General Transistor Bias Question

[Dormammu]

The R5 is not an emitter load, it's a bit of a weirdo method to bias the +input of the opamp.
The dependant load of the emitter follower is always the emitter resistor.

[antonis]

The R5 is not an emitter load

Really..??

If so, then Emitter follower gain should be constant and independent of load driven..

P.S. Plz, see page 4-3 for a better understanding..
http://www.carlomozetic.net/older/userfiles/Unit_4.pdf

[FSFX]

The R5 is not an emitter load, it's a bit of a weirdo method to bias the +input of the opamp.
The dependant load of the emitter follower is always the emitter resistor.

C2 an R5 are not really necessary as the following stage has diode clipping that restricts its dynamic range.

Neither is C10 and R12 actually needed. So if reviewed for production engineering, the cost of building that circuit could be reduced by eliminating those components. Their values add nothing significant to the frequency response or performance of the circuit.

[antonis]

Q1 & Q2 are residuals from original TS808 design..
(where they served as by-pass buffers..)

By replacing bi-polar op-amp with FET inputs one, R5 could be as big as 1M and input impedance problem solved..
By maintaining bi-polar inputs op-amp, R5 could be consisted of two discrete resistors with their junction bootstrapped via a cap from IC1a output and input impedance problem solved again..

You can find a more simplified circuit in the Son of Sreamer schematic, by Jack Orman..
(or Zombie screamer, by unknown..)

[Dormammu]

By replacing bi-polar op-amp with FET inputs one, R5 could be as big as 1M and input impedance problem solved..
By maintaining bi-polar inputs op-amp, R5 could be consisted of two discrete resistors with their junction bootstrapped via a cap from IC1a output and input impedance problem solved again..

Almost any IC opamp has enough impedance to operate on such circuit a buffer-free .
VT62, is it you ?   

[antonis]

Almost any IC opamp has enough impedance to operate on such circuit a buffer-free .

Are you trolling me or are you deperately cluless..??

[FSFX]

Almost any IC opamp has enough impedance to operate on such circuit a buffer-free .

Are you trolling me or are you desperately clueless..??

I think that there are a few that could do with a course in electronic design using op amps including those who did the original circuit.
There are many good reference books on designing with op amps from National, TI and others. 
The whole original circuit looks like something drawn up on the back of a cigarette packet in the pub or bar during a night out anyway.
Most component values and the op amp choice seem to have just been chosen arbitrarily with no consideration for whether they are actually best for the circuit, really needed or what the effect of their values has on the biasing.
The whole circuit could have been done a lot better but that is what it is and so we all have to live with it or redesign it the way it should be done. 

[Dormammu]

Almost any IC opamp has enough impedance to operate on such circuit a buffer-free .

Are you trolling me or are you deperately cluless..??

I tried a lot of opamps in such a circuit (without buffers) — they all worked fine.
Right now, I'm using a near 60k input impedance amplifier (like guitar amp), and it works pretty well, too.

[Phend]

Well I am certainly not a designer or hang out at the pub anymore I do like to make circuits that are available, then maybe have a few educational type questions. From what I have read the TS808 was/is a popular effect. Thanks to all for some insight.
http://www.geofex.com/Article_Folders/TStech/tsxfram.htm

[FSFX]

From what I have read the TS808 was/is a popular effect.

My comments were not intended as a direct criticism of the TS808 in any way, purely an observation from an electronic design perspective.
As I said, 'it is what it is' that is what makes it a TS808.
Many popular circuits are not 100% perfect electronic designs if there ever could be one. Sometimes design is an artform rather than a science.
It is just that some do have things in them that are either unnecessary or could have been done better.
My comments were more related to the rather long discussion in this thread that seemed to be getting nowhere. 

[Phend]

Question:
What should Ve Vb and Vc be using 5088 ?
And again does it matter ?

I know, believe I do, that for applications used in a Fuzz for instance you might want Vc to be 4.5 (9/2) in places. But this is a different application.

[antonis]

Question:
What should Ve Vb and Vc be using 5088 ?

Answer:
Exactly what you've measured..

P.S.
In case of Vc different of power supply, we have problem..

[FSFX]

Question:
What should Ve Vb and Vc be using 5088 ?
And again does it matter ?

I know, believe I do, that for applications used in a Fuzz for instance you might want Vc to be 4.5 (9/2) in places. But this is a different application.

There is no right or wrong answer as it really depends on a couple of things.

- Firstly do you want your buffer to be as linear as possible and therefore not introduce any significant distortion.
- What signal level do you want it to handle without clipping either at the top or bottom.
- What is the forward transfer characteristics of the particular device you are using in respect to its collector current and its collector to emitter voltage.

So this is why when building these things professionally you would determine all of those things and measure the results on an audio analyser and then adjust values accordingly.

[Phend]

Key word, Professionally, which I am not, just a lowly mechanical engineer.
In a company you might be structured like:
Think Tank Dept....secret zzuf project
Design Dept.....figures out what the heck zzuf is and draws it on a board
Engineering Dept...Corrects the Tank and Design Dept
Drawing Dept....Corrects the Eng Dept and puts it on "paper"
Checking Dept...Corrects the Drawing mistakes.
Production Dept...Sends it to Mfg Dept.
Mfg Dept....Corrects the whole mess so it can actually be built (using company change methods, ISO9001)
Shop Dept...Makes the zzuf thing
Sales and Accounting Dept's....do just that, and give themselves a raise and lay off the engineers.

[FSFX]

Well, there you have it.
If you are building a one-off for yourself then it probably doesn't matter.
If you are designing something to meet the specification requirements of a marketing department and the strict requirements of the production engineering and the QA department, then you may need to be a bit more specific in the selection of components and the test parameters for your circuit.

A this is meant to be a DIY group (as someone reminded me a few days ago) then I assume it really doesn't matter as long as it sounds good to you.   

[Phend]

Your not Diying unless you keep on Trying.

Update, 5088 is listed as a sub for 4401.
RG likes 5088.
My first post here was a little incorrect in that I used 5089 (higher gain) I'll say so.
The 5089 measured Hfe 915.
So I dug out some 5088 at Hfe 514.
Voltages dropped not much, 0.2 volts at both B and E.
But it does sound more Mellow. In fact much better.
I believe the diode switch in more prominent too.
Will make video.

[m4268588]

Neither is C10 and R12 actually needed.

No. It is necessary.

Reason?
3.9/10e3/hFE*100e3

[FSFX]

Neither is C10 and R12 actually needed.

No. It is necessary.

Reason?
3.9/10e3/hFE*100e3

Are you trolling me or just posting garbage for the hell of it?

Merry Xmas anyway. Sounds like you started already.

[m4268588]

I wrote worst case.
Frequent concern is that wiper open. What will happen when it open?


Ve(Q2) is maintained by C11.
Occurrence of crackling is difficult to predict, as it depends on quality(Wiper contact intermittent) of pot.

I have not experienced crackling by this omission. However, there seems to be some experienced people.

[FSFX]

If you are concerned about the wiper then you should be using better quality potentiometers. This is a 'set and forget' type of output volume control rather than a 'wah-wah waggle' type of potentiometer and it is carrying very little DC current.

In fact, the suggestion of mine to  drop C10 and R12 will actually improve the DC bias point on the emitter of Q2 if it is a 2N4401 being used. With C10 and R12 in place then the emitter voltage is about 2.5 volts whereas without them  (and assuming no significant offset voltages from the op amps) then the emitter voltage will be closer to 3.5 volts.

If you used a high gain transistor like a 2N5089 for Q2 then this difference in voltage is smaller with 3.8 volts without C10 and R12 and 3.6 volts with C10 and R12.

Why, you may ask?. Well it is all down to the hFE of Q2 and the base current that it is drawing through R12 which is 510k, whereas the 100k pot R21 provides a lower resistance path to Q2 base from VR. The lower base current with the higher hFE transistor also doesn't load the VR supply as much and therefore increases its value to other parts of the circuit. 

Oh, and BTW, removing C10 and R12 has minimal effect on the frequency response. All it does is increase the level at 10Hz by 0.5dB. Elsewhere throughout the audio range it is too small to measure.