[eruditass]

so this cmoy is to be used after my e-mu 0404's lineout (which i assume as no DC offset, but i probably should measure it) until it is replaced my hopefully a soha II. Also it will be used after my sansa clip's Headphone-Out (.5 mV on left channel with nothing playing). Gain will likely be 1 or 2

with my first cmoy, I used 1uF metal film caps from ratshack. I'm not quite sure if I would have been able to fit it in a altoid tin. after measuring the dc offset, well, why not forgo the caps?

With removing the caps, Is there a point to the R2 that was used to create the highpass filter? Will this just be sucking current for no reason? Can I remove it?

As for the potentiometer, I was again going to get this stereo log taper 100K pot (the only stereo one) at ratshack. However, I was debating the quality of this pot and the noise it may add. For this reason, I was considering forgoing the pot and just using the 0404's pot on the lineout that goes from -infinity to +0. With a gain of 1 on my old cmoy, it played plenty loud for my dt880's.

However, when pairing it with the clip, which is a better tradeoff? ratshack's pot or digital volume control? I could also leave the pot in and use it minimally.

Next, the power supply. I managed to find 1000uF electrolyte caps I could fit on my board, surprisingly. So with all that current stored up, I was trying to think of how to make the virtual ground a little better without an IC. although they would pair very well with the TLE 2426 with all that juice, i dont know if i want to throw a whole IC on there with space restrictions (I am going to use 2 9V batteries). So I'm looking at the resistor dividers. With the current stored up, can I get away with throwing larger resistors than the 4.7K and not have so much uneven virtual ground splits? Or are these caps not a factor at all, its simply based on the current across the resistors that divide it (i probably need to re-read that section). And is it not even a problem with 18V's of headroom? The section only mentions that resistor dividers/unbalanced ground affects the battery life, but I have heard others say it effects the SQ, is it true? i don't care too much about battery life, should i just throw in low value resistors?

lastly, opamps. I don't want to go through the hassle of 2 single channel opamps, so I need a dual channel on a single IC. I know a lot of the HQ ones are single channel. I would like to know what the top ones are without TOO much work to get them stable. I don't mind adding in a cap between the rails or anywhere else or some other discrete components, just more no IC's. What I'm looking at:


NJM 4556
AD8066
AD823
AD8620
ad8610
TLE2142
TLE2082
LM6172
AD8397
OPA2107
LM6172
ne5532
TLE2072
OPA2111

not sure if these work in a cmoy?
AD797BRZ, AD8022, AD8021, LME49720, AD8599

opinions on those or any other options? looking at 1 or 2 gain, 18V supply, will not have output buffer. Does max output current really matter if I use high impedance moderate sensitivity phones?

in other news, i am so far behind on amp building but i finally graduated and am looking for a job. and in the mean time i will build some amps. i ordered a couple things 1 year ago that never got built, and now i want to build something else.. like that soha II everyone's been raving about.

 

[rembrant]

RadioShack has a 10k dual log pot with SPST switch. At least the one near me does.

R2 is there to keep the pot from being loaded and effecting the input impedance of the amp while being moved.

If you do use the 100k pot you can remove R2 but you would need to measure the DC offset at all positions too make sure it still falls in an acceptable range.

If you find that your source has no DC offset then it is ok to remove the input caps. Just remember you shouldn't go plugging it in to sources you haven't measured.

About opamp rolling.
There are many opamps that will work in this circuit. I would reccomend that you add decoupling caps from V+ and V- PINS to ground. .1uf should do fine.(ceramic is generally known to be good for PS decoupling).
As far as the Opamp itself is concerned if the pinout is the same then from there you should check the specifications on the data sheet against Tangent's DC offset calculator to make sure you wont wind up with a high DC offset.
Electronics Calculators

Power supply questions.
As far as the TLE2426 is concerned. It is not really considered a high current device by any means. It is there to provide a more accurate voltage split and to reduce components. The resistor divider is OK if you have a fixed power supply. It tends to present a higher current draw to batteries which is why you wouldn't want to use a higher value resistor. For more accurate split tracking with increased current capability, you could buffer the TLE with a suitable opamp or a stand alone buffer.
The best you could probably do is to build a power supply for it. That would require it to be plugged into a wall socket at all times. I would say using a 470uf before the rail splitter and 220uf after on each rail to ground would be a good compromise.

 

[eruditass]

Quote:

Originally Posted by rembrant /img/forum/go_quote.gif RadioShack has a 10k dual log pot with SPST switch. At least the one near me does. R2 is there to keep the pot from being loaded and effecting the input impedance of the amp while being moved. If you do use the 100k pot you can remove R2 but you would need to measure the DC offset at all positions too make sure it still falls in an acceptable range. If you find that your source has no DC offset then it is ok to remove the input caps. Just remember you shouldn't go plugging it in to sources you haven't measured.

I just skimmed the section on cranky ampops: Balancing the Input Bias Currents

How much does this affect the operation? Tangent's schematic denotes 1K and 10K so R2 should be .9K.. yet it is 100K. depends on the opamp? I will only be using it in the clip and 0404 so I already will be removing the input caps. How will this affect the balancing? Since I may forgo the pot completely, will just throwing in the calculated parallel resistance of R3 and R4 for R2 work? Or because I am removing the cap, should I move the R2 to the input cap position?

I am still trying to understand the input loading section. So by decreasing R2, we are causing more current to go through the pot to ground and then back up R2?? What is wrong with this? Can someone clarify it for me?

"The general rule of thumb in this piece of the electronics world is that a source should have an impedance at least 10× lower than its load. "

Ok I understand this from school and impedance matching/power transfer etc, but still am confused as to how it applies.

By source are we talking output impedance of the DAC or whatever is driving it? 600ohms was common, so we would like 6k but actually gave it 10k pot. With the input filter/cap do I essentially put 100K resistor in parallel giving it a load impedance of 9K?

Even if I cut R2 in half, that input impedance of 10K pot parallel with 50K is 8K, which is > 6K based off of 600 ohm output impedance, and my 0404 is actually 150 ohms, while im sure the sansa clip's output impedance is lower than that. So will it be a problem?

I just move R2 to the input cap position that I am removing, wouldnt the impedance be very high since it is going through the opamp input which has a high impedance input?

I am not sure if I am looking at it right, since you say 100K pot will make R2 negligable since 100K is >>>> 6K, but tangent says the opposite. If you lower R2, you need to lower the pot?? 5K pot \\ 50K R2 = 4K impedance, which is less than 6K.

I am confused again. Which is the worst case we are considering these calculations for? the pot's full resistance parallel or in line with the signal? Maybe I should go back to school heh.

Lastly, what exactly do you mean measure the DC offset at all positions?

Quote:

About opamp rolling. There are many opamps that will work in this circuit. I would reccomend that you add decoupling caps from V+ and V- PINS to ground. .1uf should do fine.(ceramic is generally known to be good for PS decoupling). As far as the Opamp itself is concerned if the pinout is the same then from there you should check the specifications on the data sheet against Tangent's DC offset calculator to make sure you wont wind up with a high DC offset. Electronics Calculators Power supply questions. As far as the TLE2426 is concerned. It is not really considered a high current device by any means. It is there to provide a more accurate voltage split and to reduce components. The resistor divider is OK if you have a fixed power supply. It tends to present a higher current draw to batteries which is why you wouldn't want to use a higher value resistor. For more accurate split tracking with increased current capability, you could buffer the TLE with a suitable opamp or a stand alone buffer. The best you could probably do is to build a power supply for it. That would require it to be plugged into a wall socket at all times. I would say using a 470uf before the rail splitter and 220uf after on each rail to ground would be a good compromise.

So I should add in another V+ and V- in addition to the big ones in the power supply? I will have 1000uF's between V+ and ground, V- and ground, which comes out to 500uF's between V+ and V- which is not enough? well I will do it anyway right by the opamp because there is never too many caps between + and - until you run out of space.

As for the rail splitter, I have easy access to caps and metal film resistors so if there are no sonic advantages and only board real estate and battery life, then I will forgo it.

Quote:

Originally Posted by syllabus /img/forum/go_quote.gif Try 2x AD845K in your CMOY. They'll work well from 12V upwards.

Do not want to use 2 IC's for this particular amp. Will go to bigger and better amps afterwards instead.

 

[rembrant]

Quote:

Originally Posted by DoomzDayz /img/forum/go_quote.gif I just skimmed the section on cranky ampops: Balancing the Input Bias Currents

Problem number 1.
Go back and read it and understand it. Designing any amp is about compromise.
Voltage offset and input bias currents will determine the design from the ground up. All the answers are there. Consult the electronics calculators for a greater understanding of how theses values will apply to choosing resistors.

Feel free to use whatever rail split topology you wish. They all have strengths and weaknesses. Tangent explains this better than I ever could in..... Virtual Ground Circuits