San Pao’sHardware Development Learning Notes
Sharing knowledge to enhance skills♥Exchanging experiences to strive for excellence
Table of Contents
1. Description of Anomalies2. Investigation and Analysis Process3. Problem Identification4. Solutions5. Complaints about the validation of replacement materials
✍
Understanding the differences between crystal oscillators and crystals, design selection
Recommended previous notes👆
The specific reason for replacing the crystal oscillator model is not detailed; I just sent the original crystal model and specifications to the market manager of the replacement material.
Then I received feedback with a specific replacement model and specifications.
The performance parameters of the crystal oscillator: nominal frequency, operating frequency, temperature, adjustment frequency deviation, package size, pin information, typical application circuits, etc., are all identical; they are practically copied from the original material parameters.
After a few days, the samples were also provided.
During overtime one night, San Pao used his clumsy soldering skills to replace the crystal oscillator with the substitute, and without testing various performances, directly proceeded to board verification.
The result was abnormal. It really is only possible to work in the afternoon, as something is bound to go wrong sooner or later!!
Crystal oscillator circuit schematic
Crystal oscillator output200MHz differential LVDS clock, serving as the system clock for the FPGA.
1. Description of Anomalies
After powering on, using the emulator to download the bitstream to the FPGA prompts an Error, with the error code in Vivado being approximately ERROR:[Labtools 27-3176]
Searching online revealed that the ERROR:[Labtools 27-3176] error code has been encountered by others.
Preliminary judgment indicated a lack of clock signal.
2. Investigation and Analysis Process
Crystal oscillator circuit schematic (focused version) is as follows:
1. Using the control variable method, it was determined that the issue was with the replacement crystal oscillator (everything was normal before the replacement), but the specific cause was not yet identified;
2. Using a differential oscilloscope to measure pins 4 and 5 of the crystal oscillator (across R2) showed no clock waveform;
3. Measuring the voltage on pins 1 and 6 of the crystal oscillator was normal at 3.3V;
4. Could it be that this crystal oscillator is faulty, and replacing it still does not work (voltage and enable are normal, but there is no clock waveform)?
5. The issue must be with pin 2 being NC. I asked the original manufacturer, and the feedback was that pins 1 and 2 should not be connected, and pin 2 should be left floating.
It turns out that this replacement crystal oscillator is programmable (pins 1 and 2 can be connected), while the previous one was a fundamental design (pins 1 and 2 can be connected).
Oh no!
Why is this not indicated in the specifications as a programmable type? San Pao wasted an entire night of overtime on this!
The completeness of domestic component specifications really needs improvement, but San Pao’s experience has increased just a little bit.
Since I have already worked overtime, I might as well continue for a while.
6. Resoldering the crystal oscillator and leaving pin 2 floating allowed the program to be downloaded, and the clock waveform was also measurable.
(Unfortunately, the crystal oscillator circuit design did not initially consider the programmable type, and this board can only use the fundamental type crystal oscillator.)
3. Problem Identification
Common issues with domestic component specifications!!!(Programmable and fundamental crystal oscillator circuit designs cannot be identical; the programmable pins need to be distinguished)4. Solutions
1. The original manufacturer provided a fundamental crystal oscillator compatible with the existing circuit design;2. The schematic design of the crystal oscillator section was improved by pulling up pins 1 and 2 to 3.3V with pull-up resistors; this accommodates the differences in enable pins being pin 1 or pin 2 from different manufacturers, and this design also considers programmable crystal oscillators; the schematic is as follows:
Comparison of programmable and fundamental crystal oscillators is as follows (AI generated); for more details, you can refer to the materials yourself.
5. Complaints about the validation of replacement materials
Since San Pao entered the hardware development field five or six years ago, the work has always involved the validation of replacement materials, especially during the first two years of employment, where I frequently interacted with domestic component manufacturers. To be honest, back then, I was inexperienced and could only do menial tasks.During that time, I was validating replacement materials one after another, writing report after report;comparing performance parameters, comparing package sizes, sorting out business supply chain relationships, measuring core indicators, conducting high and low-temperature tests, ALT tests, and various other tests.Initially, the validation of replacement materials was for domestic substitution, and later it became cost reduction, becoming increasingly competitive.Some say that the validation work for replacement materials is thankless, which is somewhat true;however, as long as it is not solely focused on validating replacement materials, there are still gains; if it is solely focused on validation, then it is truly a waste of time.We just adhere to one principle: if the leader asks for validation, we will do it well; if we say the validation results are acceptable, it may not necessarily enter the supply chain; if we say the validation results are not acceptable, it basically cannot enter the supply chain.So, we have a friendly relationship with the original manufacturer’s market managers, who frequently visit us; there is no genuine concern for our well-being, to be blunt, they are just trying to gauge our opinions.Based on the principle of working for the company while being compensated; and based on the principle of being honest and having a clear conscience; and based on the attitude that a simple bowl of noodles is more comforting than a feast;I have not accepted any invitations from suppliers or market managers.In fact, I would love to go! (internally crying while refusing)It is hard to resist the temptation of delicious food.
More Notes
V7 FPGA “Serial Port” Issues: Just Simple Mistakes, Want to Curse…
Zynq7 SoC Board Fails to Start at Low Temperatures (Below -40℃): Whose Fault Is It???
RS-485
RS-232
Introduction to Common Level Shifting Circuits
Understanding SPI Interface in One Article
Understanding I2C Bus in One Article
Understanding the Differences Between Crystal Oscillators and Crystals, Design Selection
Understanding the Power-Up Timing of Xilinx Zynq7000 and 7 Series FPGAs in One Article
Understanding Bootstrap Capacitors in Buck Circuits in One Article
What Are the Differences Between LDO and DCDC?
Understanding DCDC Basics in One Article
Understanding LDO Basics in One Article
What Research Is There on Pull-Up and Pull-Down Resistors?
Why Is It Common to Place 1uF and 0.1uF Capacitors in Parallel in Circuits?
This is a soft grindof the public account, the author San Pao’s hardware development learning notes and experience sharing
Learningsharing to enhance skills, exchanging experiences to strive for excellence