go with QS1R and cut all ties to Flex.
HPSDR is well monitored by Flex to mine new products. It is a love hate because everyone used flex software. The Flex people fed me some performance BS years ago that left a bad taste.

[Steve - WB3HUZ]

Souncard based SDR is old technology, and is limited.... although interesting, and inexpensive.
(provided you already have a GOOD souncard, and relatively fast computer)

I'd say it's different technology, not necessarily old. The upside of using a soundcard and PC to do the A/D and DSP is that you get new technology at the rate of the PC industry - every 12 months or so. You are not likely to get that sort up technology update rate when your SDR is based on a proprietetary or specialized A/D front end and specialized DSP. Additionally, there are far less people available to write software for such hardware, as opposed to the more generalized PC hardware.

So, as in any engineering exercise, there are tradeoffs. New, old, better, worse and similar terms don't always fully describe the situation. A look at the over all system and the various tradeoffs relative to the end user's requirements are what is really important.

I'm wondering why no one has harnessed the incredible processing power of the PC video cards for SDR work. Although these cards are designed for video, many have high speed A/D, vast and fast memory and incredible processing power (the high end cards are almost super computers in there own right). Maybe it's just too hard to make them work with non-video signals.

[KF1Z]

Souncard based SDR is old technology, and is limited.... although interesting, and inexpensive.
(provided you already have a GOOD souncard, and relatively fast computer)

I'd say it's different technology, not necessarily old. The upside of using a soundcard and PC to do the A/D and DSP is that you get new technology at the rate of the PC industry - every 12 months or so. You are not likely to get that sort up technology update rate when your SDR is based on a proprietetary or specialized A/D front end and specialized DSP. Additionally, there are far less people available to write software for such hardware, as opposed to the more generalized PC hardware.

So, as in any engineering exercise, there are tradeoffs. New, old, better, worse and similar terms don't always fully describe the situation. A look at the over all system and the various tradeoffs relative to the end user's requirements are what is really important.

I'm wondering why no one has harnessed the incredible processing power of the PC video cards for SDR work. Although these cards are designed for video, many have high speed A/D, vast and fast memory and incredible processing power (the high end cards are almost super computers in there own right). Maybe it's just too hard to make them work with non-video signals.

Ok, I'll go along with some of that....

Though as far as SDR in general is concerned, soundcard based DSP is considered to be an older generation than direct sampling A/D. (such as HPSDR, QS1R etc).

The "older" generations still have their place, and aren't going away anytime soon however.(For the exact reasons you mention).

Performance wise though... they just can't keep up with the newer generations.




As far as the PC video cards go....
My assumption would be the newer cards are all proprietary code (in firmware on board), and not customizable. At least not easily enough to be easily accesable to the masses.

[KF1Z]

But now... with my Icom R-75 on ebay (last analog RX I own)......

Should have bought this when you offered it to me months ago.

I may end up with a QS1R yet.. also leaning towards HPSDR but I dunno.

It's still on ebay!
The prices have dropped on them recently, I'm already going to take a good hit on the price I get for it...
Oh well.... that's used equipment for ya!

[N1ESE]

As far as the PC video cards go....
My assumption would be the newer cards are all proprietary code (in firmware on board), and not customizable. At least not easily enough to be easily accesable to the masses.

Actually, NVIDIA and ATI are both embracing distributive computing.  ATI is working closely with some of the distributive computing folks who are working on protein folding.  I have a high-end gaming card on one of my boxen that protein folds via the GPU when idle. 

http://folding.stanford.edu/

- JT

[Steve - WB3HUZ]

Performance wise though... they just can't keep up with the newer generations.

I rather doubt a direct sampled system will outperform a soundcard based system with a good front-end/converter, especially at HF.

[KF1Z]

Performance wise though... they just can't keep up with the newer generations.

I rather doubt a direct sampled system will outperform a soundcard based system with a good front-end/converter, especially at HF.

Well, doubt away.....

But I know I'll never go back!

[Steve - WB3HUZ]

Back to what? Are you saying you previously had a soundcard based system with a good front-end?

[KF1Z]

Back to what? Are you saying you previously had a soundcard based system with a good front-end?

Yes,
Of course, this all depends on how YOU define "good front-end"

Pre-amp, BPF, mixer, QSD...?

Obviously, depending on how each of us define "good", we'll come up with a different answer.

====

I might also add that when I talk about direct sampling, I of course mean digitizing either "at the antenna", OR after BPF and pre-amp. (or attenuator)....
But NO mixing or detecting BEFORE digitizing.

[WA1GFZ]

Steve,
I'll let you know when Mercury shows up. I do have a good front end ahead of a QSD. Still when conditions are bad my stock RA6830 digs signals out of the noise better than the QSD plus software. We will see what the future brings.

[AB2EZ]

I finally got around to building a Softrock receiver.

Mine is a 455 kHz Softrock Lite V6.2, and I'm using it in conjunction with Power SDR software.

I highly recommend these little units as a very-low-cost way to get into i.f. - based SDR.

What is particularly fun, for me, is using the Power SDR application to demodulate (AM, synchronous AM, etc) from the I and Q outputs of the Softrock receiver.

You do, of course, need a fine-tip soldering pencil, steady hands, and decent eyesight and/or a microscope to assemble the tiny board and its surface mount capacitors and IC's. Also, the toroidal inductor and the toroidal transformer are a bit of a challenge (e.g., 89 turns of #30 wire on a core about the size of a Cheerio).

For those who prefer not to build the kit themselves... there are kit assembly services available at very low cost.

Contact

Tony Parks, KB9YIG
1344 E 750 N
Springport, IN 47386


Best regards
Stu

[W1VD]

Stu

Easier to wind inductor and transformer from a previous thread...

http://amfone.net/Amforum/index.php?topic=13686.100

[AB2EZ]

Jay

Thanks....

Stu

[AB2EZ]

So here are a couple of "provocative" questions:

Consider the case in which you are using the 455kHz i.f. output of your receiver... after the i.f. filter...

Do you really need to include a bandpass filter (a toroidal inductor in series with a capacitor + toroidal transformer with a capacitor across the primary) on the Softrock-lite board?

Putting aside the unbeatable bargain price of the Softrock kit... is there an "unmet market need" for an old buzzard's version of the kit that uses regular (non-surface mount) components... and which would work just fine at 455 kHz 

Best regards
Stu

[KF1Z]

So here are a couple of "provocative" questions:

Consider the case in which you are using the 455kHz i.f. output of your receiver... after the i.f. filter...

Do you really need to include a bandpass filter (a toroidal inductor in series with a capacitor + toroidal transformer with a capacitor across the primary) on the Softrock-lite board?

Putting aside the unbeatable bargain price of the Softrock kit... is there an "unmet market need" for an old buzzard's version of the kit that uses regular (non-surface mount) components... and which would work just fine at 455 kHz 

Best regards
Stu

Could well be that you don't need a BPF there Stu.....
Remember it was designed as a stand-alone RX... Just happens to have found an interesting use as a panadapter/ IF adapter........

The transformer may be a different story, but there's probably a way around that too....



There are some kits out there that are basically the same design that are non-surface mount jobbies.
(but the softrock is still the cheapest and most basic design)


I've found that most people find the toroid winding more of a PIA than the SMDs though.

But, ya never know....
I've built about 26 of those now, of different flavors for folks that didn't want to try it...
Either because of the SMDs, the toroids... or just plain lack of soldering skills/equipment.

I have a board layout *somewhere* on a disk that I laid out about 4 or 5 years ago using all LARGE parts.... but never did anything with it.

[Steve - WB3HUZ]

I would say not, just an impedance matching network. Caveat: I haven't actually tried this, but given the amount of filtering in most RX IF's, it should work.

So here are a couple of "provocative" questions:

Consider the case in which you are using the 455kHz i.f. output of your receiver... after the i.f. filter...

Do you really need to include a bandpass filter (a toroidal inductor in series with a capacitor + toroidal transformer with a capacitor across the primary) on the Softrock-lite board?

Putting aside the unbeatable bargain price of the Softrock kit... is there an "unmet market need" for an old buzzard's version of the kit that uses regular (non-surface mount) components... and which would work just fine at 455 kHz 

Best regards
Stu

[KF1Z]

The folks writing the software for the QS1R, have switched to C++, and using openGL for the display...

We now have FULL dual recieve... (It is still a test run..)

But really cool.... and it didn't cost an extra $600.

[WA1GFZ]

man I'm chomping at the bit to jam a pair of 8640Bs into the software and see what happens. Even with my new computer running at 10% horsepower the stock RA 6830 digs signals out of the 160M noise better than SDR.
Displays are cool but is there real performance?
Mercury has been in test a month with no reports yet.

[KF1Z]

I think it does very well as is Frank....
But, I don't have any other RX (especially high-end stuff) to compare it to.

I still believe the band-pass filter/ preamp board will make a big difference, but probably mostly on 40 meters and up.

I haven't been able to squeeze out of Phil when that's coming along.....
He had said about 4 months ago, that it would be "6 to 8 weeks.."
But we haven't heard anything since...

The computer horsepower requirements just took a big dive....
But you need a good graphics card to take the slack.

Most of the CPU load has been in the display... now going to openGL....

With the above screenshot, my CPU load was 20%.

AMD x62 dual proc.   2.7 ghz w/2gig ram.

[W3RSW]

Yeah, I'm waiting for a few more GUI and server variants to gell before springing for an updated computer.
Thinking a Mac if native code will eventually run Phils stuff. Computer selection might depend on what video cards can do with musoft. Tired of waiting for bloated musoft and all the TSR's to load everytime I turn the damn thing on.

Downloaded Phil's ver. 98.  Will be very nice when the dual rcv' stuff, etc. is incorporated in MaxII.

Operating:
Running the QS1R from the TR-7s, aux. rcvr. output after severing the '7's internal jumper between, Aux. rcv'r and ext. antenna.  Running the QS1R's hardware mute from the '7's T/R relay - ext. amp. trigger line via the unused aux. RCA jack on back.  Anyway using the TR-7s T/R circuitry gets you a built-in low pass filter before the QS1R.  Built up a BC blocking H P filter (simple 7 pole. using T-50-2 toroids and micas. )

It now seems so archaic when you can't see what's going on in the whole band, instantly go to strongest signals, see what type of modulation each sig. has before switching, tracking swishers, warbling psuedo signals, etc. Some cool stuff outside the hambands too.

Over the years I've been pretty well convinced that content has it over architecture, very similar to TV where programming still (sometimes barely) beats the HDTV delivery.  But architecture is coming on strong.  Way fun and very addictive, this SDR stuff.  Hard not to enjoy it just for SDR's sake.

After Phil and Cathy's code's incorporates most of the features we want along with a companion TX, this is going to be one hot SDR.  I too wish Phil would posit a progress time line or two. Check out the Yahoo board for latest. http://groups.yahoo.com/group/qs1r/

[W3RSW]

Oh, meant to add this message from Phil Covington.

*****************************************************************
"Hello all,

We are experimenting with dual receive capabilities in QS1R as can
been seen in Cathy's latest 2 panadapter GUI test. You can download
it from the files section of the QS1R Group. Dual receive is also
working in the new C++/Qt server as well as FMW mode which sounds very
nice.

So, unlike other SDRs and IF SDR radios where adding a second RX costs
$600+ or the initial price of the dual RX capable radio is >$5000, you
will be getting two independent (in mode, frequency (0-60 MHz),
settings, etc...) receiver capability in QS1R for an additional cost
of $0.

If we can figure out how to deal with all the receivers, two receivers
is certainly not the limit for future versions.

Regards,
--
Phil Covington
Software Radio Laboratory LLC
Columbus, Ohio
http://www.srl-llc.com "
**********************************************************

He means two or more concurrently running receivers. Already has instant selection of up to six instantly selectable freq's, any band, for working cross band, split, etc.

[KF1Z]

Yeah, I'm waiting for a few more GUI and server variants to gell before springing for an updated computer.
Thinking a Mac if native code will eventually run Phils stuff. Computer selection might depend on what video cards can do with musoft. Tired of waiting for bloated musoft and all the TSR's to load everytime I turn the damn thing on.

What waiting?

30 seconds after I hit the power switch, mine is ready to go (win XP).

Less than a minute to be listening to the RX.

Settings and file system cleaning can go a long way....

 

[WA1GFZ]

All this stuff is very cool but I prefer to see real test data. We will get there in time but as one who has had many RX design failures you can't improve unless you have real performance data.
The final test is when you have two receivers hanging off a splitter on the same antenna trying to dig out the tuff ones. So far SDR loses that race in my shack. Also having both configurations I get to enjoy the best of both.
The processing delay is a trip though. I listen to SDR audio when conditions are good but go back to analog when the conditions are poor.
 

[KF1Z]

Probably all very true Frank....

Tough having nothing to compare to.

I do know the audio quality was far better than anything else I've had... but there again, I've never owned any high-end gear... so probably not saying much.


If I knew what tests I could do with the scant equipment I have... I'd do it....


So far, I have at least been able to hear pretty much everyone in any given group that everyone else was hearing too...
Sometimes even been the only one that could hear certain people...

But of course, locations, conditions,  antennas etc are most likely the big part there.


I do not care to "try to dig anyone out.." unless it's someone coming into a qso...
I do not chase weak-signals intentionally.

So "Real performance" may not matter to me...
It's the audio quality , and noise elimination/filtering qualities that I get a kick out of.

[W3RSW]

Bruce, Frank,
ok on your comments and stuff.  Also we have to remember that Phil's board is general purpose.  It's being developed for ham rigs, test instruments, very broad band, low noise radio astronomy, moon-bounce, you name it.

Like Bruce, I'm betting that once any pre-amp, filtering, etc. ahead of the DDR is utilized, this will be one hot rig. - even after a -3db splitter to another single channel-restricted receiver. heh, heh.  Yes, comparison test specs are essential for many reasons but if it can hear 99% of what everything eles hears without filters, preamps or attenuators, and do this in a 60 Mhz broadband,
well !

Dumb as this might sound; By the same token if you don't have a QS1R to compare with other receivers then you can only guess on that side too.  Specs. should be forthcoming. 

Oh yeah, my computer.  - File clearing and stuff still doesn't help a 1.7Gig, low mem. computer loaded with (like Derb says) wifey's programs. Gotta get a new one, dedicated to me. 

I've been thinking.. uh,oh.  There's so much QS1R open code out there now along with several boards that Phil's board is not going to go away.  Up to him, his biz model, etc. to see how fast he wants to ramp up, not get bought out prematurely, what application gets developed, gets utilized first, etc.
I'm sold and just think it's really exciting.