# How to model a 2:1 center-tap transformer in spice for a Push-pull amplifier?

Discussion in 'General Electronics Chat' started by SiegeX1, Apr 27, 2011.

1. ### SiegeX1 Thread Starter Member

Mar 10, 2009
25
0
I'm designing a push-pull amplifier where the center-tap on the primary is connected to +15V and each end of the primary will go through a low-side MOSFET to ground. On the secondary, I will have a 75ohm load. The goal is to get a +/-15V pulse across the load with a source voltage of 15V. Note I am not counting for losses here, just theoretical.

I have found a datasheet (T-51117) for a "2CT:1" transformer which claims to have an "OCL" value of 5mH. I'll assume this OCL value of 5mH is the inductance across the full primary.

Since the relationship between turns and inductance is given by the equation: Ns/Np = sqrt(Ls/Lp) and we know that Ns/Np = 1/2 and Lp = 5mH. This must mean that Ls = 1.25mH.

Now here is where things fall apart for me. If we assume that the center tap cuts the number of primary windings in half, then each half-primary to secondary should make a 1:1 transformer since they both have equal turns (the full primary has twice as many turns as the secondary). Since we just calculated the secondary to be 1.25mH, each half primary should also be 1.25mH to maintain the 1:1 ratio. However, since inductance in series adds linearly, that gives us a full primary inductance of 1.25mH + 1.25mH = 2.5mH, but we already know from the datasheet that the primary inductance is 5mH, not 2.5mH!

Obviously something I am doing is not right. What am I not considering and/or what assumptions do I have wrong?

Also, what should the values be for L1, L2 and L3 in the diagram below to model this 2CT:1 transformer with an OCL of 5mH?

Thanks!

• ###### pushpull.PNG
File size:
31 KB
Views:
333
Last edited: Apr 28, 2011

Dec 26, 2010
2,147
300
Each half-primary does indeed have one quarter of the full primary inductance. This can be deduced from the fact that inductance of a winding varies as the square of the number of turns N.

Consider the whole primary as a winding of N turns, and a half-primary as a winding of N/2 turns. Then, if the specific inductance of the transformer core is AL Henry, the whole primary has inductance N$^{2}$ AL, and the half primary has inductance (N/2)$^{2}$ AL or (N)$^{2}$ AL/4.

If the two half-windings were not coupled, this would make no sense, but they are coupled, so the inductances don't add arithmetically.

3. ### SiegeX1 Thread Starter Member

Mar 10, 2009
25
0
Thanks so much, I did not take into account that there is some "overlap" due to the coupling of each half-primary which makes linearity go out the door.

So just to double check, to model this 2CT:1 with OCL of 5mH, I'm going to make all of L1, L2 and L3 1.25mH, is that correct?

Much appreciated!