Ordered an sdr-iq...

[K1JJ]

Tom,
The FT102 has a dynamic range of 85 db so S9 plus 30 is close to the level the rX starts generating internal IMD if the S meter is anything close to reality.
Any amplifier will generate IMD if you drive it hard enough.

Interesting observation, Frank.

Well, the FT-102 still has the S meter set generously, so I suspect it is really about 59 +25 when testing. I do see the improvements when making adjustments on the transmitter system, so at least it's working in a relative manner. Besides, when I measure crud it is usually below S9, so that's a good thing considering the RX's dynamic range.

T

[N2DTS]

I took off yesterday as the xyl was getting eye surgury and had some time to try the receiver tuned off to the side tests.
I had the flex into the amp, and got nothing it seems out past the high frequency passband.
Beyond that, I did not see anything on the scope, or get any reading on the s meter.

Its got to have some grunge, but the receiver did not seem to pick it up....


Brett

[w1vtp]

I took off yesterday as the xyl was getting eye surgury and had some time to try the receiver tuned off to the side tests.
I had the flex into the amp, and got nothing it seems out past the high frequency passband.
Beyond that, I did not see anything on the scope, or get any reading on the s meter.

Its got to have some grunge, but the receiver did not seem to pick it up....


Brett

Did your receiver system have greater than 60 dB dynamic range?  The 5000 has a better than -55 dB spurious / harmonic radiation spec.  I guess I need to understand what you meant by "...high frequency passband..."

If you drop the power by say 10 dB (as monitored on your power meter) does the indicated carrier drop by that amount on your receiver display indicating that there are no compression issues?  One would think that the linear would generate some additional grunge.  Or is that you have a really great system / amplifier?

Al

[K1JJ]

Brett,

In general:

Even a clean exciter will have some crud off to the sides, though way down.  Sock a nasty sounding "YAY" into the mic on LSB and you shud hear crud starting after 3kc up the band on LSB (using a ~2.5kc LSB RX), down -35db or so. It shud taper off into the RX noise floor at 5kc away, etc.  Add a linear amp and it will degrade by maybe 5db or so.  At least that's what I've found testing a number of exciters and amps.   I try to get the linear amp to add very little degradation to the exciter. The signal will never be cleaner than the worst component in the chain, of course.

A signal sounds like nice audio tones from zero beat to about + 2.5kc up the band or so  - whatever the bandpass of the TX and RX are.  The CRUD sounds like crackling up the band after that.

If you cannot hear the cracking, then you need to add more signal into the receiver until there is some crud up 5kc to hear on LSB. As Frank said, use the minimum signal to avoid IMD of overloading the RX dynamic range, but still enuff to hear your signal remnants up 5kc.  The remnants may be down -60db at 5kc, but are there, nevertheless.

[N2DTS]

I likely did not have enough gain in the receiver.
I was using the homebrew, with the pre selector mis tuned to lower the signal level into the mixer.
I had 4 divisions on the scope, about S9 on the meter, and tuning away from the signal, I only could see the high frequency stuff (sibilance), and after tuning past that (5Kc), saw nothing at all....

I doubt the amp is super clean, its four 813's!

The sdr-iq will arrive tomorrow, so I will use that for testing.
While I am at it, I should try and measure the dynamic range of the homebrew receivers.

How would I do that?

It seems to have VERY low noise, comparing weak signals on the 756ro 3 I had, the R390a I used to have, and the flex 5000, the homebrew always had readable signals the others could not copy becuase the signals were in the noise. The 5000 is very close, but i think the homebrew still wins.
I have never noticed any distortion even on the strongest signals.
The first stage is the mixer (running low voltage) which has agc on it...

Brett

[WA1GFZ]

Tom,
What are you using for a two tone generator? You want a stable input so you can actually measure the IMD level. I would use a step attenuator and put the RX into tightest filter position. If you know the two tones you will know where the imd is supposed to be. The nice thing about SDR is you will see the IMD real time with the two tones on a calibrated scale.

[w1vtp]

My mistake.  I tho't you already had your SDR Rx and was looking at a display.  I'm going to try the same thing -- I think I can do it with my second RX in the Flex going into the second Rx input with a sampler and an adjustable attenuator going so I can put the top of the Tx carrier where I want it.

At least that's what the Flex people say can be done

[w1vtp]

<snip>....While I am at it, I should try and measure the dynamic range of the homebrew receivers.

How would I do that? ....

<snip>

Brett

W1VD would be the guy for that question.

Al

[K1JJ]

Tom,
What are you using for a two tone generator? You want a stable input so you can actually measure the IMD level. I would use a step attenuator and put the RX into tightest filter position. If you know the two tones you will know where the imd is supposed to be. The nice thing about SDR is you will see the IMD real time with the two tones on a calibrated scale.

Anyone have a simple and CLEAN two-tone generator circuit?  I need to build one.   

Frank, right now I'm using relative side-channel crud levels in the RX as I make improvements.  I already know the amp chain is clean - just want to make small improvements. But when I'm done, I will need to make some final measurements with a two-tone.

T

[W3RSW]

Try standard R/C phase shift oscillators.

I borrowed mine from Heath's Mod.monitor scope/sig.gen combo.
Dug it up to look at it.

Looks like I lifted just the twin audio osc. on a new chassis and added an outboard 115vac power supply.

 It uses a twin tetrode compactron though so, on second thought, probably not what you want, -unless you have a junker that you can take parts from..

There are two pots for setting each tone generator's strengths to match. The circuit is the ol' multi stage R/C phase shift, plate to grid osc. circuit and appears to be quite stable.  The detail of the phase shift couplet is shown.

[IZ0MFI]

Hello,
My SDR-14 is 4 or 5 years old  but after some attempts to update it, I finally realized that it can't able to replace a near-serious receiver . As a low end spectrum analyzer it's dinamic range is a direct consequence of the poor selectivity of the preselector (lack of the AD conv) when it is connected to an a rod broad antenna, needs at least 10 or 20 dB of RF attenuation (10 dB externak to eventually burn replace) to turn off the oveload warning LED.
It is definitely a spectrum monitor with moderate receiving function uncomparable (to receive a single frequency) to an HRO or a lovely Super Pro  
It's very compact and I use it as a measuring receiver or as fast mouse click track milair band monitor. Some rumors say that Italian Perseus is better but the audio delay....
'73 Marco
hnwe

[N2DTS]

I wonder how the new sdr-ip is going to be, the spec's look stunning.
It should be for $3000.00, which seems way expensive for less then a fist full of chips...
80 mhz 16 bit adc, -170dbc phaze noise clock, 105 db dynamic range, 10 bandpass filters, etc

In using the regular antenna's, I never noticed the sdr-iq overloading, and I do have some very strong broadcast stations not to far away, but nothing like they have in Europe....

Still, its more a low end product, the next step up is at least twice the cost.

The sdr-14 may be worse then the sdr-iq in some respects, not sure how the 14 does the front end.

Brett

[w1vtp]

Tom said "Anyone have a simple and CLEAN two-tone generator circuit?  I need to build one."

I built one way back when.  I'll have to take it into work and check how clean it is.  If it looks good, I'll share the schematic.  It is fine for checking a SSB linear on a scope.  It's solid state

Al

[WA1GFZ]

Al I don't think a 12 volt final can do much better than -30 dB IMD3

SDR IQ sounds no better than a Perseus. HPSDR Mercury a bit newer has a 130 MHz A/D 16 bit so has a slight edge. $3K is a rip off. I would go with Perseus or QSR1. The HPSDR transceiver will also be 130 MHz. I think you can buy the 170 MHz A/D now.
It takes two clean signal generators, a power combiner and RMS volt meter to measure RX dynamic range. The procedure is in many handbooks including the ARRL solid state handbook. I use a pair of HP8640Bs a home brew combiner and a HP3400A. Minicircuits also sells them. The handbooks have designs for combiners or return loss bridges as they are called. A 1/2 inch core and 2 resistors. QEX ran an article on a good one a couple years ago. I remember I sent the editor a note on a way to improve the layout. He posted the letter a few issues later.
Tom, I thought there was a software program to generate two tone with a sound card. Check Alberto's site Weak signals. I think his will do it. Stu may have a source also. You need a stable clean source to really measure it accuratly.

[K1JJ]

"Tom, I thought there was a software program to generate two tone with a sound card. Check Alberto's site Weak signals. I think his will do it. Stu may have a source also. You need a stable clean source to really measure it accuratly. "

Yes, that would work, Frank.  Though I want to be independent of a computer if possible and build a simple SS box. The Belcher works FB as a stand-alone and I want to do the same with a two-tone.

I'll see how yours works out, Al.

T

[WA1GFZ]

I think it will be hard to build a pair of oscillators with harmonics down 50 db without distortion.

[N2DTS]

I got the sdr-iq today, and have it working in the background while I post here...
I tried the new spectraview software and sdr console.
Both have things about them I don't like, sdr console has a very nice spectrum display, but its very hard to figure out how to use it, its too small, etc....
Both don't have sync AM yet.

I still like the power sdr software best, but the sdr-iq does not run it.

Now to use it to make some tests on the transmitters....

Brett


 

[KF1Z]

Brett,

Try Winrad... it has syn AM...
I like it much better than PowerSdr...

I know, everyone likes something different   

[N2DTS]

I will give it a try!
I ran some tests, it looks like the flex barefoot does about -35 db to the next lower lobe with the 2 tone test, and through the amp, about -32.
I did not see a big difference making various adjustments except for the plate voltage on the amp, below about 2000 volts, the imd goes way up quickly.....

Brett

[K1JJ]

I will give it a try!
I ran some tests, it looks like the flex barefoot does about -35 db to the next lower lobe with the 2 tone test, and through the amp, about -32.
I did not see a big difference making various adjustments except for the plate voltage on the amp, below about 2000 volts, the imd goes way up quickly.....

Brett

Brett,

-35db IMD is a respectable number for the exciter. 

As for the amplifier, what happens when you continue to decrease C2 (load more heavily) until the power starts to drop way off?  Do you see some inprovement in IMD?   Is that the 813 amp in GG?  Interesting on the high plate voltage giving worse IMD. On my 3-500Z's, Eimac published a sheet that shows improved IMD when usiang a low 1500V. But's it's probably a different case with 813's in GG.  I run my 3-500Z's at 1500 V now, loaded heavily and get out about 250w pep, but real clean for whatever use.

T

[N2DTS]

I did not see a big difference when loading it or unloading it  past max output power, but need to study it some more.
As far as less imd with higher plate voltages, I suspect the bias set point has something to do with that.

Below about 2100 volts, I think the bias is to high.

I tried to get winrad working, but the bat file you run to get it to work....

Brett

[N2DTS]

ssb, 2 tone test, when I adjust the loading control, some intermod spikes go up, others go down, there might be a slight improvement in the highest spikes, a few db, while some lower spikes come up to the level of the next lowest spikes....

Power output and drive levels dont seem to have much impact if its not overdriven.

Overdriving the amp increases the imd 2 db.

I was on with the big rig (2x813) and it is wide.
The majority is under about 9Kc, but I have little artifacts out really far, like 20Kc.
The dual 31band EQ does not seem to cut them enough.

The flex seems to cut off right where you set it.

So I need to build some sort of cutoff filter for the audio chain.


Brett
 

[K1JJ]

Brett,

If rolling off the EQ sharply at say, 5kc and up still produces spikes out at 20kc, then adding a sharp low pass audio filter will not have much effect on these spikes.  The EQ will have rolled off big by the time it gets to 20kc.

Run some high level tones thru the system, bypassing the audio stages to see if there are audio parasitics being generated.  Think of other ways to isolate the reason for the 20kc spikes. ie, Are they RF generated spikes (RF in the audio) or audio generated spikes (audio parasitics in the audio stages)?

Well, if your 813 amp continues to stay at -32db IMD or better, that's FB -   it's never gonna have a spatter complaint, since most barefoot riceboxes are around that level today.

T

[KF1Z]

I tried to get winrad working, but the bat file you run to get it to work....

Brett

Brett,

You need to unzip the files into the winrad folder...

Next, run the batch file from that folder.

Then when you run winrad, it may ask if you want to use the file "ExtIO_SDR14"
Select it and click "open".

Then under "show options" tab...."select input"...  select SDR14/IQ

[N2DTS]

I did just that, and tried to run it as admin as well, all the files are in the winrad folder...
I need to play with it some more.

The looking wide was me overloading the RX, even though it was only reading 20 over S 9.
At S9, it looks normal.
The sdr-iq is not as good as some other sdr receivers when it comes to really strong signals...

No antenna at all makes it work well as a transmit spectrum monitor.

In a qso with HLR and a bunch of other people, Tim told me to open the eq out to 10Khz which seems crazy, but it did not look real wide on the sdr-iq...

Brett

I tried to get winrad working, but the bat file you run to get it to work....

Brett

Brett,

You need to unzip the files into the winrad folder...

Next, run the batch file from that folder.

Then when you run winrad, it may ask if you want to use the file "ExtIO_SDR14"
Select it and click "open".

Then under "show options" tab...."select input"...  select SDR14/IQ