Tag: VSD maintenance

Posted on by — Leave a comment

What happens to the VSD drive when the no-load current is greater than the load current?

Cause: A cement plant user sent a 75kW micro energy WIN-G9 frequency converter for repair. The cause of the malfunction was that during operation, the frequency converter suddenly caught fire and emitted smoke, causing it to shut down. Upon inspection, the power input circuit of the machine is a three-phase half controlled bridge. Using its controllable rectification principle, the DC main circuit energy storage capacitor is “soft charged”, eliminating the commonly used charging contactor for low-power frequency converters. In fact, the semi controlled bridge here is equivalent to a contactless soft switch. Upon inspection, it was found that there were obvious signs of arc flashover and burning on the terminals of one of the thyristor modules, but the measurement did not indicate a short circuit. During disassembly, it was found that the fixing nut was easily removed, and the cause of the flashover seems to be due to loose connection screws, causing poor contact. This module is a combination of a diode and a unidirectional thyristor. Further check the control board and inverter main circuit, and there are no abnormalities. After removing the module, the remaining two phase half control bridges were used as power inputs. After being powered on, a 2.2kW low-power three-phase motor was tested and there were no issues. After replacing it with a new module of the same model, it was installed on site.

To be cautious, first adjust the operating frequency to 5Hz, and the frequency converter is loaded with a fan. First, disconnect the motor connector and let the motor run at no load. This trial run was a shock! When the frequency is below 5Hz, the no-load operating current is 45A. Although it feels slightly higher, it may be caused by the repair of the motor winding or the adjustment of the parameters of the frequency converter, such as the starting curve or torque compensation, which was not taken into consideration. When the speed was increased to 10Hz, both the current displayed on the frequency converter panel and the output current measured with a clamp meter reached 100A! And the oscillation amplitude of the output current is very large, but when measuring the three-phase output voltage, it is about 70V, balanced and stable. Disconnect the motor connection wire and power on to measure the output of the frequency converter. The output voltage is 70V at 10Hz, 150V at 20Hz, and 250V at 35Hz. As the operating frequency increases, it reaches 400V at 50Hz. During this process, the balance of measuring the three-phase output voltage is very good. The V/F curve output by the frequency converter conforms to the quadratic load torque characteristics. No problem. The output voltage is balanced and stable, while excessive output current and severe current fluctuations are clearly caused by abnormal loads. This is a conclusion drawn from conventional judgment.
Discuss with the relevant technical personnel of the factory, attempting to identify the reasons for the motor and mechanical aspects. For example, whether the motor is newly repaired or whether the winding is poorly wound; Check for wear and unstable operation of bearings; Is there any looseness or non concentricity in the connecting shaft; The wind blades have deformation, etc. Restore the wiring of the original power frequency starting cabinet, compare the power frequency starting motors, and eliminate the above doubts one by one. According to on-site observations, the motor and connected loads are in good condition, and there is almost no electrical and mechanical noise during operation. The no-load current under full speed operation is only less than 35A, with three-phase balance and no fluctuation! There is no problem with the motor and load, and the problem is still with the frequency converter.
So where is the fault location of the frequency converter? It’s a bit scratching my head. Is the current detection inaccurate, causing erroneous output? The current value displayed on the observation panel is close to the value measured by the clamp ammeter, and there should be no problem. Is there still a problem with the CPU motherboard and the output driver waveform incorrect? It doesn’t make sense. All digital circuits, why is the waveform incorrect?
Fortunately, there is another inverter of the same model and power not far from here on site, carrying the same load. That’s great. This has brought great convenience to comparative experiments. The factory was eager to start the machine and provided active cooperation. Swapping the current transformers of two frequency converters is ineffective; Swapping the CPU motherboards of two machines is invalid. Retrieve the DC voltage display of the main circuit, which is 550V, and there is no problem with the voltage sampling circuit. I can no longer figure out which circuit the problem lies in. When the comparison machine is under load, the operating current is 75A at 10Hz. When it reaches 35Hz or above, the operating current only reaches 100A, which is smaller than the current of this motor with no load. The no-load current is much higher than the load current, so there must be a problem with the frequency converter.

During the trial operation, I occasionally measured the three-phase output current of the inverter using a current clamp meter, and even discovered an incredible phenomenon! The input and output currents of this frequency converter are completely disproportionate, with a difference of more than 10 times!
When outputting a 40A current, the input current is a few amperes, which is almost undetectable; When outputting 100A current, the input current is only below 8A! Strange, it doesn’t comply with the law of conservation of energy. Where did the 100A output current come from?! It’s like an airtight water pipe, where 1 cubic meter of water enters and 10 cubic meters of water flows out. The water inside the pipe cannot come out.
We all know that in general, the input current of the frequency converter is always smaller than the output current. The reason is that the energy storage capacitor in the DC circuit acts as if a reactive power compensation cabinet is installed at the motor end. When the frequency converter is unloaded or lightly loaded, a portion of the current is provided by the energy storage capacitor to the load, reducing the current absorbed by the frequency converter from the grid. As the load increases, the input current of the frequency converter increases proportionally. When the rated load is applied, the input current and output current of the frequency converter should be close to equal. When outputting 40A, the input is only a few amperes; When outputting 100A, the input current has reached 70A; When the output current reaches 140A, the input current has also reached this value. Under normal circumstances, there may be a difference in input and output currents, but there will not be an extremely significant difference as mentioned above. The huge difference made me doubt whether the measuring instrument was broken. After changing the watch and retesting, the same result was still obtained.
There’s no way out. Consult the manufacturer. Due to the urgent need to solve the problem and find the answer, it’s not worth considering whether long-distance phone calls are expensive at this time. The technical personnel from the frequency converter manufacturer replied that this model of frequency converter is the earliest produced frequency converter, and there are problems with slightly higher no-load current and current fluctuations, but it is a normal phenomenon and does not affect its use. After being loaded, the current will stabilize. It is best to connect a motor of the same power for testing to see if there is a problem with the motor or load. Problems with motor bearings. If all motor and load issues are eliminated, as long as the three-phase voltage output of the frequency converter is balanced and the output current does not exceed the rated current of the frequency converter, can the machine be tested under no load or on load. Can’t it break down. As for the proportion of input and output currents, it is difficult to have a fixed proportion due to different load conditions. It’s not proportional. Don’t get entangled with the issue of proportion.
Think about it too. As long as the output three-phase voltage is balanced and does not exceed the rated current, can the load test be conducted. Can’t the frequency converter break down. Perhaps after being loaded, there will be no significant fluctuations in the output current. Maybe it’s normal.
We had to conduct a load test and a miracle occurred (which was surprising): when operating at 10 Hz and outputting a current of 40A, the output current was only 7 out of 8 amperes. When operating at 30 Hz, the output current is 60A and the input current is 25A; When operating at 40 Hz, when outputting a current of 100A, the input current is 70 amperes. The operating current has decreased and the fluctuation has decreased, and it is basically stable. The three-phase voltage and three-phase current are both balanced and relatively stable. The problem was inexplicably solved.
Thank you to the manufacturer’s technical personnel for their guidance: why not try the machine on load. But due to encountering this situation for the first time, abnormal current occurs at no load, and it is not dare to increase to full speed for operation. I’m even more afraid to carry it. I always want to find out the reason before loading. I always thought this was an abnormality with the frequency converter.
After the frequency converter was put into operation and returned from the site, I am still pondering this issue.
Remembering the problem of high zero line current in a power plant during maintenance, caused by harmonic components in the transmission line. It’s harmonic current. Is there also a significant harmonic component in the output circuit of the frequency converter during no-load or light load operation? Where does this harmonic component come from? Is the measured result true?

The analysis shows that there are significant harmonic components in the output current during no-load operation. There may be two reasons for the high harmonic current: 1. The output PWM wave of the inverter is not ideal enough, and the modulation method is not optimal. The software control concept has not been optimized (the new machine must have been improved); 2. When unloaded, it is equivalent to a serious mismatch between the power supply capacity and the load capacity. The power supply capacity is much larger than the frequency converter capacity, which is also a major reason for the generation of harmonic currents. When running on load, the capacity matching situation improves, and the harmonic components are greatly reduced. The combination of these two reasons has stumped me, an old electrician. For me, during the test drive, I made an empirical mistake. I was bound by the ratio of input and output currents and almost surrendered.

Posted on by — Leave a comment

Analysis of IGBT Module and Driver Faults in Variable Frequency Drive Repairs

Analysis of IGBT Module and Driver Faults in Variable Frequency Drive Repairs

In the field of industrial automation, variable frequency drives (VFDs), as key equipment for motor speed control, are crucial for stable operation. However, due to the complex and varied working environments, faults in the IGBT modules and drive circuits within VFDs occasionally occur, causing significant impacts on production. This article delves into the identification, analysis, and handling processes of IGBT module and driver faults through three practical repair cases, aiming to provide valuable references for relevant technicians.

Case 1: Phase Deviation Fault in Dongyuan 7300PA3.7kW VFD
Fault Phenomenon: After powering on, a Dongyuan 7300PA3.7kW VFD had outputs on all U, V, and W phases, but with severe phase deviation.

Fault Diagnosis: Initially suspected as a drive circuit abnormality or IGBT module damage. Measurements revealed an open circuit in the upper arm diode of the U-phase in the inverter circuit. Typically, the IGBT transistor paralleled with this diode was also burnt due to short-circuit current, and the paralleled diode was damaged by the impact.

Repair Process:

  1. After removing the damaged SPIi12E IGBT module, clear all the module pins and prepare to test the six drive circuits.
  2. Upon powering on, the VFD immediately reported an overheating fault, with the CPU locking the drive pulse output, preventing drive circuit quality detection.
  3. Two terminals labeled T1 and T2 on the circuit board, suspected as internal overheating alarm outputs of the module, were observed. By connecting a 5V power supply through a resistor, the other end grounded. When these terminals were left open, the T1 terminal output a high-level module overheating signal through a pull-up resistor, triggering protective shutdown.
  4. After short-circuiting the T1 and T2 terminals, powering on no longer resulted in protective shutdown. It was found that the U-phase upper arm IGBT drive circuit had no trigger pulse output. After replacing the drive circuit IC/PC923, the six-pulse output returned to normal.
  5. A new IGBT inverter module was installed, the short-circuit wire between T1 and T2 was removed, and the VFD operated normally after powering on.
    Experience Summary: When an IGBT tube is damaged, the corresponding drive IC is often damaged by the impact as well. Before replacing the new module, be sure to check the drive IC in the same branch to avoid damaging the new module again due to drive circuit abnormalities.

Case 2: Repair of Lightning-Damaged Alpha 18.5kW VFD
Fault Phenomenon: An Alpha 18.5kW VFD was damaged by lightning, with the CPU motherboard reporting a 2501 error and panel operations failing.

Fault Diagnosis: Lightning caused damage to the CPU and surrounding communication circuits.

Repair Process:

  1. Temporarily ignoring the CPU motherboard issue, repair the drive board first. It was found that six A316J chips were responsible for six drive pulse outputs, with three upper arm pulse drive circuits damaged.
  2. As an alternative, three A316J chips (for three-phase lower arm drives) were used as three-phase OC signal alarm outputs, and the remaining three were replaced with 3120 (identical to PL250V) to drive the optocoupler ICs.
  3. The new ICs were adapted and soldered, and the input circuit was adjusted to ensure normal operation of the new ICs.
  4. After replacing the new CPU motherboard, the static output voltage and dynamic pulse output of the six drive circuits were tested and found to be normal.
  5. The damaged IGBT module was replaced, and the VFD resumed normal operation.

Case 3: Repair of Phase Deviation in a 7.5kW VFD
Fault Phenomenon: A user reported that a 7.5kW VFD had output but could not operate normally, with phase deviation present.

Fault Diagnosis: It was found that one of the six drive circuits was abnormal, with the drive IC model being PC929 (or A4503?). Measurements showed no pulse output at the input and output terminals of the drive IC.

Repair Process:

  1. Suspecting a fault in the CPU’s internal pin circuit, the input terminal of the PC929 was disconnected, and it was found that the voltage at the CPU’s pulse output terminal increased. However, when the drive IC was connected, the voltage dropped to nearly 0V.
  2. It was analyzed that the CPU directly drove the photoconductive tube, and long-term high-current output led to aging faults in the output stage. It was decided to enhance the signal voltage through an external amplification circuit.
  3. An amplification circuit was constructed using two NPN-type transistors and resistors to amplify the CPU’s pulse signal before inputting it to the drive IC.
  4. After powering on, the six-pulse output was normal, and the inverter module was powered on, with normal three-phase voltage output.
    Experience Summary: For aging faults in the CPU’s output stage, amplifying the signal voltage through an external circuit is an effective repair method, which not only saves repair costs but also shortens repair time.

In summary, IGBT module and driver faults are common challenges in VFD repairs. Through meticulous fault diagnosis, reasonable repair strategies, and innovative repair methods, these issues can be effectively resolved, ensuring the stable operation of VFDs. It is hoped that the sharing in this article can provide valuable references and insights for technicians.

Posted on by — Leave a comment

Handling Method for SC Jumping Fault of GS Series INVT VSD

SC: Refers to a short circuit in the output load of the frequency converter. In the state of power outage, there is no short circuit between U, V, W and between U, V, W and DC P, N. To distinguish whether it is a current detection circuit fault on the CPU motherboard or a signal returned by the driving IC, short circuit the signal to the diode side of the optocoupler device with a wire, and operate the RUN key after power on. The output frequency displayed on the panel is normal. The SC signal is returned by the driving IC, and there are two reasons for the malfunction: first, the inverter module is damaged, and second, the driving IC itself is faulty.

After cutting off the power supply to the inverter module, the six pulse signals of the drive IC were checked again. It was found that the U-upper arm drive IC had input pulses but no output pulses (static negative pressure normal). After replacing the drive IC, the output was normal. The static negative pressure of the U lower arm drive IC is only a few tenths of a volt. After replacement, the fault still persists. After welding the 100 ohm resistor connected to the trigger end of the inverter module, the static negative pressure rises to the normal value. The forward resistance of the lower arm terminal of the module U is consistent with the resistance of other trigger terminals, but its reverse resistance is slightly smaller than that of other trigger terminals, indicating that the circuit inside the U-phase trigger terminal of the module is damaged!
This is a key point in checking module faults: the resistance of the main terminal cannot be roughly measured to be normal, which means that the module is good. It is not uncommon for the circuit inside the triggering terminal to be damaged, but it is more concealed! When the frequency converter reports an SC or OC fault and the main terminal of the detection module is normal, the inspection of the forward and reverse resistance of the triggering terminal should not be missed!
The damage of modules and the damage of driving ICs are usually related: when the U-phase main circuit is damaged, the upper and lower arm driving ICs of the U-phase are often also damaged by strong voltage shocks; When the driver IC of the phase is found to be damaged, it often hides the internal circuit damage of the triggering terminal of the phase. So one of the two is that a comprehensive inspection must be conducted. Of course, there are also cases where modules and driver ICs are damaged separately, but it is extremely rare.

During maintenance, another phenomenon of drive IC failure was also discovered in the low-power model of the Envision GS: when the frequency converter is powered on and there is an SC fault signal output, the CPU performs self check, reset, and reset, and after clearing, the panel only displays H:00, and all panel buttons fail to operate. When the corresponding optocoupler circuit is short circuited to eliminate the SC signal, all panel operations are normal. If the model detects an SC fault signal output when powered on, it may perform a “program lock” and refuse to operate! When this phenomenon occurs, the fault signal output circuit should be checked and corresponding measures should be taken to temporarily eliminate the fault signal, in order to facilitate operation and judgment, and thus solve the fault.

Posted on by — Leave a comment

A Hidden Cause of Alpha Inverter OC Fault Tripping and Shutdown

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.