I sort of agree, except that there can be a significant variation due to component tolerances anyway. It seems odd to put that much effort into making sure the bias doesn't drift with temperature, if you don't know whether it was right to start with.
Something that keep's bugging me about this circuit is that there's no way to adjust the bias. I'm not familiar with these kind of output devices but the datasheet does indicate that the "turn-on" voltage can vary over a wide range. Is that not important?
I'm wondering if Goldmund doesn't have a way of setting the bias that's not shown on the schematic. e.g. R23 is shown as 330R, but maybe that's just a "typical" value, rather than cast in stone.
Scratch that, a temperature analysis would be useless since the FETs are located away from the CCS. There's no way we can expect the CCS to compensate the the FETs unless they are mechanically coupled.
With this, my opinion is that the "emitter-down" version is best, because of the lowest tempco and higher output impedance. As far as noise, we may never be able to tell.
- keantoken
With this, my opinion is that the "emitter-down" version is best, because of the lowest tempco and higher output impedance. As far as noise, we may never be able to tell.
- keantoken
I was wondering about that too. Maybe it just never got noticed. Doug Self mentioned in another thread that he once accidentally added an unnecessary (but harmless) resistor in a circuit. The cloners promptly started churning out exact copies of his amp - including the "mistake" resistor.
I'm not sure how far to trust simulators wrt upside down transistors. Is there not some simple test you could do on an actual device to check real-world behavior.
Sorry, my bad - it's R20 not R23.
They could well be avoiding trimmers for reliability reasons (this is expensive stuff, remember - it's not supposed to break).
They might use one in the factory though, to set the bias, and then replace it with a fixed resistor of the same value.
Shall I stir the pot a little more by wondering if there's any component selection involved in the input stage. (I believe jfet characteristics can be quite variable too.....)
Where's a "whistling innocently" smiley when you need one?
BTW, can someone tell me a little bit about the "fast" sound? I heard a sound I would describe as "fast" from my headphone amp, but I later discovered this was because of HF distortion coming from being overcompensated, which causes HF distortion when the frontend sees a capacitive load. I got the compensation to the right value, and the soundstage opened up, and a lot of harshness was eliminated. Thing is, my headphone amp is VERY fast. Gain-bandwidth product is 13MHz. This situation theoretically is not much different than in such an FET amp, where the FET capacitance effects the frontend. Anyone else have experience here?
- keantoken
- keantoken
Stepping back for just a second or two...
It is truly fascinating to see over the course of a few days the tremendous talent, intelligence, ingenuity and somewhat friendly give and take over what constitutes the "correct" topology of the constant current source in a circuit.
This forum, and the internet have in its own way brought together men and women from truly around the world to share and help each other learn. Having been involved in electronics for a "few" years it is amazing in more ways than one. So lets get back to the task.
Carry on LADIES and GENTLEMEN... THANKS
It is truly fascinating to see over the course of a few days the tremendous talent, intelligence, ingenuity and somewhat friendly give and take over what constitutes the "correct" topology of the constant current source in a circuit.
This forum, and the internet have in its own way brought together men and women from truly around the world to share and help each other learn. Having been involved in electronics for a "few" years it is amazing in more ways than one. So lets get back to the task.
Carry on LADIES and GENTLEMEN... THANKS
Hi Alex,
In Post#549 the pcb matches the schematic which does not match the schematic in Post#490. But the one in Post#490 does match that of the hand-drawn schematic in Post#1. So...... I am getting confused.
Other guys:
quote for PCB
new size : 205x88mm
Single side : $14 / pair
Double side : $22 / pair
I am ready. I still need the files from Alex (well, the final ones), and Alex, that won't be a problem right? . If we go for double side pcb for the main board then the daughter board will have to be made by another order.
Your can send email to tony_chan_cf@email.com, stating your forum ID, email address (which I will not post on the status report) and quantity (in pair). I will post the status report every day.
In Post#549 the pcb matches the schematic which does not match the schematic in Post#490. But the one in Post#490 does match that of the hand-drawn schematic in Post#1. So...... I am getting confused.
Other guys:
quote for PCB
new size : 205x88mm
Single side : $14 / pair
Double side : $22 / pair
I am ready. I still need the files from Alex (well, the final ones), and Alex, that won't be a problem right? . If we go for double side pcb for the main board then the daughter board will have to be made by another order.
Your can send email to tony_chan_cf@email.com, stating your forum ID, email address (which I will not post on the status report) and quantity (in pair). I will post the status report every day.
THIS is the reason I suggested everyone "slow down". A good precise DIY project will at least take a couple weeks to double check , even triple check. Without drafting software, I still have yet to have a lead spacing or electrical error. This is not as easy as some may think.
OS
Last edited:
My thought exactly , with just a slight change to , the CCS , for example .. IP considerations would be solved. Goldmund might not be too happy with it's present form. Would be the same as presenting a "NASKA" for DIY purposes.
I clone things , but always incorporate just enough change to "make it mine".
Just something to consider.
OS
That value moves with the position of the cursor on the screen, so you can point at an area of interest and measure the x and y coordinates.
It has nothing to do with -3dB point.