Kepco AC/DC inverters DC fail question

All,
I am building a DC power plant for a small telco operation, and have some questions as to
operation of the inverters.

We are using RA-62 rackmount frames that house 3-30 amp 48 volt inverters.
The documentation on them is not clear on the DC fail output lamp functionality.

Currently, I have the inverters operational, but get a DC fail output lamp even when all operating parameters seem correct.

I am operating them in both a stand-alone, and paralleled operation using the IShare bus for load sharing. Both combinations light the DC fail lamp.

At this time, the load on any one bus is probably less than 3 amps @48VDC output. I would really like to solve this before I commit the system to them.

Has anyone ever used this product and can shed some light on the operation of the DC fail detector circuit?

Thanks
Dialtone

The units don't have an option to connect a battery for AC fail conditions do they? We had Argus 48V power supplies that gave a fail warning because the batteries were not installed (backordered). I wonder if the Kepco's are the same.

As far as I can see, ther is not a requirement for an external battery to use as you suggested. Here is an excerp from Kepco on the DC fail flags for the HSP series inverters.

DCFAIL Indicator and Output Status Flags
The OUTPUT status flags and DCFAIL indicator LED are both controlled by the output fault detector circuit, which monitors both output voltage and module current to assess d-c output status. An output fault condition (DCFAIL indicator "ON") is generated if one of three fault conditions is detected: (1) Overvoltage fault or (2) Undervoltage Fault - output voltage is outside specified regulation limits, or (3) Undercurrent fault - the power supply module is supplying less than 70% of the current required by the circuit (as indicated by the load sharing signal) while the output voltage is within specification limits.

A fault condition is not generated for a combination of overvoltage and undercurrent indications, as these are mutually exclusive conditions for power supplies which are not part of a parallel-redundant configuration; this combination does, however, indicate proper operation for operational power supply modules which are part of a parallel-redundant power scheme in which one or more power supply modules are presenting overvoltage failures.

The Fault Detector Operation Table (below) provides an operating matrix of the DCFAIL status function; see timing relationships. The output voltage fault limits are ±5% of programmed output voltage, while the undercurrent fault limit is < 70% of required module current; signal reset requires output voltage recovery to within the specified ±1% regulation range and/or module current recovery to > 85% of required module current, respectively.

I know I do not have high/low voltage conditions, so that just leaves the current monitor on the ISHARE bus. They give no indication how to use it however, so I am stumped as to what it expects to see.

This may be total nonsense, but you say you are drawing less/about 3 amps off the rail. The maker states that it may show this error if the supply is not delivering 70% of the load as determined by the load share signal. Maybe under a very light load there is some sort of imbalance from one supply to the next that is "backfeeding" and confusing the load share senseing circuitry, generating an error. Slap a bigger load on the rail and see if the supplies settle down.

The reason this comes to mind is that years ago we had some of those linear Lamda ( 2 foot by about 8" square ) supplies that would go ape on each other if they varied more than a few tenths of millivolts. They would start lowering their output based on sense voltage, then get backdriven, ramp up, etc, and basically start to cycle all over the place.