API 解答论坛 - 最近的话题

TXU0204: TXU0204DTRR MSL3 防潮等級确认

Betty Song — Tue, 11 Aug 2026 09:47:05 GMT

Part Number: TXU0204

你好！

我是KAGA的采购BETTY.

目前从TI购入的物料：TXU0204DTRR ，KAGA CODE:14-10204000-000, 根据工厂的反馈和生产测试结果: 经过烘烤后，不良率有下降。因目前来料是按MSL 1交货的，如果每次使用前都要烘烤的话会额外增加费用，请帮忙确认后续可否按MSL 3包装出货（真空包装＋干燥剂＋湿度指示卡），谢谢！

RE: TXU0204: TXU0204DTRR MSL3 防潮等級确认

Links — Tue, 11 Aug 2026 09:52:50 GMT

您好，

该问题，请联系客服部门获得支持

www.ti.com.cn/.../customer-support.html

RE: MSPM0L1228: MSPM0L1228的PB2引脚未正常拉低

Eirwen — Fri, 17 Jul 2026 02:52:32 GMT

是指更换新的MSPM0L1228之后就可以正常动作吗

MSPM0L1228: MSPM0L1228的PB2引脚未正常拉低

Dong Liangrong — Wed, 15 Jul 2026 02:23:48 GMT

Part Number: MSPM0L1228

PB2引脚连接3850看门狗set1，在进入boot前本应拉低来关闭3850看门狗防止boot时被复位，实际PB2引脚未被拉低，导致进入boot后复位，boot失败。

排查后发现以下现象，请参考：

断开故障MCU_PB2与外部电路的连接，PB2可以正常拉低；连接PB2与外部电路（实际只有一个10k上拉电阻到3.3V），PB2无法拉低。
ABA交换故障件和正常的MCU，故障现象跟着MCU走。

RE: MSPM0L1228: MSPM0L1228的PB2引脚未正常拉低

Eirwen — Wed, 15 Jul 2026 02:25:58 GMT

已经收到了您的案例，调查需要些时间，感谢您的耐心等待。

RE: F29H859TU-Q1: F29H859TU8QZEXRQ1型号MCU的参考电路

Taylor — Thu, 09 Jul 2026 03:30:50 GMT

请参考

F29H85X-SOM-EVM Design Files (Rev. D)

F29H85X-SOM-EVM Schematic (Rev. C)

F29H859TU-Q1: F29H859TU8QZEXRQ1型号MCU的参考电路

ZIXIONG YU — Thu, 09 Jul 2026 03:24:38 GMT

Part Number: F29H859TU-Q1

可以提供该MCU的256pin型号的参考电路吗，仅作为设计参考，谢谢！

RE: F29H859TU-Q1: F29H859TU8QZEXRQ1型号MCU的参考电路

Taylor — Thu, 09 Jul 2026 03:26:45 GMT

您好，

已经收到了您的案例，调查需要些时间，感谢您的耐心等待。

LMZ14203H: 请提供COD for LMZ14203HTZ/NOPB

Elena Chen — Tue, 09 Jun 2026 02:31:01 GMT

Part Number: LMZ14203H

请提供COD for TI MPN:LMZ14203HTZ/NOPB

RE: LMZ14203H: 请提供COD for LMZ14203HTZ/NOPB

Lydia — Tue, 09 Jun 2026 02:40:46 GMT

您好，

TI 论坛旨在解决客户在使用TI产品过程中遇到的技术问题。关于您的COD请求，请点击下面的链接联系客户支持部门，会有客服为您提供帮助。

https://ticsc.service-now.com/csm

RE: TMS320C6678: C6678使用OPENMP编译问题

Eirwen — Thu, 14 May 2026 08:31:45 GMT

关于这个报错，请尝试参考下面链接：

TMS320F28027: undefined symbols CTRL_initCtrl first referenced in file .main.obj - C2000 microcontrollers forum - C2000™︎ microcontrollers - TI E2E support forums

TMS320F28069M: unresolved symbol _ButtonCenter, first referenced in ./Sources/Menu_Haptic_Bike/Menu_Haptic_Bike.obj - C2000 microcontrollers forum - C2000™︎ microcontrollers - TI E2E support forums

TMS320F28379D: Unresolved symbol, first referenced in cla file - C2000 microcontrollers forum - C2000™︎ microcontrollers - TI E2E support forums

TMS320C6678: C6678使用OPENMP编译问题

w lz — Wed, 13 May 2026 03:57:26 GMT

Part Number: TMS320C6678

您好！

我想要在现有的工程里添加openmp，但是编译报错：

unresolved symbol GOMP_atomic_end, first referenced in ./app/mcsDlinkTest.obj；

unresolved symbol GOMP_atomic_start, first referenced in ./app/mcsDlinkTest.obj；

unresolved symbol GOMP_parallel_end, first referenced in ./app/mcsDlinkTest.obj；

unresolved symbol GOMP_parallel_start, first referenced in ./app/mcsDlinkTest.obj；

unresolved symbol omp_get_num_threads, first referenced in ./app/mcsDlinkTest.obj；

unresolved symbol omp_get_thread_num, first referenced in ./app/mcsDlinkTest.obj

已开启--openmp选项

所用products版本如下：

代码如下：

#include <ti/omp/omp.h>

void ompTest(void)
{
int i;
int n = 0; // 初始化累加器

#pragma omp parallel for reduction(+:n)
for(i = 0; i < 1000; i++)
{
n += (i + i);
}

return;
}

请问我还需要增加什么操作，才能让我的工程顺利编译通过？

RE: TMS320C6678: C6678使用OPENMP编译问题

Eirwen — Wed, 13 May 2026 04:49:41 GMT

已经收到了您的案例，调查需要些时间，感谢您的耐心等待。

DLPC900: DLP670S: In DLPC900 + DLP670S, according to the official provided API, when performing image writing in the On-The-Fly mode, only one image can be written.

Zhenming Shen — Sat, 09 May 2026 02:16:08 GMT

Part Number: DLPC900

Here is the image download function that I wrote based on 6500. Since the code I wrote was for a single image, I first expanded the single image into a 24-bit image before proceeding with the download:

int DownloadCompressedImage(const std::string& bmpPath, unsigned char imageIndex, int dmd_width, int dmd_height)
{
//检查文件是否存在
if (!dlp::File::Exists(bmpPath)) {
std::cerr << "File does not exist: " << bmpPath << std::endl;
return -1;
}

// 读取 BMP 信息
BMPInfo info = ReadBMP(bmpPath);
if (info.width == 0 || info.height == 0) {
std::cerr << "Failed to read BMP or unsupported format." << std::endl;
return -2;
}
if (info.bitCount != 1 && info.bitCount != 24) {
std::cerr << "Only supporte 1-bit and 24-bit BMP" << info.bitCount << " bits." << std::endl;
return -3;
}

int imgWidth = info.width;
int imgHeight = abs(info.height);
if (imgWidth != dmd_width || imgHeight != dmd_height) {
std::cerr << "Resolution mismatch: " << imgWidth << "x" << imgHeight
<< " vs " << dmd_width << "x" << dmd_height << std::endl;
return -4;
}

// 如果1位图像为24位图像，如果是24位图像直接使用原始数据
std::vector<uint8_t> rgb24Buffer;
const uint8_t* imageDataForCompression = nullptr;
size_t dataSize = 0;

if (info.bitCount == 1) {
rgb24Buffer = Convert1To24Bit(info);
imageDataForCompression = rgb24Buffer.data();
dataSize = rgb24Buffer.size();
std::cout << "Converted 1-bit BMP to 24-bit RGB." << std::endl;
}
else {
// 直接使用 info.rawData（注意 BMP 存储为 BGR，Splash_createImage 内部会做 BGR->RGB 和垂直翻转）
imageDataForCompression = info.rawData.data();
dataSize = info.rawData.size();
}

// 构造虚拟的 BMP 完整文件缓冲区（用于 Splash_createImage）为了
// 使用 Splash_createImage，需要构造一个临时的 24 位 BMP 内存镜像
// 构造 BMP 文件头（14字节）+ 信息头（40字节）+ 像素数据
int bmpHeaderSize = 14 + 40;
int rowSize24 = (imgWidth * 3 + 3) & ~3;
int pixelDataSize = rowSize24 * imgHeight;
std::vector<uint8_t> fakeBMP(bmpHeaderSize + pixelDataSize);

// 文件头
uint16_t bfType = 0x4D42;
uint32_t bfSize = bmpHeaderSize + pixelDataSize;
uint32_t bfOffBits = bmpHeaderSize;
memcpy(fakeBMP.data(), &bfType, 2);
memcpy(fakeBMP.data() + 2, &bfSize, 4);
memcpy(fakeBMP.data() + 10, &bfOffBits, 4);

// 信息头
uint32_t biSize = 40;
int32_t biWidth = imgWidth;
int32_t biHeight = imgHeight;
uint16_t biPlanes = 1;
uint16_t biBitCount = 24;
uint32_t biCompression = 0;
uint32_t biSizeImage = pixelDataSize;
int32_t biXPelsPerMeter = 0;
int32_t biYPelsPerMeter = 0;
uint32_t biClrUsed = 0;
uint32_t biClrImportant = 0;
uint8_t* header = fakeBMP.data() + 14;
memcpy(header, &biSize, 4);
memcpy(header + 4, &biWidth, 4);
memcpy(header + 8, &biHeight, 4);
memcpy(header + 12, &biPlanes, 2);
memcpy(header + 14, &biBitCount, 2);
memcpy(header + 16, &biCompression, 4);
memcpy(header + 20, &biSizeImage, 4);
memcpy(header + 24, &biXPelsPerMeter, 4);
memcpy(header + 28, &biYPelsPerMeter, 4);
memcpy(header + 32, &biClrUsed, 4);
memcpy(header + 36, &biClrImportant, 4);

// 填入像素数据（需要保持每一行对齐到4字节，且BGR顺序）
const uint8_t* src = (info.bitCount == 1) ? imageDataForCompression : imageDataForCompression;
for (int y = 0; y < imgHeight; ++y) {
int srcRow = y;
// 如果原始数据是自顶向下，需要翻转？我们根据 BMP 规范：当 biHeight >0 时，数据存储为自底向上。
// 为了简单，我们假设原始数据已经是自顶向下（第一行对应图像顶部）。
// 我们构造标准的自底向上 BMP：最后一行放在数据开头。
int dstY = y;
uint8_t* dstRow = fakeBMP.data() + bmpHeaderSize + dstY * rowSize24;
const uint8_t* srcRowPtr = src + y * imgWidth * 3;
for (int x = 0; x < imgWidth; ++x) {
dstRow[x * 3 + 0] = srcRowPtr[x * 3 + 0]; // B
dstRow[x * 3 + 1] = srcRowPtr[x * 3 + 1]; // G
dstRow[x * 3 + 2] = srcRowPtr[x * 3 + 2]; // R
}
}

int32_t* pWidth = (int32_t*)(fakeBMP.data() + 18);
int32_t* pHeight = (int32_t*)(fakeBMP.data() + 22);
std::cout << "Fake BMP width = " << *pWidth << ", height = " << *pHeight << std::endl;

//获取图像splash数据
COMPRESSED_BITMAPIMAGES images = { nullptr, 0, nullptr, 0 };
int compression_format = 0xF;
bool split = true;
int ret = Splash_createImage(fakeBMP.data(), &images, &compression_format, split);
if (ret != 0) {
std::cerr << "Splash_createImage failed with error " << ret << std::endl;
return -5;
}

unsigned char* leftData = images.bitmapImage1;
unsigned char* rightData = images.bitmapImage2;
int leftSize = images.sizeBitmap1;
int rightSize = images.sizeBitmap2;
if (!leftData || !rightData || leftSize == 0 || rightSize == 0) {
std::cerr << "Compressed image data is empty" << std::endl;
// 注意需要释放可能已分配的内存
if (leftData) free(leftData);
if (rightData) free(rightData);
return -6;
}

// 上传主控制器（左半图）
bool master = true;
if (DLPC900_InitPatternMemLoad(master, imageIndex, leftSize) < 0)
{
std::cerr << "Master InitPatternMemLoad failed" << std::endl;
free(leftData); free(rightData);
return -7;
}
int left_offset = 0;
int left_remaining = leftSize;

while (left_remaining > 0) {
int sent = DLPC900_pattenMemLoad(master, leftData + leftSize - left_remaining, left_remaining);
if (sent < 0) {
std::cerr << "Master pattenMemLoad failed" << std::endl;
free(leftData); free(rightData);
return -8;
}
left_offset += sent;
left_remaining -= sent;
std::cout << "Master upload: " << (left_offset * 100 / leftSize) << "%\r" << std::flush;
}
std::cout << std::endl << "Master image uploaded." << std::endl;

// 上传从控制器（右半图）
master = false;
if (DLPC900_InitPatternMemLoad(master, imageIndex, rightSize) < 0)
{
std::cerr << "Slave InitPatternMemLoad failed" << std::endl;
free(leftData); free(rightData);
return -9;
}
int right_offset = 0;
int right_remaining = rightSize;

while (right_remaining > 0) {
int sent = DLPC900_pattenMemLoad(master, rightData + rightSize - right_remaining, right_remaining);
if (sent < 0) {
std::cerr << "Slave pattenMemLoad failed" << std::endl;
free(leftData); free(rightData);
return -10;
}
right_offset += sent;
right_remaining -= sent;
std::cout << "Slave upload: " << (right_offset * 100 / rightSize) << "%\r" << std::flush;
}
std::cout << std::endl << "Slave image uploaded." << std::endl;

free(leftData);
free(rightData);
return 0;
}

Below is the code for my main program. The current issue is that when I download these three images, only the third image can be displayed. The first two images have a display time, but no patterns are shown. There is a 4-second period where there is no image, and then the third image starts to be displayed in 2 seconds. Is this a problem with my LUT configuration or with the download function?

//只有暂停显示的情况下才允许配置 LUT 和显示图形
result[4] = DLPC900_PatternDisplay(0);
if (result[4] < 0)
{
std::cout << "Error sending pattern display command" << std::endl;
return -1;
}

std::vector<std::string> imagePaths =
{
"C:\\Users\\leaper\\Desktop\\DLPC900\\DLPC900REF-SW-5.3.0\\DLPC900REF-SW-5.3.0\\DLPC900REF-GUI\\Images and Batch files\\LCR670S_Images\\0101.bmp",
"C:\\Users\\leaper\\Desktop\\DLPC900\\DLPC900REF-SW-5.3.0\\DLPC900REF-SW-5.3.0\\DLPC900REF-GUI\\Images and Batch files\\LCR670S_Images\\0098.bmp",
"C:\\Users\\leaper\\Desktop\\DLPC900\\DLPC900REF-SW-5.3.0\\DLPC900REF-SW-5.3.0\\DLPC900REF-GUI\\Images and Batch files\\LCR670S_Images\\0303.bmp",
};

// 通用参数（可根据需要调整）
int ExpUs = 2000000; // 曝光时间
int DarkTime = 0; // 黑暗时间
int BitDepth = 1; // 图像深度
int LEDSelect = 7; // LED颜色
bool ClearPat = false; // 显示后是否清除图案
bool WaitForTrigger = false; // 是否需要外部触发（内部触发连续播放）
bool TrigOut2 = true; // 是否启用 Trigger2 输出

for(int i = 0; i < imagePaths.size(); i++)
{
result[5] = DLPC900_AddToPatLut(i, ExpUs, ClearPat, BitDepth, LEDSelect, WaitForTrigger, DarkTime, TrigOut2, i, 0);
if (result[5] != 0)
{
std::cout << "Error adding pattern " << i + 1 << " to LUT" << std::endl;
}
}

result[6] = DLPC900_SendPatLut();
if (result[6] != 0)
{
std::cout << "Error sending pattern LUT" << std::endl;
}

unsigned int numLutEntries = imagePaths.size();
unsigned int repeat = 0;
result[7] = DLPC900_SetPatternConfig(numLutEntries, repeat);
if (result[7] < 0)
{
std::cout << "Error setting pattern config" << std::endl;
return -1;
}

for (int i = 0; i < imagePaths.size(); i++)
{
result[8] = DownloadCompressedImage(imagePaths[i], i, 2716, 1600);
if (result[8] == 0)
{
std::cout << "Image " << std::filesystem::path(imagePaths[i]).filename() << " Download completed." << std::endl;
}
else {
std::cout << "Image " << std::filesystem::path(imagePaths[i]).filename() << " Download failed with code " << result[9] << std::endl;
}
}

DLPC900: DLP670S: In DLPC900 + DLP670S, according to the official provided API, when performing image writing in the On-The-Fly mode, only one image can be written.

Zhenming Shen — Sat, 09 May 2026 02:13:29 GMT

Part Number: DLPC900

// 如果1位图像为24位图像，如果是24位图像直接使用原始数据
std::vector<uint8_t> rgb24Buffer;
const uint8_t* imageDataForCompression = nullptr;
size_t dataSize = 0;

int32_t* pWidth = (int32_t*)(fakeBMP.data() + 18);
int32_t* pHeight = (int32_t*)(fakeBMP.data() + 22);
std::cout << "Fake BMP width = " << *pWidth << ", height = " << *pHeight << std::endl;

free(leftData);
free(rightData);
return 0;
}

result[6] = DLPC900_SendPatLut();
if (result[6] != 0)
{
std::cout << "Error sending pattern LUT" << std::endl;
}

TPSM82823A: TPSM82823ASILR 标签DC2548+ ,但实际零件标记59

June KE — Mon, 20 Apr 2026 09:37:17 GMT

Part Number: TPSM82823A

你好，TI团队
我们有一批TPSM82823ASILR 零件（卷轴标签日期代码2548）但实际零件标记59

我们注意到实际设备上标记的日期代码与卷轴标签不同。您能否确认这是否正常，并解释一下差异的原因？

RE: TPSM82823A: TPSM82823ASILR 标签DC2548+ ,但实际零件标记59

Taylor — Tue, 28 Apr 2026 08:37:07 GMT

请联系客服部门进行咨询

客户支持中心 | TI.com.cn

RE: TPSM82823A: TPSM82823ASILR 标签DC2548+ ,但实际零件标记59

Taylor — Tue, 28 Apr 2026 08:36:01 GMT

您好，

已经收到了您的案例，调查需要些时间，感谢您的耐心等待。

RE: TMS320F2800137: TMS320F2800137

Daniel — Mon, 16 Mar 2026 09:10:29 GMT

您好

根据目前公布的相关资料，没有具体的相关披露。

TMS320F2800137: TMS320F2800137

zs zhang — Mon, 16 Mar 2026 07:12:09 GMT

Part Number: TMS320F2800137

关于这个芯片，请问在执行FLASH写操作时，是否还会关闭关闭全局中断。

RE: TMS320F2800137: TMS320F2800137

zs zhang — Mon, 16 Mar 2026 09:03:22 GMT

这是一个封装的函数，我们无法查看这个函数内部的逻辑，所以想要咨询您那边是否可以帮忙查看，

RE: TMS320F2800137: TMS320F2800137

Daniel — Mon, 16 Mar 2026 09:01:20 GMT

您好

所有中断存在优先级，当优先级不足时，就会存在先后，类似的情况当出现reset时，其余中断介入就会存在'屏蔽'。这是个例子。建议在写入存储过程中做相关处理，防止写入错误导致系统崩溃等情况。

[quote userid="166108" url="~/support/enterprise-automation-integration/f/enterprise-automation-integration-forum/1063464/tms320f2800137-tms320f2800137/3828803"]现在就是想问Fapi_issueProgrammingCommand这个函数内部会屏蔽全局中断吗？[/quote]

请上相关库文件查找相关代码说明。

RE: TMS320F2800137: TMS320F2800137

zs zhang — Mon, 16 Mar 2026 08:52:56 GMT

我知道您的意思，我们会考虑写flash这个操作不能被打断的问题，现在就是想问Fapi_issueProgrammingCommand这个函数内部会屏蔽全局中断吗？为什么执行这个函数的时候我们的ISR中断就无法执行了，要等待几十微秒之后才能正常执行ISR中断。

RE: TMS320F2800137: TMS320F2800137

Daniel — Mon, 16 Mar 2026 08:47:58 GMT

您好

[quote userid="166108" url="~/support/enterprise-automation-integration/f/enterprise-automation-integration-forum/1063464/tms320f2800137-tms320f2800137/3828801"]在只屏蔽Fapi_issueProgrammingCommand这个函数的情况下，执行write flash操作时不会出现中断被打断的情况，请问这个函数内部是有屏蔽全局中断的操作吗[/quote]

同时写入flash这个等级比较高。您做中断时要进行考虑。

RE: TMS320F2800137: TMS320F2800137

zs zhang — Mon, 16 Mar 2026 08:39:03 GMT

我们已经尝试过了，在只屏蔽Fapi_issueProgrammingCommand这个函数的情况下，执行write flash操作时不会出现中断被打断的情况，请问这个函数内部是有屏蔽全局中断的操作吗？