Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: Prev 1 2 [3] 4 5 6 7 8 ... 11 Next   Go Down
  Print  
Author Topic: DVB-T Stick as SDR  (Read 132334 times)
KE5JPP
Member

Posts: 0




Ignore
« Reply #30 on: December 24, 2012, 08:08:33 AM »

If you compare quality and price with one of those commercially available SDRs you are getting much more per dollar.

Merry Christmas!

For $20 you getting about $20 of performance.  Glad that you are satisfied with such low performance.

Gene
Logged
KA4POL
Member

Posts: 2097




Ignore
« Reply #31 on: December 24, 2012, 09:52:19 PM »

Amen!
Logged
KN0CK
Member

Posts: 22




Ignore
« Reply #32 on: December 28, 2012, 10:01:31 AM »

You know, in life you're going to find A LOT of differing opinions - and everybody has one.  Given that, I'll weigh-in on the performance of my RTL-2832U + 'Ham-it-Up' HF Upconverter:

By design, the RTL2832U won't exhibit great receiver dynamic range since it's relying on 8-bit architecture for it's receive functionality (so I've been told), so the best you can expect is between 48 - 60dB dynamic range given that bit weighting.  While that may - to the 'SDR-purist' out there - be a nonstarter, for those who want to have a secondary receiver that can tune the bands and do an admirable job for what you DO get, I'm not complaining about not getting that extra 30 or more dB of dynamic range.  Using the same longwire antenna (I live in restrictive covenants - so don't get me started there), I can tell you that I have receivers that do a better job of pulling signals out of the noise because they have proper filtering and definitely better signal dynamic range performance according to what you pay for.  Spend low, you get marginal 'consumer product' performance.  Spend high (within reason) and you get professional quality receiver performance.  But if you're just looking to have fun with a wideband radio that can detect strong SWL signals and a few strong-signal hams while listening to nets, AND be able to tune the VHF band at the same time, this dongle+upconverter combination (for $60.00) does an ADMIRABLE job!  No kidding, I'd put it against any of my radios for casual tuning of the SWL band and it has comparable performance for stronger shortwave signals.  It undoubtedly falls short of being perfect on the Amateur bands, but it's a GREAT receiver for CASUAL tuning and listening on the cheap. 

Where the HF upconverter falls REAL short is where they upconverted the HF band to - - The FM Radio Band.  By doing a straight 100 MHz upconversion from the expected HF tune frequency, it lands you pretty much from 101.8 MHz to 130 MHz - about the lower 1/3 of the HF tuning range for the RTL-2832U and HF Converter will be subject to A LOT of interference if you have strong FM signals that land anywhere in that range.  I cannot imagine anyone living in a major metropolitan area to have this receiver be useful in the 80m, 75m, 60m, and 40m bands.  I live in a somewhat metropolitan area and I know that strong FM stations at 103.3 MHz, 104.5 MHz, and 107.9 MHz practically destroy your ability to listen to anything that falls into the lower end of the 80m band and the top of the 40m band.  The elliptical low pass filter design that they used is not effective in cutting down those strong signals.  Adjusting the RF gain of the dongle improves it some, but not enough to pull any weak signals out of the residual noise.  I would think that upconverting to 120 MHz would have been a better solution since all you have to deal with there is aircraft and public service that are, by nature, narrowband intermittent services and with A LOT less power to interfere.  If nothing else, the elliptical filter will roll off even better if the signals were upconverted to 120 MHz.  But I can understand the reasoning behind doing a straight 100 MHz upconversion:  EASE OF USE WITH TUNING.  For me, I would have just dealt with that small issue of performing the math and took the better upconversion (and receiver) performance.

I cased my RTL-2832U dongle and HF converter in a cheap black plastic Radio Shack project box, added a long (about 6') shielded USB cable to it, and drilled a hole to allow the HF Antenna connection (SMA connector) to be exposed out of the box and attached a SMA to SO239 connector to it.  Once I connected everything up I purposely hid that box in my setup (what's to look at - it's a black box) and now I have a semi-pro receiver playing through by PC speakers while I have any of the other rigs tuned elsewhere and playing, too.  Frankly, I have an older military receiver that I used to turn on and listen to as my standby receiver.  From the day I pressed this cheap SDR into service, I haven't turned on that old receiver in about a week (and it was a week ago that I cobbled this SDR together).  It plays GREAT in my setup!!

So, the final grade I give this SDR is: B- (good effort, but needs a couple tweaks to be considered a truly great receiver).

73 - de Marty, KN0CK  
« Last Edit: December 28, 2012, 10:54:47 AM by KN0CK » Logged
K9AQ
Member

Posts: 55




Ignore
« Reply #33 on: December 28, 2012, 03:54:47 PM »

I completely agree with Marty's observations. I am using the same up converter and while it does work for me on the lower bands, it would seem that it could be improved by replacing the 100 mhz oscillator with a 120 mhz one.  When I have some free time, I will give this a try.

I am really enjoying having a second reciever with the HDSDR display.  I don't know how I got along for so many years without it.  It may not be "perfect" but it is sure a lot better than not having one!  I am also really enjoying experiementing with SDR, which over time could lead me to a Flex.

Don
K9AQ
Logged
KN0CK
Member

Posts: 22




Ignore
« Reply #34 on: December 28, 2012, 04:46:37 PM »

Don,

Thanks for the nod on the observations - it's nice to have positive responses to this thread...

Before you attempt to replace the 100 MHz oscillator with a 120 MHz version, I just tried replacing the oscillator with a comparable 125 MHz oscillator that I had out of one of those AD9850 programmable oscillators and the following observations were noted:

1.) Working the math with the new oscillator frequency, you can tune and listen to stations with A LOT less interference (see point #2 directly below).  I was able to tune WWV, CHU Canada (in the 20m band) and the ENTIRE AM band (.5 to 1.6 MHz) without ANY interference from nearby FM stations.  However, the I.F. filter kicks in and does decrease the I.F. drive to the RTL-2832U some (see point #3 below).

2.) The noise floor decreased SUBSTANTIALLY (which is a good thing) since we're fighting higher noise floor with the nearby FM stations with this HF Upconverter.  It flattened the entire noise floor at least 20 dB over what it was with the 100 MHz oscillator.

3.) The IF drive to the RTL-2832U was decreased because the filter arrangement in the I.F. path cuts the power of the by about 3dB (half the signal) because the range has been shifted into the upper part of the bandpass filter in the '2932U's input.  I need to do some analysis of the I.F. filter in CircuitLab (http://www.circuitlab.com) and emulate their filter to see what the response is when the 125 MHz oscillator is plugged in.  I'm banking that the bandpass filter they're using does have the ability to cut into the I.F. power to the RTL-2832U and that's causing the issue with lower receive.  Either removing one filter section (jumpering over it) or providing new values will solve this issue.  I'm planning on doing a simulation on their filter tomorrow (12/29/12) and report back on what I see.  No kidding, plugging-in that 125MHz oscillator in place of the supplied 100MHz oscillator made a BIG difference in the noise floor (and, in effect, the performance of the receiver).

So those are my observations - stay tuned for more info as I have that available once I do the circuitlab simulations tomorrow...

73 de Marty, KN0CK
Logged
WD5GWY
Member

Posts: 406




Ignore
« Reply #35 on: December 28, 2012, 07:05:14 PM »

If you have access to the January 2013 issue of QST, there is an article there has an HF converter design that uses a 125Mhz oscillator just like you are talking about. And it moves the tuning range
to 125-155 Mhz, well away from the FM radio band. I have most of the parts to build that one.
Just cannot seem to get enough time away from work to start building it.
 james
WD5GWY

(oh, I also have the Ham it Up converter too. Waiting for the connectors I ordered to arrive)
 
Logged
KN0CK
Member

Posts: 22




Ignore
« Reply #36 on: December 28, 2012, 09:03:34 PM »

James,

Thanks for the tip - I should have noticed that one, but didn't remember until you just mentioned it here.  I do have that issue at hand (I'm a yearly subscriber) and will definitely check that out.  But I'll also report back my findings on the CircuitLab analysis of the 'Ham it Up' I.F. filter - might be interesting to check the two filter designs and see if they're similar (albeit with different component values because of the difference in upconverter tuning).

More to follow on this - do stay tuned, and thanks again for the info, James..!

Marty, KN0CK
Logged
KN0CK
Member

Posts: 22




Ignore
« Reply #37 on: December 29, 2012, 04:32:59 AM »

James,

Reviewed the design in the January 2013 QST and was certainly surprised to see that the author used an NE602 active mixer and absolutely NO filtering at the I.F. output from the mixer to the DVB-T receiver.  While this ensures that there's no loss in that path (it's straight capacitive coupling) it also ensures that there's a possibility of aliasing that can occur.  I think the author took the tact that any aliasing that shows up in the mixer's I.F. port are far enough out from the desired signal that it doesn't play an active role in the performance.  However, that's a bad assumption because those mixed products DO have an overall effect in the receiver's performance that shows up as decreased dynamic range in the desired signal because no matter what happens the power in that undesired signal is at the front end of that DVB-T stick, too, raising noise against the desired signal that could be improved through filtering out those undesired products.  He probably balanced 'good enough' performance and simplicity in design against adding complexity (and losing some I.F. power) through filtering.  There's nothing inherently wrong with that design assumption if that's his objective, but it will have a definite effect on the receiver's small signal acquisition - large signals will perform well with, again, decreased dynamic range (which is also an inherent issue with the DVB-T right out of the chute).

That said, I like the NE602 mixer design in the article and would use it as-is, but would probably add a less complex bandpass filter (fewer poles than are in the 'Ham it Up' I.F. filter) to help reduce undesired signals and improve (as best it can be) the small signal capture performance and still keep it relatively cheap.

Again, thanks for passing along the article - it does give me a new perspective on this project that's still worth pursuing (I bought a second DVB-T and will probably 'roll my own' HF upconverter as a hybrid of the 'Ham it Up' and NE602 upconverters).

73 de Marty, KN0CK
Logged
KN0CK
Member

Posts: 22




Ignore
« Reply #38 on: December 29, 2012, 07:18:39 AM »

All who own the 'Ham it Up' HF Upconverter board:

I did a circuit analysis of the HF Elliptical LPF and swept that filter from 2 to 200MHz in CircuitLab....I can assure you, THAT IS NO LPF..!  As it turns out (and I have the frequency response plots to prove it) there's a REAL GOOD reason that we're all seeing more FM in our receivers than expected: It's because the filter response has a 'pop-up' at 100MHz that doesn't REALLY roll-off at all.  About the only place that the filter does roll off is about 90MHz, but then immediately you get to 100MHz pop-up and then the filter 'sort of' rolls off at 160MHz but then again at 200MHz it's back up to the input level.  So you're going to see all sorts of FM stations in there no matter what.  The LPF needs to either have more poles to make it fall off sharper (and sooner than 90MHz), or the poles need to be retuned so the knee is more like 40 - 50MHz.

I'm going to use their existing design and tweak the poles to see if I can improve the performance and then report back.  But I have 'swept' the HF LPF and the I.F. Bandpass filter and the bandpass filter is fine - - it's capable of working properly as-designed.  But the LPF IS A MESS..!

Again - I'll tweak the values and let you know the outcome - Stay tuned...

73 de Marty, KN0CK
Logged
KA4POL
Member

Posts: 2097




Ignore
« Reply #39 on: December 29, 2012, 10:36:36 AM »

Is this the filter you are talking about?
http://www.george-smart.co.uk/wiki/FunCube_Upconverter#Low_Pass_Filter
Logged
KN0CK
Member

Posts: 22




Ignore
« Reply #40 on: December 29, 2012, 03:39:59 PM »

Dieter (and other 'Ham it Up' board users),

No, that LPF for the FunCube is quite a bit different than the 'Ham it Up' filter.  The FunCube uses a standard LPF design, while the 'Ham it Up' filter uses a Elliptical (Cauer) filter.  However, I did some more research this afternoon and came up with an updated schematic package than was posted on the open source website (http://code.google.com/p/opendous/wiki/Upconverter) a few days ago when I started to look into this issue.  As it turns out, the new schematic package more closely aligns to the design of the board I have than I first thought.  I haven't had a chance to 'run the numbers' on this new elliptical filter design (has one more pole to it than the package I downloaded 3 days ago) but if the data on the website is accurate, then this new filter should  have acceptable performance.  Again, I'm planning to run CircuitLab on it this evening and check that the design appears to match the data on the website - should, I have no reason (now since there's a more accurate schematic in the design package) to question it.

More to follow - stay tuned,

de Marty, KN0CK
Logged
KA4POL
Member

Posts: 2097




Ignore
« Reply #41 on: December 29, 2012, 09:57:10 PM »

Thanks for the link. They obviously have measured the filter using the SDR-Kits VNWA. I know that analyzer. It is a very good ham solution that delivers trustworthy data if the calibration has been done well enough.
Logged
K9AQ
Member

Posts: 55




Ignore
« Reply #42 on: December 30, 2012, 06:25:08 AM »

Has any one had any experience with the upconverter from 9A4QV http://upconverterup-100.blogspot.it/ ?  I am currently using an RTL with the Ham it Up upconverter and while I am happy with it, there are some areas that could be improved.  9A4QV's upconverter has an LPF between the oscillator and the mixer and an LPF on the input.  It also has a build in preamplifer to compensate for the conversion losses.  It is supposed to have an overall conversion gain of 10 db.

I have ordered one and when I get it will run some comparison tests.

Don
K9AQ
Logged
KN0CK
Member

Posts: 22




Ignore
« Reply #43 on: December 30, 2012, 07:23:33 AM »

Don,

I think I've seen these HF upconverters listed on EBay at one time and then they disappeared for awhile - haven't seen them in a few weeks.  The design looks REAL interesting against everything I've seen (the NE602 QST article upconverter, 'Ham-it-Up', and now this one), but the only drawback I see is that it requires an additional 9 - 12V supply to power it, as well as RF amps to recover conversion losses that often comes with added noise figure - - but then again we're not talking about high performance when the primary receiver (the dongle) has a 40 - 50dB dynamic range to work with.  The only advantage that the 'Ham-it-Up' and the NE602 HF upconverters have is that they can be powered from USB and makes for a pretty simple application to augment anyone's setup (USB and an antenna is all that's needed).  Other than that, they come with conversion losses that make for lesser receive sensitivity - which this UP-100 seems to have the leg-up over everything else out there.

Thanks for passing this one along and do check back with performance data and observations when you have it - I'd be particularly interested in knowing how it compares to the 'Ham-it-Up' board.

73 de Marty, KN0CK
Logged
KN0CK
Member

Posts: 22




Ignore
« Reply #44 on: December 30, 2012, 08:04:34 AM »

I just did a circuit analysis this morning of the updated Elliptical LPF that's in the schematic of the 'Ham-it-Up' board.  It's a great improvement over an earlier schematic that I acquired off the open source website, and did an analysis on a few days ago.  I'm convinced that the LPF that's on the current 'Ham-it-Up' board is acceptable and should provide decent FM band rejection provided that you're not situated within line of sight of the commercial FM transmitter.

On the flip, I did a circuit analysis of the M1GEO LPF that's been applied to the FunCube dongle and can report that at 88MHz it provides 39dB of rejection and when at 108MHz it provides 52dB of rejection - - the rolloff is pretty sharp starting at 75MHz.  From 500kHz to over 30MHz it shows no perceptible loss and at a couple of points in its range (15MHz and 42MHz) it shows an some modest gain.  Frankly, I liked this filter design - A LOT.

With all the different methods of filtering and upconverting I've come to the conclusion that there's a hybrid design out there that employs the best of all three <known> upconverter designs out there.  I'm about to go apply them to my own HF upconverter design.  I'm planning to design it this week, 'run the numbers' on it in CircuitLab, and then I might set out to design a board for this if I have time (the prototype will be on perfboard).  My design criteria is pretty simple:

1.) Has to operate on USB power ONLY
2.) Has to be simple in design (nothing exotic)
3.) Two, or less, active components
4.) Passive filtering with no RF amps to recover losses
5.) Connectivity with wires to the board or easily found (and/or inexpensive) connectors
6.) Has to fit on a board no larger than 1.5" x 2.5"
7.) Uses readily available 'RTL' apps (HDSDR and Zadig)
8.) Price target of $20 or less (it's just my own personal objective)

I'll keep you posted with my progress - stay tuned...

73 de Marty, KN0CK
Logged
Pages: Prev 1 2 [3] 4 5 6 7 8 ... 11 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!