maybe add an weird bootstrap circuit or an led constant current source on the input?
Indeed, although simplified it would hold current constant.
Under operation with linear power supply the rails will vary in voltage.
So yes definitely, DC offset wont dance around on the output.
As noticed a set resistor vary like crazy. Great idea.
Also second gain stage current was cranked up 9ma
Only needs to be 6ma. So the bootstrap with cap resistor values can go up.
Higher value resistors are more desirable to maintain current and raise impedance.
Actually since we are at low voltage, 12 to 15v feedback current source be ideal.
With simple Zener we could potentially loose to much voltage.
You can check, bring amp up to clipping.
And determine the max voltage swing of the differential.
led could work or similar voltage zener.
You loose the voltage swing to the zener or led.
Its fine why its common for differential.
For second gain stage we want full swing to rails.
So bootstrap and feedback types more common for that.
Bootstrap will swing pretty good, feedback types only loose .6 to .7 volts
basically the common drop of silicon at temp
Under operation with linear power supply the rails will vary in voltage.
So yes definitely, DC offset wont dance around on the output.
As noticed a set resistor vary like crazy. Great idea.
Also second gain stage current was cranked up 9ma
Only needs to be 6ma. So the bootstrap with cap resistor values can go up.
Higher value resistors are more desirable to maintain current and raise impedance.
Actually since we are at low voltage, 12 to 15v feedback current source be ideal.
With simple Zener we could potentially loose to much voltage.
You can check, bring amp up to clipping.
And determine the max voltage swing of the differential.
led could work or similar voltage zener.
You loose the voltage swing to the zener or led.
Its fine why its common for differential.
For second gain stage we want full swing to rails.
So bootstrap and feedback types more common for that.
Bootstrap will swing pretty good, feedback types only loose .6 to .7 volts
basically the common drop of silicon at temp
Last edited:
Assuming wanting something simple with no extra transistors.
Just zener with 2 resistors. Show you in a min.
Can also add another transistor and zener.
Which yields far far better impedance.
Was just hoping youd be able to check differential voltage.
To make sure or just basic exercise to see its output swing.
I already know were fine.
Full power before clipping 4ohm
differential on 100mV scale = VM2. Amp output 5 volt scale = VF1
diff is under 100mV or 200mV peak to peak
At absurd amounts of clipping 500mV scale
still under 800mV or 1.6 volts peak to peak.
So more than fine with 6 or 9 even 12 volt
zener. Not worry about limiting the voltage swing.
Or losing rail voltage of the differential with the zener
Simplified 9 volt zener be fine, with transistor usually
3 to 6 volt is fine. No worry with 14 volt rails or 28 pp
Just zener with 2 resistors. Show you in a min.
Can also add another transistor and zener.
Which yields far far better impedance.
Was just hoping youd be able to check differential voltage.
To make sure or just basic exercise to see its output swing.
I already know were fine.
Full power before clipping 4ohm
differential on 100mV scale = VM2. Amp output 5 volt scale = VF1
diff is under 100mV or 200mV peak to peak
At absurd amounts of clipping 500mV scale
still under 800mV or 1.6 volts peak to peak.
So more than fine with 6 or 9 even 12 volt
zener. Not worry about limiting the voltage swing.
Or losing rail voltage of the differential with the zener
Simplified 9 volt zener be fine, with transistor usually
3 to 6 volt is fine. No worry with 14 volt rails or 28 pp
Last edited:
Getting there. Were just trying to get a basic amplifier with differential
without excessive DC offset on output.
Least we established the whole heart of the amplifier the differential.
Needs to be balanced, and the current must be constant, even with
rail voltage that varies in operation.
The amplifier is simplified. Distortion will be typical.
Once you try to improve that. Simplified designs no longer
remain simple. After I show him a simple current source for the Diff.
I can show Fourier for expected distortion.
Yes we will check open loop/closed loop phase margin / Gain Margin
And determine stability
Im redesigning the amp 4 or 5 times. Wont be doing extra work for open loop test.
Until the design is final.
without excessive DC offset on output.
Least we established the whole heart of the amplifier the differential.
Needs to be balanced, and the current must be constant, even with
rail voltage that varies in operation.
The amplifier is simplified. Distortion will be typical.
Once you try to improve that. Simplified designs no longer
remain simple. After I show him a simple current source for the Diff.
I can show Fourier for expected distortion.
Yes we will check open loop/closed loop phase margin / Gain Margin
And determine stability
Im redesigning the amp 4 or 5 times. Wont be doing extra work for open loop test.
Until the design is final.
Last edited:
geniuenly intrested if i could copy this current source that uses an led instead of a zener becuse while i have zeners theyre hidden among regular diodes.
That is one way to do it. Somewhat conventional
Usual high parts count, but normal.
Combination of Led Current source for differential, and feedback current source for 2nd gain stage not fully visible
You would have to calculate proper values depending on rail voltage.
Like any other amplifier.
When using approx 1.8 volts LED voltage ref as shown with transistor.
Output impedance is around 300k
When shooting for low distortion amplifiers, and tossing in current mirrors / beta enhancement etc etc
You can raise the voltage ref to 3.3 volts or better 6.8 volts with a zener.
Output impedance goes way up to around 2M
So beneficial for low distortion amplifiers. 2M is about as good as you'll get
and feedback style needed to not reduce voltage swing is about 1M impedance.
We can use led or zener reference current sources for differential.
Since the 1.8 to 6.8 volt rail loss is no concern.
And the 2M impedance gained from 6.8 volts ref is beneficial for lower distortion
Aside from assumption the zener maintains well over wide temperatures
For the simple zener is shown.
The rail is simply regulated by the zener to around 8 volts. And since we obtained constant voltage.
Now a set resistor will have constant current. Simply zener regulation will reduce noise or rail ripple as well.
Could probably lower the current of the zener even more. Half watt not to worried.
Rail voltage can change from 12 to 18 volts or more and current will remain constant at 2mA
Usual high parts count, but normal.
Combination of Led Current source for differential, and feedback current source for 2nd gain stage not fully visible
You would have to calculate proper values depending on rail voltage.
Like any other amplifier.
When using approx 1.8 volts LED voltage ref as shown with transistor.
Output impedance is around 300k
When shooting for low distortion amplifiers, and tossing in current mirrors / beta enhancement etc etc
You can raise the voltage ref to 3.3 volts or better 6.8 volts with a zener.
Output impedance goes way up to around 2M
So beneficial for low distortion amplifiers. 2M is about as good as you'll get
and feedback style needed to not reduce voltage swing is about 1M impedance.
We can use led or zener reference current sources for differential.
Since the 1.8 to 6.8 volt rail loss is no concern.
And the 2M impedance gained from 6.8 volts ref is beneficial for lower distortion
Aside from assumption the zener maintains well over wide temperatures
For the simple zener is shown.
The rail is simply regulated by the zener to around 8 volts. And since we obtained constant voltage.
Now a set resistor will have constant current. Simply zener regulation will reduce noise or rail ripple as well.
Could probably lower the current of the zener even more. Half watt not to worried.
Rail voltage can change from 12 to 18 volts or more and current will remain constant at 2mA
Last edited:
Of course the differential has no mirror
So your still at the disposal of R2 value maintaining balance.
Why im still using 1 to 5% value.
Any small jump of 50 to 200 ohms can cause DC offset to climb 20 to 40mV
Suitable for some, garbage to me.
Being aware with quasi combining CFP and E2 stages.
Any DC offset can effect stability and more so even with the dropping diode
in the CFP. DC bias will be unbalanced for the power transistors and again.
Becomes more likely for T8 to conduct more AC current.
I usually accept no more than 3 mV DC offset for any amp.
But important for Quasi.
Regardless 10 to 15mV is about all you get with simplified Lin topology.
Usually worse from 20 to 40mV over temperature ranges and circuit values
And includes the use of 1 to 5% resistors in certain locations.
R9 , R15 degen. And R2
1% ideal for degen, unless made adjustable. Ideal for many reasons.
So your still at the disposal of R2 value maintaining balance.
Why im still using 1 to 5% value.
Any small jump of 50 to 200 ohms can cause DC offset to climb 20 to 40mV
Suitable for some, garbage to me.
Being aware with quasi combining CFP and E2 stages.
Any DC offset can effect stability and more so even with the dropping diode
in the CFP. DC bias will be unbalanced for the power transistors and again.
Becomes more likely for T8 to conduct more AC current.
I usually accept no more than 3 mV DC offset for any amp.
But important for Quasi.
Regardless 10 to 15mV is about all you get with simplified Lin topology.
Usually worse from 20 to 40mV over temperature ranges and circuit values
And includes the use of 1 to 5% resistors in certain locations.
R9 , R15 degen. And R2
1% ideal for degen, unless made adjustable. Ideal for many reasons.
Last edited:
Dual Supply Rail no output cap.
The voltage is fine.
What do you think it was when there is normally a 4.7k resistor there?
Nothings changed just the current is constant when the rail voltage changes under load
Diff operates under 2 volts regardless
Without Zener and just resistors, its the same
The voltage is fine.
What do you think it was when there is normally a 4.7k resistor there?
Nothings changed just the current is constant when the rail voltage changes under load
Diff operates under 2 volts regardless
Without Zener and just resistors, its the same
Last edited:
Would have to see the datasheet for voltage tolerance and current.
Usually not much useful for more than 1ma to 500uA assuming.
They do have dual matched 3904/3906 on a package as well.
They are gain matched and thermally linked on same wafer. So would work
for this project for a differential, or even current mirrors.
likewise may other NPN PNP dual matched packages.
TI also has current mirrors and differential packages on one chip.
Often include 1ma current sources as well.
Usually not much useful for more than 1ma to 500uA assuming.
They do have dual matched 3904/3906 on a package as well.
They are gain matched and thermally linked on same wafer. So would work
for this project for a differential, or even current mirrors.
likewise may other NPN PNP dual matched packages.
TI also has current mirrors and differential packages on one chip.
Often include 1ma current sources as well.
the ul1202 can dissipate around 300mw and idles at around few ma so it could definitly output 2/5ma.
i also have the ul 1111.
i also have the ul 1111.
also how much would the distortion improve if i switched to two darlington pairs made of (to-220) mje3055/mje2955? or other overall faster transistors such us ones from an smps power supply.
couls i substitute the zener diode with an led or just use an "basic" single transistor constant current source?