lumanauw said:
I'd also like to know.
Yes, a resistor makes designing the circuit fairly simple, but also a current source is nice because the current in the amp stays almost the same, even with different supply voltages.
I was considering using a current source for both the diff pair, and the VAS load for my sub amp I'm building in the future, but what do you mean by "better sound"? Clearer treble, tighter bass, or what?
THE BEST CURRENT SOURCE......
hi guys,
for me this is the best of best current source ever made be man...........
pls. check on the link below.
audio quality? this is also my best choice.....
I think Japanese are also some of the best amp builders in the world!
http://www.eserviceinfo.com/downloadsm/5170/rotel_rb1090.html
hienrich
hi guys,
for me this is the best of best current source ever made be man...........
pls. check on the link below.
audio quality? this is also my best choice.....
I think Japanese are also some of the best amp builders in the world!
http://www.eserviceinfo.com/downloadsm/5170/rotel_rb1090.html
hienrich
I cannot answer exactly why. I just came up with subjective opinion, dont have objective data.
The simple R ccs gives sound more complete sonics than if using ccs. But I remind you, I found this in low biased class AB (cold amp). Could be different in classA or heavy biased classAB.
Some people like AKSA and Grey Rollins have experimented with this also. Maybe they can came up with objective facts?
If you talk about measurement, I'm quite sure, the semiconductor CCS will gives better result. But it sometimes, what is measured is oposition to what the sonics/your ear tells you. From technical point of view, yes, simple R ccs is bad. Very dependant on voltage rail fluctuation/not constant current there is. You can try it, very simple, just calculate the appropriate R size and watt rating, and intercange them with semiconductor ccs.
The simple R ccs gives sound more complete sonics than if using ccs. But I remind you, I found this in low biased class AB (cold amp). Could be different in classA or heavy biased classAB.
Some people like AKSA and Grey Rollins have experimented with this also. Maybe they can came up with objective facts?
If you talk about measurement, I'm quite sure, the semiconductor CCS will gives better result. But it sometimes, what is measured is oposition to what the sonics/your ear tells you. From technical point of view, yes, simple R ccs is bad. Very dependant on voltage rail fluctuation/not constant current there is. You can try it, very simple, just calculate the appropriate R size and watt rating, and intercange them with semiconductor ccs.
Resistive feed to a LTP front end will generate a signal modulated current through the LTP, leading to assymetrical distortion.
This distortion is euphonic, low order, mostly H2, like a tube.
A very good CCS has a LED as voltage reference fed from a bleed resistor to ground (LED is green, about 2V), and a 2K2 resistor leading from the voltage reference to the CCS transistor base. The green LED is cheap, very quiet, and shares the same tempco as the transistor, keeping current constant with temperature changes. The transistor base is then decoupled with a large electro to the rail. This is as good as it gets, though some will say cascoding the CCS device with another like device is superior again.
Cheers,
Hugh
This distortion is euphonic, low order, mostly H2, like a tube.
A very good CCS has a LED as voltage reference fed from a bleed resistor to ground (LED is green, about 2V), and a 2K2 resistor leading from the voltage reference to the CCS transistor base. The green LED is cheap, very quiet, and shares the same tempco as the transistor, keeping current constant with temperature changes. The transistor base is then decoupled with a large electro to the rail. This is as good as it gets, though some will say cascoding the CCS device with another like device is superior again.
Cheers,
Hugh
Yugo,
Now, surely a resistive feed to a long tailed pair ain't so complex it needs a schematic!!
But there is always more to this that meets the eye, and unfortunately I'm not prepared to place my company intellectual property into the public domain. After all, I have built a business on my work, and seek to protect it.
Or do you perhaps mean the schematic for the complete amplifier?
If this lack of hard evidence concerns you, you could always dismiss it. OTOH, if you are genuinely interested, email me privately.
Cheers,
Hugh
Now, surely a resistive feed to a long tailed pair ain't so complex it needs a schematic!!
But there is always more to this that meets the eye, and unfortunately I'm not prepared to place my company intellectual property into the public domain. After all, I have built a business on my work, and seek to protect it.
Or do you perhaps mean the schematic for the complete amplifier?
If this lack of hard evidence concerns you, you could always dismiss it. OTOH, if you are genuinely interested, email me privately.
Cheers,
Hugh
CCS
Hugh,
no of course I did not mean that!I did not wanted you to post the schematic of your excelent sounding amp,but simply to share your point of view regarding good sounding CCS which was BTW the title of my post.In other words to draw the schematic of that CCS(the best one for that purpose from your opinion) wheather is it with NPN or PNP transistors,which transistors are used,how to calculate for higher currents,(simple formulas)....Being not so good in SS electronics I was trying to find easier way for implementation. Hugh, I am very sorry if there was misunderstanding!
Regards,
Yugovitz
Hugh,
no of course I did not mean that!I did not wanted you to post the schematic of your excelent sounding amp,but simply to share your point of view regarding good sounding CCS which was BTW the title of my post.In other words to draw the schematic of that CCS(the best one for that purpose from your opinion) wheather is it with NPN or PNP transistors,which transistors are used,how to calculate for higher currents,(simple formulas)....Being not so good in SS electronics I was trying to find easier way for implementation. Hugh, I am very sorry if there was misunderstanding!
Regards,
Yugovitz
Green LED
Hugh,
Any particular reason to use Green ?
Any experiences or comments about red or yellow, or even blue ?
Red for example has a steeper current vs voltage curve. I have also used blue here and there when I need a bit more voltage, but my oscilloscope is not good enough to measure any differences in noise. So your experiences would be a real great help.
(If you do not want to discuss this openly, we could email each other. I have just enabled private message. Thanks.)
Patrick
Hugh,
Any particular reason to use Green ?
Any experiences or comments about red or yellow, or even blue ?
Red for example has a steeper current vs voltage curve. I have also used blue here and there when I need a bit more voltage, but my oscilloscope is not good enough to measure any differences in noise. So your experiences would be a real great help.
(If you do not want to discuss this openly, we could email each other. I have just enabled private message. Thanks.)
Patrick
Hi EUVL,
Red is typically 1.6V, a little low, as this tends to worsen noise performance because the emitter resistor is very small; noise on the collector depends heavily on the size of this resistor.
Blue is expensive, though much higher voltage, typically 2.2V or more.
Green is a very cheap LED, and coincidentally the tempco is almost identical with the PN junction of the CCS transistor. This is important to stability, and worth having.
Hope this is helpful,
Cheers,
Hugh
Red is typically 1.6V, a little low, as this tends to worsen noise performance because the emitter resistor is very small; noise on the collector depends heavily on the size of this resistor.
Blue is expensive, though much higher voltage, typically 2.2V or more.
Green is a very cheap LED, and coincidentally the tempco is almost identical with the PN junction of the CCS transistor. This is important to stability, and worth having.
Hope this is helpful,
Cheers,
Hugh
Hugh,
Thanks for the info.
I take it that you have no objections to blue other than costs.
The tempco is a good point but then only really valid if the transistor and the LED both have similar thermal resistance chip to ambient, AND see same amount of dissipation, and / or they are very well thermally coupled.
Patrick
Thanks for the info.
I take it that you have no objections to blue other than costs.
The tempco is a good point but then only really valid if the transistor and the LED both have similar thermal resistance chip to ambient, AND see same amount of dissipation, and / or they are very well thermally coupled.
Patrick
Bricolo,
Not too sure about the math; but I would have thought that the different doping levels for different colours would affect the tempco??
No objection to blue aside from cost. The more the ererence voltage, within the limiations of the circuit, the better is the noise performance (and the impedance charactireistics at the collector).
Cheers,
Hugh
Not too sure about the math; but I would have thought that the different doping levels for different colours would affect the tempco??
No objection to blue aside from cost. The more the ererence voltage, within the limiations of the circuit, the better is the noise performance (and the impedance charactireistics at the collector).
Cheers,
Hugh
Hugh, I'll try to check for the tempco. But zeners with different voltages have different tempcos, so you're certainly right.
But what about the emitter resistor? Why does a smaller resistor give more noise? You're not ralking about thermal noise, since this one increases with the resistance.
But a larger resistor would give a higher output impedance, since it is multiplied by the transistor's gain, if my maths are correct
But what about the emitter resistor? Why does a smaller resistor give more noise? You're not ralking about thermal noise, since this one increases with the resistance.
But a larger resistor would give a higher output impedance, since it is multiplied by the transistor's gain, if my maths are correct
Bonjour Bricolo,
I'm dancing on my toes today......
A CCS is in fact an open loop amplifier, with any AC voltage at the base with respect to the supply amplified at the collector.
A larger voltage reference gives us a large emitter resistor for the same output current. The gain of this common emitter configuration is a function of the collector impedance to the emitter impedance, and we know that the emitter resistor, in this case necessarily viewed as degeneration, has profound influence because it swamps re, which is generally given by 26/mA. If we use a 100R emitter resistor for a 10mA CCS, the emitter impedance (total) Re + re, is 102.6. If we increase the size of the voltage reference from say 1.6V (red) to 2V (green), then the emitter resistor increases from 100R to 140R, and Re + re increases to 142.6, commensurately decreasing open loop gain and hence noise.
The other advantage is impedance seen at the collector by the voltage amplifier beneath (or other active circuit block), which is often significant. Degeneration increases this impedance markedly, but more importantly, linearises it with frequency.
Hope this is helpful,
Cheers,
Hugh
I'm dancing on my toes today......
A CCS is in fact an open loop amplifier, with any AC voltage at the base with respect to the supply amplified at the collector.
A larger voltage reference gives us a large emitter resistor for the same output current. The gain of this common emitter configuration is a function of the collector impedance to the emitter impedance, and we know that the emitter resistor, in this case necessarily viewed as degeneration, has profound influence because it swamps re, which is generally given by 26/mA. If we use a 100R emitter resistor for a 10mA CCS, the emitter impedance (total) Re + re, is 102.6. If we increase the size of the voltage reference from say 1.6V (red) to 2V (green), then the emitter resistor increases from 100R to 140R, and Re + re increases to 142.6, commensurately decreasing open loop gain and hence noise.
The other advantage is impedance seen at the collector by the voltage amplifier beneath (or other active circuit block), which is often significant. Degeneration increases this impedance markedly, but more importantly, linearises it with frequency.
Hope this is helpful,
Cheers,
Hugh
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.