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

[Tom WA3KLR]

Yep, not far from being just another video game.

[Steve - WB3HUZ]

LOL. I've always found it amusing the apparent aversion some have to using the term Hertz, but regularly use Volt and Amp(ere).

You have to special order it, but you can get one with a 455 kc input. I have one awaiting assembly (yea, I'm lazy and scroteless). If you plan to hook it up to an older receiver, you must order a 455 kc input. If you plan to use it with a newer receiver, you must order a 455 kHz input.

   I'm slow today, just caught that one. I used to always type kc, now I'm using khz! Oh no, I'm modern! 

Mack

[WA1GFZ]

It's just a cycle you are going through Mack

[WA1GFZ]

Mack, You don't need to go crazy to get started invest $15 on a softrock and use your stock sound card. FPGA is a logic array that you can program to do a function. Think of it as a board full of logic chips that you program rather than wire wrap the logic.
The flex guys will move on to the real cash cow the USG after they get the hams to work out all their bugs for them. They keep a close eye on HPSDR for ideas to mine. HPSDR is where the real brains hang out. I'm impressed by some of the guys who do hardware and software equally well.

[WA1GFZ]

Mack,
I used a color burst crystal in my set up and it gives me a 7.625 KHz IF low side convert. (divided by I'm thinking of using my old Novice crystal 3.725.
I also changed the input transformer design to a two hole type 77 bead. Tony uses a low perm core with lots of turns which adds a lot of leakage inductance.
I have the part number at work. I can help you with the transformer if it is a problem. Look back through the old postings on the BB. A number of guys are using these boards as IF converters.

[WA1GFZ]

You want a bit of an offset because there is a pile of noise at zero IF. The DSP can handle the offset as long as it isn't too far away. Flex uses 11 khz in the SDR1K and I think 10 KHz in the new one. I got a copy of SDR1K schematics. Not much hardware for $1300. There is a bit more in the 5K. They put the sound card function in the box. I suspect a bit of HPSDR copy there but used fire wire to move data since it is faster than usb2. External sound card made their phone ring off the hook with interface problems with users. They actually started a BB to address issues. Must have made them crazy.
My set up is 7.625 KHz and it allows me to run up to about 14.5 KHz BW. I may change to my old novice crystal in the future to move it further away.
I'm sure Flex would love a nice fat contract from the USG. I still think they need a better preselector based on my testing. I could imagine it on a ship full of RF emitters saturating it. Heck the boston buzzie does a good job on the software.

[VE7 Kilohertz]

Hi Mack,

Well at least you have heard from Tony. I haven't heard a thing from him or anyone else on where/how to order. Could you please send me his contact info?

I want to get something here by Christmas. I should have some colorburst xstals around here.  3.5759XXX right?

Cheers

Paul

[WA1GFZ]

That the crystal i used. I also changed the input transformer design for better operation at 455 KHz. I also changed the input transformer to three turns #30 trifilar on a Fair-Rite 2873002402 2 hole core. I would get the 160 meter kit and modify it. The stock transformer has lower perm material and lots of turns of wire. This could effect phase shift when looking for best performance. fc

[WA1GFZ]

The filters seem ok in the flex software but dealing with high noise another issue. Gates bloat wear is the big issue. I2PHD is a c program and works with noise better. 

[Steve - WB3HUZ]

I2PHD is a c program and works with noise better.

How does the fact that it's a C program affect its ability to deal with noise? It still has to run on some operating system. What is it?

[WA1GFZ]

Yes but it doesn't need 60 meg of gates bloat code to run like Flex does.

[Steve - WB3HUZ]

I2PHD is a c program and works with noise better.

How does the fact that it's a C program affect its ability to deal with noise? It still has to run on some operating system. What is it?

I thought C was just a programmer's tool to compile lines of code into a working program for any OS?

OK. So my question still remains.

The filters seem ok in the flex software but dealing with high noise another issue. Gates bloat wear is the big issue. I2PHD is a c program and works with noise better. 

Gates 'bloat ware' is looking like the correct description of a limiting factor in all this.

So are you claiming Windows creates noise? Or just that it slows down processing/processes that deal with noise? How does I2PHD get around either or both of these if it is running on Windows?

Using less RAM does not address noise issues.

[WA1GFZ]

No windows does not create noise flex software does not handle it as well as I2phd program.

[Steve - WB3HUZ]

Then that's a problem with the Flex application, not the OS it's running on (in this case Windows).

[WA1GFZ]

Yes, When you download Flex software the first thing you have to do is go to MS and download another 60 meg of bloat to get it running. With all that bloat you need a hot rod computer to keep up with all the ones and zeros going by after each input sample.

[Steve - WB3HUZ]

Given all that, how does I2PHD deal with noise in ways the Flex doesn't. In other words, if I have a hot rod computer, will I still see less performance with Flex as compared to I2PHD?

[Steve - WB3HUZ]

If you are just messing with the soft rock (which is what I was planning to do) you can DL the Flex software at no charge. That's why I'm trying to get some info on whether it's worth my while or not.

I agree, Window out of the box runs a ton of stuff most people don't need. Some even negatively impacts security. That said, judicious use of control panels (e.g. turn off some of the GUI "eye candy"), and msconfig to shut off unneeded services and control what happens at boot, and you can improve the performance of any Windows machine. There's lots of good guidance at site like Tech Republic, Black Viper, TweakXP.com and Windows Annoyances. Only you can decide if it's worth the effort though.

I use a system at work that runs under Windows on a laptop. It is required to do much "heavy lifting" as well as control some hardware. Almost all the services are turned off on Windows on this system. I should ask the developer sometime how much difference this makes.

The FlexRadio SDR software seems like it is capable of much more in it's present form but appears to be 'held back' by the environment of being used by most on the 'average' Windows OS PC, just not enough hardware & software resources are left available by Windows for the program to really perform to it's highest potential. I think it's a purely commercial reason that FlexRadio is a Windows OS based piece of gear, it's made for mass marketing and compatibility with what's 'out there' now.

I completely agree. I wonder if they could have done their SW dev so that they could also have released Linux/Unix versions. My guess is they would have gotten for more open source input this way.

It just appears to me that a UNIX/Linux version of the FlexRadio SDR software would be a real killer. Or using the Windows OS version on a much more powerful PC would accomplish the same thing.

That's why I use a Unix machine.

[VE7 Kilohertz]

Thread hijack warning!!!!

Can you imagine how an IBM PC would run Windoze 1 on todays 3GHz machines....oh boy.

If you were to compare the computer industry (Steve Jobs vs Bill (I wanna kickhiminthenuts Gates) to the auto industry, it would be like FORD saying, well our engines are going to be 2 times more powerful, so we are going to need twice as much octane. Chevron, can you make twice as much octane? Oh yes, we can, but it will cost twice as much. Oh, no problem. The price of FORD steel is dropping and we can drop our prices....if we sell enough of them, and need enough high octane fuel, can you drop your prices too?  Oh sure. What is your next engine going to look like? Will it need higher octane fuel? We should start developing a new fuel to meet your requirements. Yes, please do. It will need at least 2 times more octane. Oh, and BTW, our old engines don't seem to run on your new fuel. Seems it causes pre-ignition and our engines keep crashing. Can you add something to make the old engines work again........SORRY....BUY A NEW ENGINE....  (hey Steve, here's your $50M kickback for this year..thanks Bill.......keep 'em beggin for more)

Ohhhhh..again..sorry, soapbax mode off..

Paul

[WA1GFZ]

Guys the Flex spectrum display rocks it is very accurate. The S meter tracks an HP8640D attenuator. It is well worth doing. I have observed I2phd software doing a better job than the stock Racal sync detector when conditions are poor. When I got the HPSDR interface I was forced to change over to Flex Software. One night under poor conditions Audio was coming out of the same Racal RX while nothing was coming out of the SDR set up. I was very surprised and reported my findings to Alberto. All he told me was he was not impressed with Flex Software. My Motorola KISS converter with a  few lines of simple code provided by WA1SOV and hacked by be is still the best but it was a real pain to change operating conditions because you had to reload the program. No GUI. The more crap you have to wade through between samples the faster you need to go. bits is bits.
But as i said the spectrum display rocks even with a stock sound card.. 

[Steve - WB3HUZ]

Frank it would seem that synchronous AM detection is an area where SDR would be hard to beat but since the FlexRadio people seem to have left out a lot in the area of AM processing completely, understandable for their marketing strategy.

While Flex isn't playing up AM in their marketing, they have been working with several AMers, including Bill, W3DUQ. Improvements in both the transmit and receive sides for AM have been made. And the Flex software does include synchronous detection for AM. Bill/DUQ seems quite happy with it and he's been using sync detectors since he built one back in the 60's. You may want to talk to him.

The ability to split the USB & LSB into seperate bit streams should give the software some great advantages in processing and comparing them.

Indeed. In a perfect world, you would chop up the spectrum of both sidebands and select segments from each based on which yields the best SNR. Bacon, WA3WDR conceptualized and/or built something like this many years ago. Seems like a very cool idea to me.

Frank, looking around at the standard 'off the shelf' variety of PC available for this is starting to look like a waste of time for me.

I've found that too. I wanted to buy a low priced unit dedicated to the SDR stuff. The problem with these low priced unit is that the sound card is integrated on the motherboard. It's hard to determine the capability of these type of sound cards and how well they would work for SDR. Anyone have any experience?

The FlexRadio SDR GUI does look cool and I can see how it's S meter functin probably offers a ham the truest signal strength indicator that he's ever had available.

Probably true but still a rather meaningless number on HF. What is important is SNR.

[WA1GFZ]

Flex has a wide capture range while I2PHD is quite small. Neither one is perfect. I would think the capture range could be modified to small range after a lock is established producing the cleanest audio. I2PHD sounds cleaner.
A stock sound card will give you a noise floor around -125 DBM well suited for a IF converter and low hf band operation. HPSDR is better than -140 dBM at 500 Hz. if you set it up properly. I'm running around -130 dBM due to long poor audio cables between RX and HPSDR set up.
I can't open ity up till I clean up my set up.
A stock card sound card is very usable as an IF converter. You need at least a 1 GHz machine with 512 meg of ram to get in the game.
I have also had a few emails with Flex software guys. I tell them real test results that they don't really want to hear.
I2PHD claims you need at least a 400 MHz machine to run his stuff.
I ran I2PHD on my 1 Gig Dell for a year with a stock built in sound card and got plenty of use out of it.

[VE7 Kilohertz]

I'm going to use my 2.4GHZ puter as described above, or somewhere else on the board, with 1G RAM and I am looking for an M Audio sound card, either Delta 44 or 66.

I have emailed Tony and am awaiting reply for the SDR kits.

Going to start looking at my HF-2050 manual to see if there is anything I can play with on the radio.

I also have boxes of DSP and misc. kits and boards from ADI, MOT and TI and am going to get a list together to see if there is anything usable in there. These are full blown board level DSP EVAL boards and may be usable with what we are doing. Frank, you mentioned hacking one of these in another thread.  I have found 2 of the boxes so far but there is more. I know I have some of the ADI 985X series of DDS's too. Cool stuff.

Cheers

Paul

[WA1GFZ]

Paul,
That should work fine for you. sounds like Tony missed your email. fc

[KF1Z]

Trying to buy a softrock kit???

just use the price list, and send tony the money...paypal is the quickest...

announcements, availability, and prices at:

http://softrockradio.org/node/1

[WA1GFZ]

Every now and then Tony runs out of stock and there is a 3 or so week void where orders are stacked up. He will let you know if there is an issue.
Alberto sent me his latest Beta software to test....man I hope he included the hpsdr interface. He did modify the sync detector.