28 volt military installations

[izak]

I am looking for someone who has done some DC work, and, possibly some 400hz work as well... i am working on an air force base, and am a little confused about how we are installing our wire for the 28 volt DC.

thank you

 

[dereckbc]

I do a lot of DC design work, and know a little about 400 HZ. Have you got a question?

The biggest difference in both applications the conductors will be larger than what is called for in the NEC applications.

 

[izak]

ok this is what i got....

i am on a job at an air base putting in a PMEL Lab

dont know what that means but...

for the DC work, on the prints, we will have... lets just say.. circuits 1,3,5,7,9,11 blah blah blah

now if we group 3 circuits together.. 1,3,5 for a home run.... how many wires do i need to make it work? 4 or 6?

am i right in assuming that DC CAN NOT SHARE a Negative like ac can share neutrals?


there is alot more to this and i dont want to get off on too many tangents, but i have a feeling that what we are putting in is NOT GOING TO WORK

i need someone to tell me how these systems are different from the common ac systems that i am more familiar with

 

[dereckbc]

now if we group 3 circuits together.. 1,3,5 for a home run.... how many wires do i need to make it work? 4 or 6?

am i right in assuming that DC CAN NOT SHARE a Negative like ac can share neutrals?

You are correct, you cannot share returns on DC like you can AC.

In AC systems like single phase, the currents are 180 degrees out of phase and subtract when shared on a common neutral. It is possible to have 0 current on a common nuetral assuming the loads are equal.

Not possible on DC, the two load currents will add up. For example if you had 5-amps load on two circuits sharing a common return, the return would have 10-amps.

 

[winnie]

DC systems can most certainly share a common conductor, in a fashion similar to single phase AC systems.

However you can only do this if you have a center tapped DC supply. In other words, if you had a 56V system with a center tap, then you could have -28V on one bus, +28V on another bus, and 0V on your neutral. Some circuits would be -28V and ground, other circuits would be +28V and ground. Edison did this for lighting...but I doubt that it is what is expected in your installation.

If all the circuits are -ground, then neutral current will be additive.

-Jon

 

[petersonra]

you can share commons, you just have to be careful about overloading the common.

in fact, it is not unusual to have a common bus where all the commons are tied together.

 

[dereckbc]

Jon,
can you name one premises or vehicle power system that uses bipolar DC power. I cannot think of one. Only thing I can think of off the top of my head is an internal dc power supply for things like a high power audio amplifier and electronic equipment. The three systems I know of are 24 + VDC, 28 + VDC, 48 –VDC, 130 VDC, and a couple of other odd non referenced higher voltages used in UPS applications. None of these are bipolar, and none share returns on the branch distribution..

Bob,
So how would you size that circuit per NEC of say a 30-amp breaker placed on the hot conductor, say with 50 feet one-way? What rules apply?

All AC and DC systems share common return at some point. For a DC system that point is the DC return bus located inside the Breaker Distribution Frame, it’s AC counter part is the Main Distribution Panel, Breaker or Sub Panel, ect. However once you get to branch distribution each circuit has it own dedicated circuit conductor for each polarity. The only thing in a DC system that is shared is a single wire equipment frame ground.

 

[dereckbc]

Bob,
So how would you size that circuit per NEC of say a 30-amp breaker placed on the hot conductor, say with 50 feet one-way? What rules apply?

I will give you the answer. A #6 AWG on the hot polarity, and a #2 on the common return assuming the common is sharing two hot conductors. I will let you think about this a while before explaining how I came up with these numbers.

Hint: Throw away the code book, and think about voltage drop.

 

[winnie]

Jon,
can you name one premises or vehicle power system that uses bipolar DC power. I cannot think of one. Only thing I can think of off the top of my head is an internal dc power supply for things like a high power audio amplifier and electronic equipment. The three systems I know of are 24 + VDC, 28 + VDC, 48 ?VDC, 130 VDC, and a couple of other odd non referenced higher voltages used in UPS applications. None of these are bipolar, and none share returns on the branch distribution..

I cannot come up with a DC power distribution system _in current use_ that uses bipolar DC. As you note, bipolar DC is pretty common inside systems; the internal example that I think of is in a VFD, where the DC rail will be pretty well balanced relative to the supply neutral. I bet that somewhere there is some large DC system that uses rectified AC and is thus bipolar DC as a result, but I couldn't point you to it.

If I recall correctly, when Edison was building the first electrical lighting systems, he used bipolar DC with a common return, in exactly the same way we use MWBCs today. This was specifically to eliminate voltage drop on the common, and to reduce the number of conductors required. MWBCs are still sometimes called 'Edison circuits', and see 250.162(B) for a bit of code still linking this bit of history.

-Jon

 

[petersonra]

The biggest difference in both applications the conductors will be larger than what is called for in the NEC applications.

I don't do much work on distribution of 28VDC and 400 Hz power, but do a fair amount of design work on equipment for these applications. I am always amazed at how small designers make these wires inside equipment. I have seen 8A circuits that used #20 wire, using teflon insulated wire. The stuff must glow.

I think you will find that the little bit of extra voltage drop at 400 hz is mostly noticable with larger conductor sizes. Below about #6, it is usually ignorable. Above 1/0 it can become a major pain.

 

[petersonra]

I will give you the answer. A #6 AWG on the hot polarity, and a #2 on the common return assuming the common is sharing two hot conductors. I will let you think about this a while before explaining how I came up with these numbers.

Hint: Throw away the code book, and think about voltage drop.

Another hint. The voltage tolerance specs still have to be met, and 3% won't cut it.

 

[cpal]

If this is a Gov't installation are there not spec's and prints???

 

[petersonra]

If this is a Gov't installation are there not spec's and prints???

If I had to guess, I would say there are probably drawings and specs a foot high.

 

[dereckbc]

Another hint. The voltage tolerance specs still have to be met, and 3% won't cut it.

Actually 3% is just about right. On 24 VDC (27 VDC operating) is designed for .75 to 1 volt drop, and 48 VDC (54 VDC operating) for 1.5 to 2 volts. So with those figures the range is 2.7 to 3.7 percent loop voltage drop.

Here is how I came up with my cable sizes. I calculated the cable sizes using a modified formula used by DC power engineers that you will recognize:
CM = (22.2 * I * D) / VD
Where:
CM = Circular Mills of copper cable needed.
22.2 = Constant of a copper loop.
I = Max load current.
D = One-way distance.
VD = Voltage drop.

So from my example using a 30-amp breaker and 50-feet, I have to figure max current which is .8 * 30 = 24-amps for the hot conductor, and 48-amps on the common return conductor. I used 1-volt for voltage drop to simplify the equation to CM = 22.2 * I * D. Now it is just a matter of plugging in the values of the variables.

The whole point here is this is not an NEC application. You cannot use the code or an EC to determine cable sizes. It has to be designed. My guess is most EC's would have determined the wire size by the breaker and would have installed a #10 AWG for both hot and common return.

 

[petersonra]

The milspec stuff I have worked on does not allow 3% voltage drop in the distribution network.

 

[dereckbc]

The milspec stuff I have worked on does not allow 3% voltage drop in the distribution network.

OK, so what do they spec. It just means the wire gets larger.

 

[izak]

the chart we are going by tells the allowable distance each conductor can travel... 8 feet for a #12 on 28 volt DC... somewhat more for a #12 on 120/208 400hz single phase, a little more for 120/208 3 phase..

but, as far as i can tell, the prints only show... get circuits 1 or 26 or whatever to THIS point, and this is how far you can take a given conductor...

on the DC work, there are runs over a hundred feet that we are pulling 1/0 wire to... and DONT say anything about box fill.... i still dont know how we are going to splice #1 down to #10 ... in a 4sq deep box... twice.... and then install the twistlok device...

but i havent seen anything ANYWHERE that says " pull one positive conductor and one negative conductor per circuit"

NOTHING

just circuit numbers on a peice of paper

bear in mind... i am not running this job, nor have i been on it much longer than a month... but i have asked ALOT of questions and im pretty sure that its NOT going to work...

 

[dereckbc]

the chart we are going by tells the allowable distance each conductor can travel... 8 feet for a #12 on 28 volt DC... somewhat more for a #12 on 120/208 400hz single phase, a little more for 120/208 3 phase../QUOTE]

So what is the OCPD size? That is a clue.

 

[izak]

oh sorry its been so long.... got caught up splicing #1 wire to #10 with BLUE WIRENUTS!!!

gawd this stuff pisses me off......

the OCPD is a 30 amp SQD Iline single pole breaker...
we are into the trim out stage now, and a little closer to )MELTDOWN(
I am eager to see what happens....

 

[tom baker]

Didn't edison invent the three wire Edison system and it must of been DC?
It did it to save copper