Hi Marco & Chris

apologies for the delayed reply.
There was indeed a typo.
It should heave read 12.5Ohm because the amp is designed to work into
a 50:4=12.5 Ohm load.
Then there was another wrong entry:2000W instead of 1000 W as each 1kW amp has 
its
own 1:4 transformer.
And there is another problem which I have detected meanwhile:
The program calculates only for 50Ohm primary Z not e.g. 12.5Ohm.
One could circumvent this shortcoming by entering 250W instead of 1kW for 
calculating the
correct flux density because 250W generate the same voltage over the primary 
winding at
50 Ohm as 1000W at 12.5Ohm.But then the other parameters are calculcated 
incorrectly.
Maybe I'll contact the author if he can implement some changes/improvements  in 
his tool.

73
Clemens
DL4RAJ 





 

>-----Original Message-----
>From: [email protected] 
>[mailto:[email protected]] On Behalf Of 
>[email protected]
>Sent: Tuesday, May 16, 2017 4:42 PM
>To: [email protected]
>Subject: R: Re: LF: RE: Combining LF amps revisited
>
>Chris,
>
>I guess there is a little misstyping in the calculation performed from 
>Clemens on the DJ5IL tool....
>he entered 2.5 instead of 25 in the impedance box: the primary 
>winding after the combiner should work on 25 ohms. If you enter 25 ohm 
>the flux rise a lot beyond the saturation of the 3 stacked cores and 
>therefore they have all the rights to warmup!
>On the other hand if you apply the Bmax formula given from Andy, the 3 
>stacked cores have A of 471mm²
>and with 2kW of power on 25 ohms the minimum number of turns on the 
>primary should be 8 not 5.
>So one more turn to keep B below 0,1T and you should run more 
>confortable with N1/N2=9/13
>73 de Marco, IK1HSS
>
>----Messaggio originale----
>Da: [email protected]
>Data: 15-mag-2017 14.42
>A: <[email protected]>
>Ogg: Re: LF: RE: Combining LF amps revisited
>
>For transformers the core is not terribly critical provided you use a 
>grade
>of ferrite designed for the frequency of interest.  And at LF that 
>means
>SMPSU cores
>
>If I see a certain size core being used in a 500W SMPSU then there's a
>pretty good chance a core of the same size is going to work OK 
>in an LF 
>Tx
>of the similar rating.   No data sheets or trying to work out 
>temperature
>rises.
>
>I just use whatever feels right and apply the V < 4.44F.N.A.B  
>absolutely
>rigorously, keeping B to less than 0.1 usually
>
>As for impedances:  well they're  just Z1/Z2 = SQRT(N1/N2)
>Use V = 4.44.F.N.A.B to give you turns min, then calculate N1 and N2 
>while
>meeting that criteria
>
>Andy  G4JNT
>
>On 15 May 2017 at 11:59, Chris Wilson <[email protected]> wrote:
>
>> Hello Clemens,
>>
>>
>> Thanks  for  the  reply, the 3 stacked cores will rise to 70 degrees 
>C
>> plus  even with a small computer processor fan blowing through them. 
>I
>> am  investigating  ways  of  reducing  this, the cores are shown as 
>77
>> material  in  jay's  schematic,  but  he  and I are using the newer 
>78
>> material  for the PA output transformer cores. I may try using this 
>76
>> material  in  the  combiner transformer. Someone else suggested 
>making
>> the   PA   output   transformers   work   at  100  ohms  and  using  
>a
>> transformer less  zero  degree  hybrid  coupler.  I'd  try  Andy's 
>twin
>> primary,  one  secondary  transformer instead of the coupler if i 
>knew
>> what winding ratio and what core to start off with. Thanks again.
>>
>> > Hi Chris,
>>
>> > I have a tool by DJ5IL here which predicts around 60°C for the 
>cores
>> > (excluding heat caused by the winding)
>> > and ca. 480 Gauss flux density.
>>
>> > 73
>> > Clemens
>> > DL4RAJ
>>
>>
>>
>> --
>> Best regards,
>>  Chris                            mailto:[email protected]
>>
>>
>>
>
>
>
>