Using the HYPNOTIZE as a general shunt reg PCB

[Salas]

Isn't there a higher beta bigger package you may got for the output BJT so to handle more current linearly and dissipate better? If you will allow just 100mA extra the MJE340 will already droop its hfe curve.

 

[Rodeodave]

Maybe one of the D44H series? Has a somewhat higher hfe and should be good up to 1A. No published info on noise though.

http://www.newark.com/pdfs/datasheets/spc/01H0544.pdf
http://www.farnell.com/datasheets/296708.pdf
http://www.fairchildsemi.com/ds/NZ/NZT44H8.pdf
http://www.onsemi.com/pub_link/Collateral/D44H-D.PDF

 

[Mark Johnson]

Isn't there a higher beta bigger package you may got for the output BJT so to handle more current linearly and dissipate better?

Here is a transistor housed in the supremely macho TO-264 package (aka "TO-3P"). Max power is 200 watts. Beta is 100 from Ic=0 to Ic= 8 amperes, see Figure 4. LTSPICE models for this transistor were lovingly created by Bob Cordell and are on his website. http://www.onsemi.com/pub_link/Collateral/MJL3281A-D.PDF

The transistor is in stock at the usual online distributors, price is reasonable but not Dirt Cheap. DIYAers sometimes have difficulty locating heatsink insulators for the TO-3P package, but now there are several eBay sellers offering 100 mica insulators sized for TO-3P, very cheaply. Operate the googles to learn more.

If speed (fT) is what you crave, consider the 2SC5171. ft=200MHz, max power = 20 watts, all-plastic TO-220AB package. http://www.amplimos.it/images/2SC5171.pdf 12,000 of them are in stock, on the shelf at Mouser.

 

[Salas]

The MJL is very nice especially for heavy hot rod or for T-Amp. I had used the Toshiba in V1.1 once at 200mA and went alright. Some third party tries to measure its Ft could not find it more than 60-70MHz if I remember well. Although with some valid method still not with dedicated lab gear config for the exact purpose. D44H11 and MJE15030 had worked in circuits of mine like this one reg here too.

 

[back]

hi i want to power a sound card with hypnotize.it`s the version by crt and can use mur820 so i think no problem there.the sound card draws 12v-900ma.what current should i adjust it and what transformer should i use(volts and va)?capacitor bank 2x4700uf is enough?

thanks.

 

[Salas]

1.1A maybe. 15VAC 3A transformer (50VA).

 

[back]

hi i want to power a sound card with hypnotize.it`s the version by crt and can use mur820 so i think no problem there.the sound card draws 12v-900ma.what current should i adjust it and what transformer should i use(volts and va)?capacitor bank 2x4700uf is enough?

thanks.

i started building this and i noticed it`s slightly different.the 10r at the vref leg at the first picture at my pcb is marked 1r.should i replace it with 10r?

 

[Salas]

No, just calculate the current via voltage drop for 1R.

 

[back]

No, just calculate the current via voltage drop for 1R.

thanks salas

 

[back]

the first three leds light but not the other two with the 220r resistor.
maybe this don`t need to solder the two legs together?

i have the first one from here.

http://www.diyaudio.com/forums/swap-meet/197590-dcb1-black-white-red-kits.html

 

[Salas]

In that Hypno board you linked its taken care already.

 

[back]

nothing. the 3 leds light up and the 2+test resistor don`t.i took some measurements and i have 16+volts across the current resistor(12ohms) with 18v transformer.the current resistor starts to smoke after a couple of minutes.i have no output voltage.didn`t have time to measure anymore before it start to smoke.i uninstaled the test resistor and tried with 5 leds but its the same.

 

[Salas]

You probably got a problem with the Mosfets or something is wrong in the assembly. They need to show a few volts Vgs when in circuit when the Mosfets are healthy. Make checks to all semis. BJTs need show about 0.6V Vbe. Use double the CCS target current fuses whenever you hook up testing.

 

[back]

must be the mosfets i can`t find a mistake.i will desolder them tomorrow because they seem ok in circuit.

 

[Salas]

Check to have a rightly oriented leds path and test resistor for sure. Used the DMM diode position if it can light up leds or a battery that can light them up, but do it quick.

 

[back]

ok fixed.a bad mosfet.works like a charm.i used a scope and it great.i have 2 generic leds and 3 high brightness leds and i have 13volts output.since i need 12volts maybe i can take 12volt by using 2 high brightness and 3 generic instead of a resistor?

 

[Salas]

High brightness and blue can be suspicious for intrinsic noise. Prefer generic to be on the safe side. You can make a string of as many you need on a small matrix board cut-out. The size of a postal stamp or there about. You can fix a couple of leads to the mini led ref board and piggy back it on the "mother" board. And use only leds that way.

 

[back]

great thanks again for the support.

 

[back]

An externally hosted image should be here but it was not working when we last tested it.

Uploaded with ImageShack.us

is this acceptable or i have oscillation?

time/div is 5us/cm and volts at 5mv/cm.

it was much smaller with leds but i couldn`t adjust at 12volt and added resistor as described.

 

[Salas]

Its no oscillation. Oscillation is periodic all over the visible line and sinus or sinus like, it also has tens or hundreds of mV. What you show its what noise bursts you can pick up from everywhere in the DUT & measurement loop really. It was smaller because you had less impedance in the Vref so it was less prone to picking things in that part. Driving 1Meg impedance of a scope input from ~0 out of a PSU makes a first class antenna. Is it there if you run the scope slower, say at 100nS? Which is more like in the serving band for audio purposes? Also press the 20MHz bandwidth limit if the scope is a fast model.