Laboratorijski SMPS

[sonny]

Nije baš direktno vezano za audio, ali nije ni daleko...

Već duže vrijeme se bavim mišlju napraviti nekakav laboratorijski izvor napajanja...
U principu mi najviše treba single supply za 12-14V sa što više ampera i tu bi u principu SMPSi iz rashodovanih računala/servera i sličnog trebali odraditi posao i više nego dovoljno dobro... S@ki je kemijao nešto i ima više nego pozitivna iskustva... Ja međutim nemam nijedno napajanje a da iole vrijedi... Imam jedno 12V/10A (kao) i to u principu ne može van dati ni 7 a da se ne ugasi S druge strane nema regulaciju napona...
Imao sam par reguliranih ispravljača s klasičnim trafoom i serijskim tranzistorima u regulaciji i u pravilu nisam bio zadovoljan... Ili bi trebao ići na prilično visok napon sekundara pa da se na tranzistorima "ruši" prilično napona što pri jakim strujama nije nimalo jednostavno za ohladiti bez da mi ispravljač zauzme pola stola, ili su redovito izgarali neki od tranzistora pri jačim strujama ( i relativno malo "srušenog" napona )...

Pa sam naletio na :

http://danyk.wz.cz/reg60v_en.html

OK, ne treba mi uistinu toliki raspon regulacije napona i tolika struja, tako da se neću slijepo držati ni svih naputaka...
Svima su problem feritne jezgre...
Ja na svu sreću čuvam 2 koje sam davno kupio u Chipoteci i nisu koštale malo...

IMG_1706.jpg

Ako se ne pokažu adekvatnima, imam još zamjena ( no o tome kasnije )...
Sve ove "AUX" i ostale se nađe po izgorenim napajanjima od računala i sličnog i gotovo od svake imam po dvije iste
U principu, trenutno (pre)crtavam shemu, pa ću probati što je moguće bolje nacrtati pločicu, da bude što kompaktnija, iako se to kosi s disipacijom i veličinom hladnjaka...
Radit' će je, pa zna se tko

Jedino što trenutno vidim kao "problem" su IGBT tranzistori...
Makar, kao što rekoh, profesija mi je zavarivanje... U servisu imamo preko nekoliko pločica iz invertorskih izvora struje gdje je "odlazilo" svašta... Od IGBTa do dioda u ispravljaču, kontrola razno raznih i tako... Pločice su mijenjane novima, tako da su stare za rashod... Jezgri i ostalog ima na bacanje... 50A novih duplih brzih dioda imam barem 50 komada... Tako da vjerujem da ću među svim tim elektroničkim "otpadom" i pronaći 2 adekvatna IGBT tranzistora...

Probati ću to sastaviti uz što manje troškove, odnosno korištenjem "rashodovanih" i novih dijelova složiti ovo i vidjeti da li uopće funkcionira kako je prikazano...
Iskreno se nadam da da... Da ne bih trebao zvati vatrogasce u pomoć

Mišljenja ?

EDIT : Gledao sam ja i gotova rješenja po ebayu, ali cijene su visoke (iako ne toliko previsoke ako to radi kako treba) i mahom je regulacija plus minus max. 10% od nazivnog napona... Što mi je premali opseg...

[the_bebe]

Nemogu otvoriti link...blokira mi ga google????

[Schultz]

Ako ovog borca nije tresnula struja onda neznam koga će.

Pa gori je od onog famoznog kasapina Fikusa Lampizatora....

[Schultz]

High power adjustable switching power supply (SMPS) 3-60V 40A

This switching power supply was built because I needed powerfull adjustable power supply for laboratory purposes. Linear topology would be unusable for this power (2400W - 2.4 kilowatts!). So I chose the switching topology of two switch forward (half controlled bridge). In my article about SMPSs it is a topology II.D. The switching upply uses IGBT transistors and is controlled by the UC3845 circuit. The schematic of my switching power supply you can see below. Mains voltage goes first through EMC filter. It is then rectified and filtered on capacitor C4. Due to high capacity, there's an inrush limiting circuit with Re1 switch and R2. Relay coil and fan (from PC power supply AT / ATX) are powered from 12V, which is obtained via resistor from 17V aux supply. Select R1 so that voltage at the relay coil and the fan was 12V. The auxiliary supply circuit uses TNY267. It is similar to described supply described here. R27 provides undervoltage protection of auxiliary power - will not turn on when less than 230V DC. UC3845 control circuit has the output frequency of 50kHz and a maximum 47% duty cycle. It is powered through the zener diode, which reduces the supply voltage of 5.6 V (ie to 11.4 V) and also shifts the UVLO thresholds of 7.9 V (lower) and 8.5 V (upper) to 13.5 V and 14.1 V. Source then begin work on power 14.1 V and never run below 13.5 V, which protects the IGBT from desaturation. The original thresholds of UC3845 were set too low. The circuit controls MOSFET T2, which drives gate drive transformer Tr2. It provides galvanic isolation and floating drive for upper IGBT. Via forming circuits with T3 and T4 it drives gates of IGBTs T5 and T6. They then switch rectified line voltage (325V) to power transformer Tr1. Its output is then rectified and finally averaged and smoothed by coil L1 and capacitors C17. Voltage feedback is connected from the output to pin 2 of IO1. The output voltage of the power supply can be set using the potentiometer P1. Galvanic isolation of the feedback is not necessary because the control circuit is connected to the secondary SMPS side and thus separated from the network. Current feedback is fed through the current transformer TR3 into 3 pin IO1. The threshold current of the overcurrent protection can be set by P2.
Transistors T5 and T6, diodes D5, D5 ', D6, D6', D7, D7 ', and the bridge should be placed on the heatsink. Diodes D7, capacitors C15 and protective RDC snubbers R22 + D8 + C14 should be placed as close as possible to the IGBT. LED 1 indicates operation of the supply, LED 2 indicates the current mode (overload / short circuit) or error. It lights up when the supply is not working in voltage mode. In voltage mode, on the IO1 pin 1 is 2.5 V, otherwise it is about 6V. LED lights can be omitted.
Inductance: Power transformer Tr1 I salvaged from an old powerfull 56V switched power supply. Transformation ratio of primary to secondary is about 3:2 to 4:3 and the ferrite core (EE shape) has no air gap. If you like to wind yourself, use a similar core as in the inverter So about 6.4 cm2 (acceptable range is 6-8 square cm). The primary is 20 turns of 20 wires, each has diameter of 0.5 to 0.6 mm. The secondary 14 turns of 28 wires together of the same diameter as primary. It is also possible to make windings of copper strips. On the contrary, the use of a single thick wire is not possible for reasons of skin effect (as it works with high frequencies). Halving the winding is not required, you can for example wind primary first and then secondary. Forward gate drive transformer Tr2 has three windings, 16 turns each. All winding are wound at once, by three twisted insulated bell wires. Is is wound on the ferrite core EI (EE may be) without air gap. I salvaged it from the main power transformer from a PC ATX or AT. The core has a cross section of about 80 to 120 mm2. Current transformer TR3 is 1 turn to 68 turn on the ferrite ring and the size or number of turns is not critical. In case of different number of turns you have to adjust R15. Auxiliary power transformer TR4 is wound on ferrite EE core with air gap and the cross section about 16 to 25 mm2. It comes from the auxiliary power transformer gutted from ATX. Orientation of winding of transformers (marked by dots) must be followed. Double choke EMI filter can be eg from microwave. Output coil L1 also comes from a 56V SMPS I gutted. It consists of two parallel inductors 54uH on the iron powder rings and the total inductance is therefore 27uH. Each coil is a wound by two magnet copper wires 1.7 mm diameter. The total cross section of L1 is then approximately 9 mm2.
L1 is connected to the negative branch, so the cathodes of diodes have no RF voltage and can be mounted on heat sink (or part of it) without insulation. Maximum input power of this switched power supply is about 2600W and efficiency at full load more than 90%. In this switching power supply I used IGBTs of type STGW30NC60W. They can be replaced by the types IRG4PC40W, IRG4PC50W, IRG4PC50U, STGW30NC60WD or similar enough powerful and fast. The output diodes can be any ultrafast with enough current rating. The upper diode (D5) sees 20A average current in the worst case, for the lower diode (D6) it 40A. The upper diode thus can be selected for half current than the bottom one. The upper diode can be eg two HFA25PB60 / DSEI30-06A or single DSEI60-06A / STTH6010W / HFA50PA60C. Bottom diode may be two DSEI60-06A / STTH6010W / HFA50PA60C or four HFA25PB60 / DSEI30-06A. Heatsink of diodes must dissipate a loss of approximately 60W. Loss of IGBTs can be up to 50W. The loss of D7 is difficult to predict because it depends on the properties of Tr1 (its inductance and coupling). Loss of the bridge is up to 25W. Source uses similar schematic as welding inverter since it works well. Switch S1 allows standby shutdown, because frequent mains switching of this powerful supply would not be appropriate (especially when used as laboratory power supply). Consumption in the standby state is only about 1W. S1 can be omitted. The supply is also possible to construct for a fixed voltage source. In this case it is advisable to optimize the transformation ratio of Tr1 for the best efficiency (eg, primary 20 turns and sedondary 1 turn for every 3.5 - 4V).
Attention!!! Switching supply is not for beginners, because most of its circuits are connected to mains. The mains voltage can occur on the output when improperly constructed! The capacitors can remain charged to dangerous voltage even after disconnected from mains. The output voltage can be above the safe touch voltage. This is the switchning power supply of high power. The AC input must have proper fuse, power outlet and the cable must be dimensed for the drawn current, otherwise there is a risk of fire. Everything you do, you do it on your own risk.

[the_bebe]

Na kraju sam uspio otvoriti ali stvarno ima mesara okolo...jasno mi je da S@ki presup da mu radi pločice ali mogao se malo potruditi oko tih pločica...

[Schultz]

A s@ki mora za Golfa zaraditi... Nemoj tako...

[Dean]

Sonny.....SMPS

[the_bebe]

Sonny.....SMPS

[sonny]

Baš sam htio vidjeti tko će prvi skužiti mesarenje...
Fikus je mala beba za ovo...
Ako ovo radi ovako kako radi ( ima par filmića ), onda stvarno ne smije biti problema kad se to odradi "kako treba"...
Makar, ima puno stvari koje će morati van pločice, tako da...

[s@ki]

reci ti šta oćeš, al treba ono znat onak sfušat