sdr-widget vs USB2SDR
139 views
Skip to first unread message
Alex Lee
Jan 3, 2012, 5:21:13 AM1/3/12
to MOBOKITS
Hi all,
I am not commenting on sdr-widget vs USB2SDR in terms of performance,
pros and cons, etc. :-)
There are plenty of choices as to the device you would like to use to
connect your sdr hardware to the PC, including commercial sound cards
galore, or just the PC's internal sound card (especially if you have
stereo input).
The sdr-widget BETA (factory built units) and sdr-widget kits have all
been sold out. So even if you are interested in one you cannot buy
one :-)
To those lucky owners of sdr-widgets, thanks for your interest and
support and I hope you enjoy using the sdr-widget and have plenty of
fun :-)
73 Alex
I am not commenting on sdr-widget vs USB2SDR in terms of performance,
pros and cons, etc. :-)
There are plenty of choices as to the device you would like to use to
connect your sdr hardware to the PC, including commercial sound cards
galore, or just the PC's internal sound card (especially if you have
stereo input).
The sdr-widget BETA (factory built units) and sdr-widget kits have all
been sold out. So even if you are interested in one you cannot buy
one :-)
To those lucky owners of sdr-widgets, thanks for your interest and
support and I hope you enjoy using the sdr-widget and have plenty of
fun :-)
73 Alex
Andrew Lees
Jan 4, 2012, 7:18:21 AM1/4/12
to mobo...@googlegroups.com
As a lucky owner of an SDR widget, I can attest that it was a real pleasure to build, and very confidence building for attempting any other SMD project - having hand soldered that CPU, I feel up to anything!
The performance is brilliant, I've dropped the softrock op-amp gain by 20dB to get the noise floor to a sensible level, and it could handle a bit more gain reduction yet. Also reduced the roll-off, to improve the phase response over the full frequency range.
So I offer my thanks to the developers of these boards, they've produced a real gem.
Andy, VK2DQE
The performance is brilliant, I've dropped the softrock op-amp gain by 20dB to get the noise floor to a sensible level, and it could handle a bit more gain reduction yet. Also reduced the roll-off, to improve the phase response over the full frequency range.
So I offer my thanks to the developers of these boards, they've produced a real gem.
Andy, VK2DQE
John Williams
Jan 4, 2012, 7:25:11 AM1/4/12
to mobo...@googlegroups.com
Andy,
Can you describe the softrock changes in detail? I'd like to give them a try...
John
Can you describe the softrock changes in detail? I'd like to give them a try...
John
-- John Williams KE5SSH - ham since 2007 WQKA523 - GMRS for family use on the farm
Oliver Goldenstein
Jan 4, 2012, 8:03:21 AM1/4/12
to mobo...@googlegroups.com
Hi John !
I have a new SDR-Widget wihich am driving with a UHFSDR.
I noticed very high levels also -18 dB at the RX IQ.
I built a cable like described here (see page 20):
http://www.ntms.org/files/KC4YOE_UHFSDR-USB2SDR.pdf
that did the trick on my side.
73 Oliver DL6KBG
--
Oliver, DL6KBG, JO61UB
Blog: http://dl6kbg.blogspot.com
John Williams
Jan 4, 2012, 9:25:50 AM1/4/12
to mobo...@googlegroups.com
Oliver,
Interesting -
On the softrock I would prefer not to attenuate then amplify, that could
lead to a higher noise floor. I would prefer to reduce amplification at
the op-amps to also reduce any noise that is also present on input.
John
Roger Critchlow
Jan 4, 2012, 2:46:16 PM1/4/12
to mobo...@googlegroups.com
The mod is to replace R31 and R32 on an Ensemble RX II, or R56 and R59 on an Ensemble RXTX, which set the output opamp gain. They are 4k99 in the standard build, reducing to 499 should give a 20dB reduction. But experimentation has been urged rather than just hacking and forgetting.
Roger Critchlow
Jan 4, 2012, 2:49:11 PM1/4/12
to mobo...@googlegroups.com
I knew I needed to recheck that as soon as I sent it.
That should be R32 and R33 on the Ensemble RX II.
-- rec --
VK5ZAG
Jan 4, 2012, 6:36:47 PM1/4/12
to MOBOKITS
Since this thread is USB@SDR vs Widget I feel qualified to comment.
I own a delta 66,EdirolFA-66,USB2SDR and a widget.
I have built 3 MoBo3.4s with RXTX6.3s
I have been able to compare the RX performance of all using WSPR
and splitting the antenna.(I have found this the best comparison)
I can say that the widget had the best RX performance with adjacent
large signal interference and inter-modulation effects..
followed by the USB2SDR, FA-66 and Delta 66
All of the tested units had Audio artifacts (visible when terminated
but not connected to the MoBo) the number, frequencies and level of
these
artifacts varied between units. If anything the USB2SDR had the least
number
of the lowest level. and the Delta 66 the most and highest.
I tested all units on more than one SDR program with Panoramic
displays.
Some programs have artifacts with some units and some with others
suggesting that these artifacts are software generated. I counted only
those artifacts visible on all software and a Marconi TF 2330A Wave
Analyzer. (Old slow but accurate).
In terms of usage: The widget missed out using PSDR compared
with all three other units because of being 2in 2 out rather than
4in 4out. The FA-66 and the USB2SDR are by far the easiest
and most fully featured to use for Amateur SDR radio The USB2SDR
having been designed for SDR and (Rather than high end audio)
and having a wide variety of interface options and CW keying options
wins out head and shoulders above anything else I have used.
Also its compatibility with PSDR IQ which is without any doubt
at this time the best and most fully featured SDR PC program
makes it taking into consideration its high introductory pricing
The best bang for the buck. It is also available which at this time
the widget is not. It is to be hoped that further development of the
widget into 4in 4out will have it up there too since having 2 "sound
card"
drivers installed at once has proved to be problematic. The USB2SDR
will have an add on ADC board which may give it equal RX performance
with the widget. Also the designer has a RX and TX unit in the
pipeline.
Also work is to be done providing external interface with the MoBo 4.3
and PSDR IQ control of the MoBo. PSWR etc.
The MoBo has a few artifact/noise problems which may well be minimized
by externalizing the processor and minor modification of I2c
isolation.
All the above having been said The SR6.3/MoBo4.3/USB2SDR/PSDR-IQ
combination can make a high performance fully functional SDR
transceiver
at low cost, but high personal labour time. My concept of true Amateur
radio
has always been one of self construction rather than appliance
operation.
Without prejudice Eric VK5ZAG
On Jan 5, 7:20 am, Oliver Goldenstein
> before_100k_to_10k_mod.png
> 595KViewDownload
>
> after_100k_to_10k_mod.png
> 451KViewDownload
I own a delta 66,EdirolFA-66,USB2SDR and a widget.
I have built 3 MoBo3.4s with RXTX6.3s
I have been able to compare the RX performance of all using WSPR
and splitting the antenna.(I have found this the best comparison)
I can say that the widget had the best RX performance with adjacent
large signal interference and inter-modulation effects..
followed by the USB2SDR, FA-66 and Delta 66
All of the tested units had Audio artifacts (visible when terminated
but not connected to the MoBo) the number, frequencies and level of
these
artifacts varied between units. If anything the USB2SDR had the least
number
of the lowest level. and the Delta 66 the most and highest.
I tested all units on more than one SDR program with Panoramic
displays.
Some programs have artifacts with some units and some with others
suggesting that these artifacts are software generated. I counted only
those artifacts visible on all software and a Marconi TF 2330A Wave
Analyzer. (Old slow but accurate).
In terms of usage: The widget missed out using PSDR compared
with all three other units because of being 2in 2 out rather than
4in 4out. The FA-66 and the USB2SDR are by far the easiest
and most fully featured to use for Amateur SDR radio The USB2SDR
having been designed for SDR and (Rather than high end audio)
and having a wide variety of interface options and CW keying options
wins out head and shoulders above anything else I have used.
Also its compatibility with PSDR IQ which is without any doubt
at this time the best and most fully featured SDR PC program
makes it taking into consideration its high introductory pricing
The best bang for the buck. It is also available which at this time
the widget is not. It is to be hoped that further development of the
widget into 4in 4out will have it up there too since having 2 "sound
card"
drivers installed at once has proved to be problematic. The USB2SDR
will have an add on ADC board which may give it equal RX performance
with the widget. Also the designer has a RX and TX unit in the
pipeline.
Also work is to be done providing external interface with the MoBo 4.3
and PSDR IQ control of the MoBo. PSWR etc.
The MoBo has a few artifact/noise problems which may well be minimized
by externalizing the processor and minor modification of I2c
isolation.
All the above having been said The SR6.3/MoBo4.3/USB2SDR/PSDR-IQ
combination can make a high performance fully functional SDR
transceiver
at low cost, but high personal labour time. My concept of true Amateur
radio
has always been one of self construction rather than appliance
operation.
Without prejudice Eric VK5ZAG
On Jan 5, 7:20 am, Oliver Goldenstein
<oliver.goldenst...@googlemail.com> wrote:
> Many thanks John and Roger !
>
> That did the trick finally with my uhfsdr and the widget:
>
> R21, R22 changed from 100k to 10k (on UHFSDR opamps stage)
>
> 73 Oliver DL6KBG
>
> On 4 January 2012 20:49, Roger Critchlow <r...@elf.org> wrote:
>
>
>
>
>
>
>
>
>
> > I knew I needed to recheck that as soon as I sent it.
>
> > That should be R32 and R33 on the Ensemble RX II.
>
> > -- rec --
>
> > On Wed, Jan 4, 2012 at 12:46 PM, Roger Critchlow <r...@elf.org> wrote:
>
> >> The mod is to replace R31 and R32 on an Ensemble RX II, or R56 and R59 on
> >> an Ensemble RXTX, which set the output opamp gain. They are 4k99 in the
> >> standard build, reducing to 499 should give a 20dB reduction. But
> >> experimentation has been urged rather than just hacking and forgetting.
>
> >> See http://groups.google.com/group/sdr-widget/msg/c46b463dede0fd3b
>
> >> and http://groups.google.com/group/sdr-widget/browse_thread/thread/fb1583...
> Many thanks John and Roger !
>
> That did the trick finally with my uhfsdr and the widget:
>
> R21, R22 changed from 100k to 10k (on UHFSDR opamps stage)
>
> 73 Oliver DL6KBG
>
> On 4 January 2012 20:49, Roger Critchlow <r...@elf.org> wrote:
>
>
>
>
>
>
>
>
>
> > I knew I needed to recheck that as soon as I sent it.
>
> > That should be R32 and R33 on the Ensemble RX II.
>
> > -- rec --
>
> > On Wed, Jan 4, 2012 at 12:46 PM, Roger Critchlow <r...@elf.org> wrote:
>
> >> The mod is to replace R31 and R32 on an Ensemble RX II, or R56 and R59 on
> >> an Ensemble RXTX, which set the output opamp gain. They are 4k99 in the
> >> standard build, reducing to 499 should give a 20dB reduction. But
> >> experimentation has been urged rather than just hacking and forgetting.
>
> >> See http://groups.google.com/group/sdr-widget/msg/c46b463dede0fd3b
>
> 595KViewDownload
>
> after_100k_to_10k_mod.png
> 451KViewDownload
Andrew Lees
Jan 4, 2012, 7:10:42 PM1/4/12
to mobo...@googlegroups.com
Hi John,
On examination I find I increased the op-amp feedback resistor to 82R, so not quite 20dB gain reduction. I recall I decided to start out conservatively, and reduce the gain further if it was needed (it is), but major renovations got in the way - the shack is currently packed up in boxes - so later! My softrock is installed in a Mobokit, and I plan to build a preamp for 30m and up, so I'll sort out the final op-amp gain at that point. A value of 100R would be better even without a preamp, but you should probably a look at your own noise floors with and without power applied to the softrock - I feel that any more than 10dB noise lift on a dummy load with power applied is a waste of dynamic range. The value should really be based on "quiet band" days in your location on the bands you use.
I also reduced the bypass capacitance, but I cannot immediately determine to what value, as I installed SMD items on the back of the board, and I don't have the values to hand. I recall that I intended the 3dB point to be around 200kHz, so the caps should be around 20pF I believe.
Andy
On examination I find I increased the op-amp feedback resistor to 82R, so not quite 20dB gain reduction. I recall I decided to start out conservatively, and reduce the gain further if it was needed (it is), but major renovations got in the way - the shack is currently packed up in boxes - so later! My softrock is installed in a Mobokit, and I plan to build a preamp for 30m and up, so I'll sort out the final op-amp gain at that point. A value of 100R would be better even without a preamp, but you should probably a look at your own noise floors with and without power applied to the softrock - I feel that any more than 10dB noise lift on a dummy load with power applied is a waste of dynamic range. The value should really be based on "quiet band" days in your location on the bands you use.
I also reduced the bypass capacitance, but I cannot immediately determine to what value, as I installed SMD items on the back of the board, and I don't have the values to hand. I recall that I intended the 3dB point to be around 200kHz, so the caps should be around 20pF I believe.
Andy
Reply all
Reply to author
Forward
0 new messages