Simplistic NJFET RIAA

Jump to Latest
#2,147
RCruz said:
I did a new layout so I can fit the individual heatsinks....

I calculated a led string dropout of 6,39v, using 14r I have ICCS with load = 174 mA

Using R1 = 10, I get 259mA... is this viable ?

Ricardo
Click to expand...

Depends on the Vgs of the exact part, but with such Leds I would surely use the 14R so not to overshoot 200mA.
 
#2,148
The green leds I used presented an average of 2,1v drop.

I now found some smaller yellow leds with 1,9 drop. Using 3 leds I get 5.7v... if my calculator is correct, admiting Vgs 3.4v, i can use R1=10r and get away with 190mA ICSS (230-40)... am I right ?:apathic:

Ricardo
 
#2,149
Yes if Vgs will be like that. You just measure the drop across the resistor in the end, and if its non usefully excessive, you make it higher R. I prefer the yellows slightly for best noise also. Its that the greens are easier to predict for those they don't match them.
 
#2,150
I will match the leds before use so I will know vdrop in advance.

Using your formulae: ICCS=(Vdrop-VGS)/R1 I can find R1.

After building the shunt I will measure R1 drop and using ohm´law I will verify the current.

Looking very good already.😀

Ricardo
 
#2,156
RCruz said:
Hi Salas

After several hours designing.... I hope you agree with this layout.
Click to expand...

It will surely work. Its better than the previous one. With very short and twisted cables to the load. But what Andrew says is that if you are going to make a PCB, you can make it with nodes ready for ''remote sensing''. The benefits are, that the reference and driving circuit can ''sense'' and correct for the cable to the load impedance. That becomes more critical if you put the shunt in another box with a transformer, or even for vaporizing the short length if near. There is some benefit in perfection, if near the load just incremental, but you are to decide. Here is the remote sensing with 4 wires for the V1 shunt.
 

Attachments

  • 28vrmsens.gif
    28vrmsens.gif
    7.2 KB · Views: 550
#2,158
Yes, W1,2,3,4. They go to the same places on the riaa really, they just help the shunt to ''see'' the length and compensate, by returning to different points on it. If you gonna have it just next to the riaa board, the benefit is going to be small I remind you.
 
#2,160
newbie need help

Hello to all,
I built one channel of these 2 circuits:
An externally hosted image should be here but it was not working when we last tested it.

By net_urbano

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

By net_urbano

the rectifiers diodes, the bulk C 10000uF, the shunt and the RIAA are together all on a single perforated board.
Before to build the second channell, for try, i connect the output of the riaa to a stax srm-313 headphone amplifier. What I hear in headphones is a very thiny music and a big hum. Reading the thread I saw that Thomo had a similar problem to mine and Salas suggested to measure some voltages because it is possible that last Q (buffer NJFET) has higher than calculated Idss and drops its Vd too much, Vg under its Vs, so it won’t work.
First of all I checked all the wiring, and all the parts are correctly oriented.
Than i measured some voltages:

Q5: VD = 17,7V; VS = 6,83V (OK?); VG = 6,80V (BAD);
Q3 BASE = 7,62V (OK?); Q3 COLLECTOR = 15,07V (OK; PSU = 24,6V)
Q1: VD = 6,96V (OK)

The Idss of Q5 is 9,7mA so i changed it with another one with Idss = 8mA but the problem remains the same. Big hum, no music. Q5 Vs is now = Q5 VG: 6,72V.

Can someone help me ?

Luca
 
You must log in or register to reply here.
Top Bottom