Root/prod/doc/analysis.hmac

Source at commit 29ff186882c0bb78138bfb32e30b686bec44824e created 12 years 10 months ago.
By Werner Almesberger, prod/doc/: make END macro generate the date and autor line; use current date
1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
2<HTML>
3<TITLE>Production and testing: Fault analysis</TITLE>
4<BODY bgcolor="#ffffff" link="#000000" vlink="#404040">
5
6<INCLUDE file="style.inc">
7
8<PAGE_BAR title="Production and testing">
9  <PAGE_ITEM href="setup.html">Software setup</PAGE_ITEM>
10  <PAGE_ITEM href="flash.html">Flashing</PAGE_ITEM>
11  <PAGE_ITEM href="test.html">Functional test</PAGE_ITEM>
12  <PAGE_CURR href="analysis.html">Fault analysis</PAGE_CURR>
13</PAGE_BAR>
14
15<SECTION_BAR>
16  <SECTION_ITEM href="#orientation">Component orientation</SECTION_ITEM>
17  <SECTION_ITEM href="#voltages">Voltages</SECTION_ITEM>
18  <SECTION_ITEM href="#clock">Clock frequency</SECTION_ITEM>
19</SECTION_BAR>
20
21
22<!-- ====================================================================== -->
23
24
25<SECTION ref="orientation" title="Component orientation">
26
27
28<!-- ====================================================================== -->
29
30
31<SECTION ref="voltages" title="Supply voltages">
32
33The transceiver has three voltage domains:
34<UL>
35  <LI>The supply and I/O voltage, which is nominally 3.3 V in
36    <B>atben</B> and <B>atusb</B>,
37  <LI>the digital (core) supply, which is nominally 1.8 V, and
38  <LI>the analog (RF) supply, which is nominally 1.8 V.
39</UL>
40<BR>
41On <B>atusb</B>, there is also the USB voltage domain at nominally 5.0 V.
42<P>
43Voltages should be tested in the following order: USB, then I/O, then
44digital, and finally analog. The table below gives the permissible
45ranges. Any voltages outside of these ranges indicate a problem.
46<P>
47<TABLE frame="border" cellpadding="2">
48  <TR><TH>Domain<TH>Nominal<TH>Minimum<TH>Maximum
49  <TR><TD>USB<TD>5.0 V<TD>4.5 V<TD>5.25 V
50  <TR><TD>I/O<TD>3.3 V<TD>3.0 V<TD>3.6 V
51  <TR><TD>Digital<TD>1.8 V<TD>1.7 V<TD>1.9 V
52  <TR><TD>Analog<TD>1.8 V<TD>1.7 V<TD>1.9 V
53</TABLE>
54<P>
55
56
57<!-- ---------------------------------------------------------------------- -->
58
59
60<SUBSECTION title="Measurement setup">
61
62The measurements should be performed with a digital multimeter.
63The transceiver's analog and digital supplies (1.8 V) are only
64activated when sending or receiving.
65To enable all voltage domains, put the transceiver in receive mode:
66<PRE>
67atrf-txrx
68</PRE>
69or
70<PRE>
71atrf-txrx -d net:ben
72</PRE>
73Exit with Ctrl-C.
74<P>
75To produce periodic transmissions in addition to enabling all voltage
76domains, use
77<PRE>
78atrf-txrx -p 3 -E 0
79</PRE>
80or
81<PRE>
82atrf-txrx -d net:ben -p 3 -E 0
83</PRE>
84Again, exit with Ctrl-C. Note that the transmissions may disturb nearby
85equipment operating in the 2.4 GHz band, such as 802.11 networks. This
86can be prevented by shorting the antenna to ground.
87<P>
88In case the board does not accept commands, only the USB and I/O voltage
89can be checked. If they are correct, proceed with checking the clock.
90
91
92<!-- ---------------------------------------------------------------------- -->
93
94
95<SUBSECTION title="Measurement points on atben">
96
97The supply voltages on <B>atben</B> can be measured at the terminals of
98components as shown in this table:
99<P>
100<TABLE frame="border" cellpadding="2">
101  <TR><TH>Domain<TH>Voltage<TH>Component
102  <TR><TD>I/O<TD>3.3 V<TD>C3, C6
103  <TR><TD>Digital<TD>1.8 V<TD>C5
104  <TR><TD>Analog<TD>1.8 V<TD>C4
105</TABLE>
106<P>
107Ground can be accessed at the cover of the crystal.
108Note that the fiducials are not connected to ground.
109<P>
110This image shows the location of the measurement points:
111<P>
112<A href="atben-B.png"><IMG src="atben-B-small.png"></A>
113
114
115<!-- ---------------------------------------------------------------------- -->
116
117
118<SUBSECTION title="Measurement points on atusb">
119
120The supply voltages on <B>atusb</B> can be measured at the terminals of
121components as shown in this table:
122<P>
123<TABLE frame="border" cellpadding="2">
124  <TR><TH>Domain<TH>Voltage<TH>Component
125  <TR><TD>USB<TD>5.0 V<TD>C1
126  <TR><TD>I/O<TD>3.3 V<TD>C2, C10, C13
127  <TR><TD>Digital<TD>1.8 V<TD>C12
128  <TR><TD>Analog<TD>1.8 V<TD>C11
129</TABLE>
130<P>
131Ground can be accessed at the cover of the crystal, at the shield of the
132USB connector, or at the test point P11.
133Note that the fiducials are not connected to ground.
134<P>
135This image shows the location of the measurement points:
136<P>
137<A href="atusb-B.png"><IMG src="atusb-B-small.png"></A>
138
139
140<!-- ====================================================================== -->
141
142
143<SECTION ref="clock" title="Clock frequency">
144
145The precision of the crystal oscillator is crucial for
146operation. Anomalies are easy to detect with even a low-cost oscilloscope
147and pinpoint specific problems and help to select further analysis steps.
148<P>
149The crystal used in <B>atben</B> and <B>atusb</B> has a nominal tolerance
150of +/&minus; 15 ppm at 22-28 C. Low-cost oscilloscopes typically have a timing
151accuracy of
152+/&minus; 100 ppm, which means that only major excursions can be detected by
153measuring the clock output with such an instrument. Full-speed USB only
154requires an accuracy of +/&minus; 2500 ppm.
155We can therefore consider all results within a range of +/&minus; 1000 ppm as
156sufficient, and perform more precise measurements by other means. This
157applies to <B>atben</B> as well as to <B>atusb</B>.
158
159
160<!-- ---------------------------------------------------------------------- -->
161
162
163<SUBSECTION title="Measuring the clock on atben">
164
165<B>atben</B> normally does not output a clock signal. A 1 MHz clock
166can be enabled with the following command:
167<PRE>
168atrf-txrx -d net:ben -C 1
169</PRE>
170This configures <B>atben</B> as a promiscuous receiver. The reception
171of any IEEE 802.15.4 frame or pressing Ctrl-C will terminate the command.
172<P>
173The clock signal (CLKM) is available on the test pad shown here:
174<P>
175<A href="atben-A.png"><IMG src="atben-A-small.png"></A>
176<P>
177<TABLE frame="border" cellpadding="2">
178<TR><TH align="left">Clock<TH align="left">Action
179<TR><TD>0 Hz<TD>Check voltages; check that the clock is enabled;
180  check for shorts around crystal; check connectivity of crystal
181<TR><TD>0.999-1.001 MHz, ~3.3 Vpp<TD>Perform precision measurement with
182  <B>atrf-xtal</B>
183<TR><TD>Other<TD>Check voltages; check for contamination around crystal
184</TABLE>
185<P>
186
187
188<!-- ---------------------------------------------------------------------- -->
189
190
191<SUBSECTION title="Measuring the clock on atusb">
192
193The transceiver provides the clock for the microcontroller in <B>atusb</B>.
194A clock signal is therefore always available. Immediately after reset,
195the transceiver generates a 1 MHz clock. When the microcontrolled comes out
196of reset, it raises the transceiver's clock output to 8 MHz and then
197enables USB.
198<P>
199The clock signal is available at the terminals of several components,
200either as the direct output from the transceiver (CLKM) or after passing
201a low-pass filter (CLK):
202<P>
203<A href="atusb-A.png"><IMG src="atusb-A-small.png"></A>
204<P>
205<TABLE frame="border" cellpadding="2">
206<TR><TH align="left">Clock<TH align="left">Action
207<TR><TD>0 Hz<TD>Check voltages; check for shorts around crystal; check
208  connectivity of crystal
209<TR><TD>0.999-1.001 MHz, ~3.3 Vpp<TD>Check presence of firmware; check for
210  shorts on SPI signals; check connectivity of SPI signals
211<TR><TD>7.992-8.008 MHz, ~3.3 Vpp<TD>Perform precision measurement with
212  <B>atrf-xtal</B>
213<TR><TD>Other<TD>Check voltages; check for contamination around crystal
214</TABLE>
215<P>
216Note that, if testing a board into which no boot loader has been flashed
217yet, the clock frequency should be 1 MHz. If an unsuccessful attempt has
218been made to flash the boot loader, the frequency may be 1 MHz or 8 MHz,
219depending on how much code was successfully flashed.
220
221
222<!-- ---------------------------------------------------------------------- -->
223
224
225<SUBSECTION title="Precision measurements">
226
227The clock frequency of <B>atben</B> can be measured with an accuracy
228of about +/&minus; 100 ppm using the program <B>atrf-xtal</B>. <B>atrf-xtal</B>
229runs directly on the Ben and measures the duration of packet transmissions.
230The transmission time depends on the bit clock which is in turn derived
231from the oscillator.
232<PRE>
233atrf-xtal 100
234</PRE>
235The number reported is the number of poll loops the CPU counted. This
236value should be compared to a reference count obtained with a known to
237be good <B>atben</B> board on the same Ben at a comparable temperature.
238<P>
239<TABLE frame="border" cellpadding="2">
240<TR><TH align="left">Difference<TH align="left">Action
241<TR><TD>&gt; +/&minus; 50 ppm<TD>Correct operation
242<TR><TD>&lt; &minus;80 ppm<TD>Check soldering of capacitors;
243  check for contamination around crystal
244<TR><TD>&gt; +120 ppm<TD>idem
245<TR><TD>Other<TD>Divergence can be compensated by adjusting trim value
246</TABLE>
247
248<END author="Werner Almesberger" date="<GEN_DATE>">
249</BODY>
250</HTML>
251

Archive Download this file



interactive