完善资料让更多小伙伴认识你，还能领取20积分哦， 立即完善>
3天内不再提示
电子发烧友论坛|
问候语，我使用PIC16F1827 8B MCU和CCP3 PWM输出来控制功率MOSFET的栅极。该MOSFET驱动大开关的分流跳闸线圈；相当大的感性负载。这个想法是，当一个“绊倒”事件发生时，MOSFET关闭，完成一个24V的电路跨越线圈。为了更仔细地匹配开关的跳闸轮廓，而不是仅仅在开关上开/关——我想在100毫秒的间隔内使占空比从0到100%倾斜。我已经能够相当容易地做到这一点，然而，当行程周期大约为60ms时，MCU的电源被中断（不知何故），它复位，结束行程周期并返回到MCU的正常启动序列。非PWM解决方案（“哑铃ON/OFF”）工作正常，不重置MCU。问题表明，有一些切换瞬变到达MCU。我已经阅读了一些要尝试的东西，所以我将覆盖已经做的事情：-0.1UF旁路帽从Vdd到VSS复位引脚配置I/O，不浮动，所以不应该容易受到无意重置信号反激二极管跨24V分流线圈-1500瓦TVS二极管在我的24V电源- MCU是战俘从一个5V稳压器连接到24V电源，在调节器的输入端上有一个0.33μF的上限，在它的输出端上有10uF盖。-开关FRQ＝20千赫（CCP3开关频率，16MHz时钟的7位分辨率）：下一步我要考虑的是：缓冲器电路穿过线圈（R~120欧姆，串联）。W / 0.33 UF上限；尝试各种值-降低PWM开关频率（1-2 kHz）-？？？？附上一个粗略的示意图以供参考。感谢大家帮助我解决这个难题。谢谢，戴夫。
以上来自于百度翻译 以下为原文 Greetings, I am using a PIC16F1827 8b MCU and the CCP3 PWM output to control the gate of a power MOSFET. This mosfet drives a shunt-trip coil of a large switch; a fairly large inductive load. The idea is that, when a "trip" event happens, the MOSFET closes, completing a circuit with 24V across the coil. In order to more carefully match the tripping profile of the switch--instead of just slamming it ON/OFF--I want to ramp the duty cycle from 0 - 100% over a 100 ms interval. I have been able to do this fairly easily, however, when the trip cycle gets about 60ms in, the power to the MCU is disrupted (somehow) and it resets; ending the trip cycle and returning to the MCU's regular startup sequence. The non-PWM solution ("dumb ON/OFF") works fine, and does not reset the MCU. The problem suggests that there are some switching transients that are reaching the MCU. I have read about things to try, so I'll cover what has already been done: - 0.1uF bypass cap from Vdd to Vss - RESET pin is configured I/O, not floating, so should not be susceptible to inadvertent reset signal - flyback diode across the 24V shunt coil - 1500W TVS diode across my 24V supply - MCU is powered from a 5V regulator connected to the 24V supply, with a 0.33uF cap on the regulator's input, and 10uF cap across its output. - switching freq = 20 kHz (CCP3 switching frequency, 7 bit resolution at 16MHz clock) What I am considering next: - snubber circuit across the coil (R ~ 120 ohm, in series w/ 0.33 uF cap); try various values - reducing PWM switching frequency (1-2 kHz) - ??? A crude schematic is attached for reference. I appreciate any and all help in solving this puzzle. Thank You, Dave Attached Image(s) 
0
|
|
相关推荐
8个回答
|
|
|
CAP和PIC之间的电线是否尽可能短？小于5mm？这是一个两层的板，有地面轨道，还是多层的接地平面？在这个语句中有一些不正确的假设。在MCLR之上的VDD上的任何故障都可以触发编程模式，不管I/O设置，它看起来完全像一个重置。为了获得最佳的抗扰性，应该有一个大约4.7K的Vdd的上拉。
以上来自于百度翻译 以下为原文 Are the wires between cap and PIC as short as possible? Less than 5mm? Is this a 2-layer board, with ground tracks, or multi-layer with a ground plane? There's some incorrect assumptions in this statement. Any glitches above Vdd on MCLR could trigger programming mode, regardless of the I/O setting, which looks exactly like a reset. For best immunity, there should be a pullup to Vdd of about 4.7k |
|
|
|
|
|
你好，丹，谢谢你的回复。我不确定关断线圈是否是“弹簧复位螺线管”，但我敢打赌它是！我们的跳闸机构工作的一般方式是分流线圈/螺线管将机械杆推到弹簧上，直到达到临界阈值为止。一旦超过阈值，一些东西释放，卸载弹簧，迫使开关的主电源触点非常快地打开，以防止电弧和损坏触点。PWM的原因是，将更多的电流传递到螺线管，并因此在跳闸周期结束时跳闸单元的弹簧——例如当螺线管的弹簧力处于其最高点时（F＝Kx）。我正在考虑这一切，因为MFG推荐了一个200瓦的电源，并且由于成本和其他各种原因，我们只限于40W电源和一个大的旧电容器。PS——如果你有不同的建议，我会洗耳恭听！因此，已经有了一些进步-一个0.1欧姆系列与150欧姆电阻跨线圈形成一个“缓冲”电路似乎允许完整100Ms的行程周期完成。所以一些早期的成功，我会努力优化RC值。令人惊讶的是，只使用0.1μF旁路电容跨越线圈+和-实际上使情况更糟-跳闸序列只会得到几个周期之前，复位MCU。还有一个有趣的和意外的副作用是快速切换分流线圈，它看起来像是我的电压正在提升-而现在看到的是一个周期为100-50V的周期电压为100Ms的周期！不管怎么说，我认为电源链中的所有部件都被额定为50V，但我会确认一下。
以上来自于百度翻译 以下为原文 Hi Dan, Thanks for the reply. I am unsure about whether the shut trip coil is a "spring returned solenoid," but I bet that it is! The general way that our trip mechanism works is that the shunt coil / solenoid pushes a mechanical rod against a spring that loads until a critical threshold is reached. Once the threshold is exceeded, something releases, unloading the spring and forcing the switch's main power contacts to open very quickly to prevent arcing and damage to the contacts. The reason for the PWM, then, is to deliver more current to the solenoid, and consequently the trip unit's spring at the end of a trip cycle--e.g. when the spring force against the solenoid is at its highest point (F=kx). I am going about all of this because the MFG recommends a 200W supply, and for cost and various other reasons we are limited to a 40W supply and a big old capacitor. PS--If you have any suggestions about a different way to do this, I am all ears! So there has been some progress--a 0.1 uF in series with a 150 ohm resistor across the coil forming a "snubber" circuit seems to allow the complete 100ms trip cycle to complete. So some early success, and I'll work to optimize the RC values. Surprisingly, just using a 0.1 uF bypass capacitor across the coil's + and - actually made the situation worse--the trip sequence would only get a few cycles in before resetting the MCU. Also one interesting and unintended side-effect of fast-switching the shunt coil is that it looks like my voltage is being boosted--and the coil is now seeing an ON cycle voltage of 30-50V for the 100ms period! No matter, I think all of the components in the power chain are rated >50V, but I'll check to be sure ;) |
|
|
|
|
|
QHB的答案：1。盖跨越MCU封装，引线很短，因为我可以使它们为18-DIP封装。2。无焊面包板（原谅我犯了罪）。3。MCLR引脚接地，我用它作为I/O——&，我将它拉到5V，并使用不同的引脚I/O功能。谢谢非常实用的提示！戴夫
以上来自于百度翻译 以下为原文 Answers to qhb: 1. The cap straddles the MCU package, with the leads as short as I could make them for a 18-DIP package. 2. solderless breadboard (forgive me for I have sinned). 3. The MCLR pin was grounded, and I was using it as I/O --> I'll pull it up to 5V and use a different pin for that I/O function. Thanks for the very practical tips! Dave |
|
|
|
|
|
那么你需要多注意你的地电流。
以上来自于百度翻译 以下为原文 You need to pay a LOT of attention to your ground currents then. |
|
|
|
|
|
对于这种设置，我宁愿使用一个单独的5V电源，而不是从24V的高功率轨道上下来。但是这是我最好的选择。
以上来自于百度翻译 以下为原文 Hi For this kind of setups, I would rather use a separate 5V power supply, than step it down from the 24V high power rail. But that its me... Best regards Jorge |
|
|
|
|
|
豪尔赫，好点。如果情况允许的话，我会深入研究这个设计。UPDATE-QHB可能会赢得这一轮琐事。我通过一个上拉电阻将MCLR的引脚固定到VDD上，尽管在切换过程中杂乱无章的噪声看起来毛病已经消失了。在这个陈述中有一些不正确的假设。MCLR之上的VDD上的任何故障都可以触发编程模式，而不管I/O设置是哪一个。KS完全像一个重置。为了获得最佳的免疫力，应该有一个拉到Vdd的4.7Ki移除缓冲电路来确认，这个问题仍然是固定的。除了MCLR PIN的明显教训之外，还有一个：你的假设常常是你错误的根源。至少在我的情况下是这样。谢谢大家的帮助！谢谢！
以上来自于百度翻译 以下为原文 Jorge, good point. If the situation allows, I will look into that to enhance the design. Update--qhb probably wins the this round of trivia. I pinned the MCLR' pin to Vdd through a pull up resistor and it appears the glitch has gone away, despite the mess of noise during switching... There's some incorrect assumptions in this statement. Any glitches above Vdd on MCLR could trigger programming mode, regardless of the I/O setting, which looks exactly like a reset. For best immunity, there should be a pullup to Vdd of about 4.7k I removed the snubber circuit to confirm, and the issue remains fixed. Besides the obvious lesson surrounding MCLR' pin, there's another one: your assumptions are often the source of your mistakes. At least they are, in my case. Thanks for everyone's help. Thx! |
|
|
|
|
|
戴夫，所以你把你的PIC电路建立在一个无焊点的电路板上，并使用一个MOSFET来驱动一个螺线管，制造商建议使用24 VDC，200瓦的电源。似乎你已经建立了一个故障工厂。一个24伏特，200瓦电源建议螺线管需要超过8安培的电流在24伏特激活。我不记得曾经见过一个分流跳闸线圈需要这么多的权力激活。你肯定这是什么数据表为您的SH？脱扣线圈需要吗？
以上来自于百度翻译 以下为原文 Dave, So you built your PIC circuit on a solder-less breadboard and are using a MOSFET to drive a solenoid that the manufacturer recommends using a 24 VDC, 200 watt power supply for. Seems like you have built a glitch factory. A 24 volt, 200 watt power supply suggests that the solenoid needs over 8 amps of current at 24 volts to activate. I do not recall ever seeing a shunt-trip coil requiring this much power to activate. Are you certain that this is what the data sheet for your shunt-trip coil requires? |
|
|
|
|
|
哈哈！是的，丹，条件不太理想。我会得到一个由专业人员组成的真正的电路板，但是概念证明很重要。至于线圈，它上面标有“180 VA”。我可以确认，所需的电流和总能量都接近于该数字，但不完全是因为，由于线圈的内部电阻，它不能从24V电源中汲取超过6A。不管怎样，这种特殊设备的可靠发射证明了它自己的工程挑战。多亏了这里的友好人士，我有一些新的工具可以尝试。
以上来自于百度翻译 以下为原文 Haha! Yeah, Dan, conditions are not quite "ideal." I'll get the real boards made up by a professional, but proof of concept is important. As for the coil, it's marked "180 VA" right on it. I can confirm that the current and total energy needed are both indeed close to that figure, but not quite because, due to the coil's internal resistance, it can't draw more than 6A off of a 24V supply. Whatever the case, reliable firing of this particular device has proven its own engineering challenge. Thanks to the friendly folks here, I have a few new tools to try. |
|
|
|
|
只有小组成员才能发言，加入小组>>
5150 浏览 9 评论
1994 浏览 8 评论
1923 浏览 10 评论
请问是否能把一个ADC值转换成两个字节用来设置PWM占空比？
3164 浏览 3 评论
请问电源和晶体值之间有什么关系吗？PIC在正常条件下运行4MHz需要多少电压？
2221 浏览 5 评论
716浏览 1评论
602浏览 1评论
有偿咨询，关于MPLAB X IPE烧录PIC32MX所遇到的问题
488浏览 1评论
PIC Kit3出现目标设备ID(00000000)与预期的设备ID(02c20000)不匹配。是什么原因
616浏览 0评论
515浏览 0评论
/7 
关注我们的微信
下载发烧友APP
电子发烧友观察
版权所有 © 湖南华秋数字科技有限公司
电子发烧友 （电路图） 湘公网安备 43011202000918 号 电信与信息服务业务经营许可证：合字B2-20210191
工商网监
湘ICP备2023018690号
1320
评论
淘帖
1196
