The plant I work at, had a 200 HP VFD Powerflex 750 make a loud bang then trip a 400 amp breaker. I replaced the VFD with one we had in storage. Not sure how long it has been on the shelf but at least 3 years. I searched through the manual and did not see anything about reforming the capacitors, so I decided to apply power (480v) and wait a few hours before applying a load. This is the method listed for the Automation Direct VFDs we normally use.
I did some more searching and found this Allen Bradley service bulletin:
Table 3 - Drive Storage Duration and Reforming Recommendations.Duration Reforming GuidelineUnder 2 years No reforming required.
2 …3 years Apply rated voltage, per the normal method, for 30 minutes under no load.
Over 3 years Using a DC power supply connected directly to the DC terminals of the product, ramp-up the voltage from 0…100% of DC bus voltage (as per Table 4) in increments of 50%, 75%, and 100% rated voltage, under no load, for 30 minutes at each increment. See Figure 5 on page 4 for an illustration of this method
Does anyone actually use the last method? I dont have this type of equipment and would have to hire someone to reform the caps.
This drive costs over $21k. Does not reforming the capacitors properly effect the longevity of the VFD? Or is it only a danger of capacitor failure on startup?
The issue is not "longevity", it is a lot more immediate in nature. If the capacitors have already de-formed (more below), just applying full voltage to the line terminals will blow the caps apart, often violently to the point of taking out other things in the drive. I had one blow the doors off of a 40HP drive once., it's pretty spectacular (unless you have to pay the bills). The issue is that the way electrolytic caps work, these are layers of aluminum film line foil, inside with an electrolyte gel in between them. When first made, a current is passed through the film layers and a layer of aluminum oxide, an insulator, forms on the surface of each layer. That insulating layer is what makes it a capacitor. When sitting unpowered for a long time, some of that aluminum oxide dissolves back into the electrolyte and the layer gets thin, then cracks, referred to as the oxide layer "de-forming" (because it was "formed" in the first place).
If full voltage is applied, those cracks allow full current to jump the gaps between layers and burn holes in them, destroying them and often simultaneously vaporizing the electrolyte, which is what make it explode. By applying LOW VOTAGE DC to the caps and keeping the current below 1/2A (500mA) however, the same process that MADE it a cap in the first place, the growing or "re-forming" of that aluminum oxide layer, takes place, so the capacitor self-heals by "sealing" the cracks in the layer and is returned to normal operation. That cannot be rushed, because it is a "growing" process. If you simply energize a stored drive every so often (I've always used once per year, A-B says 2 years), any small cracks that STARTED to form will self-heal before they can become destructive.
The re-forming process differs for different manufacturers based on the construction of the drive, you really can't compare one to another. Also the time frame for when it is necessary tends to vary as well, even within the same manufacturer. I have another document from Rockwell that is close, but not the same, as what you read :
Disconnect the Drives control boards and I/O boards prior to applying the DC voltage. Otherwise, the external
power supply will turn on the Drive, and in the process draw too much current and overload the supply.
< 1 year and stored at rated temperature and humidity conditions (No reforming required)
1 – 2 years: Apply rated voltage, per the normal method, for 60 minutes under no load.
2 – 3 years: Using a DC power supply connected directly to the DC terminals of the product, ramp-up voltage
from 0-100% of DC bus voltage (see table below) in steps of 25%. Dwell at 25%, 50% and 75% steps for 30
minutes each. At 100% voltage, dwell for 60 minutes, all under no load.
> 3 years: Using a DC power supply connected directly to the DC terminals of the product, ramp-up voltage
from 0-100% of DC bus voltage (see table below) in steps of 25%. Dwell at each step for 120 minutes
The final forming voltage shall be 1.35 – 1.45 times the rated AC system voltage
Power supply current draw must not exceed 500 mA.
You CAN, by the way, connect the DC to the DC terminals of the VFD if it has them, but they still want you to unplug any control boards on the larger drives (PF700 / 750). Some people say it's OK to apply power to the AC line terminals through a variac (variable transformer), and I have done that many times, but in some drives, doing so is also then giving low voltage to the
control system for the drive and can damage some of them, so it's kind of a craps shoot. The trick there though is to use a current meter with it and adjust the AC to ensure that the current flow is never more than 500mA. (Edit: I have built my own capacitor re-former for this with a 120V variac, a current limiting resistor, a 1A ammeter and a CPT that steps the 120V to 240 or 480 as the case may be, and 1/2A fuses on the HV side.)
I have had several people tell me that they just hook up some light bulbs in series to the AC terminals to drop the voltage to the drive, then unscrew bulbs one at a time over a few hours to increase the voltage. I don't recommend this method as it is a lot less accurate, and I have seen one of these setups fail to protect the drive, so they lost it anyway.
)
