Amplificator Alpha Nirvana 4R sau 8R detalii despre constructie

[cdinu332]

Acum 17 ore, carageae a spus:

Asta ar fi ok? 143 kB · 0 descărcări

 

Acum 17 ore, hpavictor a spus:

Curentul din sarcină se închide prin ... sursă .

Sursa este la fel de importantă ca și amplificatorul propriu zis .

Traseele circuitului imprimat sunt esențiale  .

Încearcă să iei lumină de la oricine altcineva decât de la electroniști închipuiți ...

Iată întrebarea,  poftim răspuns. 

Dacă schema aia nu e bună,  modificați ceva pe ea,  scoateți un condensator,  o rezistență,  schimbați o valoare unde este necesar.  Mai puneți componente....ceva acolo. 

Nu! 

Rezolvarea este simplă,

"ești prost " sau îți iei  " lumină "  de la lumânare. 

În concluzie,  face fiecare cum poate,  forumul rămâne,  în cea mai mare parte, pentru contre și a te da rotund fără a contribui concret. 

 

Editat Februarie 18, 2024 de cdinu332

[cdinu332]

Am intrebat, referitor la schema pusa de carageae, pe diyaudio.  Eu de acolo imi iau ''lumina''.

Iata raspunsurile:

 

 

From a capacitance standpoint, you should be fine. But the resistor values imho are too large, you will lose a lot of voltage off of your rectified rails. Remember this is not a Class AB build, it’s a Class A build, where a constant bias current of at least 2A (since this is a 4 ohm build) is going through these resistors. So you lose 0.4V at R1, 1V at R2, and 1V at R3 = 2.5V. This will raise the impedance of the supply which can affect bass response. The resistors will also need to be higher wattage, i.e. P = I^2*R. So R1 will need to absorb 0.88 watts, R2 will need to absorb about 2 watts and R3 will need to absorb 2 watts. You will need to upsize those resistors by 3 times that value, so 3 watts, 6 watts and 6 watts because they will get very warm and will be prone to fail if undersized in power. I would not recommend your supply at all.

Please consider the schema published by Randy T elsewhere on diyaudio and also his build guide:
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&ved=2ahUKEwjUu_Wi_LKEAxXsMlkFHSABBv0QFnoECBAQAQ&url=https://www.diyaudio.com/community/attachments/v8-crcrc-build-guide-1-0b-pdf.1157033/&usg=AOvVaw2bfkbGwUfmmG3v8WvYtNce&opi=89978449

You can use one per channel and point to point wire it, design your own pcb or just purchase the boards from Randy himself (easiest and most optimized option!). Notice his design is CCRCRC which has it’s advantages. This power supply has been used numerous times by Aleph 60 and Aleph 2 builders and is pretty much bullet proof. It also includes the bleeder resistors so the capacitors don’t hold charge at switch off. It also includes a simple softstart using thermistors and also a ground loop breaker (using a thermistor). In the future, if you want to you can replace the rectifier portion with active rectification if you want since the rectifier boards are completely separate from the main PS board.
....................................................................................

Anand has put it very well; caps are fine, I'd suggest reducing all the resistors to 0.15R, all rated at 3W.
With a 4R load the 2A quiescent (ideally it should be 3A actually) each resistor will dissipate 0.6W, and 3A would dissipate 1.35W. So 3W will give a moderate temperature rise and long life.
Be aware that more power supply capacitance will draw higher surge currents from the transformer; there is a balanced here. My thoughts are that about 30mF is about right for each rail on a 40W Class A.
......................................................................................
 

Be aware that more power supply capacitance will draw higher surge currents from the transformer; there is a balanced here.

Yes. It’s one of the reasons I use softstarts in all my builds that use a transformer of 300VA or more.

Tom Christiansen of Neurochrome Audio has written a rather comprehensive summary of all the factors involved in designing a softstart. An important factor is the energy storage of each capacitor in the powersupply. For the newbie diy builder (not Hugh of course!) It can be calculated rather easily…E =0.5 * C * V * V; where C is the capacitance in farads, and V is the DC rectified voltage. That’s the storage of EACH capacitor which you have to add together so as to understand the energy load imposed on the transformer! The peak currents these transformers endure are pretty amazing, and that happens every time you switch the amplifier on!

This huge current surge repeatedly stresses the primary of the toroid and over time the enamel and insulation of the primary breaks down eventually short circuiting to the core. At this time transformer demise and failure occurs.

The energy storage of each capacitor in your power supply becomes important in selecting an inrush thermistor that can withstand that joule rating. The maximum (for example) of the ubiquitous CL-60 is only 36 joules. That’s usually not my first choice unless low currents and low rectified dc voltages are involved (see the equation above for why). My first choice is typically Ametherm SL22-20007 which has a rating of 125 joules.

Thankfully, these thermistors have to endure this surge for only a short period of time (less than 1 second). In the previous example where I mentioned R.Thatcher’s V8 power supply, I would substitute the thermistor with at least a 125J rating since the energy to magnetize the core of the transformer has to also be taken into consideration. For example the energy to magnetize a 400VA core is just under 20 joules.

For example if one were to use a V8 power supply for EACH channel, you have (16) 22,000uf capacitors in the stereo amplifier. Assuming dc rails at +/- 20V (for example), then each capacitor would hold 4.4J of energy. 16 of them would be 70.4J of energy. And if you were using two 400VA transformers, you would add 40 more joules. The total is 110J. Thankfully that’s not super high because the voltage rails are only +/- 20V. Change that to +/- 28V as is seen in the 8 ohm Alpha Nirvana and the energy storage is much much higher since voltage is squared in the equation. Of course, the dc voltage only holds true if the mains stays at the nominal level. If your mains varies, so will your dc voltage .

I’ve rambled on enough but I think y’all get the idea!

 

Editat Februarie 18, 2024 de cdinu332

[dumitrumy]

Pai o pagina A4 de raspuns este sintetizata in simpla simulare facuta de mine. In tabelul de sub grafic sunt tensiunile pe condensatori, deci se vede exact cat se pierde ca putere.

Folosesc MC12 din comoditate. Cu LTspice poti afisa puterea disipata pe fiecare componenta.

[cdinu332]

1 oră în urmă, dumitrumy a spus:

Pai o pagina A4 de raspuns este sintetizata in simpla simulare facuta de mine. In tabelul de sub grafic sunt tensiunile pe condensatori, deci se vede exact cat se pierde ca putere.

Folosesc MC12 din comoditate. Cu LTspice poti afisa puterea disipata pe fiecare componenta.

Vă spun cu toată sinceritatea,  nu mă pricep să interpretez ce stimulați dvs.  Cunoștințele mele în electronică sunt la nivel de lipit piese pe cablaj.  Mai mult nu pot,  recunosc. 

[Shambala]

editat - se pare ca am gresit aici. Scuze.

[Shambala]

editat - si aici la fel. 

[dumitrumy]

1 oră în urmă, Shambala a spus:

Este nevoie ca C2 sa fie de 3-5 ori mai mare decat C1, iar C3 de cca 5-10 ori mai mare decat C1.

**********************************

Sa vedem si de unde ia curent sarcina la 500Hz:

*******************************

Si la semnal de 25Hz:

Editat Februarie 18, 2024 de dumitrumy
completare continut

[cdinu332]

Si fix invers cum arata? C1-30.000uF, C2... etc.

[dumitrumy]

[Shambala]

Din simulari se pare ca am gresit eu in cele afirmate.  Puteti pune va rog si rezultatul riplului de tensiune la acelasi simulari ?

 

Mai demult am facut teste practice (pe sarcina statica) si pe osciloscop riplu de tensiune era mai mic cu C2 mult mai mare decat C1.  Si la CRC  si la CLC.

[dumitrumy]

*************************************************

Editat Februarie 18, 2024 de dumitrumy

[AlpHs00]

Se poate face si o simulare CLC cu doi condensatori de 150 000 uF si o bobina de 5mH, 0,2 ohm rdc?

Multumesc.

[dumitrumy]

***************************************8

Editat Februarie 18, 2024 de dumitrumy

[Shambala]

Va multumesc.

Nu pare sa fie mare diferenta intre cele doua situatii simulate (rip25 si rip26). in testele pe care le-am facut diferentele erau mult mai mari.

 

Daca tot sunteti atat de amabil, puteti sa mai faceti o simulare, filtru RCLC,  sarcina 6A,  cu R1=0,22ohm, C1=C2=60mF si L=10mH. 

 Si eventual R1=0,22ohm,  C1=C2=30mF,  L=20mH, tot pe 6A  sarcina.

[dumitrumy]