Guys, i found some interesting design notes on the internet while looking some class D amps.
https://www.infineon.com/dgdl/irs20965spbf.pdf?fileId=5546d462533600a4015356761d8b279b
https://www.infineon.com/dgdl/Infin...N.pdf?fileId=5546d46269e1c019016a4a5a8fa04a9b
Both chips are PWM input and capable of driving IGBTs. Last chip should allow us to draw current directly from AC without violating isolation hazards.
All the power supply we need is some tiny isolated 2-3€ switching DC to DC converter for the chip. Everything else will be connected directly to rectified AC, ground will be neutral, floating design, some diodes here and there, some zeners to make sure, some bootstrap, few resistors, nothing serious.
In order to isolate speaker for safety reasons or go "tube", we could build up multiple chips to create conditions for centertapped transformer.
PRO thing is that we loose that bulky high-expensive transformer from the powersupply and use 2 diodes only for RECT. There will be only 2 diode sound on the speaker.
Can be made all digital too. Analog only in speaker cables.
So my point is that all this costs nothing and should be researched into.
Approx size could be 15x15cm for PWM.
A very attractive and fancy design, suitable for 2022. Super green too 🙂, maybe sound close to tubes. You can put it into ur pocket, take a walk to friends house and easily compare amps.
To style it up, use this box to build it into.
https://img1.etsystatic.com/000/0/5624852/il_fullxfull.150654955.jpg
Also, can be used as a battery powered amp with a boost converter, for headphones.
Is someone interested on designing it ?
https://www.infineon.com/dgdl/irs20965spbf.pdf?fileId=5546d462533600a4015356761d8b279b
https://www.infineon.com/dgdl/Infin...N.pdf?fileId=5546d46269e1c019016a4a5a8fa04a9b
Both chips are PWM input and capable of driving IGBTs. Last chip should allow us to draw current directly from AC without violating isolation hazards.
All the power supply we need is some tiny isolated 2-3€ switching DC to DC converter for the chip. Everything else will be connected directly to rectified AC, ground will be neutral, floating design, some diodes here and there, some zeners to make sure, some bootstrap, few resistors, nothing serious.
In order to isolate speaker for safety reasons or go "tube", we could build up multiple chips to create conditions for centertapped transformer.
PRO thing is that we loose that bulky high-expensive transformer from the powersupply and use 2 diodes only for RECT. There will be only 2 diode sound on the speaker.
Can be made all digital too. Analog only in speaker cables.
So my point is that all this costs nothing and should be researched into.
Approx size could be 15x15cm for PWM.
A very attractive and fancy design, suitable for 2022. Super green too 🙂, maybe sound close to tubes. You can put it into ur pocket, take a walk to friends house and easily compare amps.
To style it up, use this box to build it into.
https://img1.etsystatic.com/000/0/5624852/il_fullxfull.150654955.jpg
Also, can be used as a battery powered amp with a boost converter, for headphones.
Is someone interested on designing it ?
Attachments
![IMG_20220214_154054[1].jpg](/community/data/attachments/933/933381-39064a66ad783f151ce47717a6ef3d8a.jpg?hash=OQZKZq14Px)
Last edited:
Than i found a this one too
It contains everything the body needs. Use of high-voltage circuit, isolation, using a CPU as PWM output.
And some other interesting buffer ideas.
It contains everything the body needs. Use of high-voltage circuit, isolation, using a CPU as PWM output.
And some other interesting buffer ideas.
Directly powered class D amps have been made decades ago. The isolation problems are about commercial liability, so I don't know of any commercial products. An amp that connects directly to the line could be perfectly safe if the installation was done by a certified electrician or equivalent, but when you sell a product to the public, you have no control how they (mis)use it. The output would have to meet class-A wiring code. The input would have to be isolated or otherwise protected from hot ground voltage. Such an amp is very much like VFD motor drive controllers, which are now common, but it is understood that "laymen" do not mess with large (~3 phase) electric motors. Some VFDs operate at an audible frequency that would be unsuitable for an audio amplifier, but issues with bearing damage are pushing the switching frequencies up. Before I retired, I wrote a lot of test software that used an "AC source". The HP/Agilent/Keysight 6813 operated at a switching frequency of 80KHz and used MOSFETS, while other larger AC sources used IGBTs and operated at a much lower switching frequency. The schematic and service manual for a 6813 are available on-line. It does not power directly from the AC mains because,
1. It needs to work from 90 to 240 VAC input.
2. The DC supply Voltage needs to be higher than the peak output, which is about 300VAC x 1.414. The 6813 is a bridge pair running off 500VDC.
3. As a service instrument, it is constantly being connect/reconnected, so the odds of a safety issue are high.
It should be noted that similar technology is used in electric vehicles. I heard that Tesla went with MOSFETs rather than IGBTs for reason of efficiency. IGBTs are rugged but have a very high on state voltage drop.
So if you need massive power in an institution, installed and operated by professionals, a directly powered amp would be fine. But such institutions are not inclined to skimp and therefore spend whatever it costs to buy professional amps.
1. It needs to work from 90 to 240 VAC input.
2. The DC supply Voltage needs to be higher than the peak output, which is about 300VAC x 1.414. The 6813 is a bridge pair running off 500VDC.
3. As a service instrument, it is constantly being connect/reconnected, so the odds of a safety issue are high.
It should be noted that similar technology is used in electric vehicles. I heard that Tesla went with MOSFETs rather than IGBTs for reason of efficiency. IGBTs are rugged but have a very high on state voltage drop.
So if you need massive power in an institution, installed and operated by professionals, a directly powered amp would be fine. But such institutions are not inclined to skimp and therefore spend whatever it costs to buy professional amps.
Last edited:
Take a look at a drawing and think, can audio be amplified using only the 15V power section, low count IGBTs involved, only that one IGBT ? And grounding to neutral possible ?
We can use IGBTs to make a DC voltage as high as any can desire.
Smallest power amp would include only 1 IGBT. Transformer core will determine sound signature.
We can use IGBTs to make a DC voltage as high as any can desire.
Smallest power amp would include only 1 IGBT. Transformer core will determine sound signature.
All i am saying is that neutral can be our ground, do audio differently... some experiments are needed.
Playing with the 15V regulator circuit, distortion and noise could be measured differently. Just some experiments in that area.
Playing with the 15V regulator circuit, distortion and noise could be measured differently. Just some experiments in that area.
Don't forget that there are a lot of 'faulty installations' where neutral and phase have been interchanged/swapped over by mistake.
I don´t see the advantage if you need an output transformer for the audio.
Why not use a tiny SMPS transformer in the supply to get the same safety?
Jan
Why not use a tiny SMPS transformer in the supply to get the same safety?
Jan
I had some strange experience with some of my builds, but there is something that have always pushed me further, instead of my own creativity i tend to stay with the trend, circuits that are available to public. For example, i was testing a mosfet speaker protection for naim 140 amp and i disliked it immidietly, not because of the mosfets but the layout and wiring was wrong, i felt like a peace of wires, installation distance and material reflection was wrong. It could be fixed but finding a good layout, hmmm you know, questionable. I used that protection earlier outside amp case with different wires and layout, i liked it very much. After this dislike, i just accidently dropped my speaker ground on the aluminum amplifier case and i had sound, it seemed that it was different, but good, loudness same. almost same as if i connected speaker ground between capacitors. My case was grounded to a transformer shield, but my outlet was not. The current path was a transformer ground which is floating, or throught the computer output socket which was connected to the computer case and it had a ground connection to the outlet. So, multiple paths can be good, there was no humm too. I dont know if i have time to test things again. Also I had some ac filters that "pushed " some content in music out of the speaker, but its not working with every amplifier, only with certain schematics. Ac is everything guys. We can do it by pumping our dc psu units but doing it directly in ac could be more rewarding and no doubt about cost effectivness
Last edited: