Date:2010-09-10 01:56:58 (9 years 3 months ago)
Author:Werner Almesberger
Commit:d3bc274dab1ae58e5e5ad741111a63d9b6e5e72a
Message:Moved the ECNs from atrf/ecn to the top-level.

Files: atrf/ecn/INDEX (1 diff)
atrf/ecn/ecn0001.txt (1 diff)
atrf/ecn/ecn0002.txt (1 diff)
atrf/ecn/ecn0003.txt (1 diff)
atrf/ecn/ecn0004.txt (1 diff)
ecn/INDEX (1 diff)
ecn/ecn0001.txt (1 diff)
ecn/ecn0002.txt (1 diff)
ecn/ecn0003.txt (1 diff)
ecn/ecn0004.txt (1 diff)

Change Details

atrf/ecn/INDEX
1Number Status Description
20001 Edit Adjust balun component values
30002 Done Add load capacitors to 16 MHz crystal
40003 Edit Replace balun and filter with integrated balun
50004 Edit Take into account layout considerations for RF
atrf/ecn/ecn0001.txt
1Adjust balun component values
2
3
4Some of the components specified in the schematics were not at hand
5in my lab and were thus replaced with similar parts. Furthermore, the
6LED color was changed to ease visual identification:
7
8Component Norminal Board 1 Board 2
9LED (D1) red red red-orange
10C6/C7 0.82 pF 0.7 pF 0.9 pF
11L2/L3 4.3 nH 4.7 nH 4.7 nH
12
13The resulting spectrum should be examined to determine which variant
14performs better. Note that the discrete balun circuit shall be
15replaced with an integrated balun later on.
atrf/ecn/ecn0002.txt
1Add load capacitors to 16 MHz crystal (C14, C15)
2
3
4The crystal has a specified load capacitance of 8 pF. The AT86RF230 has
5an internal capacitor array that can be trimmed in 16 steps from 0 pF to
64.5 pF.
7
8The crystal has a nominal tolerance of 15 ppm at 25 C plus a temperature
9tolerance of 15 ppm. The following deviations from 1 MHz at CLKM were
10measured with the CNTR board, roughly at 20-25 C:
11
12Board Error Meas. accuracy C14/C15 Trim
13        (f, ppm) (ppm, nom.) (pF) (pF)
142, "orange" +268 99.6 NC 0
15        +266 99.5 NC 0
16        +266 40.0 NC 0
17        +148 99.8 NC 3.0
18        +147 99.7 NC 3.0
19        +147 40.0 NC 3.0
20        +111 100.0 NC 4.5
21        +110 99.8 NC 4.5
22        - 13 99.9 15 0
23        - 14 100.0 15 0
24        + 11 99.9 12 0
25        + 11 99.6 12 0
26(cleanup) + 14 100.0 12 0
27        + 14 100.0 12 0
28
291, "red" +263 99.7 NC 0
30        +263 100.0 NC 0
31        +107 99.3 NC 4.5
32        +108 100.0 NC 4.5
33        + 24 99.8 10 0
34        + 24 100.0 10 0
35        - 9 99.7 10 4.5
36        - 9 100.0 10 4.5
37(cleanup) + 14 100.0 12 0
38        + 14 100.0 12 0
39
40(Multiple measurements to assess CNTR performance and drift. Note that
41CNTR measurements are repeatable within 1 ppm, so the nominal accuracy
42appears to be far too pessimistic.)
43
44
45Conclusion: these measurements suggest that, combined with parasitic
46capacitance, load capacitors of 12 pF make the crystal roughly half the
47trim range faster than 16 MHz.
atrf/ecn/ecn0003.txt
1Replace balun and filter with integrated balun
2
3
4We consider the balun needs for the AT86RF230 and the TI/Chipcon
5CC2520 we may try as a design alternative.
6
7For the AT86RF230, Atmel recommend baluns with integrated filter,
8namely the Wuerth 748421245 and the Johanson 2450FB15L0001, both in the
9AR86RF230 data sheet.
10
11For the CC2520, TI recommend a microstrip-based design for the balun,
12both in the CC2420/CC243x/CC2480 application note [1] and the reference
13design [2], without fully characterizing neither the transceiver's RF
14output nor all the components in the balun.
15
16Johanson lists the 2450FB15L0001 [3] and even a 2450BM15B0002 [3] for
17the TI/Chipcon CC2520, but not even Octoparts is able to find a
18distributor for these parts.
19
20Digging deeper, TI reveal more information about the balun in [5].
21
22Finally, TI somewhat hesitatingly acknowledge that the Murata balun
23LDB182G4510C-110 can be used for the CC2520 [6], with a 3.9 nH inductor
24connecting RF_P and RF_N, the balun coupled to GND not directly but via
2510 nF, and finally an LC low-pass filter with 1.5 nH and 2.2 pF at the
26output, for EMI compliance.
27
28Note that this also means that CC2520 and AT86RF230 both have an
29impedance of 100 Ohm.
30
31A balun without filter similar to the Murata part would be the Johanson
322450BL15K100.
33
34Manufacturer Part number Package Digi-Key Price/Qty
35Wuerth 748421245 0805-6 732-2230-1-ND 2.15/25
36                        732-2230-2-ND 0.753/4000
37Johanson 2450FB15L0001 0805-6 - -
38Johanson 2450BM15B0002 0805-6 - -
39Johanson 2450BL15K100 0805-6 712-1045-1-ND 0.488/10
40                        712-1045-2-ND 0.225/4000
41Murata LDB182G4510C-110 0603-6 490-5023-1-ND 0.325/10
42                        490-5023-2-ND 0.114/4000
43
44[1] http://www.ti.com/litv/pdf/swra098d
45[2] http://focus.ti.com/docs/toolsw/folders/print/cc2520em_refdes.html
46[3] http://www.johansontechnology.com/images/stories/ip/baluns/Balun_Filter_Combo_Matched_2450FB15L0001_v11.pdf
47[4] http://www.johansontechnology.com/images/stories/ip/baluns/balun_filter_combo_matched_2450bm15b0002_v2.pdf
48[5] http://www.ti.com/litv/pdf/swra236a
49[6] http://e2e.ti.com/support/low_power_rf/f/155/t/15910.aspx
50
51
52Conclusion: the Wurth balun appears to be the safest choice for
53prototyping the AT86RF230. Due to its high cost, a circuit with a
54discrete filter may be considered for larger quantities.
55
56For the CC2520, it's probably safest to directly try the Muarta balun
57with the recommended discrete filtering circuit.
atrf/ecn/ecn0004.txt
1Take into account layout considerations for RF
2
3
4There are a number of layout considerationg when designing RF systems
5that were not taken into account or not quantified when making the
6first design.
7
8- transmission line width
9
10  The microstrip [1] transmission line connecting the balun and filter
11  circuit with the antenna must be impedance-matched with the antenna.
12  The rule of thumb according to [2] is to make its width twice the
13  board thickness, in this case 0.8 mm or 31.5 mil.
14
15  The microstrip calculator at [3] also takes into account the
16  thickness of the copper, 1 oz, and yields a slightly narrower 57.5
17  mil or 1.46 mm.
18
19  A more elaborate calculator can be found at [4].
20
21- via spacing
22
23  Section 4.2 of [5] recommends a via spacing of no more than
24  Lvia = C/sqrt(Er)/Fres
25  where
26  C = the speed of light, 3*10^8 m/s
27  Er = the board's dielectric constant, 4.5 for FR-4
28  Fres = the resonance frequency, at least 24.5 GHz
29
30  We thus obtain Lvia = 5 mm.
31
32- component placing
33
34  [5] places DC blocking, balun, and filter close to the transceiver,
35  with only the feed line between the RF circuit and the antenna. Thus,
36  no changes are needed.
37
38- feed line termination
39
40  Point 12 of [6] warns us that we may need to terminate the
41  transmission line if it is longer than 20% of the signal's rise time.
42
43  Point 1 of [6] gives the rise time as 1/(10*Fclk), which looks as if
44  it's meant for digital signals. But we'll use it anyway.
45
46  [2] gives us the typical propagation delay for a microstrip as
47  150 pS/in.
48
49  This means that Lmax = 0.2*tr*v
50  with
51  tr = 1/24.5 GHz
52  v = 1 in/150 pS
53
54  We thus obtain Lmax = 1.4 mm
55
56  [2] suggests that the maximum unterminated stub is L(in) = tr(nS).
57
58  With tr = 1/(10*Fclk), we thus obtain Lmax = 1.04 mm.
59
60  Not sure if all this even applies to antennas. This needs looking to by
61  someone who understands about RF.
62
63[1] http://en.wikipedia.org/wiki/Microstrip
64[2] http://www.hottconsultants.com/techtips/rulesofthumb.html
65[3] http://www.cepdinc.com/calculators/microstrip.htm
66[4] http://mcalc.sourceforge.net/
67[5] http://www.ti.com/litv/pdf/swra236a
68[6] http://www.pcbmotif.com/home/index.php?option=com_content&view=article&id=104&Itemid=137
69
70
71Conclusion: the antenna feed line needs to be revised. The via spacing
72of the RF area needs to be examined. The recommended spacing may be
73beyond the capabilities of a DIY process, though.
ecn/INDEX
1Number Status Description
2------ ------- ---------------------------------------------------------------
30001 Edit Adjust balun component values
40002 Done Add load capacitors to 16 MHz crystal
50003 Edit Replace balun and filter with integrated balun
60004 Edit Take into account layout considerations for RF
ecn/ecn0001.txt
1Adjust balun component values
2
3
4Some of the components specified in the schematics were not at hand
5in my lab and were thus replaced with similar parts. Furthermore, the
6LED color was changed to ease visual identification:
7
8Component Norminal Board 1 Board 2
9--------------- --------------- --------------- -------------------
10LED (D1) red red red-orange
11C6/C7 0.82 pF 0.7 pF 0.9 pF
12L2/L3 4.3 nH 4.7 nH 4.7 nH
13
14The resulting spectrum should be examined to determine which variant
15performs better. Note that the discrete balun circuit shall be
16replaced with an integrated balun later on.
ecn/ecn0002.txt
1Add load capacitors to 16 MHz crystal (C14, C15)
2
3
4The crystal has a specified load capacitance of 8 pF. The AT86RF230 has
5an internal capacitor array that can be trimmed in 16 steps from 0 pF to
64.5 pF.
7
8The crystal has a nominal tolerance of 15 ppm at 25 C plus a temperature
9tolerance of 15 ppm. The following deviations from 1 MHz at CLKM were
10measured with the CNTR board, roughly at 20-25 C:
11
12Board Error Meas. accuracy C14/C15 Trim
13        (f, ppm) (ppm, nom.) (pF) (pF)
14--------------- --------------- --------------- ------------ --------
152, "orange" +268 99.6 NC 0
16        +266 99.5 NC 0
17        +266 40.0 NC 0
18        +148 99.8 NC 3.0
19        +147 99.7 NC 3.0
20        +147 40.0 NC 3.0
21        +111 100.0 NC 4.5
22        +110 99.8 NC 4.5
23        - 13 99.9 15 0
24        - 14 100.0 15 0
25        + 11 99.9 12 0
26        + 11 99.6 12 0
27(cleanup) + 14 100.0 12 0
28        + 14 100.0 12 0
29
301, "red" +263 99.7 NC 0
31        +263 100.0 NC 0
32        +107 99.3 NC 4.5
33        +108 100.0 NC 4.5
34        + 24 99.8 10 0
35        + 24 100.0 10 0
36        - 9 99.7 10 4.5
37        - 9 100.0 10 4.5
38(cleanup) + 14 100.0 12 0
39        + 14 100.0 12 0
40
41(Multiple measurements to assess CNTR performance and drift. Note that
42CNTR measurements are repeatable within 1 ppm, so the nominal accuracy
43appears to be far too pessimistic.)
44
45
46Conclusion: these measurements suggest that, combined with parasitic
47capacitance, load capacitors of 12 pF make the crystal roughly half the
48trim range faster than 16 MHz.
ecn/ecn0003.txt
1Replace balun and filter with integrated balun
2
3
4We consider the balun needs for the AT86RF230 and the TI/Chipcon
5CC2520 we may try as a design alternative.
6
7For the AT86RF230, Atmel recommend baluns with integrated filter,
8namely the Wuerth 748421245 and the Johanson 2450FB15L0001, both in the
9AR86RF230 data sheet.
10
11For the CC2520, TI recommend a microstrip-based design for the balun,
12both in the CC2420/CC243x/CC2480 application note [1] and the reference
13design [2], without fully characterizing neither the transceiver's RF
14output nor all the components in the balun.
15
16Johanson lists the 2450FB15L0001 [3] and even a 2450BM15B0002 [3] for
17the TI/Chipcon CC2520, but not even Octoparts is able to find a
18distributor for these parts.
19
20Digging deeper, TI reveal more information about the balun in [5].
21
22Finally, TI somewhat hesitatingly acknowledge that the Murata balun
23LDB182G4510C-110 can be used for the CC2520 [6], with a 3.9 nH inductor
24connecting RF_P and RF_N, the balun coupled to GND not directly but via
2510 nF, and finally an LC low-pass filter with 1.5 nH and 2.2 pF at the
26output, for EMI compliance.
27
28Note that this also means that CC2520 and AT86RF230 both have an
29impedance of 100 Ohm.
30
31A balun without filter similar to the Murata part would be the Johanson
322450BL15K100.
33
34Manufacturer Part number Package Digi-Key Price/Qty
35--------------- ----------------------- ------- --------------- ---------------
36Wuerth 748421245 0805-6 732-2230-1-ND 2.15/25
37                        732-2230-2-ND 0.753/4000
38Johanson 2450FB15L0001 0805-6 - -
39Johanson 2450BM15B0002 0805-6 - -
40Johanson 2450BL15K100 0805-6 712-1045-1-ND 0.488/10
41                        712-1045-2-ND 0.225/4000
42Murata LDB182G4510C-110 0603-6 490-5023-1-ND 0.325/10
43                        490-5023-2-ND 0.114/4000
44
45[1] http://www.ti.com/litv/pdf/swra098d
46[2] http://focus.ti.com/docs/toolsw/folders/print/cc2520em_refdes.html
47[3] http://www.johansontechnology.com/images/stories/ip/baluns/Balun_Filter_Combo_Matched_2450FB15L0001_v11.pdf
48[4] http://www.johansontechnology.com/images/stories/ip/baluns/balun_filter_combo_matched_2450bm15b0002_v2.pdf
49[5] http://www.ti.com/litv/pdf/swra236a
50[6] http://e2e.ti.com/support/low_power_rf/f/155/t/15910.aspx
51
52
53Conclusion: the Wurth balun appears to be the safest choice for
54prototyping the AT86RF230. Due to its high cost, a circuit with a
55discrete filter may be considered for larger quantities.
56
57For the CC2520, it's probably safest to directly try the Muarta balun
58with the recommended discrete filtering circuit.
ecn/ecn0004.txt
1Take into account layout considerations for RF
2
3
4There are a number of layout considerationg when designing RF systems
5that were not taken into account or not quantified when making the
6first design.
7
8- transmission line width
9
10  The microstrip [1] transmission line connecting the balun and filter
11  circuit with the antenna must be impedance-matched with the antenna.
12  The rule of thumb according to [2] is to make its width twice the
13  board thickness, in this case 0.8 mm or 31.5 mil.
14
15  The microstrip calculator at [3] also takes into account the
16  thickness of the copper, 1 oz, and yields a slightly narrower 57.5
17  mil or 1.46 mm.
18
19  A more elaborate calculator can be found at [4].
20
21- via spacing
22
23  Section 4.2 of [5] recommends a via spacing of no more than
24  Lvia = C/sqrt(Er)/Fres
25  where
26  C = the speed of light, 3*10^8 m/s
27  Er = the board's dielectric constant, 4.5 for FR-4
28  Fres = the resonance frequency, at least 24.5 GHz
29
30  We thus obtain Lvia = 5 mm.
31
32- component placing
33
34  [5] places DC blocking, balun, and filter close to the transceiver,
35  with only the feed line between the RF circuit and the antenna. Thus,
36  no changes are needed.
37
38- feed line termination
39
40  Point 12 of [6] warns us that we may need to terminate the
41  transmission line if it is longer than 20% of the signal's rise time.
42
43  Point 1 of [6] gives the rise time as 1/(10*Fclk), which looks as if
44  it's meant for digital signals. But we'll use it anyway.
45
46  [2] gives us the typical propagation delay for a microstrip as
47  150 pS/in.
48
49  This means that Lmax = 0.2*tr*v
50  with
51  tr = 1/24.5 GHz
52  v = 1 in/150 pS
53
54  We thus obtain Lmax = 1.4 mm
55
56  [2] suggests that the maximum unterminated stub is L(in) = tr(nS).
57
58  With tr = 1/(10*Fclk), we thus obtain Lmax = 1.04 mm.
59
60  Not sure if all this even applies to antennas. This needs looking to by
61  someone who understands about RF.
62
63[1] http://en.wikipedia.org/wiki/Microstrip
64[2] http://www.hottconsultants.com/techtips/rulesofthumb.html
65[3] http://www.cepdinc.com/calculators/microstrip.htm
66[4] http://mcalc.sourceforge.net/
67[5] http://www.ti.com/litv/pdf/swra236a
68[6] http://www.pcbmotif.com/home/index.php?option=com_content&view=article&id=104&Itemid=137
69
70
71Conclusion: the antenna feed line needs to be revised. The via spacing
72of the RF area needs to be examined. The recommended spacing may be
73beyond the capabilities of a DIY process, though.

Archive Download the corresponding diff file



interactive