DC voltage measured on AC supply (or so the story goes)

[wirenut1980]

Good morning! I have a customer with a technician who claims to have measured 7-8 V DC on an AC utility supply. They measured line to line on a single phase 120/240 V service with a Fluke handheld meter (the model # escapes me). A week earlier, this customer had a large UPS fail several batteries. The UPS technician is the one who originally measured the DC voltage. He is claiming that the DC voltage failed the batteries.

A few days later, I was called out to investigate. The technician and I were unable to measure the DC voltage at this time.

Some additional background information: this is a communications tower with a few companies that have equipment attached. AT&T is one company and they have some DC equipment.

Also, a couple weeks ago, there was some sort of surge event that took out some surge protectors. I don't know if it was lightning or not. I think it is more likely this surge is what is responsible for failing the batteries rather than 7 V DC.

My first question is it possible to measure DC voltage on an AC supply using something like a Fluke 87? Next, any thoughts on where the DC might have come from if it was really there? One theory is a rectifier on AT&T equipment went bad and was dumping DC voltage back on the supply. AT&T comes out and replaces their equipment without telling anyone and the voltage is gone when I get there later.

I do have a PQ monitor that can measure 0-600 V DC and record over time. I think I am going to hook it up to the incoming 240 V AC and see if it measures anything.

Any thoughts are appreciated.

 

[kwired]

I'd say it is possible to read a DC voltage on an AC line.

I'll also say the rectifier on the front end of his UPS wouldn't care if that voltage is there, unless it were at a high enough voltage to be above it's normal operating voltage. Rectifier puts out DC voltage whether input is AC or DC.

 

[GoldDigger]

I'd say it is possible to read a DC voltage on an AC line.

I'll also say the rectifier on the front end of his UPS wouldn't care if that voltage is there, unless it were at a high enough voltage to be above it's normal operating voltage. Rectifier puts out DC voltage whether input is AC or DC.

One consequence that would affect the rectifier rsther than the batteries is that it would see higher current on one or more of the pulses instead of a uniform current distribution. Worst case it could reduce the maximum current the rectifier bridge can supply, just as operating on single rather than three phase input would.

 

[Jraef]

What he said. The UPS technician is just clutching at straws and looking for something else to blame. I think I can measure 8VDC with a Fluke meter by waving the probes around in the air...

Whenever there is rectified AC line voltage, there is the possibility of having what is called a "DC offset", meaning that because of the voltage distortion that a rectifier creates on the AC line, the peaks of two lines are not the same so the zero crosses might not be the same either and the difference in the zero crosses looks like DC when you measure line to line. Voltage distortion is caused by current distortion, which is related to the ratio of available current to load current at any given moment. This just means that if, the day you went to check on it the equipment was pulling less current than on the day the technician was there, you may not have seen it.

But a line rectifier for a good quality battery charger should not care. Only if the charger circuit is old dirt cheap junk and uses SCRs on the front end for rectification, instead of a simple diode bridge, it MIGHT care because the zero cross points are used to determine the firing angles of the SCRs for voltage control. Old battery charger designs had this issue, back in the days when batteries were just car batteries; cheap and robust, so a little extra voltage float was fine, and there were not a lot of other rectified power supplies like there are now. Even then, an 8VDC offset would not have caused much of a problem on the battery side.

But modern versions of chargers now use a Switch Mode Power Supply (SMPS) concept which makes that irrelevant anyway because the voltage is controlled on the DC side, not the AC side. Any DC offset on the AC side becomes irrelevant because DC offset cannot pass through the transformer that is on the front end of the SMPS. A high DC offset might eventually cause the SMPS transformer to overheat, especially a cheaply designed one, but that would only affect the batteries down stream when the charger itself completely fried, which would be obvious.

Your theory on the surge event is much much more plausible.

 

[iceworm]

Good morning! I have a customer with a technician who claims to have measured 7-8 V DC on an AC utility supply.

... A week earlier, this customer had a large UPS fail several batteries. The UPS technician is the one who originally measured the DC voltage. He is claiming that the DC voltage failed the batteries.

A few days later, I was called out to investigate. The technician and I were unable to measure the DC voltage at this time.

Some additional background information: this is a communications tower with a few companies that have equipment attached. AT&T is one company and they have some DC equipment.

Also, a couple weeks ago, there was some sort of surge event that took out some surge protectors. I don't know if it was lightning or not. I think it is more likely this surge is what is responsible for failing the batteries rather than 7 V DC.

My first question is it possible to measure DC voltage on an AC supply using something like a Fluke 87? Next, any thoughts on where the DC might have come from if it was really there? One theory is a rectifier on AT&T equipment went bad and was dumping DC voltage back on the supply. AT&T comes out and replaces their equipment without telling anyone and the voltage is gone when I get there later.

I do have a PQ monitor that can measure 0-600 V DC and record over time. I think I am going to hook it up to the incoming 240 V AC and see if it measures anything.
.

Give us some context:
Since it is telephone, I'm guessing 48V, positive ground.

What size batteries and charger? 20AH, 200AH, 2000Ah?
Is the UPS an inverter with an AC output, or just a DC output?
What type batteries? VRLA, AGM, Wet lead acid (Calcium or Antimony)?
What was the battery failure mode? Did they internally short, boil out, go dead?

As for a DC offset on the incoming voltage to the charger causing failing the batteries - A few questions for the battery tech:
The DC offset caused the charger output to change to where it either overcharged, under charged, or increased ripple? If so, which was it?

Here is an example:
If the batts are VLRA, or AGM, and the charger failed full on, overcharging, that will murder the batts.


As for a charger failure mode that would cause the charger to put DC back on the incoming line - Of the charger circuits I've seen, I don't know of any failure mode that would inject a DC offset back on the line.

I really don't know how a surge event would cause the batteries to fail - Unless the surge event burned up the charger, allowing the batteries to discharge completely.

Not enough inforation from the battery tech. All we got is:
There is an intermittent 7VDC offest on the incoming line
There was a surge event some time previous
The batteries failed (unknown failure mode)
Comm tower, so there is likely positive ground 48VDC system(s)

PS: What is a surge event?

ice

 

[iceworm]

...But a line rectifier for a good quality battery charger should not care. Only if the charger circuit is old dirt cheap junk and uses SCRs on the front end for rectification, instead of a simple diode bridge, it MIGHT care because the zero cross points are used to determine the firing angles of the SCRs for voltage control. Old battery charger designs had this issue, back in the days when batteries were just car batteries; cheap and robust, so a little extra voltage float was fine, and there were not a lot of other rectified power supplies like there are now. Even then, an 8VDC offset would not have caused much of a problem on the battery side.

But modern versions of chargers now use a Switch Mode Power Supply (SMPS) concept which makes that irrelevant anyway because the voltage is controlled on the DC side, not the AC side. Any DC offset on the AC side becomes irrelevant because DC offset cannot pass through the transformer that is on the front end of the SMPS. ...

Hummm ... There are new, high quality rectifiers out there using mag-amps and SCRs on the input. I have not seen switch mode power supplies used for rectifiers once one gets above maybe 75A. He asks - Why would one still use mag-amps? Because they are bullet-proof. Switch modes are not. And the OP did say, "this customer had a large UPS fail several batteries.

What ...Your theory on the surge event is much much more plausible.

What is the theory on the surge event?
Is this a sustained over voltage on the rectifier line side caused batteries to fail? This is not a household plug-in UPS - I'm not seeing this.

... The UPS technician is just clutching at straws and looking for something else to blame. .

I agree with this. Time to pin him down and request information above hand waving "DC offset" or the elusive "SURGE".

What failed?
What was the failure mode?

ice

 

[Jraef]

Hummm ... There are new, high quality rectifiers out there using mag-amps and SCRs on the input. I have not seen switch mode power supplies used for rectifiers once one gets above maybe 75A. He asks - Why would one still use mag-amps? Because they are bullet-proof. Switch modes are not. And the OP did say, "this customer had a large UPS fail several batteries.
ice

I was thinking float charger, not a cycle service charger. You're right, this is more likely a cycle service charger and the loads are likely DC directly, no inverter back to AC, so all of the power probably has to come through the rectifier, not just the float power. So it's going to more likely be a mag-amp or ferroresonant unit anyway.

What I meant by SCRs was an old phase angle controlled front-end rectifier, not a mag-amp. The SCR in a mag-amp is controlling the mag-amp itself, so being on the DOWNSTREAM side of the saturable reactor it's not going to see a DC offset on the AC side either. The same would be true of a ferro.

Bottom line is that no matter how you get there, an 8VDC offset would not even show up very high on my list of things to worry about.