Fluke Fluke 900a 电路.pdf

《Fluke Fluke 900a 电路.pdf》由会员分享，可在线阅读，更多相关《Fluke Fluke 900a 电路.pdf（8页珍藏版）》请在收音机爱好者资料库上搜索。

1、收音机爱好者资料库 Ra d i o Fa n s .CN The automated troubleshooter for dynamically testing ICs at speed, while still in circuit Tests digital ICs, at speed, while still in circuit Detects static, dynamic a known good IC serves as the reference for each test. This eliminates the need for general test soft- w

2、are, and even for device timing models. The 900 automatically analyzes the performance of the device under test, which can be operating at data rates of up to 20 MHz. Ten nanosecond timing resolution, over a full 20 - 200 nanose- cond range, ensures the 900s precision in fault capturing. Input signa

3、ls for the device under test can also be checked with the 900. Since it doesnt require any compli- cated set-ups or interpretation of results, you can begin using the 900 right away. The Fluke 900 isolates thefunctional board failure to the failing component. It captures dynamic and intermittent mal

4、functions as well as static component failures. It helps you isolate faults rapidly and accurately. And the 900 is effective as a stand-alone tool, or as a complement to an emulative tester or hot mockup. An emulative tester, such as the Fluke 9010A, 91OOA, or 90 Series, will quickly test the kernel

5、 of your UUT, and localize the fault to a subsection of the circuitry. The 900 is then used to isolate faults down to the defective component. Any of our emulative testers can also be used to generate the required logic activity on the UUT while the 900 tests each component. The interface buffer is

6、connected to the main- frame via a four-foot ribbon cable. A full set of clips is available to accommodate DIP style as well as surface mount small-outline-style digital components. Test esult.s are displayed in easy-to-interpret for- mats. Test conditions for each pin of the device under test may b

7、e reviewed in detail. 2 Fluke 900 收音机爱好者资料库 Ra d i o Fa n s .CN Fluke 900 Dynamic Troubleshooter Power to solve the tough problems Proprietary technology has been com- bined with practical features to give the 900 some unique capabilities. Among these are: Functional test quality-with in-circuit tes

8、t resolution Because each component is tested while operating in circuit, the 900 can detect a wide variety of faults: Timing-Related Faults. The 900 identi- fies a failure when the response of the device under test (DUT) differs from that of the reference part. The timing toler- ance on this compar

9、ison can be adjusted to compensate for normal circuit loading conditions which might affect the response of the DUT; its adjustable with 10 nanosecond resolution, across the full 20-200 nanosecond range. And fault cap- ture resolution is independent of the sig- nal speed on the DUT, effective from D

10、C to 20 MHz. Logic voltage levels can also be set by the user. Static Device Faults. The 900 detects logic faults, whether due to blown out- puts or internal chip failures. The real- time comparison test ensures that failures are detected immediately. Intermittent and Thermal Faults. The 900 can con

11、tinuously test the DUT while the circuit board is thermally cycled. Every pin on the DUT is monitored asynchronously, without clocking or sampling. Intermittent or transient faults are thus captured with the same 10 nsec resolution as timing and static faults. Short circuits and open circuits. The 9

12、00 monitors every pin on the DUT for logic status, logic activity, or an expected fre- quency. It flags any pin when there is no activity -whether due to a shorted node or an open lead. (The actual physical location of a short can then be found with a conventional shorts finder.) The 900 can also be

13、 set to ignore a failing pin, so the performance of the rest of the DUT can be verified. Isolates faults in “untestable” circuits Some testers have difficulty with com- ponents in “untestable” circuits, such as ICs with internal memory that cannot be reset. The 900 readily handles many such common t

14、esting problems. Feedback Loops. Fault isolation in feed- back loops can be difficult, since the fault seems to be everywhere at the same time. But the 900 verifies the operation of a chip in a loop, by using whatever inputs are present. Thus, working chips Upon detection of a mismatch between the o

15、utputs of the Reference Device and the cone- sponding output pins of the device under test, the tester ends the comparison and displays the failing output pin(s). that simply have garbled inputs will pass, but defective chips will be caught. The 900 also measures the amount of time elapsed until a f

16、ault occurs. This can be used to identify the first chip to fail in cases of apparent multiple faults. Bus Devices. Testing bus devices with conventional tools is difficult, because any device connected to the bus can gen- erate activity. The 900 readily tests bus devices, since it automatically gat

17、es its testing with the Output Enable signal at the DUT The 29-channel Gate function can be used to further ensure that testing is performed only when the DUT, itself, is active on the bus. Components with Internal Memory. The 900 tests devices with internal mem- ory, such as latches and counters. I

18、t has built-in automatic synchronization rou- tines to put the reference device and the DUT in the same state before the test begins-even when the DUTs internal memory cannot be initialized because control pins are strapped in an inactive state. Hardwired Inputs data rates are programmable, up 25 MH

19、z, accurate to within 0.5 % . Duty to 19,200 baud. Test Sequences can be cycle is also measured to characterize downloaded and test results can be clocks and other signals at the UUT. uploaded to your PC. An optional IEEE-488 interface is also available. This results screen indicates that thefrequen

20、cy expected at pin 2 of this bus controller chip falls well within thespecified tolerances. User-definable trigger. A two-word trig- ger function can be defined from a com- bination of signals on the DUTs 28 leads plus an external node, to start a test (sim- ilar to a conventional logic analyzer). T

21、rigger event detection can be used to start the comparison test, or as a power- ful diagnostic tool for finding complex data patterns. User-definable gate. A one-word gate function can be defined from a combina- tion of the DUTs 28 pins plus an external node. This can be used to mask out inde- termi

22、nant board conditions, such as bus contention. It can also be used with the frequency counter to measure complex functions, such as the duty cycle and fre- quency of accesses to memory. User-definable test cycle. The 900 will issue an automatic RESET signal to the UUT at the start of the test for ea

23、ch device. The user can specify the polarity, duration, and timing, relative to the comparison test. The duration of the syn- chronization, initialization, and com- parison phases can also be specified by the user. This ensures that the 900 is applicable to the widest variety of devices and circuit

24、configurations. Nonvolatile RAM cartridges. Removable memory modules will hold predefined Test Sequences and test results. You can The 900 provides the ability to exchange files with a (personal) computer via its standard RS-232C port. A Test Sequence may be uploaded to facilitate editing and to add

25、 operator instruc- tions as well as comments. Automatic logging of test results. The 900 can automatically save test results, including a record of each step in the diagnostic process. You can use a mem- ory cartridge, or a device connected to one of the remote interfaces. This is useful for subsequ

26、ent analysis of failures, and refinement of Test Sequences. Friendly operator interface. The 900 has a soft-key based menu-driven interface. This supports complete operation of the unit, plus creation and editing of test sequences. The built-in clock and calen- dar can be used to date stamp test res

27、ults. Principles of operation Dynamic comparison tests The Fluke 900 operates on the princi- ple of Dynamic Hardware Reference Comparison. This means that the real- time operation of the DUT is compared to that of a known good Reference Device (RD) , which is inserted in a zero- insertion-force sock

28、et on the tester. The 900 continuously monitors every pin on the DUT, asynchronously, without clocking or sampling. TESTRESULTS FLUKE900 The known good reference device in the tester is subjected to the same input signals as those observed at the inputs of the device under test; the output signals g

29、enerated by both devices are compared in real time. The 900 determines which pins are inputs and which are outputs, at any given moment; tri-state and bidirec- tional pins are handled automatically. It is thus able to apply the same signals that are on the input leads of the DUT to the inputs of the

30、 RD. Simultaneously, the 900 compares the outputs of these two chips, and flags any differences as an error. Comparisons are made while the DUT is operating at speed, for data rates of up to 20 MHz. Timing resolution on this comparison is 10 nsecs, over a 20 - 200 nsec range. 4 Fluke 900 收音机爱好者资料库 R

31、a d i o Fa n s .CN Some logic activity is necessary on the UUT in order for the 900 to test each IC . The 900s reset function can be used with a UUTs built-in self test routine to force a few machine cycles of activity under the control of the tester. A hot mockup can also be used to sustain activit

32、y on subassemblies which do not function standing alone. Alternatively, an emu- lative tester, such as the Fluke 9010A, 9100A, or 90 Series, can be used to stimu- late activity on the UUT. An emulative tester is also effective at first localizing the fault to a subsection of the UUT, especially when

33、 the UUT locks up with a kernel fault. In general, an RD will not experience the exact same loading as the DUT does in - circuit. This typically results is slight differences in the response times of the RD and the DUT. The 900 accommo- dates this by using a specifiable value for the allowable diffe

34、rence (a “fault win- dow”) in response times for the two chips-adjustable from 20 nsec to 200 nsec, in 10 nsec steps. Logic threshold levels can also be spec- ified, at any level between 0 and 5 volts, with 100 mV resolution. The fault window and logic threshold voltage levels together define a Perf

35、omn- ance Envelope against which the DUT is compared to the RD. To determine the Performance Enve- lope for various DUTs, a known good board should first be checked. Any ICs on this good board which require adjust- ments in the default test parameters should be noted. Usually less than 10 % of the D

36、UTs will require any adjust- ments to the 900s default test param- eters. These adjustments can be saved in the corresponding step of a test sequence for the UUT. v( THRESHOLD MLTS REFERENCEDEVICE I II w - TIME FAULT WINDOW CHECKEDAGAINST ACTUALTIMING DISCREPANCY The timing discrepancy between the o

37、utput responses of the Reference Device (RD), operat- ing in ideal conditions, and the Device Under Test (DUT) is masked using a window, called the F-Mask. The output of the DUT is expected to ;zzkoith the output of the RD at the end of Testing, the complete cycle The test for an IC includes the fol

38、low- ing steps: (1) the RD is tested -using predefined test patterns built into the 900, or optional user-specified test patterns; (2) the DUT clip is checked to verify that it is properly connected; (3) a RESET pulse is issued to the UUT, at some point during the synchronization interval; (4) the c

39、omparison test is performed; (5) test results are displayed. TIME k_-+-+-+-_+_- RO TEST CLIP TEST SYNCHRO- COMPARISON RESULTS NIZATION, TEST DISPLAYED you can add test patterns for your custom parts. Any RD may also be checked at speed, by clip- ping onto a good board to compare against the device i

40、n the ZIF socket (a reverse comparison). This shows the test varameters to test a complex device like a PAL. Only two parameters need to be modified from the “default”settings: (1) the trigger condition (TRIG), and (2) the syn- chronization time parameter STIME. Param- eters can be modified using th

41、e soft-labeled function keys. 7. Can the RD and board activity be stored on a cartridge so we dont need RD hardware? This is undesirable because most stored-pattern testers operate in the kilo- hertz range of speed. By using actual hardware RDs, the 900 achieves full 20 MHz operation. Specifications

42、 8. Are surface mounted devices testable? Yes. We offer clips for J-lead SO ver- sion devices that have dual sided nar- rowly spaced pins. The equivalent DIP version IC is used as the Reference Device. 9. Does clipping onto the DUT affect its performance? The loading per pin is 10 KQ and 30 pf. This

43、 normally does not affect TTL or CMOS devices. Direct clipping on a crystal oscillator will adversely affect its signal. 10. How do I test boards that operate faster than the Fluke 996 20 MHz specification? Most high speed clocks are immedi- ately divided down to lower frequencies for routing around

44、 a board. A device counting down the high frequency will have outputs that are half this value, or less; they can be measured with the 900s built-in frequency counter. Dusing the execution of a Te.st Sequence called F900, the tester is testing a PAL with reference designation FU3. The 900 tests PALS

45、 and other custom and semi-custom components without requiring knowledge about the internal programming of these types of devices. Il. What types of devices can I test on a board? The 900 tests devices with digital inputs and outputs, up to 28 pins with 5V Vcc. For devices with larger pin counts, va

46、rious signal condition tests pro- vide indirect verification of their opera- tion. These tests include high, low, and active status, frequency measurement, and event detection (two-word trigger on any combination of DUT pin states plus one external node). 12. How many devices are in the built-in dev

47、ice library? Approximately 350 device numbers and their standard comparison test parameters are stored in the 900. These include 7400 series TTL, most RAM and PROM families, plus many micro- processor support devices of 28 pins or less. You can easily add devices, such as PALS, to this reference lib

48、rary. Troubleshooting 13. Can the 900 troubleshoot various microprocessor architectures such as the 8085, 68000 or multiprocessor boards? Yes, this comparison test technique is applicable to virtually any architecture. The 900 looks at each device indepen- dently; signals from each pin on the DUT ar

49、e monitored asynchronously, without clocking or sampling. 14. How do I handle devices over 28 pins? Use the Fluke 900 to evaluate up to 28-pin devices and determine where the fault is not. Larger devices can be indi- rectly evaluated with the 900s condition tests, as noted in the answer to question 11. Larger chips which are in sockets or on the microprocessor bus can usually be tested with another approach, such as using the Fluke 9010A or 90 Series. 收音机爱好者资料库 Ra d i o Fa n s .CN 15. Can the 900 detect differences between S, LS and HC type parts? Can it distinguish between