Tag: Variable speed driver

During the maintenance of the Alpha low-power inverter, it was discovered that the inverter has a common problem – it is prone to OC failure. It manifests as: often jumping faults during start and stop operations, but sometimes also jumping faults during operation; Sometimes it inexplicably improves and can run for varying lengths of time. When I thought there was no problem, I started to frequently skip OC faults again; When measuring the output voltage of U, V, and W with a probe under no load, it is prone to tripping faults. However, after connecting the motor, it starts running and does not jump again. After a while, the OC fault still jumps when connecting the motor.

The handling of such faults is quite tricky, and it is possible that the fault has been eliminated during the testing process, resulting in no evidence for investigation. Even when testing the hardware circuit (protection circuit) during frequent malfunctions, I couldn’t find any problems and couldn’t figure out the root cause of the problem. This problem puzzled me for over two months.
The hardware protection circuit is mainly completed by two LM393 dual operational amplifier circuits, U22 and U24. The signal is then inverted by a first stage inverter and sent to pin 16 of the CPU. U22 and U24 jointly input two output current signals, one overload OC signal returned by the inverter drive IC, and one DC voltage detection signal, which are respectively added to the input terminals of four operational amplifiers. After open-loop amplification processing (the operational amplifier circuit is actually used as a switch circuit here), Four fault signals were connected in parallel, and after undergoing a first level of phase inversion processing, they were sent to pin 16 of the CPU. I first cut off the overload OC signal returned by the inverter drive IC, and then cut off the “total” fault signal of the inverter output, but both were ineffective, and the fault phenomenon still persists. Is there another way to string in OC signals elsewhere? Impossible!
There may be some inexplicable interference in the circuit, but the source and cause of the interference are difficult to identify. We racked our brains and exhausted all means to install capacitor and resistor filtering elements in the fault signal circuit to improve the anti-interference performance of the circuit, but to no avail. Could it be that during the loading and unloading process of the inverter drive module during the start/stop moment, the fluctuation of CPU power supply caused the malfunction? The measured CPU power supply is 4.98V, which is very stable and meets the requirements.
Without any reason, I had a sudden inspiration and adjusted 4.98V to 5.02V. After conducting a start/stop test, the fault was surprisingly eliminated!
Analyze and speculate the cause of the malfunction as follows: the setting of the static voltage working point outside or inside the CPU is improper or too low, which is exactly at the critical point of signal interference level, making it easy to experience random OC faults that are confusing. After slightly increasing its 5V power supply, the voltage value at its operating point also increases accordingly, avoiding the critical point of interference level, and the frequency converter changes from “neural” to “normal”.

When the machine leaves the factory, if the CPU power supply adjustment value is slightly higher, the machine can run normally for a long time. If the adjustment value is slightly lower, or if there is a slight decrease in 5V due to some reason (such as component variation, temperature drift, etc.) during use, frequent OC tripping faults may occur. Adjusting the 5V power supply can easily solve the problem while ensuring that there are no issues with the hardware protection circuit. If it is not due to an accidental factor, then the depth of the concealment of this fault makes it difficult to “adjust” it properly.