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

[w3jn]

Naw, I got back Wed nite.  A FB 26 hour trip including the 4-1/2 hour drive from JFK.  Considering all that was going on I didn't have time to do much of anything other than the family thing.

[WA1GFZ]

Oh well you have to pay for that kitchen pass once and a while.

Got to drive a Cubic CDR 3280 today cool radio 1/2 the size of a 6830. Didn't have a good generator to do tests but looks like a nice performer.

[WA1GFZ]

We need a full report. I wish I could have put on my torture bench.

[WA1GFZ]

Please don't send it to me. I want to buy hardwood floors for the new QTH not another radio and it might suck me in. R3030 has this cool meter function that centers the carrier in the bandpass. The little display on the right side of the 3280 display duplicates that function. I can give you some simple tests that tell a lot about a radio. All you need is a crystal oscillator and step attenuator.
The S meter starts function above -115 dBm and goes up to -17 dBm. I had the S meter pegged and tuned off center frequency to listen to phase noise. Based on a crude test it looked similar to the RA6830 but filters had better ultimate rejection close in. I don't think it outperformed my hot rod but it had many more choices for bandwidth. It had no problem handling over 100 dB of input range.

[Mike/W8BAC]

Hay Frank,

Would running the 3038 through it's paces be of any interest to you? I could forward it to you for the bench torture test. It would be interesting to see the true specs. You'll need a 600 ohm line amp to listen. I guess that's standard.

Mike

[WA1GFZ]

Mike,
I wonder if you could send me a list of the module part numbers in that radio.
They are on top of the modules. Also what filters does it have. That should make a nice diversity radio. I have a R2307 mil R3030 and a R3090 VLF radio with HF modules installed so it is a carbon copy of the R2307 except for the name tag.
I'll convert it to VLF if I ever find the missing modules. They are pretty easy to repair so if one side works different from the other you can swap modules to find the problem. I can help you if you ever have a problem.
I'm sure it performs as well as the 3030 and may have the same modules in the front end. I will be glad to share the test data I have. Also have some 3030 mods to reduce background noise in the audio if you have the same detectors.
The 3030 is very sensitive maybe 6 to 8 dB better than the Racal since it has an RF stage preselector. A bit more close in phase noise than the RA6830 though.
Too much risk / bucks to ship the whole radio but I can do repairs on modules. 

[WA1GFZ]

Yup, That fan moves some air but the package is so small I would be careful turning it off. I have a couple 6830s with fans and they are a lot quieter in a rack.
Filters. I noticed the same thing with my TCI/BR 8174 radio. Roofing filters were kind of wide. I suspect this was done for digital modes. I would put a 16 KHz filter in the 456, Heck a 455 KHz filter will work. The filter skirts in the 3280 are quite nice you would need cascaded filters in an analog radio to match that performance. I was able to hear the effects of the wide roofing filters by pumping a strong signal into the input to hear the blow by. Also I would find something tighter for the first IF. Bummer these guys shifted the IF by 1 KHz. The advantage of my homebrew SDR is I can select any filter I want ahead of the QSD. Still the 3280 is quite cool and I got used to the keys quickly. Hope you can get a service manual. Yes, give them a call they never answer email. fc

[WA1GFZ]

Dallas sent me his article a couple weeks ago but I agreed not to post it.
Notice the A/D is saturated at -16 dBM 

[WA1GFZ]

Mack,
That serial stream is no big deal to a good software guy. All it is is raw A/D serial data. Demodulation is demodulation and filters are filters. I would think by '98 that would have been sorted out. I think Cubic has been doing DSP demodulation since the late '80s. The options I saw made me think the only thing I might want is a spectrum display. it would be no big deal to strobe that serial data into shift registers and drive a pair of D/As to shoot it into a sound card. A smart software guy could take it into a fast parallel port. Or another option is pick off the 456 kHz and send it through a softrock and do the same thing I'm doing. I bet all you need is a cable to connect it internally to the rear panel. My guess a DB insert at one end and a BNC at the other. I had to peg the S meter at -17 dBM input just to hear a couple spurs as I tuned off frequency. Just use it and enjoy your cool new toy.

[WA1GFZ]

Looked at the 3280 manual and convinced all you need is a DB coax insert at one end of a cable and a BNC at the other to get 456 KHz out the back. Last time I looked a DB coax connector with 5 inserts was $25 at Newark. I might even have some in the cubic stash.

[WA1GFZ]

The HPSDR Mercury is supposed to be the same radio but interfaced through ozy. Phil's radio has everything on one board. Function should be the same. $1k by the time your done getting it into a chassis. Not sure I want to jump ship from HPSDR after investing in the hardware. If I was to start over I might go after Phil's board. Both use the same A/D. Phil designed both of them.

[WU2D]

You guys seen this new Ten Tec? We are looking at this at work.

http://radio.tentec.com/Commercial/RX400

We have been looking at alternatives to our sister companies stuff.

Mike WU2D

[WA1GFZ]

RX 400 site says IP3 of 12 dBM that kind of sucks. I'm shooting for better than 30 as a base line for the future.
I see the trend for GUI toys replacing real performance. I want both or I will stick with performance and use the GUI toys as an add on.
Mack, I guess I will stick with HPSDR  and wait for Mercury.  Phil claims his software will work with HPSDR so I'll give it a try when he releases it. Only because I started there and he designed both.

So how is that 3280? Ready to swap it for a nice analog radio?

[WA1GFZ]

Mack,
I'm not sure but new releases of FPGA code will need to be downloaded directly into the prom he stores the code in. New FPGAs are like RAM  so code is loaded when power is turned on. This is different than the program in the computer that communicates with the board and processes data. I have not checked his site in a few days to see if he has posted his code. 

[WA1GFZ]

TAPR is taking orders on the next module the 1/2 watt exciter. Spectrum plots show third order crud down 50 dB. It is supposed to be all mode.

[WA1GFZ]

They come assembled and tested.

[WU2D]

Got a couple of these cuties at work: 

http://www2.rohde-schwarz.com/en/products/radiomonitoring/product_categories/receivers/EB200.html

Mike WU2D

[WA1GFZ]

These results are impressive but look closer. The noise floor is up almost 20 dB at that power level and he is doing the test so 30 dB above the noise floor.
When I test dynamic range I consider spur or noise increase of 3 dB as the trip point. I consider these results impressive but a bit inflated.

[WA1GFZ]

Phil sent me a screen shot of 2 signals closely spaced that also showed the noise floor increase and claimed it was his generator.
I reference everything to the noise floor and never make claims unless I can demonstrate performance. Yup you saturate an A/D you are done so don't know how he gets around the problem.
Maybe the DDC deletes saturated samples???

[Steve - WB3HUZ]

They've been touting that thing for over three years now. Is it actually something real now, as opposed to the vaporous item it has been?

You guys seen this new Ten Tec? We are looking at this at work.

http://radio.tentec.com/Commercial/RX400

We have been looking at alternatives to our sister companies stuff.

Mike WU2D

[WA1GFZ]

Hey Mack,
Sorry but 40% jump in A/D speed is news.......

[Steve - WB3HUZ]

No, it's not. They had black box receivers, but they were HF only.

I suspect the RX-400 is a rework of their original "black box" rcvr that they did well on selling under Gov. contracts years ago. I recall them bringing that one to the Chattanooga Hamfest many years ago for display only, since the sign said individuals could not purchase them, top open for view inside and nice functional block diagram display. So I look at the TenTec guy and ask how happy the DOD might be over this display? Five minutes later the block diagram was gone and the lid was back on the box! No more detailed descriptions of digital sampling were being offered either! Funny stuff.

While I'm typing here:

Frank/'GFZ, how about you slow down on introducing any more new info to the SDR user groups, you're responsible for a big jump in email volume lately. Like there isn't enough volume there already?

[WA1GFZ]

Phil told me the new A/D will drop right in. Same package. I suspect the oscillator would need to be changed.

[WA1GFZ]

I was involved in a company standards program on PC board design covering the subject because we have way too many experts suffering from cranial rectal inversion. I'm had a number of good PMs on oscillators and I'm sure glad I have a good stash of SAW VCSOs.
My best are a couple 1 GHz units that hit the trash because the welds leaked 3 atoms of hydrogen during a leak test. I wonder why nobody ever thought to solder the seam over the weld....that was what I did.

[WA1GFZ]

It wasn't three but it was low. I just fished them out of the trash and ran a bead across the bad section of weld. I knew the day would come when they would be useful. BTW using the internal PLL of a DDS to multiply clock generates plenty of phase noise as confirmed by a couple of the smart guys so the guy who made the claim was quite wet.