Clarity 2x300W: Class-T amplifier （3）

Published:2011/8/3 2:20:00 Author:Li xiao na From:SeekIC

By Ton Giesberts

Output filter

Thanks to the high switching frequency, here it is only necessary to use a second-order filter with a relatively high corner frequency (resonant frequency 101 kHz). To dampen the Q factor of the filter, which is primarily important if no load is connected, a Zobel network is placed at the output, since otherwise resonance currents and ringing signals at the output could reduce the reliability of the amplifier. As the comer frequency of the filter is higher than for conventional Class D amplifiers, the connected speaker system has a considerably smaller effect. In light of the large currents involved here, an ordinary noise-suppression choke cannot be used for the output filter. A special core material must be used to maintain low distortion and high efficiency. We have more to say on this subject in Part 2 of this article, which will appear in the September 2004 issue.

Amplifier configuration

The gain of the modulator is set using feedback resistor R15 (R36) and voltage divider R18/R20 (R39/R41). These components must be dimensioned according to the value of the supply voltage that is used. This is necessary to make the amplifier independent of the behavior of the power supply (such as voltage fluctuations due to the output amplitude, mains voltage variations, etc.). Additional reverse feedback to counter ’ground bounce’ is provided by R16 (R37) and voltage divider R17/R19 ((R38/R40). These two networks must be identical! The resistor values can be calculated quite easily. A value of 1 kΩ is generally used for R17 and R18, so the value of the other resistors only depend on the value of the supply voltage VPP (assuming a purely symmetric power supply) and the value selected for R17:

R19 = R17 X VPP÷(VPP - 4)

This yields an E96 resistance value of 1.07 kΩ. This value is reasonably independent of the supply voltage. If a maximum supply voltage of 51 V is used, it only increases to 1.10 kΩ. Finally, the value of R15 determines the gain of the modulator:

R15 = R17 X (VPP ÷ 4)

We have assumed a maximum supply voltage of 62 V (the special decoupling electrolytics on the amplifier board are 63-V types). This yields a value of 15.4 kQ for R15. The gain of the modulator can then be calculated in the same manner as for a standard non-inverting amplifier:

A_modulator = (R15 * Rp) + 1

where R_p is the parallel resistance of R18andR20.

Capacitors C11 and C24 filter and delay the feedback signal to the modulator. They have different values. These capacitors prevent RF noise with very high frequencies from penetrating the feedback network, and using different values causes the modulators to have different switching frequencies. This prevents mutual interference between the modulators. The values are chosen to cause the difference to be greater than 40 kHz.

Reprinted Url Of This Article: http://www.seekic.com/blog/project_solutions/2011/08/03/Clarity_2x300W__Class_T_amplifier_（3）.html

Print this Page | Comments | Reading(303)