OK, here it is with a few changes. & finally got it working in the sim! Now on to the breadboard!
Effective OP-AMP Input Impedance, JFET vs. BiPolar
-
phatt
- Transistor Tuner
Matt,matt239 wrote:Ah! Thanks!
OK, here it is with a few changes. & finally got it working in the sim! Now on to the breadboard!
Remember that witho t t e Buf er you loose out when it comes to S/n
Now unless I'm from lala land
You *DID* just comment on how much you Hate noisey circuits? 
Anyway,,On your drawing;
*If* R1 is big and R5 is big ,,, Then expect NOISE and lots of it.
(Maybe go back to page one, as it's already been said)
BTW, Can I suggest you Loose the diodes I doubt they do much.
(better minds may like to pass comment on using diodes)
R6 and 7 need only be 10k. I work on the 1k per volt rule of thumb for this stuff.
So a 10 Volt supply use 5k dividers. I use 10k ,, 0ver 20 volts use 22k divider.
(I think if these are BIG then the bias is no longer stiff and that may cause issues,,, again need better minds to comment.
C3 will kill a fair bit of hi freq ,,, which will impact the result but that depends greatly on what you want from the circuit and where it is in the signal chain of events.
As too R5 (The 1 meg) well that has been done to death,,,
![[smilie=18_13_1.gif]](./images/smilies/18_13_1.gif "18_13_1"){kind=small}
With no buffer you may need to keep it fairly large ,, but if you used a buffer it can be MUCH lower without any loss. Hint,,(An I'm not going to tell you that anymore,, you will just have to learn that one the hard way,
Sims for all their flaws give a clue as to what is happening inside but the most common mistake is to assume that the mojo comes from ONE only circuit.
Hence after reading the writeups on the latest TS9999999 model stomper the poor kids assume the new one is better.
In reality you need to sim the whole darn circuit path as tone shaping happens ALL the way through the amplifier stages. a stomp box can aid the final result but if the rest of your system has tone shaping problems then it may prove futile no matter what Mojo pedal you try.
I used Sims to develop an understanding of the tone shape changes as the signal passes from each stage and how tweaking of these parameters can make big inroads to the final outcome.,, and also handy when trying to work out which part of the circuit is distorting.
Oh and Proto Everything,, preferably at concert level.
Testing on headphones late at night might get your circuit tweaked faster but as they say; we'll see what happens on Race Day.
Have fun,, Phil.
-
matt239
- Solder Soldier
Yeah. It's still not the FINAL version..
My values for R6,R7 weren't carefully chosen, I was still playing with that. I had thought that since we don't need much bias current we could save the battery a bit using larger resistors, & they wouldn't really add noise because Vr is held @ AC ground by C2... but I wasn't too sure about this... I see many folks here use 10k.. 250k probably doesn't allow enough current, & might not keep the voltage stable, I'm sure I should make them smaller..
D1,D2 are intended not to do anything under normal conditions. There's been differing opinions on the need to protect op-amp inputs in this thread, but if you read R.G. Keen's "when good op-amps go bad." etc. apparently there is at least some possibility of damage from spikes etc. The op-amp may not fail, but suffer gradually degrading performance. Protecting them this way is somewhat common practice.
I figure even if such damage is rare/unlikely, diodes are cheap insurance. Who knows what might get plugged in to a pedal input? Of course I could be wrong to worry about this.. It's just more stuff I'm trying to figure out..
R1 isn't that big, it's 5.1k, R2 is small, 1k.. - R5 yeah, done to death.. Some say; use lower resistor value, don't sweat some loading, some say; keep input impedance high, don't sweat some noise..
Some folks say use 2M, 4M, 10M! some say use 100k. Seems like it's a compromise, I figure with two 1M resistors I present a total impedance of 500k, which will then be paralleled with guitar controls..
Seems like a decent compromise.. I reserve the option to change my mind later..
Someone here said; (paraphrasing) If you've chosen to use an op-amp for buffer, & got the input impedance you want, then it's just as good, possibly less noisy, to go ahead and have it have some gain.
Is this wrong?
If I use an op-amp stage just as a buffer won't it also have to have a "large" input/bias resistor? -So the next stage can use a smaller resistor to ground on input, but we've already used a big one in the circuit? I could be all wet here.. Maybe you can explain it to me a bit more?
Thanks again for your, & everyone's help!
My values for R6,R7 weren't carefully chosen, I was still playing with that. I had thought that since we don't need much bias current we could save the battery a bit using larger resistors, & they wouldn't really add noise because Vr is held @ AC ground by C2... but I wasn't too sure about this... I see many folks here use 10k.. 250k probably doesn't allow enough current, & might not keep the voltage stable, I'm sure I should make them smaller..
D1,D2 are intended not to do anything under normal conditions. There's been differing opinions on the need to protect op-amp inputs in this thread, but if you read R.G. Keen's "when good op-amps go bad." etc. apparently there is at least some possibility of damage from spikes etc. The op-amp may not fail, but suffer gradually degrading performance. Protecting them this way is somewhat common practice.
I figure even if such damage is rare/unlikely, diodes are cheap insurance. Who knows what might get plugged in to a pedal input? Of course I could be wrong to worry about this.. It's just more stuff I'm trying to figure out..
R1 isn't that big, it's 5.1k, R2 is small, 1k.. - R5 yeah, done to death..
Some say; use lower resistor value, don't sweat some loading, some say; keep input impedance high, don't sweat some noise..Some folks say use 2M, 4M, 10M! some say use 100k. Seems like it's a compromise, I figure with two 1M resistors I present a total impedance of 500k, which will then be paralleled with guitar controls..
Seems like a decent compromise.. I reserve the option to change my mind later..
Someone here said; (paraphrasing) If you've chosen to use an op-amp for buffer, & got the input impedance you want, then it's just as good, possibly less noisy, to go ahead and have it have some gain.
Is this wrong?
If I use an op-amp stage just as a buffer won't it also have to have a "large" input/bias resistor? -So the next stage can use a smaller resistor to ground on input, but we've already used a big one in the circuit? I could be all wet here.. Maybe you can explain it to me a bit more?
Thanks again for your, & everyone's help!
No, you are right. A buffer is just another non-inverting amplifier. If you had a buffer and another amplifier afterwards, then you have at least 3dB more noise than with the amplifier alone. Common sense.matt239 wrote: Someone here said; (paraphrasing) If you've chosen to use an op-amp for buffer, & got the input impedance you want, then it's just as good, possibly less noisy, to go ahead and have it have some gain.
Is this wrong?
-
phatt
- Transistor Tuner
No offence to *Merlin*
But I've got a whole bin full of common sense circuits that drove me nuts until I started researching the less common stuff. winky.
I don't have the edumukation to ezplain it all but my advice to folks like *matt* is build it both ways (Breadboard testing not those simulations)
and test it for yourselves.
One will certainly make a LOT more noise than the other. LOL.
Cheers Phil.
But I've got a whole bin full of common sense circuits that drove me nuts until I started researching the less common stuff. winky.
I don't have the edumukation to ezplain it all but my advice to folks like *matt* is build it both ways (Breadboard testing not those simulations)
and test it for yourselves.
One will certainly make a LOT more noise than the other. LOL.
Cheers Phil.
-
matt239
- Solder Soldier
Well, can a unity gain buffer present a high input impedance w/o using a high value input resistor?
I may be able to be convinced that an input buffer will give us better noise performance, but I really would like to know WHY??
NEXT QUESTION:
Is there any reason not to use even smaller value resistors in the feedback loop?
Say, 1k for "R1," 250R for "R2?"
Similar gain, much less resistor noise. Is there a downside?
I may be able to be convinced that an input buffer will give us better noise performance, but I really would like to know WHY??
NEXT QUESTION:
Is there any reason not to use even smaller value resistors in the feedback loop?
Say, 1k for "R1," 250R for "R2?"
Similar gain, much less resistor noise. Is there a downside?
-
phatt
- Transistor Tuner
As to WHY???,, well yep that is something that I'm still not able to nail down myself....matt239 wrote:Well, can a unity gain buffer present a high input impedance w/o using a high value input resistor?
I may be able to be convinced that an input buffer will give us better noise performance, but I really would like to know WHY??
NEXT QUESTION:
Is there any reason not to use even smaller value resistors in the feedback loop?
Say, 1k for "R1," 250R for "R2?"
Similar gain, much less resistor noise. Is there a downside?
But hey do we all have to know why????
Heck for me chum ,,, It works!!!! and I have no noise issues with my gear,,, I have more sustain than I'll ever need,,,,I have no treble loss,,, and I'm loud as hell if I need it.
I have some chaps coming over this W/end just to hear my system because they heard one of my setups the other day and could not believe what they heard. And NO I'm not bragging ,, I'm sharing
I have no problems with input imp.
Matt the problem (Answer you seek) is likely to do with the **Internal Z** of the active device that is used.
As one of Australia's top pedal builders commented a while back *the average opamp input Z* even with a 10 meg resistor to ground may only deliver a 300k Z at the input.
Sadly this stuff I also struggle with as it is very complex Maths stuff.
In my world,, I don't have enough time left in one lifetime to understand the intricately complex.
So I tend to accept the limitations of these things and work with it.
Some good reading can be found in books like *Art of Electronics by Horowitz an Hill* but those books are not cheap.
OR,,, just read Five seven's comments on your other posting for clues;
https://www.freestompboxes.org/viewtopic ... 22&t=15305
cheers, Phil.
-
matt239
- Solder Soldier
NEXT, NEXT QUESTION:NEXT QUESTION:
Is there any reason not to use even smaller value resistors in the feedback loop?
Say, 1k for "R1," 250R for "R2?"
Similar gain, much less resistor noise. Is there a downside?
What is a good final noise figure??
I've been using some sims, & calculators, but I don't really know what to make of the results; how many dB is a micro-Volt of noise?
What is a good SNR?
By the way "PHATT" I did really like the sound of the clips you posted, they sound great!
-
phatt
- Transistor Tuner
Hi Matt sorry if I'm late,,,
Look you really need to just proto a few of these circuits and you will soon learn that no matter how good the sims might look the result is not always as clever as the software may suggest.
Be warned;;; the best laid plans of Men and Mice often go haywire.
All the most expensive simulations on the planet could not accurately identify the destruction of the space shuttle upon re-entry in less than 60 seconds.
(if Ya catch my drift,, winky)
In your case no one is going to die if you get it wrong
So go build stuff,,, THAT is the best way to learn it.
Go plug the best design Hi Gain pedal into a crap Amp and you will soon learn that good gain into really bad gain is the same result. At some point you run out of options,,, if you want tons of gain you have to loose something somewhere in the signal path otherwise it all turns into crud.
Some folks say max Z on each input for best transfer,,,
Now go look at the grid R value of say the Carvin Legacy Valve Rig,,, (It's only 220k)
Yet that Amp has treble that can rip your ears off.
ANSWER,, The Amp has so much gain inside they had to turn it down somewhere otherwise it's all noise if you crank it up.
The PI grid R further into that circuit is also only 100k (not the normal 1Meg !!!!) Yet it's as loud as hell and suffers no Bass loss. Those Amps where built specifically for *STEVE VAI* I'm not about to write and tell him his power transfer is way off. 
Aussie legend,,*Phil Emmanual* (The brother of Phil Emmanual) also uses a Carvin Legacy.
Phil E is a local in my town and I get to hear him play his *Carvin Lagacy* often.
So I know darn well it wants for nothing,,, trust me it makes a Fender Dudville sound like a toy wannbe.![[smilie=pope.gif]](./images/smilies/pope.gif "pope")
Yes sadly sims only help if you know what the hell you are looking to find.
As to your Q,, I have no doubt that better Q'ed folk here know the Answer to that but it won't help much unless you have built enough to know what the answer is worth. (What it sounds like in a real working circuit)
I'm far from the expert but the few repairs I do to Amps and gear is often just a simple matter of turning gains down which mostly results in,, Wow that's much better from the clients. HINT.
A lot of gear sold today is so bad that SNR is likely not even measurable over the noise produced.
Kids today must have tin ears. (I shake my head at some gear in shops today)
Phil.
Look you really need to just proto a few of these circuits and you will soon learn that no matter how good the sims might look the result is not always as clever as the software may suggest.
Be warned;;; the best laid plans of Men and Mice often go haywire.
All the most expensive simulations on the planet could not accurately identify the destruction of the space shuttle upon re-entry in less than 60 seconds.
(if Ya catch my drift,, winky)
In your case no one is going to die if you get it wrong
So go build stuff,,, THAT is the best way to learn it.
Go plug the best design Hi Gain pedal into a crap Amp and you will soon learn that good gain into really bad gain is the same result. At some point you run out of options,,, if you want tons of gain you have to loose something somewhere in the signal path otherwise it all turns into crud.
Some folks say max Z on each input for best transfer,,,
Now go look at the grid R value of say the Carvin Legacy Valve Rig,,, (It's only 220k)
Yet that Amp has treble that can rip your ears off.
ANSWER,, The Amp has so much gain inside they had to turn it down somewhere otherwise it's all noise if you crank it up.
The PI grid R further into that circuit is also only 100k (not the normal 1Meg !!!!) Yet it's as loud as hell and suffers no Bass loss. Those Amps where built specifically for *STEVE VAI*
I'm not about to write and tell him his power transfer is way off. Aussie legend,,*Phil Emmanual* (The brother of Phil Emmanual) also uses a Carvin Legacy.
Phil E is a local in my town and I get to hear him play his *Carvin Lagacy* often.
So I know darn well it wants for nothing,,, trust me it makes a Fender Dudville sound like a toy wannbe.
Yes sadly sims only help if you know what the hell you are looking to find.
As to your Q,, I have no doubt that better Q'ed folk here know the Answer to that but it won't help much unless you have built enough to know what the answer is worth. (What it sounds like in a real working circuit)
I'm far from the expert but the few repairs I do to Amps and gear is often just a simple matter of turning gains down which mostly results in,, Wow that's much better from the clients. HINT.
A lot of gear sold today is so bad that SNR is likely not even measurable over the noise produced.
Kids today must have tin ears. (I shake my head at some gear in shops today)
Phil.