Simple +/-15V shunt regulator, with the two zener diodes. Most PA amps which use op amp input stages do that. Then follow the op amp with a gain stage or two.
In that schematic shunt regulators with the 4744 zener diodes provide 15 volts to the opamp
*( looks like wg_ski beat me too it, didnt see the initial post)
*( looks like wg_ski beat me too it, didnt see the initial post)
I had already noticed the Zener diode and the two capacitors in parallel with the 4700Ω resistor in series with the power supply, but is that alone really sufficient to drop the voltage entering the opamp? I thought the process seemed too simple... So simple, yet it works. Thanks for the explanation!
By the way:
I'm looking for opamps for a project, and the NE5534 was one of the few I could find in a DIP package. I'd like to know how good it is compared to more modern components. Are there significant performance differences for audio applications, especially in terms of noise and distortion? Additionally, in this kind of package, is there anything substantially better that is relatively easy to find?
I'm looking for opamps for a project, and the NE5534 was one of the few I could find in a DIP package. I'd like to know how good it is compared to more modern components. Are there significant performance differences for audio applications, especially in terms of noise and distortion? Additionally, in this kind of package, is there anything substantially better that is relatively easy to find?
OPA134. I am changing the TL071 in my Quad 405 to these. A NE5534 was an alternative.
Although the NE5534 is a venerable device, it really isn't a bad one. With line level signals, which are in question here, it is hardly to beat, unless there are special requirements, such as zero to no input current etc.
Best regards!
If it is of interest, this is type of circuit that was very popular in professional amps from 90's, reaching several hundred watts stereo and over 1kW in bridged-mono. I have one in garage collecting dust, Solton Craaft execution with 4 Toshiba transistors in parallel (like on picture, but 2SC5200/2SA1943). Theoretically, if I ever make space on my bench, I could run RTA on it to see how it behaves....
That’s your basic transnova circuit. Hafler started it with mosfets, QSC followed using bipolars, then everyone and their dog copied the QSC circuit. By moving the power supply around (effectively swapping the roles of the power supply ground and the speaker output) the output follower becomes a common emitter stage. It biases like driving a CFP follower with an op amp, but because of where the output it taken, works like a common emitter stage and provides darn near all the voltage gain.
The catch is now the speaker output is taken from the node in between the two supply caps. Two or more channels can not share the same power supply. It can share common transformer core with separate windings but to improve channel separation to the levels you get with other topologies totally separate transformers are required.
The catch is now the speaker output is taken from the node in between the two supply caps. Two or more channels can not share the same power supply. It can share common transformer core with separate windings but to improve channel separation to the levels you get with other topologies totally separate transformers are required.
Totally agree, and all executions I have seen have single toroid with separate windings for each channel. It is very popular design of the era, Please don't take me that I advocate this circuit, only what I can say is that I not critically listened to those amps, no much complains, and it delivers humongous power, if that's needed...
Years ago we had a thread about these common collector power amplifiers. It started with a QSC schematics that also featured an opamp input, but unfortunately ran out of juice somewhen and vanished silently. Does anybody remember?
Best regards!
Best regards!
They sound better than they have a right to. Professional amplifiers use them for two reasons: cost and fewer points of failure. For someone like QSC, getting a toroid made in China that has 8 isolated windings is only marginally more than one with only two - IF you order enough of them. Go to a Toroidy and it will exceed the entire BOM cost, including the chassis and China transformer. At one time it would have exceeded the entire retail price of these amplifiers - but an RMX2450 in those days was $500 out the door. Prices have more than doubled, accelerating the transition to class D with SMPS.
In between all this, QSC also started using a modified version of the circuit using a more “normal” common emitter output stage. In their PLX and smaller Powerlight models. It’s sort of reminiscent of the SWTPC “Tiger” output stage, but has all the necessary safety features in place so nothing gets accidentally reverse biased or overdriven by too much during clipping. With a normal speaker output, it can run both channels off a common supply - Which is essential if using an SMPS. It still retains the advantages of low (amplifier circuit) BOM cost and few points of failure. Of course, other manufacturers copied it - including Crest, who in the early days wouldn’t copy anybody.
In between all this, QSC also started using a modified version of the circuit using a more “normal” common emitter output stage. In their PLX and smaller Powerlight models. It’s sort of reminiscent of the SWTPC “Tiger” output stage, but has all the necessary safety features in place so nothing gets accidentally reverse biased or overdriven by too much during clipping. With a normal speaker output, it can run both channels off a common supply - Which is essential if using an SMPS. It still retains the advantages of low (amplifier circuit) BOM cost and few points of failure. Of course, other manufacturers copied it - including Crest, who in the early days wouldn’t copy anybody.
Its the most fault tolerant way to get the +/-15V. PA amplifiers (and ones often used in that role) get beat to hell, and often operated with dead channels. You want a circuit which minimizes collateral damage if it gets severely upset by some event. This is it.
Thanks man
Another quick question:
Could I swap the MJE340 and 350 with 2N5551/ 5401, the 2SC5171 with MJE340/50 and leave the Toshibas's as it is?
Oh yes, they sound decent and these are workhorses, robust at least! For home use they do have few caveats do:
- nobody needs several hundred Watt
- Heat sinks are made for active fan cooling and that's noise. It can be fixed by fiddling with super low noise fans and regulators for low noise PC's.
- Huge trafo's used there are prone to buzz especially since mean grid voltage went up in last 2-3 decades, yet another unnecessary noise. It also can be fixed by reducing mains voltage feed by using yet another bucking transformer, 10 times smaller in size than main one : https://www.sound-au.com/articles/buck-xfmr.htm
There is usually a knee-jerk reaction that “nobody needs that kind of power” when they see a +/-120V circuit being discussed. In the sound business, everyone needs several hundred watts. Sometimes several thousand, sometimes tens or hundreds of thousands. That sort of thing isn’t DIY’ed very often, but it can be. Yes, it is a hobby rather than a way to make a living, but that hobby happens to be building PA equipment. Many times what you get though, is some newbie who has never picked up a soldering iron in his life, or studied electronics, wanting to build one of these monstrosities (perhaps to use in his club or bar) out of some simplistic circuit he found on the internet, using parts totally unsuited to the application. That never works.