Member
Joined 2009
Paid Member
I don't think you understood again.
My point is that if you are striving for miniscule dc offset you have to pay attention to the LTP balance and this balance in some designs is affected by a change in the tail current. A CCS helps stabilize the tail current. I have never used a CCS in any design I've built, not sure why, just don't like 'em for some reason. I have always used a resistor. A resistor does OK, but if you change rail voltage you change LTP current and this can affect dc offset.
I don't intentionally unbalance any LTPs, but I also don't care too much if they aren't perfectly balanced since I don't care too much about dc offset being better than 50mV.
I guess, this is very strong dependend of the gain mode; both non inverted and inverted mode is possible, in last case the influence of the input capacitance variation is lower, because the right half works together with the left half now as emitter coupled cascode and not like a differential amp.
But what happens, if I compare non inverted gain mode with J-FET's vs. such, where are BjT's instead jFET's in use, but with additional jFET single buffer (source follower) stage ?
Last edited:
If it is unbalanced, it is unbalanced. If it is balanced, and you raise the tail current, it should still be balanced, i.e., the tail current will be equally shared.
Just speaking off the top of my head with no calculations attempted, perhaps what you are seeing it that by raising the tail current in an unbalanced stage, the currents go up in the stage differently, and the higher the current, the "more different" they are.
Really?
Gareth, it has nothing to do with "minuscule" amounts - the LTP balance will not be grossly affected by small fluctuations in tail current. A CCS maintains constant current, even if the supply drops - yes. A resistor maintains constant current if the supply voltage remains constant. If it drops, so does the current, but it's through both legs of the LTP - balanced. How much do you expect your supply voltage to drop by anyway? 2 or 3 volts? That would cause how much deviation in the LTP current? Something on the order of .2mA sound reasonable?
Why argue?
Member
Joined 2009
Paid Member
I agree, if the LTP is closely balanced then changes in the tail current is not a real concern but I am thinking of porting an amp without a CCS from one power supply to another with different rail voltages so it's something I need to check. I don't match devices in my amps, simply too lazy and don't care about matching 
Absolute balance in a LPT is an almost impossible task even if you match the transistors unless you draw the exact same currents from the collectors, by another differential pair that is matched. But if you are only drawing from one of the collectors the LPT will most certainly be unbalanced. Some designers artificially balance the current by placing higher imbalanced loads on the collectors. See KRELL and ESP.
Balancing with an external pot feeding each of the pairs is only forcing more current through one than the other so they are not in equilibrium anyway. As Gareth rightly comment was is the point to balance them, just a waste of time and components, unless someone can swear that the sound is markedly superior.
Balancing with an external pot feeding each of the pairs is only forcing more current through one than the other so they are not in equilibrium anyway. As Gareth rightly comment was is the point to balance them, just a waste of time and components, unless someone can swear that the sound is markedly superior.
Do you really think that would be a problem, around here.
I don't get my shorts in a bunch if I see 10mV or so of offset. 50mv or more would throw up a red flag though. My experience (and it should be known, my amps don't have a CCS on the LTP) is that it's not difficult to get the offset down to 10mV or less and that is without the use of a trimmer.
As far as not using caps in the feedback path or on input or wherever - to each his own but I think this 'avoid caps or other baddies in the signal path' is just bunk. Expanding the signal path to include the power supply is bunk too. Of course, this is my opinion, totally unsolicited and (I'm sure) unappreciated.
Build an amp that works correctly, has reasonably low distortion and has a decent power supply, and it will sound just like the other one that works correctly, has reasonably low distortion and has a decent power supply. Build one that does silly things like distort the signal and amplify DC and is on the brink of oscillation due to deficient compensation and of course it will sound different. It no longer an amplifier in the strictest sense of the work - it's an equalizer.
Last edited:
Member
Joined 2009
Paid Member
I don't want to start an argument of any sort, but as stated by one esteemed member, if two different amps built "correctly" sound the same, then 'n' number of amps, based on different topologies, parts and layouts but built "correctly" should 'subjectively' sound the same. Then we at diyaudio are on to quite a vain pursuit and not very different from "Sisyphus on a roll".
For the record, I clearly hear sonic differences, as and when they are introduced by 'whatever' and can demonstrate it to any common man in the street. Of the many hundreds with who I have auditioned both commercial and diy equipment in the past quarter century, I have so far come across only one gentleman who could not perceive even large sonic differences. He confessed that his hearing ability had been severely affected by occupational hazards, he having worked in the power-print and tea-blending industries, being subject to continuous high db ambient noise.
For the record, I clearly hear sonic differences, as and when they are introduced by 'whatever' and can demonstrate it to any common man in the street. Of the many hundreds with who I have auditioned both commercial and diy equipment in the past quarter century, I have so far come across only one gentleman who could not perceive even large sonic differences. He confessed that his hearing ability had been severely affected by occupational hazards, he having worked in the power-print and tea-blending industries, being subject to continuous high db ambient noise.
Were you listening to amplifiers or equalizers? Are the amplifiers themselves making the 'sonic' differences or is it their interaction with other components? Is mood a factor? Time of day?
If there are "large sonic differences" from one 'amp' to the next, they are processing the signal differently. How else would they sound different? Which one is reproducing the signal more accurately (AKA closer to what an amplifier is supposed to be).
I have heard many amps, built many. I hear no differences between competently made amplifiers - that is my honest assessment. It has absolutely nothing to do with hearing acuity or lack thereof.
If there are "large sonic differences" from one 'amp' to the next, they are processing the signal differently. How else would they sound different? Which one is reproducing the signal more accurately (AKA closer to what an amplifier is supposed to be).
I have heard many amps, built many. I hear no differences between competently made amplifiers - that is my honest assessment. It has absolutely nothing to do with hearing acuity or lack thereof.
Member
Joined 2009
Paid Member
My esteemed member, the kind of difference I have heard are sometimes heard only on specific music types, other differences can be more of an overall 'feel' to the sound that I notice more when I 'downgrade' to a lesser amp than when I first listen to the new one.
For example, one of the first high end amps I ever heard was a KRELL playing through JBLs. I noticed some glare on violins. At the time I didn't know what to call it, but I heard something that wasn't right and only from talking to others did I learn the common name for it.
It maybe that some people are more sensitive to these things - a mixed blessing perhaps ?
For example, one of the first high end amps I ever heard was a KRELL playing through JBLs. I noticed some glare on violins. At the time I didn't know what to call it, but I heard something that wasn't right and only from talking to others did I learn the common name for it.
It maybe that some people are more sensitive to these things - a mixed blessing perhaps ?
Same here.
But how safe is to go without a LP and/or HP input filter?
And btw, back to original topic, how high can be made the impedance with an all-BJT input stage?
I voiced my opinion, which I feel is my right. It is based on MY experiences and observations and I'm not trying to convince anyone one way or the other. It may be a refreshing change to hear someone affirm what many quietly believe though.
If you feel you hear a difference and if that makes you happy, more power to you.
If you feel you hear a difference and if that makes you happy, more power to you.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.