How to analyze this overcurrent protection design?

普拉卡图

How to analyze this overcurrent protection design? [Copy link]

 

Image Description:

The above picture shows the overcurrent protection design. It is divided into 3 parts.

1. After IV conversion, the current enters the inverting terminal of the comparator. When overcurrent occurs, the comparator outputs a low level (overcurrent protection signal).

2. When the EN pin 4 of the analog switch U16 is high, pins 1 and 2 are turned on, and the overcurrent protection signal is sent to the latch U15 behind for processing.

3. The latch latches the overcurrent protection signal.

My questions:

1. Why is there an analog switch in the diagram? Wouldn't it be better to have the signal go directly to the latch? This path seems a bit redundant? What is its purpose?

2. Could you please help me check if my understanding of the latch is correct?

My analysis on latch:

1. In the figure, LE is pulled up and OE is pulled down, which corresponds to rows 1 and 2 of the truth table. Under normal circumstances, it is equivalent to direct connection between D and Q.

2. When a low level enters D, Q is also low and then LE is clamped and pulled low through the D15 diode.

At this time, corresponding to the third row of the truth table, the output is also determined only by Q0 during latching, that is, a low level. No matter how the input D changes, the output will remain at a low level.

I don't know if this is the analysis?

普拉卡图

Regarding question 1, it is necessary to explain the application scenario. This current detection is to detect the motor current. When powered on, there may be a large current causing the latch to output a low-level protection signal. So this analog switch may prevent the overcurrent protection from being triggered by power-on. But I don’t quite understand what the principle is? It looks like charging through C40 of the EN pin. The low level output of the comparator must reach a certain time, and the C40 voltage is charged to the threshold before the signal is transmitted to the next stage, which plays an integral role?

maychang

[1. Why is an analog switch added to the diagram? Wouldn't it be better to have the signal go directly to the latch? This path seems a bit redundant? What is its purpose? ]

Guess, just guess: before power on, 3V3_IN is zero, and the voltage across C40 is zero (because it has been out of power for a long time). After power on, C40 is charged through R47, and this time constant is relatively large, so the EN end of the analog switch is low level for a short time after power on, and it takes a short time to be high level to enable the analog switch. During this short period of time, even if there is overcurrent, the latch will not latch.

maychang

Placardo published on 2022-12-13 13:31 Regarding question 1, it is necessary to explain the application scenario. This current detection is to detect the motor current. When power is turned on, there may be a large current causing the latch to output a low level...

That's what I mean.

However, it is a bit wasteful to use analog switches, small MOS tubes and large capacitors. Comparator or latch chips with enable terminals are very common.

maychang

【2. Can you help me check whether my understanding of the latch behind is correct? 】

correct.

In fact, this is the "D flip-flop" in the digital circuit, which feeds back the Q end to the LE end to form a latch.

普拉卡图

maychang posted on 2022-12-13 13:55 That's what I mean. However, it is a bit wasteful to use analog switches, small MOS tubes, and large capacitors. A comparator or lock with an enable terminal (enable terminal)...

Okay, thank you teacher.

maychang

Plakatu published on 2022-12-13 14:04 OK, thank you teacher. Is there a specific model for this? I want to find a specification sheet to see what it is like.

Your latch model is SN74LVC1G373. The leftmost column of the truth table below is the OE terminal. It can be seen from the truth table that when the OE terminal is low, the output is controlled by the LE terminal to be "transparent" or "latched", and when the OE terminal is high, the output is in high impedance state. In high impedance state, if there is a pull-up resistor, it is a high level, and if there is a pull-down resistor, it is a low level. The figure also explains the function of welding or not welding R66R68. Using the OE terminal, it is possible to achieve that the overcurrent protection does not work for a period of time after power-on.

maychang

Plakatu posted on 2022-12-13 14:04 OK, thank you teacher.

Traditional latches, such as 74HC3773 and 74HC573, have LE and OE terminals. However, they are used to latch the address of the microcontroller bus, and there are 8 signals in the chip. You only use 1 channel, so it is a waste to use an 8-channel chip.

普拉卡图

maychang published on 2022-12-13 14:48

OK, thanks.

maychang

Plakatu posted on 2022-12-13 15:08 OK, thank you.

If you want to simplify the circuit, you can add some positive feedback to the comparator chip to achieve "latch", that is, when there is no overcurrent, the comparator outputs a certain level (high or low). Once there is overcurrent, the comparator flips, the current returns to normal, and the comparator output does not flip, and the overcurrent state remains. To restore the original state, you must manually or use another circuit to make it return to the initial non-overcurrent state. In this way, the work of the comparator and the latch is combined.

maychang

Plakatu posted on 2022-12-13 15:08 OK, thank you.

Such a comparator with positive feedback is called a Schmitt circuit, also known as a hysteresis comparator.

To make such a comparator inoperative for a period of time after power-on, of course, an RC charging circuit can be used to achieve this.

普拉卡图

maychang published on 2022-12-13 15:49

This design idea is great. It turns out that Schmitt can be used as a comparator + latch. I've learned something!

maychang

Placatto posted on 2022-12-13 16:44 This design idea is great. It turns out that Schmidt can be used as a comparator + latch. Learn something!

In fact, these are all composed of amplifiers, which can be said to be branches of amplifiers. Usually, when we talk about amplifiers, we are discussing amplifiers that work in a linear state. But amplifiers can add positive feedback, and after adding positive feedback, they can form Schmitt circuits (hysteresis comparators), oscillators (astable circuits), monostable circuits, bistable circuits... What you want is a bistable circuit, which requires that it is in a certain stable state when powered on, and immediately flips to another stable state after the current reaches the threshold.

laker2008

I've learned a lot. Thank you for your selfless sharing. I haven't been very busy at work recently, so I often browse the forum to recharge. Many old posts are definitely valuable.