Choke power supplies for Class AB Power Amplifiers - which Rules for Calculating ??

[tiefbassuebertr]

Choke power supplies (not regulated) for Class AB Power Amplifiers - which Rules for Calculating the Values ??

In general there is to read anything arround this topic about the URL
Resonant-choke Power Supply

By loads with very constant current requirements this is easy for me (e. g. preamps and pure class-A amps like ZEN).
But it is an issue for me, if I want to realize such a thing for a Class AB amp, where the idle current runs arround 20 mA for the output buffer (50-100 mA at whole) and the peak current is according the speaker load between 5 and 50 A.

Have a look to figure 5 about
Resonant-choke Power Supply
(Distributed Filtering Approach)
How I must calculate the values with the respect to large jumps in the the dynamic current and power consumption (naturally by Class B and AB power and integrated amps) ??

There are also approaches, where is introduced an additional serial rectifier diode between the already present bridge rectifier and the first capacitor C. Means this additional advantages resp. benefits ?
Additional there are low cost versions with only a resistor and/or rectifier diode in place of the inductor choke. Also for such a topology I want to know the rules for calculating.

Perhaps there are already a thread here on diyaudio, where this is descripted in detail (unfortunatey I have found various threads only in combination with tube amps).

Thank you for your comments and advices to exist papers and reports.

[AndrewT]

Originally Posted by tiefbassuebertr

........loads with very constant current requirements this is easy for me (e. g. preamps and pure class-A amps like ZEN).

most ClassA stages do not draw constant current from their supply rails.

They draw a quiestcent current that varies only slightly with temperature.
But as soon as you ask for output current the rail currents vary and vary very markedly so.

Just measure the AC voltage across a low value resistor placed in the +ve and/or -ve supply rail/s. An oscilloscope will do even better, since it will show the phase of the rail currents exactly matching the phase of the output currents.
If you have a 1Apk output current then the variation in the +ve and -ve rail currents will be 2App, i.e. total rail current = Iq +-1Apk

[ingenieus]

From the link

It should however be noted that the critical value Lc is not a fixed parameter, it varying with Iout and being approximately given (in Henry) by:

• Lc=Vout/Iout for 60 Hz operation (as given in the ARRL Handbook)
• Lc=1.2*Vout/Iout for 50 Hz operation

where Vout is expressed in V and Iout in mA.

And there lies the rub - Iout is not constant in an audio power amplifier.

Interesting idea though. A resonant choke might not be of much use, but an input choke would add extra regulation. The cost would work against the idea, but apart from that - what can go wrong?

[AndrewT]

what can go wrong?

The voltage rises by ~50% when the output current falls to a low value. Blown capacitors?
There will be ways to insure against this, but first one must be aware of the risk.

[davidsrsb]

When the current in the choke falls to zero, there is a very high spike on the rectifier diodes. If you build a choke filter psu, you must maintain a minimum current. means that mosfet outputs with high quiescent current are a more practical match to chokes

[ingenieus]

Thanks for the heads-up guys. Now we know why there are no class B/AB amps with chokes in the power supply. None that I have seen anyway. The requirement for a minimum current to be maintained can be met with a class A circuit, but that's probably not the answer that tiefbassuebertr was looking for.

[AndrewT]

Originally Posted by ingenieus

Now we know why there are no class B/AB amps with chokes in the power supply. None that I have seen anyway. The requirement for a minimum current to be maintained can be met with a class A circuit,

No & no.
There are choke regulated PSU in ClassAB amplifiers. they are expensive and they are heavy. The MF that I am aware of are a two man lift.
ClassA does not solve the problem. They too have rail currents that can fall to near zero, just like ClassAB.

[tiefbassuebertr]

Originally Posted by davidsrsb

When the current in the choke falls to zero, there is a very high spike on the rectifier diodes. If you build a choke filter psu, you must maintain a minimum current. means that mosfet outputs with high quiescent current are a more practical match to chokes

Yes, and for this reason I haven't work without capacitor in front of the inductor coil.
The main problem is in the moment, that I don't find the appropriate paper, which was released from the AES. Additional I don't know the right english terms to find this paper about google arround this low pass " π " - filter topology.
"Choke power supply" cannot be an often used term therefore (sorry - English is just not my native language).

Originally Posted by ingenieus

Thanks for the heads-up guys. Now we know why there are no class B/AB amps with chokes in the power supply. None that I have seen anyway. The requirement for a minimum current to be maintained can be met with a class A circuit, but that's probably not the answer that tiefbassuebertr was looking for.

There is still a minimum current, because both the input stage (single ended or LTP) and the VAS of the power amp unit runs usually in pure class-A. If there is an integrated version of amplifier, additional current runs at idle, because the preamplifier section runs also in pure class-A.

I have some outdoor power supplies scheduled (mostly with four or six typical 12VAC white lighting transformers, which have low noise character), in which the first capacitor value has approximately 30-50% from the value of the second ones behind the inductor (sometimes only resistor). For the resistor value I calculate values between 0,33 ohms and 0,68 ohms, both at inductor dc resistance and resistors only. And if I choice an inductor (only air coil), the values are between 2,2mH and 10mH (independend of the wanted costs). But nevertheless I want to know the exact way for calculating.

BTW - the audible differences by listening tests was always extremely large, both by cheap integrated amps below 20W output power and expensive typical high end audio amplifier components, independend of the brand.

[a.wayne]

Interesting discussion .....

[AndrewT]

There is a full explanation/calculation paper linked on this Forum.

I cannot recall the author nor the Thread.