THD Cancellation in SEPP Stages

Another renowned exponent of the Push-Pull Society is the circuit known as SEPP (Single Ended Push-Pull). In  Fig.27 you can see the simplified electric circuit.

        Likewise a true push-pull it requires two out-of-phase driving signals nevertheless cause of its structural asimmetry  (the lower tube behaves like a

Fig. 1 The SEEP

common cathode and the upper as a common plate ) the input signal must have different amplitudes in order to produce currents with the same amplitude. From a point of view of the harmonic spectrum a SEPP behaves similar to a SRPP that to a true push-pull: like to a SRPP the theoretic minimum for the THD is obtained when current ip1 and ip2 are out-of-phase with the same amplitude; practically  the different “modus operandi” of VT1 and VT2 could require different amplitudes for ip1 and ip2 in order to minimize the THD therefore, differently to the true push-pull, the second order cancellation is not obtained perfectly, besides the currents equilibrium is dynamic.

Fig.2 SEEP Test Bench

     As an example take a look at Fig. 2. This circuit is drived by two  out-of-phase signals whose relationship is:

Vin2= 1.30*Vin1;                  (31)

this condition allows a good balancing until the amplitude of Vin1 is not greater than 60Vp. I have stressed this circuit by varying Vin1 between 5Vp and 60vp with steps of 5Vp, then Vin2 for the (31) will vary between 6.5Vp and 81Vp. This variations permit even the Class AB operation. Fig. 3 shows the trend of the 2nd, 3rd, 4th, 5th harmonics and THD as function of Vin1.The second harmonic results always increasing with a relative minimun approximatively at Vin1=40Vp. Here a good simmetry exists between ip1 and ip2. Beyond this value, an abrupt increase in the 2nd harmonic happens, mainly for a loss of simmetry  . The 2nd and 3rd harmonics

        


 

          

Fig. 4 Harmonic Analysis  Results
 

are similar in value but the latter is  greater  while 4th and 5th have an eccentric behaviour . THD increases evenly with  the input signal.

Fig. 5 The Class A Operation Mode

        Very interesting appears to observe the harmonic spectra in two different condition. When Vin1=10Vp, Fig. 5, the circuit is in Class A operation and you can see a spectrum very close to that of a Single Ended stage; although from the graph barely appears, the symmetry in the currents is not perfect. Fig. 6 shows the result when Vin1=40Vp (Class AB operation). An excellent symmetry exists and therefore the spectrum emphasize odd harmonics and when Vin1 further increases you have a loss of symmetry that return to enrich the spectrum, Fig. 7.

Fig. 6 The Class AB (soft-mode) Operation Mode
 

Fig. 7 The Class AB (strong-mode) Operation Mode

Fig. 7 The FDH for Vin1=20Vp
 

 

 
 

Fig. 8 The FDH for Vin1=60Vp

The FDHs of a MonteCarlo Analysis with RL=10% Dev/Uniform when Vin1 is 20Vp and 60Vp rispectively don’t show meaningful results, Fig. 7 and 8.

 

What did you think of this article?
Click here to send us your comments, feedback and suggestions