Modeling Pull-Push transformer in LTspice

Status
Not open for further replies.
Jump to Latest
P

pha0001

Member
Joined 2011
#1
I am having trouble modelling an ideal Pull-Push center tap 8 ohms transformer with a 6.7k primary anode-anode impedance in LTspice.

Formula:

8*(Np/Ns)^2=7.6k
Thus, Np/Ns=30.82/1

Here are my LTspice schematic. Could someone please help me out?

pull-push LTspice.png

Thanks
 
#5
including series R and parallel C help get more realistic as well as the right K less than 1.0 to include leakage L - can then see more of the (high) frequency response

if you include the parasitics in the L def in LTspice you should check the box for them to be displayed on the schematic so people can see what your using
 
Last edited:
#11
if you include the parasitics in the L def in LTspice you should check the box for them to be displayed on the schematic so people can see what your using
Click to expand...

oops, doesn't look like there's a box for R, L, C - the V, Isource parts do have a "make information visible" check box

maybe a request to Mike?

certainly the schematic isn't adequate documentation with the information only visible when running LTspice and clicking on the component

so to share as a drawing documenting your circuit you would need to put the info in text comment lines

adding explicit Rseries, Cparallel works but LTspice claims the solver works better with the internal device spec'd parasitics
 
Last edited:
#14
For the Marshall JTM45 (based on the Bassman), Hammond offers this replacement OPT. At 240V and 50Hz, they spec the input inductance at 136H (open circuit on secondary). I would presume that is across the full primary.

I'm curious why they rate the inductance at such a high voltage and low frequency? I'm guessing the rating at such a low frequency is important for low frequency drive considerations. Does the inductance increase with voltage? If so, why?
 

Attachments

#16
> how to calculate the screen grid tap inductance (40%)?

40% is 0.4

Inductance goes with Square of turns.

So 0.4*0.4 is 0.16 of inductance.

The G2-P part of the winding is obviously 0.84 of inductance.
 
#17
PRR said:
> how to calculate the screen grid tap inductance (40%)?

40% is 0.4

Inductance goes with Square of turns.

So 0.4*0.4 is 0.16 of inductance.

The G2-P part of the winding is obviously 0.84 of inductance.
Click to expand...

Also

60% is 0.6, so 0.6*0.6=0.36

Then, deff (*)

L1 = 0.16 L, L2 = 0.36 L

Then

L = L1 + L2 + 2 M

M = k √(L1 L2)

For k=1 (this was already assumed)

M = √(L1 L2)

Et voilà

L = 0.16 L + 0.36 L + 2 L √(0.16 0.36 )

Hint: Assume (*) as an approximation and make

k = √(1 - Lleak / L)
 
Last edited:
#19
PRR said:
> how to calculate the screen grid tap inductance (40%)?

40% is 0.4

Inductance goes with Square of turns.

So 0.4*0.4 is 0.16 of inductance.

The G2-P part of the winding is obviously 0.84 of inductance.
Click to expand...

Thanks so much for that.... but still there is a confusion on it. ...I think the picture can be more clear. My question is:
L10 = 1.025, L11 = 0.475 and L15 = 0.008 is how to calculate it?
 

Attachments

  • Capture.JPG
    Capture.JPG
    65.6 KB · Views: 752
#20
Although I use TINA rather than LT Spice, I have always found that first you put in an ideal transformer, with the correct ratio.
Then you add an inductor of the real primary inductance across the primary of the ideal transformer. If you have measured the leakage inductance (measured at the primary with the secondary short-circuited) then you add two series inductors (assuming a P-P output in each half of the primary windings. You can of course improve the model by adding the values for the winding resistances and stray capacitances.
Doing things this way, starting with a perfect transformer model, you can see if the basic design is OK, and how much it degrades as the model is made more like the real thing.
 
Status
Not open for further replies.
Top Bottom