Update on CMoy using OPA637

[OhYeah 0]

When using OPA637, I realised that full input volume will result in distortion. Meaning to say, there must be a volume pot or a 10k resistor along the input before the high pass cap.

 

Wanted to save space by doing without the volume pot. Apparently it will not work for the OPA637.

[OhYeah 0]

[jasonhanjk 0]

Try removing R2 and see does it improves...

[Firefox 0]

Try removing R2 and see does it improves...

Removing R2 in his diagram would not be a good idea.

[tropicalrips 0]

i hav some high pass caps that causes distortion at higher volumes, they are MMK 1uf (white color)

the distortion is gone when i change them to .22uf Wima (red)

 

removing the R2 in the drawing above means no feedback rite?

[hacknet 0]

i`d wouldn`t recommend removing R2. R2 is the feedback resistor, without this the amp has no determined gain. it could go into osscilation.

 

i`d recon that the distortion that you get is not caused by the cap but by the opamp not being able to swing the output voltage, thus causing it to clip.

 

i could be wrong but it is very rare that a coupling capacitor causes noticable distortion especially if it is used within its means.

 

another note, those evox rifa MMK 1uf caps are cool cheap caps, i`ve used them to great sucess in many projects, they even work well with high voltage. i`ve dumped them in a tube preamp but they don`t sound expecially wonderful in comparison to the Wima`s going at the same price.

[hacknet 0]

i`d went to check my list of formulas and found the opamp to have little too much gain.

 

this is why.

 

CD players max output is around 2v RMS. say, we listen to music at levels of 1v RMS. 1v RMS equates to 1.414v peak to peak.

 

the opamp above is configured in a non-inverting fashion. with a gain of 11. this is found through the formula, Gain = 1 + (Rfbk/Rg). in this case, 1 + 10k/1k.

 

with a input signal of 1.414v

and a gain of 11

you will get 15.554v at the output.

 

 

since it is powered with +/-9v rails. in theory with an ideal opamp, you should be able to swing 18v P2P. sadly, we live in a unforgiving world, we loose about 2-3 volts of voltage swing in the silicon die of the opamp. say 2v, 18 - 2 = 16. we should be able to swing 16v P2P. one possiblitly is that the signal gets larger that 1vrms at the input which causes the distortion.

 

another possiblity is the opamp not being able to drive the cans.

 

the buf634 was designed to deliver an abmax of 250ma. with cans of 32ohms, and a swing of 15v P2P from the previous stage, the cans will draw 0.468 ampere. this is more that what is was ment to do, thus causing it to clip.

 

it works out that the output cannot exceed 8 volts for a 32ohm pair of phones. (V=IR, 0.25*32=8).

 

well, theres certainly more that can happen.

 

REGARDS

[huangyong 0]

i suggest that you refer to META and PPA topology for this problem.

 

0.22uF is no good for OPA637.(personal opinion)

 

keep on working....

 

[jasonhanjk 0]

Oops. see it in a hurry. It should be reconnecting the R@ to the pin 6 of OPA637...

[hacknet 0]

eh... pin 6, isn`t that the output pin? i think that too isn`t too good an idea. i`d recommend putting a 100R resistor on pin3 after the cap and reducing R2 from 10k to maby 6.8k.

 

this reduction of R2 will reduce the gain, hopefully solving the problem.

 

 

if it still doesn`t solve the problem, try changing the pot, it could be a bad contact along the track that is causing the distortion.

[OhYeah 0]

Thanx for all the replies, just came back from a one week overseas trip. Very helpful indeed, especially hacknet's comment. Will try it out again. Thanx again!

[JTT 0]

S'cuse me butting in late in this thread, but it's not clear to me what you're trying to achieve. Why are you using 2 opamps? I am doubtful about the wisdom of using either the 634 or the 637 as a power amp, even for the low powers you are probably after, and there is nothing useful to be gained by cascading two of them in the way you are trying. And including them in a single feedback loop is definitely asking for for oscillatory trouble. Even if it would be stable,

the gain would be 10 (R2/R1), and with only 18v to play about with, 2v rms input from your CD player would drive it into immediate clipping on peaks.

 

2v rms is approx 6v pk-pk, so to accomodate the max signal with an 18v supply you can't afford a gain greater than 2 - 2.5.

 

I would try using a single opamp, with R1=2k2 and R2=4k7.

 

JohnT

[hacknet 0]

the gain would be 10 (R2/R1)

 

 

 

 

have a look...

 

Why are you using 2 opamps?

 

 

you need to cascade them because the 637 can only pump out 7ma. this would certainly clip with cans.

 

 

 

 

Edited March 6, 2004 by hacknet

[Firefox 0]

you need to cascade them because the 637 can only pump out 7ma. this would certainly clip with cans.

The OPA637's can pump out way more than 7mA of current.

 

Second, your calculations are not exactly very true to life. A CDP may output 2.2v RMS but the potentiometer would reduce this to under 1v RMS for typical listening volumes.

 

the buf634 was designed to deliver an abmax of 250ma. with cans of 32ohms, and a swing of 15v P2P from the previous stage, the cans will draw 0.468 ampere. this is more that what is was ment to do, thus causing it to clip.

 

Not true. The opamp doesn't force a 15v swing into the cans. You can calculate the power required using the sensitivity of the cans. The voltage swing can then be worked out. Note that once the voltage swing is met (volume turned up enough), it's just up to the opamp to output the current required of it.

 

Worked example:

 

Grados with 32ohm impedance and 98dB @ 1mW. Normal listening volume of 106dB.

 

dB gain = 8 dB

Using: dB gain = 10 log (Power gain)

8 dB = 10 log (Power req'ed / 1mW)

Power required to drive cans to 106dB = ~6.31mW

 

Using: P = V²/ R

Voltage required = Sqrt (6.31mW * 32 ohms) = 0.449v

 

Using: P = I² * R

Current required = Sqrt (6.31mW/ 32ohms) = 14mA

 

[jasonhanjk 0]

The first opamp is for voltage amplifier and the second opamp is act as an buffer.

 

Buffer always has higher current ability as not to distort the given voltage signal when current surge appear.