Negative 48 Volts DC

On 1/26/20 8:22 PM, gray_wolf wrote:
> I think you have a poor understanding of basic current flow theory and 
> likely very little practical experience.

I /thought/ I had an acceptable understanding, particularly for a server 
/ network administrator.

> This may be a good place to start.

> So your job is connecting equipment to other equipment and you have 
> no idea how to do this?

That is a bit ambiguous.

My job is working with server and network equipment, including racking / 
stacking / cabling (data and power).  AC power is quite simple and to be 
honest, hard to mess up.  DC power is much more rare and requires being 
more careful.  Hence this thread.

Copper data connections can be tricky, particularly if equipment is 
powered from different legs / phases and something about the power is 
not correct.  That's a great way to burn out ports.  Sneak current's are 
a bitch.  Ground loops can be a problem if you're not careful.  Optical 
cables are safer to work with.

So, "connecting" is a bit ambiguous and can involve a number of 
different things.



-- 
Grant. . . .
unix || die

On 1/27/20 9:46 AM, Whoey Louie wrote:
> Why is it confusing?  One terminal is the negative polarity, the other
> the positive.  On your supply, one terminal is marked -48V, the other
> return.  I would also make the -48V black, the return one red.

Having looked at the equipment earlier today, you'd be wrong.

Negative had the red wires.
Positive / ground had the black wires.

The confusion is understanding why things are wired the way that they are.

> but even with just -48V on one and return on the other, it's clear 
> which is negative, which is positive.  If you had a 9V battery and 
> one terminal was marked -9V, or 'neg", would you be confused about 
> the polarity?

In a negative ground system, no, I would not be confused.

In a positive ground system, I apparently am.

> Woah there Pilgrim!  Are you saying on that -48V supply has a
> RED terminal or wire for the -48V and BLACK for return?

Yes!!!

That is EXACTLY what I'm saying.  (See above.)

> Now that I would agree would be confusing.

Hence this thread.

> I would expect black to be on the -48V, red on the return.

Nope.

I don't know what I would have expected / guessed this time last week. 
I'm loath to guess, lest I confuse myself more this week.

> It's either one polarity or the other, positive is never called negative.
> What is called the "hot", the "source" or the "return" could be either
> depending on the system polarity.

I'm slowly coming to the conclusion that I have naively assumed that red 
≈ positive.  Which is obviously wrong for positive ground systems.

I think I need to retrain myself that red = hot and that what is hot is 
dependent on the type of system it is; negative ground or positive ground.

> With one voltage, two wires like you have, it can't.

I think that I need to reword my statement to be "a hot wire has a 
negative voltage … in a positive ground system".

> But it does make sense.  That power supply is designed to be used in a
> positive ground system.  So the supply terminal is labeled -48V.

I checked, and one terminal is labeled "negative" and the other terminal 
is labeled "positive".

> So what color is the -48V terminal?  Is there a color?  I would expect it
> to be either black or no color identification and the return to be either
> red or no color.

There is no color (save for polished / chrome steal).

The red wire was connected to negative.  The black wire was connected to 
positive.

I've not had the opportunity to measure the voltage yet.  I'm planing on 
doing that tomorrow.

> Why are you now adding neutral to this?  Neutral is on the AC side.

I'm fairly certain that neutral and ground are bonded together somewhere 
upstream.  Thus I typed "neutral / ground" as "neutral or ground". 
Because, from an electrical conductivity standpoint, they are at the 
same potential and in some ways the same.



-- 
Grant. . . .
unix || die

On 1/26/20 3:36 PM, Rick C wrote:
> I don't assume something I don't know about.  You'll have to measure.

I'll try to make a point to do so tomorrow.

> Hot or power.

Okay.

Unless anyone objects, and provides an explanation of why they object, 
I'm going to try to make a point to use the term "hot".

> Ground/common is typically used for the "reference" of the power.

I'm learning that.

Previously, I had assumed that ground / common also made reference to 
the negative side.  I'm having to unlearn that assumption as I go 
through my Negative 48 Volt DC journey.

> So the other lead will be positive in this system and so, hot.

I want to agree, but I'm reticent because of the word "positive" in there.

I agree that hot will be positive in a negative ground system.

I'm learning that hot will be negative in a positive ground system.

> Just that the definition of -48 volt power is that the negative is 
> the power or hot lead and the positive is common.

Thank you.

> I really shouldn't use "ground" for the common lead since ground 
> is often used for "protective earth" which is a different concept.

Agreed.

I think chassis is another similar thing that can be used as a 
reference, which may also be distinct from ground (protective earth).

> They may be connected, but are not the same thing, just as neutral and 
> protective earth are connected (if all is well) but not the same thing.

Agreed.

> If I mean protective earth I will say protective earth, otherwise 
> ground will mean common.

Okay.

> Not really.  I think you are making this a bit too complicated than 
> it is.  I'm not complaining, just stating an observation.

Fair.

> I expect that by the time you get to the equipment this will all be 
> much more clear to you.

It's starting to.



-- 
Grant. . . .
unix || die

On Monday, January 27, 2020 at 9:09:27 PM UTC-5, Grant Taylor wrote:
> On 1/26/20 3:28 PM, Rick C wrote:
> > That's why I would look at the documentation for the equipment/supply.
> 
> That is the plan.  I've not had the opportunity to do so yet.
> 
> > If it is a -48 volt system it certainly would be ok to call the 
> > positive terminal "return" or "common".  It would be ok to call the 
> > negative terminal "hot".
> 
> Okay.
> 
> > I have seen some systems use a negative ground with a 48 volt power 
> > source.
> 
> Fair enough.  I've not had the exposure to (positive) 48 volt power.
> 
> > Now you are talking color codes.  I can't tell you what the color 
> > code is for a -48 volt system.
> 
> I wasn't meaning to focus on the color code.  Allow me to modify my 
> statement slightly by removing the colors from it.
> 
> I believe that if I hook the positive lead of a voltmeter to the 
> positive lead of a power supply, and the negative lead of the volt meter 
> to the negative lead of a power supply, I expect the following voltages:
> 
> Standard bench top 12 VDC power supply:  + 12 VDC
> Negative 48 VDC telephone power supply:  - 48 VDC

Positive and negative are electrical terms of physics.  It has nothing to do with convention of colors or which polarity of common you use.  Connect the positive meter probe to the positive power line and you will always measure a positive voltage no matter what.  


> > "Called" vs. "looks" I can't say.  I never talk about the "red" lead 
> > when dealing with power circuits.  I use the other terms listed above.
> 
> Okay.
> 
> > Now you have me confused again.  Who talked about a positive wire with 
> > a negative voltage???  Positive/negative are electric polarities and 
> > don't change when your definition of the common or ground changes.
> 
> I believe it was a permutation of "hot" -> "red" -> "positive" in a 
> system that has negative voltage.

Let's forget color.  If it is a -48 volt system the positive wire will be the common, but it will still be positive.  


> > Ok, if they say it's a -48 volt supply with a minus sign in front 
> > of the voltage, then it is safe to believe the hot wire will be the 
> > negative the return/common wire will be the positive.
> 
> Okay.
> 
> > Ok, because I'm willing to bet the supply is not grounded, but it 
> > may be.  If not grounded they are only calling it a -48 volt supply 
> > because that is how most people use it.  There is nothing inherent 
> > that makes it minus vs. positive.
> 
> I had a few minutes to look at the -48 VDC power supply today.
> 
>   · The negative terminal was connected to the red wires.
>   · The positive terminal was connected to the black wires.
>   · The positive terminal was connected to the chassis.
>   · The chassis will make electrical contact with a rack that is 
> connected to earth ground.

Ok, so they are using the convention that red is "hot" and black is "common".  They are also grounding their common to earth ground which is likely connected to a protective earth wire. 


> > The volt meter will always tell you the voltage between the two 
> > probes assuming the red will be positive and the black negative.
> 
> The implication of polarity with volt meter leads doesn't help this 
> thread.  But I generally agree with your statement.
> 
> > If that is correct the meter will show a positive voltage.  I the 
> > red lead is negative it will show a negative voltage.
> 
> I believe the red lead does show a negative voltage.

I was talking about the voltmeter leads.  But whatever. 


> > That will give you a positive voltage on the positive lead with 
> > respect to (wrt) ground.
> 
> I still need to measure, but I believe the positive lead has 0 volts to 
> ground.

You just said above the positive power lead is common and connected to a grounded chassis.  You can't get any more zero than that.  Of course you have to use common or ground as your reference (e.g. connect the black meter lead to common). 


> > That will give you a negative voltage on the negative lead wrt ground.
> 
> Agreed.
> 
> > BTW, none of these supplies should have any connection to power line 
> > neutral.  That is dangerous situation and can kill.  Neutral is not 
> > ground (protective earth).
> 
> I agree that ground and neutral (in typical 120/240 VAC 1ɸ wiring) are 
> separate conductors and should not be used interchangeably.  However, 
> ground and neutral are bonded together in the main electrical panel. 
> (At least every main panel that I've ever looked in.  Sub-panels are 
> decidedly different with ground and neutral isolated from each other.) 
> So, there is a conductive path between ground and neutral.

And that is the hazard.  If the neutral connection to the panel is lost, there is no return path for the power circuit and anything touching neutral will be hot.  The bond between protective earth and neutral must be connected in a way that no failure can cause the protective earth to be hot. 


> > Why would you bond any voltage line to earth?  Do your equipment 
> > systems use actual grounds anywhere?  It's easy enough to measure 
> > the voltages and see if the power is connected to ground.
> 
> Ground is quite prevalent in networking equipment and servers.  Some 
> vendors refuse to honor warranty if the equipment is not properly 
> grounded.  Static is a (big) concern.  So ground is used to dissipate 
> static before it builds up and causes problems.

That would imply the equipment cabinet be grounded and the equipment connected to that.  The power source can be floating and the equipment should not care.  Do the equipment vendors define "properly grounded", that is, any specific instructions? 


> > I'd say a measurement is in order... with the resistor I mentioned.
> 
> After looking at the -48 VDC power supply earlier today, I am quite 
> convinced that the positive terminal has electrical conductivity with 
> the building's earth ground.  (Positive terminal strapped to chassis 
> which is bare metal connecting to the rack it mounts in (also bare 
> metal) which is earth grounded.)

Is this strap at the power supply or the equipment being powered?  Braided copper strap or just a regular wire?  Just curious.  As I've said, I've not worked with this sort of equipment. 

-- 

  Rick C.

  +-+ Get 1,000 miles of free Supercharging
  +-+ Tesla referral code - https://ts.la/richard11209

On Monday, January 27, 2020 at 9:37:34 PM UTC-5, Grant Taylor wrote:
> On 1/26/20 3:36 PM, Rick C wrote:
> > I don't assume something I don't know about.  You'll have to measure.
> 
> I'll try to make a point to do so tomorrow.
> 
> > Hot or power.
> 
> Okay.
> 
> Unless anyone objects, and provides an explanation of why they object, 
> I'm going to try to make a point to use the term "hot".
> 
> > Ground/common is typically used for the "reference" of the power.
> 
> I'm learning that.
> 
> Previously, I had assumed that ground / common also made reference to 
> the negative side.  I'm having to unlearn that assumption as I go 
> through my Negative 48 Volt DC journey.

Yes, that is the only real difference. 


> > So the other lead will be positive in this system and so, hot.
> 
> I want to agree, but I'm reticent because of the word "positive" in there.

I'm talking about negative ground systems which I consider "typical".  


> I agree that hot will be positive in a negative ground system.

Yes. 


> I'm learning that hot will be negative in a positive ground system.

Yes. 


> > Just that the definition of -48 volt power is that the negative is 
> > the power or hot lead and the positive is common.
> 
> Thank you.
> 
> > I really shouldn't use "ground" for the common lead since ground 
> > is often used for "protective earth" which is a different concept.
> 
> Agreed.
> 
> I think chassis is another similar thing that can be used as a 
> reference, which may also be distinct from ground (protective earth).

No, any exposed metal is always connected to protective earth.  That way no matter what happens with the power circuit (AC or DC) the exposed metal parts won't become "hot".  Either a fuse will blow or the power source will blow up.  lol 

That may also be connected to the common power rail, but the point is it MUST be protective earth. 


> > They may be connected, but are not the same thing, just as neutral and 
> > protective earth are connected (if all is well) but not the same thing.
> 
> Agreed.
> 
> > If I mean protective earth I will say protective earth, otherwise 
> > ground will mean common.
> 
> Okay.
> 
> > Not really.  I think you are making this a bit too complicated than 
> > it is.  I'm not complaining, just stating an observation.
> 
> Fair.
> 
> > I expect that by the time you get to the equipment this will all be 
> > much more clear to you.
> 
> It's starting to.

Yeah, it's not hard, but takes a bit of work to get the mental connections correct. 

-- 

  Rick C.

  ++- Get 1,000 miles of free Supercharging
  ++- Tesla referral code - https://ts.la/richard11209

On 1/27/20 8:17 PM, Rick C wrote:
> Positive and negative are electrical terms of physics.  It has nothing 
> to do with convention of colors or which polarity of common you use. 
> Connect the positive meter probe to the positive power line and you 
> will always measure a positive voltage no matter what.

Presuming the negative meter probe is connected to negative power line, 
and I agree.

The thing that I've learned is that there is no correlation between 
"hot" / "common" / "return" / "ground" with positive or negative like I 
have naively thought for the last 30 years.

> That would imply the equipment cabinet be grounded and the equipment 
> connected to that.  The power source can be floating and the equipment 
> should not care.  Do the equipment vendors define "properly grounded", 
> that is, any specific instructions?

Yes.  The better documentation calls out very specific directions on how 
to bond ground to the equipment.  They are quite specific about earth / 
safety ground.

> Is this strap at the power supply or the equipment being powered?

It is part of part of the power supply.

> Braided copper strap or just a regular wire?  Just curious.  As I've 
> said, I've not worked with this sort of equipment.

It is a piece of solid (what looks to be) steal that fits over the bolts 
of the positive terminal where wires get bolted on top, thus sandwiching 
the steal between the terminals and the restraining bolts.  The other 
end is bolted to the metal chassis.



-- 
Grant. . . .
unix || die

On 1/27/20 8:39 PM, Rick C wrote:
> Yes, that is the only real difference.

*nod*

> I'm talking about negative ground systems which I consider "typical".

Fair enough.

I agree, negative ground has been about 99% of what I've seen and worked 
around.

> No, any exposed metal is always connected to protective earth. 
> That way no matter what happens with the power circuit (AC or DC) 
> the exposed metal parts won't become "hot".  Either a fuse will blow 
> or the power source will blow up.  lol

Sorry I mistyped, I was thinking where common and hot were floating 
inside of and insulated from the chassis and the chassis itself is 
grounded to earth.  Thus chassis and and common can be at two different 
potentials.

> That may also be connected to the common power rail, but the point 
> is it MUST be protective earth.

*nod*

> Yeah, it's not hard, but takes a bit of work to get the mental 
> connections correct.

Yep.  Now that I've bifurcated ground / common from the negative 
polarity, along with some better understanding of terms, I think I'm 
starting to understand.



-- 
Grant. . . .
unix || die

On Monday, January 27, 2020 at 11:20:11 PM UTC-5, Grant Taylor wrote:
> On 1/27/20 8:17 PM, Rick C wrote:
> > Positive and negative are electrical terms of physics.  It has nothing 
> > to do with convention of colors or which polarity of common you use. 
> > Connect the positive meter probe to the positive power line and you 
> > will always measure a positive voltage no matter what.
> 
> Presuming the negative meter probe is connected to negative power line, 
> and I agree.
> 
> The thing that I've learned is that there is no correlation between 
> "hot" / "common" / "return" / "ground" with positive or negative like I 
> have naively thought for the last 30 years.

That's because positive ground systems are the exception rather than the norm.  Telephony related gear is an exception... when powered from 48 volts, but not always even then.  I have seen gear that didn't have an internal connection to common and so could be used as positive ground or negative ground. 


> > That would imply the equipment cabinet be grounded and the equipment 
> > connected to that.  The power source can be floating and the equipment 
> > should not care.  Do the equipment vendors define "properly grounded", 
> > that is, any specific instructions?
> 
> Yes.  The better documentation calls out very specific directions on how 
> to bond ground to the equipment.  They are quite specific about earth / 
> safety ground.
> 
> > Is this strap at the power supply or the equipment being powered?
> 
> It is part of part of the power supply.

That surprises me a bit.  A AC/DC power supply output is typically isolated from the input power line, so common is whichever terminal you connect to... common.  Having a connection internally to the AC protective earth limits the applications with no added advantage. 


> > Braided copper strap or just a regular wire?  Just curious.  As I've 
> > said, I've not worked with this sort of equipment.
> 
> It is a piece of solid (what looks to be) steal that fits over the bolts 
> of the positive terminal where wires get bolted on top, thus sandwiching 
> the steal between the terminals and the restraining bolts.  The other 
> end is bolted to the metal chassis.

That sounds very external then.  That makes sense.  That's not really part of the power supply.  They just provide the strap and you likely can use it on either output.  But since the positive terminal is labeled "common", maybe not. 

-- 

  Rick C.

  +++ Get 1,000 miles of free Supercharging
  +++ Tesla referral code - https://ts.la/richard11209

On Sunday, January 26, 2020 at 4:30:12 PM UTC-5, Jeff Liebermann wrote:
> On Sun, 26 Jan 2020 10:16:41 -0800 (PST), mpm <mpmillard@aol.com>
> wrote:
> 
> >On Sunday, January 26, 2020 at 2:10:41 AM UTC-5, Michael Terrell wrote:
> >> A communications system that I worked on for the International Space Station ran on -48VDC.
> 
> Hmmm...  I guess the ISS doesn't need to worry about ground leakage.
> 
> >I don't know if it is still the case, but that last (and probably only) time 
> >I had to install a two-way radio into a backhoe or bulldozer - that was 
> >a "positive ground" system.
> 
> In 1971, after graduating college directly into a recession, the only
> job I could find was working for the local 2-way radio shop.  I spend
> 2 months installing Motorola Motrac radios in concrete trucks.  I
> think they were either Mack or Peterbilt but I'm not sure.  Most had
> positive ground system.  There were also some 24 volt systems (2
> batteries in series) but we were able to tap off one battery.  


There were plastic cases with Mobile CB radios mounded inside for Truckers, back in the '70s. They had a 6A or 25A full wave bridge on the DC input, and a mag mount antenna with a thicker than normal insulating pad on the mag mount, or a 1:! RF transformer to isolate the output for Mirror mount antennas. They were popular with Fleet truckers who never knew if the truck would have positive or negative ground. They were a pain in the ass on either grounding system, but truckers bought them.

On Sunday, January 26, 2020 at 4:30:12 PM UTC-5, Jeff Liebermann wrote:
> 
> >On Sunday, January 26, 2020 at 2:10:41 AM UTC-5, Michael Terrell wrote:
> >
> >A communications system that I worked on for the International Space Station ran on -48VDC.
> 
> Hmmm...  I guess the ISS doesn't need to worry about ground leakage.


They used some COTS -48VDC items, so the entire station was designed to operate off the batteries and solar cells at that voltage. 

The equipment that we supplied used a pair of Vicor 120/240 input supplies. One +5VDC and one +12/-12V dual output supply. It was a simple modification to replace them with a single, triple output,-48VDC input power supply. The biggest problem was that we only had one power supply at the factory that could power it, so it had to be moved from station to station through final test with the electronics.