HF converter for FunCube Dongle test report and ... - Page de F5RCT
Feb 23, 2012 - It was shown a high level of local oscillator at output and filters not well matched to the mixer. This report ... Local oscillator (HCMOS) and NE602 LO input are not 50 Ohms ! ... Let's go deeper in analysis ! .... http://www.hsmicrowave.com/FCD-5-55-UC_102711.PDF. Better, but needs a notch on LO at output ...
HF converter for FunCube Dongle test report and improvement solutions
Many users of the kit of the HF converter for the FunCubeDongle complain on poor receiving quality. The receiver shows a full spectrum of spikes and strong signals as broadcast stations saturate the FCD receiver. Reducing the gain of the FCD enhances a little the performance, but the sensitivity is very poor compared to a classical HF receiver. It was shown a high level of local oscillator at output and filters not well matched to the mixer.
This report concerns test of HF converter for FunCube Dongle sold in kit version by several suppliers : -
CT1FFU: http://www.ct1ffu.com/site/ then “KITS” http://www.ct1ffu.com/site/index.php?option=com_content&view=article&id=178&Itemid=104
(pictures of the converter taken from CT1FFU website) F6BZG gave me this converter for tests because he got poor receiving conditions with FCD receivers. He was very disappointed by results compare to FCD on VHF and UHF bands. When antenna is connected all FCD spectrum is covered by spikes even with LNA gain at 0 dB on FCD input. The converter works better with tuned loop antenna and with 10 to 20 dB attenuation at FCD receiver input (remote supply not possible in that case). We tried with separated supply in case of noise in coming from USB port, but no better results ! This converter is very poor in performances and not suitable for 30 and 50 MHz band due to poor sensitivity.
Original schematic :
Zin, Zout=1500R
Lo leak -30dBm !
The LNA is not mounted and not necessary on HF band.
Schematic expertise gave impression of a circuit constructed with simulated block but never completely measured, neither totally checked ! -
Input filter designed under 50 Ohms not matched to NE602 input (Zin 1500 Ohms).
-
Output filter designed under 50 Ohms not matched to NE602 output (Zin 1500 Ohms).
-
FM band notch C19 // L11 not efficient ! calculated resonant frequency 100µH (low Q ferrite inductor) with 25 pF is about 3.2 MHz ! May be a calculation mistake, inductor should be 100 nH !
-
Local oscillator (HCMOS) and NE602 LO input are not 50 Ohms !
-
Poor supply decoupling from USB !
-
L10 at 1 mH is more capacitive than inductive from 106 to 136 MHz ! It should be 1 µH !
-
C1 and C12 are too low value and limits frequencies under 20 MHz as high pass filter ! better with 1 nF.
We measured a strong LO leak of -30 dBm at output. This will block FCD receiver and saturate its LNA, even with 0dB gain configuration !
The spectrum here at converter output in max hold mode, with a swept signal of -30 dBm at converter input. -
LO leak : -30dBm
-
Conversion gain is not flat and results of a loss from -2 to -10 dB Conversion loss related to -30 dBm at input
Poor gain for freq < 10 MHz
Let’s go deeper in analysis !
Input filter simulation : Input filter designed under 50 Ohms, not matched to NE602 input (Zin 1500 Ohms). Simulation with 1500 Ohms load : power loss of -10.5 dB at 30 MHz !
Schematic of simulated filter :
100pF
Port1
Port2
0
L0805CS_221XJBC 220nH
0
L0805CS_470XKBC 47nH
0
56pF
100pF
L1008CS_271XJBC 270nH
100pF
L0805CS_221XJBC 220nH
56pF
L0805CS_470XKBC 47nH
0
Measurement :
Network analyzer measurement . - Input at 50 Ohms - Ouput measured with HiZ probe at NE602 input 12 dB loss at 30 MHz (M1)
INPUT FILTER : Input filter introduce more than 10 dB loss. It is not possible to match 50 Ohms antenna input to 1500 Ohms of the mixer over a wideband from 2 to 50 MHz. A compromise of low pass transformation filter can be found to get about -5 dB loss at 30 MHz and -10 dB at 2 MHz. Better give up 50 MHz band due to poor gain and noise figure of this mixer ! Here below, a proposal of a better filter with only 3 inductors !
Port2 470pF
Port1
0
0
0
47pF
150pF
L1008CS_821XJBC 820nH
150pF
L1008CS_471XJBC 470nH
68pF
L1008CS_331XJBC 330nH
0
OUTPUT FILTER : Same situation with output filter ! It introduces more than 16 dB loss at 136 MHz! It is not possible to match 1500 Ohms mixer to 50 Ohms output over a wideband!
0
20pF
L0805CS_470XJBC 47nH
0
15pF
L0805CS_820XJBC 82nH
0
20pF
82nH L0805CS_820XJBC
91pF
Port2
L0805CS_470XJBC 47nH
Port1
91pF
0
Simulation results for 1500 Ohms source impedance at input and 50 Ohms at output
Here a simulation with a filter based only on 3 inductors : - Better performances at 136 MHz (Rx = 30 MHz band) - Better attenuation under 106 MHz. Note : - C3 = 100 pF is connected from mixer to port 1 of filter - Port 2 goes to FCD input
0
12pF
L0805CS_470XJBC 47nH
Port2
L0805CS_820XJBC 82nH
6.8pF
12pF
L0805CS_820XJBC 82nH
680
Port1
33pF
0
0
OSCILLATOR matching : It is not necessary to overdrive the mixer and have a heavy load by attenuator. Attenuator values are changed as follow: R2 not mounted ! R3 = 150R R4 = 33R LO leak at output is now -35 dBm Conversion gain -10 to -2 dB L10 (1mH) replaced by 100µH (L11) for DC from FCD. C3 = 100pF Oscillator notch : to improve LO leak, but layout is so bad that oscillator is radiating all over the board ! The layout and the schematic need to be completely revised. L11 : 2 turns of 0.5 mm silver wire on 4 mm diameter strap, adjust spacing to tune to 106 MHz C19 : 100pF SMD 1206
Final results : -
Gain -8 to -1.5 dB. Flat response; higher gain at 28 MHz rather than 7 MHz. Lower LO level at output. Measured compression at input -23 dBm.
Now we get better performances with modified version and 0 dB LNA gain on FCD, but it is still a “compromise” ! We did performance measurement of the FCD receiver, it was shown poor dynamic performance and IP3 of this receiver : any signal greater than -45 dBm reduce the sensitivity by blocking receiver chain. LO leak of this converter is too high for FCD receiver. This converter can be improved but, the combination of this converter with FCD receiver is very bad. For HF band, it is better to use ring diode mixer for easier matching and higher intermodulation level. Don’t expect performance on this kit, a good “softrock” receiver is better for HF band.
73s de F5RCT Jean-Matthieu STRICKER
revision date : 23/02/2012 http://www.f5kav.org/
FCD list : http://uk.groups.yahoo.com/group/funcube/ more serious converter but gain seems to be too high for HF band : http://www.hsmicrowave.com/FCD-5-55-UC_102711.PDF Better, but needs a notch on LO at output : http://www.george-smart.co.uk/wiki/FunCube_Upconverter Oscillator can be found here as spare part : SARONIX Oscillator 106.25MHz STAA9FC-106.2500T http://www.ebay.com/itm/SARONIX-Oscillator-106-25MHz-STAA9FC-106-2500T-SMD-Qty-10-/400272682219
The measurements are confirmed by blocking level and third intercept point levels. .... by adhesive cooper foil coating and connections to ground shield of SMA ... The Y-Factor is the ratio of Hot and Cold noise powers (in watts) and is ... This need
Mar 30, 2003 - uninhabited area of desert south of California City, California provided an ideal ... The circuit schematic used to construct these .... As in the case for data collected for the study of Azimuth radiation, no significant difference in
Mar 30, 2003 - proponents claim a very favorable omni-directional azimuth radiation pattern .... calibration point between the test and reference antennas.
Mar 17, 2017 - 90 kg. 125 kg. 2. 2. 2. B. 0. 0. -. Test-criteria. Minimum take off weight. Maximum take off weight. Special take off technique required. A. A. A. A. A.
Mar 30, 2017 - DHV certifies that the model compliance with all criteria defined in CCC, February 2016. Place, date, stamp and signature (manufacturer):.
Sep 14, 2007 - A Yes. A. Minimum speed. Less than 25 km/h. A Less than 25 km/h. A. 4. Control movement. Max. weight in flight up to 80 kg. Symmetric control ...
Flight test report: PG_0645.2012 / page 1 of 3. 1. Inflation/Take-off. A. Rising behaviour. Smooth, easy and constant rising A Smooth, easy and constant rising. A.
Inflation/Take-off. A. Rising behaviour. Smooth, easy and constant rising A Smooth, easy and constant rising. A. Special take off technique required. No. A No. A.
Apr 18, 2007 - Sink rate after two turns. Up to 12m/s. A More than 14 m/s. B. 10. Symmetric front collapse. Entry. Rocking back less than 45°. A Rocking back ...
Oct 10, 2016 - C. 0. 0. -. Test-criteria. Minimum take off weight. Maximum take off weight. Special take off ... Smooth, easy and constant rising, no pilot ...
Oct 11, 2016 - 2. 3. 3. C. 0. 0. -. Test-criteria. Minimum take off weight. Maximum take off weight. Special take off ... Landing - 4.4.2. EAPR- ... 0° bis 30°. A.
Spontaneous in less than 3 sec. 25 km/h to 30 km/h. Special landing technique required. Speed range using the controls larger than 10km/h. Smooth, easy and ...
Nov 8, 2006 - 89259-700 Jaraguà do Sul, S.C. Brasil. Representive None. Classification B. Type of glider Prymus 2 S. Trimmer not available. Test Pilot Seiko ...
Nov 8, 2006 - Rua Walter Marquardt, 1180 cp 370. Place of test. Villeneuve. 89259-700 Jaraguà do Sul, S.C. Brasil. Representive None. Classification B.
B. Sink rate after two turns. Up to 12 m/s. A More than 14 m/s. B. 10. Symmetric front collapse. A. Entry. Rocking back less than 45°. A Rocking back less than 45°.
Date of flight test. 24. 04. 2009. Representative None. Place of test. Villeneuve. Glider model. Sprint S. Classification. B. Trimmer no. Test pilot Thurnheer Claude.
Special take off technique required. No. A No. A. 2. Landing. A. Special landing technique required. No. A No. A. 3. Speed in straight flight. A. Trim speed more ...
86-031 Osielsko. Poland. Date of flight test. 02. 12. ... Niviuk Gliders - Hamak 2 M. Total weight in flight (kg) 70. 95 ... 12 m/s to 14 m/s. A More than 14 m/s. B. 10.
Flight test report: PG_0351.2010 / page 1 of 3. 1. Inflation/Take-off. A. Rising behaviour. Smooth, easy and constant rising A Smooth, easy and constant rising. A.