Direct drive amp schematic

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Thanks for the schematics and article, Lukas.

Also, what does it mean "Passive load" and "Active load" ?

The "load" referred to here is a method of setting the output voltage to 1/2 the supply voltage with no signal. The "Active Load" version does it with U5, Q3 and their associated components. The "Passive Load" version does it with R15, a power resistor.

I believe that this schematic is for headphone use only...

It looks like it is for full size ESL panels. From the article:
The purpose of this article is to describe a universal amplifier capable of driving a number of electrostatic speaker ( E.S.L ) panels from various designs, without the use of coupling transformers.

I'm still trying to figure out how you would get it all to work. From the article:
Each module requires a 1000V D.C supply at 40 mA. So the total current for each channel is 80 mA since module drives one side of an ESL plate i.e 2 modules per mono channel.

Maybe this:
 

Attachments

  • mckean_amplifier.jpg
    mckean_amplifier.jpg
    51.6 KB · Views: 957
Hi,

maybe the schematic is understood easier, if You think in ´tube-terms´. MOS-Fets and Triodes have many similarities. Replace Q2 by a Triode. R15 is then the anode load resistor (passive load). It´s purpose is to work as a constant current source (CCS). It´s dimensioned so that app. half of the supply voltage is dropped over it when the idle current flows (500V=10kOhmx50mA). It has to be a high wattage type, because even at idle 25W of power develops across it.
While using a simple resistor works, the result is quite a bit away from the optimum. A simple resistor is not very good as a CCS. The ohmic value is very low. Rising the ohmic value of the resistor leads to a much higher V supply. The Triode/MOS-Fet´s linearity is not the best and output resistance will be quite high. When replacing the non optimal CCS with an active one the performance will be much better. Here a second MOS-Fet together with the OP-Amp and associated components works as a voltage controlled CCS. Letting the output voltage of the amp control the CCS is a feature that gives the application a push-pull character (SEPP). The dynamic resistance of the CCS is very high, so that the MOS-Fet Q2 (or a Triode) works very linear with less distortion at higher levels. The output resistance drops to small values, giving good control over the speaker by a high damping factor. Output current to the load can be twice the idle current.
Especially at higher frequencies You need some current to load the ESL-capacity. The more bandwidth and the greater the capacity the higher the needed current. 50mA is a quit low value and in the description a bandwidth of just 5kHz is claimed. Its not a huge problem because the content of highest frequencies in music and their voltage need is small. But You should keep this in mind when You plan to drive larger panels.
Having a iddle voltage centre of 500V will allow probably for +-400V peak-peak swing, which is just 280Vrms. This can be quite loud with an highly efficient panel like MLs, but most of those panels have capacities well above 1000pF. So I would use this stage for ESL-headphones but would thoroughly check if its capable to drive a bigger speaker panel.

jauu
Calvin
 
Hi,

For easier understanding You may read the articles by Alan Kimmel (µ-follower) that describe this concept for low-voltage applications.

To drive larger panels with high capacitance You would have to redesign the output stage for higher supply voltages and different transistors. But it will be hard to find the right devices. There are companies that manufacture high voltage transistors like Powerex (CM100-CM200-CM400 IGBT-series; maybe someone would like to comment on these components?) or Sensitron (SCP-4926/4979 IGBT-series)

The schematic itself is imo a very clever. It should give a cleaner, less distorted output and works in true push-pull mode, what SRPP designs like the Strickland/Beveridge etc. don´t. Even running the SEPP-concept without global feedback should work. Because of the similarity between MOS-Fets and tubes You could redesign the output stage with Powertubes (e.g RCA 4-125..4-1000 or similar)


jauu
Calvin
 
Hi,

"..mostly current driven" I say no! They are not dynamic drivers, where a current provides for the magnetic field, but ESLs where the voltage provides for a electrical field. Imagine how much the current varies at constant voltage over the audio frequency band. 1:1000. To achieve a constant current over the whole audio band You´d need excessively high voltages at low freqs.

yap
Calvin
 
Hi,

@Bazukaz

It won´t work with this schematic, because the output Trannies work as Buffers with an voltage amplification of ~1. Therefore the driver circuit (the IC) has to supply all the needed voltage which it doesn´t.
The thing could work with added circuitry, that steps up the low output voltage of the IC to the desired high level. But adding steps means adding probs with phaseshift and stability and..and....and........

jauu
Calvin
 
I think that direct drive amps are generally better then output trannies.
Transformer gives much more distortion than amplifier itself.

ESL's are very efficient.Almost all power is dissipated in amplifier.
So , a new type of amplifier that could use this feature of ESL is required .

I don't even imagine how such an amplifier should look like ...
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.