discrete op amp

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
QSerraTico_Tico said:
lineup, your circuit will work better with a dual JFET like 2N5912.

My circuit:
------------------------
1. Supply 9.0 Volt DC (for one battery)
2. Gain x100
3. -3dB at 410.000 Hz
4. less than -110dB second harmonic distortion in 1mV in - 100mV out into 22kohm
5. stable into 22nF capacitance across 22 kohm
6. supply current < 2.5mA per channel


.... works good enough :cool: at least in my opinion
using simply BC550C transistors

Maybe you need better amplifier.
But for me this one is really okay.



My pre amplifier use an old NPN Only based design.
( to improve this old model, I have added two BC560C, PNP for use in current sources )
It is intended as elektret microphone amplifier for 9V battery power.
Typically input from such a capsule can be 0.5-2 mV.



But thanks anyway.
I may design some other JFET input pre amplifiers.
I will keep 2N5912 N-FET in mind for this.
I will download datasheet now!
And also try to find a good Spice model for 2N5912.

:)

http://www.interfet.com/pdf/DS_2N5911_12.pdf

Thank you.
lineup
 
2N5911 2N5912 Dual JFET matched pair

QSerraTico_Tico said:
lineup, your circuit will work better with a dual JFET like 2N5912.


I found and downloaded SPICE MODEL for 2N5911 and 2N5912, Dual JFET.
Very good FET!
Think they are for low voltage applications.
What I can find out, they are Max 25 Volt
( Gate Source Breakdown Voltage: Min -25V )

The only difference is that 2N5911 has 10mV max VGS match
while 2N5912 has got max 15mV difference between the JFET Pair.
So slightly better match precision granted in 2N5911.

I hope these spice models will work!
.model J2N5911 NJF(Beta=2.203m Betatce=-.5 Rd=1 Rs=1 Lambda=12m Vto=-2.694
+ Vtotc=-2.5m Is=94.35f Isr=907.9f N=1 Nr=2 Xti=3 Alpha=13.92u
+ Vk=88.13 Cgd=2.822p M=.3744 Pb=1 Fc=.5 Cgs=3.65p Kf=36.4E-18
+ Af=1)
* National pid=93 case=TO78
* 88-08-05 rmn BVmin=25

.model J2N5912 NJF(Beta=2.203m Betatce=-.5 Rd=1 Rs=1 Lambda=12m Vto=-2.694
+ Vtotc=-2.5m Is=94.35f Isr=907.9f N=1 Nr=2 Xti=3 Alpha=13.92u
+ Vk=88.13 Cgd=2.822p M=.3744 Pb=1 Fc=.5 Cgs=3.65p Kf=36.4E-18
+ Af=1)
* National pid=93 case=TO78
* 88-08-05 rmn BVmin=25

I think it is from PSPICE library.
This jfet.lib with 2N5911
can be download from a Greece link:
OrCad Libraries/Library for Capture/PSPICE/



Regards & Enjoy!
lineup
 
tnx, i will try that model out too

i'm glad this got people thinking.......

i tried removing the cap, got an offset of 13V...... i might use a DC servo.

a lot of preamps use pots in the feedback as volume controls with no problems, but i wouldn't use one in a power amp feedback loop. i just set the gain at some predetermined level so i get full power with 1.5Vrms input. for a 100w amp, that is about a gain of 30. that's one of the reasons the IHF recommends that manufacturers publish their amp sensitivity specs, so that if preamp "A" with the volume set at 10 gives you full power with amp "B" (a 100w amp), and you buy amp "C" (400w), you already know from the specs whether setting the volume at 10 on the preamp will give you full power, or over/under drive the amp.
 
ok, as i suspected, my sim has a "dirty" signal source..........
i added a 5 section resonant low pass filter with a 100db/octave slope between my sim sine wave source and the amp. the sim sine wave source has -60db 3rd harmonic in it!!!!!!

after adding the filter, i get -80db 2nd harmonic, and -90db 3rd at the amp output.

btw remember to put a resistive load on the filter, otherwise your signal into the unit under test will begin to climb. i think the proper value to use is slightly less than Xc.

i'll have to do a spice model of the filter network, so i can use it as a 3 terminal device, but i have to find a way to tune it, as the one i have is a 1khz cutoff. who woulda thought that a sim sine source would have so much distortion?....... lets see..... -80db would be .0001% distortion? did i get that calculation right?
 
with 1.5Vrms input. for a 100w amp, that is about a gain of 30
your use of about = very approximate.
I reckon the gain is more like "about" 18.9
Times 30 will give about 250W from 1.5Vrms.

Lineup.
your use of a decimal point in the middle of long numerical values is very confusing. Please try to be less ambiguous.

410.000Hz is I guess 410kHz or did you really mean 410Hz?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.