Francis,
If we took a transformer say designed to run at 10 kilogauss, but the
iron would run okay at 12 kilogauss, then dropped the frequency down
to 50 Hz, it would probably be okay since the flux density would rise
up to an acceptable amount at 12 kilogauss. However if we had one
designed around say 15 kilogauss and dropped the frequency, the
transformer could go into saturation and the current really shoot up
if it's max flux density was lower like say M-22 or M-27, etc. When
looking at the B-H curve for a type of iron, one supposed to pick a
spot just below the knee, to just maybe up on it a small amount. On
up the knee though is going into the saturation region, where the
higher the flux, magnetizing current raises rapidly, and
expotentially. If one could find one where the flux density was ran a
bit below the knee, and then try the lower frequency, it should just
raise up on the knee to where it's still acceptable. They do wind
some this way for heavy duty service, adding more iron than needed.
In reality though, one would still be using what would have been a 50
Hz transformer at 60 Hz though they spec it as a 60 Hz transformer.
The power capability would drop though I guess by 1.2 times. But,
since it was over-sized already like above, it would still be okay at
the projected power level, just derated to a different service
factor. The thing is, most off the shelf transformers are designed
right close to the edge for normal duty cycles. One would have to buy
one with a higher power rating to achieve this result.
They'll all draw some magnetizing current, but after coming into the
saturation region, it raises rapidly as compared to smaller increases
in flux density. The saturation region can be seen on a scope as the
waveform will become distorted. If the magnetizing current does
raise, it doesn't necessarily mean the core is saturated. It really
shows that the flux density has raised because of the lower
frequency. If it raised and the waveform is still normal, it should
be okay to run as long as the wire size will handle the increased
current. Of course we would be derating it anyhow by about the same
amount.
If so, one ought to be able to buy a transformer with about 1.2 times
the power rating needed to achieve the same result. This since the
only things that effect flux density is voltage, core area,
frequency, and the number of turns. That is as long as the iron will
handle the increase in flux with the voltage, number of turns, and
core area the same. The frequency being the only changing factor. One
fly in the ointment would be that most 50 Hz applications uses a
lower voltage.
If designing a 50 Hz to set in place of a 60 Hz exactly, and the iron
they used was like M-19, one could design the new one with M-6. It
will raise the flux density limit a good bit. However, when you bring
a different voltage into the equation along with lower frequency, I
don't know if it would work or not.
Best,
Will
--- In ham_amplifiers@..., FRANCIS CARCIA <carcia@...>
wrote:
Will,
I lived it once and was unable to make the loss equal for both 50
and 60 Hz. I got close but the 50 Hz high line started to suck
current. frank wa1gfz
craxd <craxd1@...> wrote:
Francis,
Yup, your right. I was going by the difference in the two formulas
between 50 and 60 Hz. Both sets have a difference of 1.2. See below;
For 60 Hz
TPV = 4.85 / A
A = 0.1725 x sq rt of P
P = ( a / 0.1725 )^2
--------------------------------
For 50 Hz
TPV = 5.82 / A (5.82 / 4.85 = 1.2)
A = 0.207 x sq rt of P (0.207 / 0.1725 = 1.2)
P = ( a / 0.207 )^2
If you notice, the difference between each formula from 60 Hz to 50
Hz is 1.2. That's why after I looked, I assumed both raised. The
TPV
had to be changed to make it come out correctly. One can either
adjust the number of turns or the core size to achieve the goal,
but
not both together as it would be off (the formulas above have to
though). I guess one could by splitting the difference, but the
formulas above don't take that into account. When I did both and
checked them with the long formulas, they came out even on the
turns.
The resistance then would only raise over the extra wire length.
Since the core area is 1.2 times greater, I'd say the wire length
will be 1.2 times longer.
So void what I said earlier, it's not correct. I'm glad you caught
this as I'm sure not at myself today.
Best,
Will
--- In ham_amplifiers@..., FRANCIS CARCIA <carcia@>
wrote:
Will, untrue all you need to do is make the core bigger to reduce
the flux at 50 Hz. Volts per turn stay the same. Wire will be a bit
longer due to bigger window. So bigger core makes more resistance
when the 50 hz transformer is at 60 hz but core loss is very
slightly
less.
Whenever you make the core bigger you can get away with lessturns
of wire for the same flux density. Double the core area= 1/2 the
turns for constant flux
craxd <craxd1@> wrote:
Actually everything goes up by a factor of 1.2. Not only
does the
core have 1.2 times the iron, the turns per volt is 1.2 times
higher.
Since the turns are higher, the resistance will go up 1.2 times
unless a larger diameter wire is used to drop it back down.
Whoever
designed the transformer should know this I would think, and
account
for it so as to acchieve the same power output.
Generally what is done is to use a larger laminate size and stack
them to get the right thickness. The larger lam size has larger
windows which allows a larger diameter wire to be used. This has
to
be done anyhow to hold the extra wire over the higher turns so it
has
to be stepped up even larger to increase the wire diameter. If
the
turns are dropped to just accept the larger wire, the flux
density
and magnetizing current will raise. One could play with the lam
material here, and use one that operates at a higher flux
density,
but the cost will go up. Another alternative would be add more
iron
which would lower the flux density, but again, cost goes up.
If one was in production, it would be best just about it to use a
50
Hz transformer for both 50 Hz and 60 Hz. The plus would be for 60
Hz
users as it would run cooler with a lower flux density. This
would
make a one size fits all transformer so the chassis wouldn't have
to
be modified between the two.
Best,
Will
--- In ham_amplifiers@..., FRANCIS CARCIA <carcia@>
wrote:
Lower frequency you need more primary turns so a higher tap to
reduce heating but this will reduce the output voltage due to
turns
ratio change, gfz
pentalab <jim.thomson@> wrote: --- In
ham_amplifiers@..., "Hsu" <Jbenson@> wrote:
Thanks,Will
Could tell me the power rating of TL922 HV transformer?
Thanks again!
73! Hsu
----- Original Message -----
#### HSU.... as far as I know.... from memeory... the plate
xfmr in a TL-922 is rated at aprx 1.2 KVA CCS. The RL Drake
plate xfmr, used in the L4/L4-B/ L-7 is also rated at 1.2 KVA
CCS.
#### In the case of the RL drake xfmr.... that rating is for 60
hz only..... and like Will sez.... it MUST be de-rated for 50
hz. Several VK/ZL's I have spoken too over the years have
complained of over heating the RL drake xfmr, when run on 50
hz.
### Dunno about the TL-922 xfmr. On my old yaesu FL-2100 B
[1977] , it had primary taps for 100/120 200/240.... in Japan
they use 100/200 V and 50 hz. So In that case, I'd assume
the yaesu plate xfmr would run on 50 hz.
### IF the TL-922 xfmr has pri taps for 100/200v.... you can
asume it will run ok on 50 hz. IF it only has taps for
120/240... then asume it's 60 hz only. It seems to me that
the TL-922 is popular in the UK... and I believe they use 220 v
50
hz.
### In normal operation... the TL-922 xfmr would be good for
600 w output RTTY/ FM CCS..... and 1200 w out pep on ssb....
and
maybe a little less on CW.
### I agree with Will. You can't go by weight alone. Case in
point, the Hammond 795 series plate xfmr's weigh in at 80 lbs
[36 kg] , are rated for 2.2 kva CCS.... and run VERY hot when
used in a C input filter.... and when running 1900w PEP output
from the linear.
### In the case of the Hammond, it was a high reactance type,
with a center tap, designed for tube rectifier's... and had a
high
68 ohm sec DC resistance. The drake L4B plate xfmr has a 10
ohm DC sec... is made for a C input filter.. and is of the LOW
reatance type.... the TL-922 and the SB-220 both have 10 ohm dc
resistance secondaries.
### These Peter Dahl hypersil C core plate xfmr's flat out
won't
blow up, or overheat, doesn't matter what you do to em. They
are
either the most underated things, or the greatest things since
sliced bread... take ur pick.
Later... Jim VE7RF
