Naval AC power 1

[ScottI2R]

Hi again all,
After talking with a few co-workers that were in the US Navy, I have learned that on board ship, a whole lot of the equipment used was powered by 400Hz ac. I was wondering why. Can anybody help?

BTW, one guy was an EWO, man, I couldn't believe some of the unclassified "techspecs" and capabilities he told me about! ((Unfortunately, he STILL couldn't answer 75% of the questions I asked (Still Classified!)...Guess I'll try "Janes"..))

Thanks,
Scott

In a hundred years, it isn't going to matter anyway.

 

[Warpspeed]

If I may put in my two cents worth here.

I am not a military man, but it seems to me that most military power has to be self contained, mobile, light weight and compact. Airborne, marine, or vehicle mounted, the one aspect is that it must all operate well away from the national power grid.

400 Hz has a lot of advantages mostly for wound components, alternators, transformers and motors. All will be much smaller,lighter and more compact.

A very good example of this is the ordinary humble automotive 12v car alternator. They are very compact for something that may be rated at 14v and maybe 60 amps. Not a lot of copper and iron in there either. They do it by running at a much higher than "normal" operating frequency.

But 400Hz has one big disadvantage over 50/60Hz when it comes to very long range power distribution. Problems of surge impedance and resonance, standing waves and other nasty effects can become problematic when you start trying to send 400Hz power over tens or hundreds of miles.

For aerospace 400Hz is definitely the best way. Aboard ship weight is not going to be such a big factor, and I would expect normal commercial 50/60Hz standard voltage machinery would be more economic to buy and install.

 

[TheBlacksmith]

Aboard ship weight is not going to be such a big factor, and I would expect normal commercial 50/60Hz standard voltage machinery would be more economic to buy and install.

Warpspeed - that's it, in a nutshell.

Blacksmith

 

[SolderingGunslinger]

Blacksmith,

Aboard ship, weight may not be a huge factor, however space certainly is. Every cc of space taken by a transformer is one less cc of fuel, cargo, ammunition, food, living space...ad nauseum.

Back in the good (bad) old days, the same radio which sat in a PBY was probably in the radio shack on a Cruiser as well. Either a DC to DC motor-generator (Dynamotor) was used for the mobile plate supply (we’re talking about hollow-state technology here) or the radio was designed to operate off 400 Hz aircraft power. Often the equipment was designed to have external supplies (the old TCS transmitter comes to mind) for each application.

When it comes to tubes, some like the 4CX250 came in a couple of flavors. One had a 24V heater for aircraft use and the other with the 7 V heater. Again, the 24V tube would draw less current than the 7 V unit.


Wrong100, thanks for the wonderful explanation. I reminded me of myh father's explanation nearly a half century ago when I asked why the surplus selsyns ran on 400 Hz. Dad taught Electronics for the Navy.

I remain,
The Old Soldering Gunslinger

 

[zeusfaber]

Thanks Wrong100.

Having been the first one to mention that higher frequency meant less peak flux, it was a bit dim of me not to recognise that that meant fewer turns. Guess I've forgotten more than I'd realised in the 20 years since I last worked on power systems.

A.

 

[Warpspeed]

I have just thought of something else, skin effect.

With dc current the current flows through a conductor evenly through the whole cross sectional area. With ac current the continuously changing magnetic field causes the current to crowd towards the outer "skin" of the conductor. This effect is highly dependant on frequency, and it becomes a limiting factor with high currents at high frequencies.

That is another good reason why it is not practical to transmit thousands of amps of 400 Hz power over very long distances and why 50/60hz is better for that.

 

[wrong100]

Thank you for your "yea".
glad it served. makes me happy to see that you understand.

I just spent 2 hrs writing a thing on pk current,
than I got shut down.
tried to recover, "No".

Is this normal? Not trying to be flip,
but some of the concepts need time to explore
and I am a very slow hin pecker (typer).

Nutshell ;
C=i*t/v been told no, (but math is yes) it works and is valid.
Use the falling voltage of the rectified form

I hate to do this but (short answer) results for
pk current are 60/400 = .15% for a given
ripple voltage with the same cap value and esr.
So acording to the above math, for a given C
if you increase the fc at the same load, ripple
decreases. If you do it the other way, for a given
ripple voltage and current the amount of cap decreases.

Understanding is a time and absorbing thing (remember all things in elec are related to all other things).

If you do not understand I will try to make
it clear.
Now almost 4 hrs on line (typing and
clearing thoughts so people can understand, I do need to
make my thinking into a more specific order)
"go to bed".

I do enjoy this, but I am "upset" about being cut off.
The only way I know, is to explane in a logical manner.
This takes time, (if you cannot relate to your subject,
and person "people" every thing else is for not),
now you can see why and how I call myself a teacher
and why I was so upset
(about what I had written before).
All I do is give them the links to knowladge they already have (how to put it together).
Good will to all.

 

[davidbeach]

wrong100,

If you are having problems maintaining a connection with the site long enough to form your message, perhaps you could type it in Word/Word Pad/Notepad, some other text program, and then do a copy and paste into the forum message box. By the way, you can skip using the 'Enter' key except at the end of your paragraphs. The software will take care of wrapping the words onto the next line.

 

[itsmoked]

wrong100, davidbeach is correct!! If you are going to embark on a relatively long post by all means write it in WordPAD or Notepad or any other word processor. Then you can spell check it or abandon it for the night so you don't muddle it up. You can come back to it the next day. Re-read it, touch it up. Then SAVE IT. Then just copy and past it into here.

ALL of us have had our posts evaporate.

It is very annoying, almost shocking.

Remember that as you are typing your post in the window you are really just typing it into your WEB browser. When you submit it or preview it then it goes to the web... Well the site can go down for maintence or because its buggered or whatever over that long period you are typing and then your work is piffed.

Thanks BTW for your efforts.

 

[wrong100]

Hi people
I am trying the suggestion of wordpad. Thanks.

I will do the last item on the list , plus two others.
If you start with a wire in free space it will generate
a magnetic field around it. As you get futher away
from the center of the wire the mag field is less,
(the same rules of radio transmittation)
To make a very very long story short, the very
center of the wire (due to the induced mag field)
opposes the flow of current. This makes the center
of the wire not as conductive as the outer edges.
The result of this is most of the current flows on
the outside of the wire (on the "skin"). Sence the
conductivy (call it effective area of wire at the fc) of a given wire,
is dependent on the "active (cross section)" wire. The higher
the frequency, the more loss in the wire, compaired
to its dc conductivity component.

At one time I used a "MS BASIC" program that I
had written to calc skin effect, yes that important.

Form my previous comparesion of 60hz to 400hz transformers
the numbers are for 60hz = .358 inches, and for 400hz
= .1377 inches (this is the number which states the "ratio of loss")
(this is about the 50% loss nunber ? level if I remember).
While this is not much for home use
(most home wire is less than .133in in dia) too small
at 100 amp in out, think of power companies).
Lets see .358/.1377 = (1.599-) call it 1.6 times. This loss ratio is
per inch, or foot, or mile or 1000 miles. If I could reduce my losses
to 1/1.6 = 1/.625 = .375%. I know I would do it without thinking.

Put it another way, if you were designing a power system, ( for a power co.)
and had to put 10K or 100M amps (in a wire) this would be very important in
your transformer design and wiring. That is why power conpanies
try to do the best answer (for them the answer is 60hz).
For them 60hz is the best answer, of the tradeoff of total energy
transfered verses size, fc. etc.

Skin effect is related to fc numbers but it is also by nature
related to current, for losses (I^R) it is a resistance rato at best.
1 amp (losses) are dfferent than 100 amp losses, (the square factor of current).
What is not said, is if you are designing a 60hz or 400hz
transforner, unless you get to the .1inch wire dia do not worry (min effect)
unless you are pushing it to the limit, where 10% is a make or break thing.

How about a new concept (problem) with wires.

If you extend the concept of (skin effect) to related wires in a coil,
you end up with a concept of wires carrying currents next to each other.
Start over except that all of the concepts about skin effect apply.
Proof is, what is the difference in mag flux and its control of resistance
based on the center of a wire (take a number .1(skin )),
verses a wire placement .02inches, removed from the center of the wire.
This is called the proximinity effect. The result is it still pushes the wire
current into a smaller portion of the wire, only now to one side.

The third major effect is due to ac currents generating voltage (and therefore)
current loops within the magnetic material as well as in the wiring.
This is eddie current effects, and can be very lossy.

The result is, if a wire has current
in it, it is subject to many various effects, (these effects can be most dominant).

Just to make sure that you understand the results of the concepts.

A design for a 1000 watt transformer "can have (10watts 100watts 1000watts loss)."
It all depends on the base magnetic losses as well as the, fc, wire size,
nunber of strands, size of strands, and on and on, than you get into the mag material,
where it starts all over again.

Yes I understand , the original question was answered.

Question is did you learn anything, or was it too much and became
just words.

Have expanded it with your approval ?

My concept is to let you be exposed to the problems involved
in a 60hz verses 400 hz transformer, and the base concepts that
drive these designs (which are not throught without an unstandering
of the higher fc components, as well as the low fc components).
Like most things, there is no set answer .
I have had eng tell me that they have had two inch dia
bolts melt due to eddie current losses in large 60 hz units.

I am just trying to expose you to something more than the
"yes, no" answer
Did any of this make sense?.

I think my head hurts, not a normal ake, it just hurts thinking
about all of the concepts explained, and not explained.

One thing to the general public (and you mil people in specific)"I design to the mil specs, but am not in the mil".

"LOOSE LIPS SINK SHIPS".
I have been waiting 50 yrs to be able to put this in contex.

I think this runs over 3 or 4 days, hope it is not too much,
I still do not know how much you want, please let me know.

Oh I almost forgot, more detailes can be found if you do a
specific search for "UNITRODE" added words "SEM",
go directly to the "TI\FOCUSE" website, (HANG ON TO
YOUR MIND), it gets very complex very fast, but many good
things can be understood (pick and chose).


Next or not, what to go to now.

 

[itsmoked]

Thanks yet again wrong100. Seems the wordpad method worked!

This stuff came up in college. It's great to get a reminder and the applied angle as apposed to the theoretical angle prevalent in school. Skin effect is generally taught about a wire running through free space NOT in the context of a coil as you have done.

Now you're making me think.. If you have a multilayer coil of wire (like 30 layers) and you are operating at a frequency where skin effect is significant are you saying the skin effect normally present on a single wire is now different on the wires in the middle of this coil verse the outer wraps and the same with the inner most layer?

 

[Warpspeed]

Skin effect exists in all conductors carrying ac current the wire does not have to be formed into a coil. The mechanism is the building up and collapsing magnetic field around a conductor with each half cycle. This drives the bulk of the current to the outside edge.

Skin depth is generally measured as the depth above which 95 % of the current will flow. Skin depth in mm is roughly 66/Hz^0.5 This effect may be completely ignored if the cable diameter is less than twice the skin depth.

Skin depth at 60Hz = 8.5mm
Skin depth at 400Hz = 3.3mm
Skin depth at 20Khz = 0.46mm

If you wanted to send vast amounts of electrical power over a long distance you would try to use as high system voltage as feasible, but also the conductor diameter should not be appreciably more than twice the skin depth. That limits you to the size and current carrying capacity per conductor.

In practice the ac resistance of a thicker conductor is higher than the dc resistance because the whole cross sectional area may not be efficiently used to carry the current. For very high frequency conductors, flat strips and foils are often used, or sometimes even hollow tube, or bundles of thin wires insulated from each other (Litz wire). All these have a large surface area. For high powered radio transmitters silver plated parts are often used. The silver plating carries most of the current.

 

[wrong100]

Yes
squared I think, also prox is squared per layer if I
remember (it is squared for the second layer). I just kind of push buttons now.
The first hf trans took about 2wks (1 pass needed 20).
I got smarter, rearanged some gen math,
got it down to only two days (1 pass).
Did that for a couple of years, than bought a
program.

"confirmed" the program and all following updates.

You must know how the program is doing math.
Kind of like using a slide rule, if you do not
know where your answer should be, how do you know
if it is correct.
If it seams too good it probably is. Aside from
program updates I have only had to do about 1 a yr
for the last seven or eight years (most not critical).
That being said I have about 3 or 4 that need to be done
in the next week,"checks"(do not like the computer answers).
I just noticed it is about 4am must go.

I did manage a couple of other posts,but cannot do long
answers for a week or two.
later