[jeffreyj]

Quote:

Originally posted by ofb doh! ah, well, some things changed and other didn't. .... technics sa-200, R 66.0 L 79.0. stable. ... sony wm-af54, ... starting at -340mv and slowly decending, both channels negative this time, and decending in sync. note, this unit doesn't have a 'pause'. it's technically 'off'. .... bummer about the sa-200; i still use it. just how dangerous is that for my headphones?

The results for the wm-af54 might not be valid if the unit doesn't have a pause function, or some other "mode" where the unit is on but a disc is not playing.

The offset on the sa-200 might be from leakage current through the output coupling capacitors (if it has them), in which case placing a resistor from each output to ground (~1k) ought to swamp it out. If this doesn't, then the amp has a direct coupled output stage with a minor dc offset problem that should be corrected. The magnitude of the offset is not likely to be damaging, but what it will do is push (or pull) the voice coil of the drivers away from their resting point. Part of the dynamic range the drivers would otherwise be capable of is lost, then, because their travel in that direction has been reduced. But the big problem with this offset comes when it gets amplified by a direct coupled amplifier with the ever popular Av=11. Now you have up to 900mV of offset, which into 32 ohm headphones is 24mW of power! Will it hurt them? Still, not likely, amazingly enough, but neither will they be likely able to play music at more than a very low volume setting without seriously distorting!

 

[ofb]

thanks, tangent & jeffreyj.

jeffreyj: the wm-af54 is a tape walkman, hence it's 'play' or 'off'. just tested again with no tape and got 0mV, both at play and off. then played a tape and got the slow descent again afterwards, on play without tape, and off, starting at over 600mV this time.

as for the sa-200, i can't tell if it has output coupling caps. just had a long look and i got lost. headphone output is a simple pair of 330 ohm resistors, but then things get complicated as the paths snake through the psu region and back again. i can't even make out two channels of the amp. i'll have to try again when i have more experience. meanwhile i have more excuse to replace it with a gainclone. thanks, tho.

 

[Vdubjunkie]

Quote:

Originally posted by jeffreyj So, a DC blocking capacitor is required on the input as well. Fortunately, because the size of the capacitor necessary is proportional to the load impedance, and the input impedance of most amplifiers is high, said capacitor won't need to be so large in value

Hello, and please excuse my ignorance. I'm slowly but surely locating and placing the many puzzle pieces of this relatively new hobby I've chosen. When I see something I sort of understand and can form a coherent (relative) question, I will

With that said.. You say that [/i]most[/i] amps have high input impedence, and the capacitor should be proportional to this load impedence. I don't get why that would equate to a not so large value cap.

I would appreciate if you would entertain my noobish ignorance.

 

[yejun]

Quote:

Originally posted by Vdubjunkie Hello, and please excuse my ignorance. I'm slowly but surely locating and placing the many puzzle pieces of this relatively new hobby I've chosen. When I see something I sort of understand and can form a coherent (relative) question, I will With that said.. You say that [/i]most[/i] amps have high input impedence, and the capacitor should be proportional to this load impedence. I don't get why that would equate to a not so large value cap. I would appreciate if you would entertain my noobish ignorance.

there's an artical about coupling capacitor in tangent's website.

 

[tangent]

This is what yejun is referring to:

http://tangentsoft.net/audio/meta42/advice.html#C1

You will also want to look at the schematic while you read that if you're not familiar with the circuit used on the META42 board. The input impedance that JeffreyJ is referring to is set by R2 in that circuit. The input impedance of the whole amp is set by the pot, but R2 is what matters here, since the cap is followed by R2, not the pot.

 

[jeffreyj]

Quote:

Originally posted by Vdubjunkie .... You say that most amps have high input impedence, and the capacitor should be proportional to this load impedence. I don't get why that would equate to a not so large value cap. ...

This series capacitor forms a high-pass filter with any resistances present from it to ground (including op-amp input impedance, stray capacitances between traces, volume pot, etc.). The cutoff frequency - that is, the frequency at which the signal is cut by 3dB - occurs when the capacitor's reactance - 1/(2*pi*f*C) - is equal to the shunt impedance. The input impedance of the non-inverting amplifier configuration (consider this as applying to regular op-amp circuits, for the nit-pickers out there) is determined by the shunt resistance intentionally put there from input to ground; e.g. - a volume pot. If the value of the pot is 50k, then the cutoff frequency formed by the typical 0.1uF blocking capacitor would be:

fc = 1/(6.28*50000*(1*10^-7))

Or, 31.8Hz.

This is a bit high of a cutoff, not only because it is well into the audible range, but because of the phase shift which occurs from the RC network (45 degrees at fc, approaching 90 degrees deep into the stopband, or close to DC in this case).

If we ditch the volume pot and go with the inverting op-amp configuration, which strives to keep both of its inputs at zero volts through feedback (i.e. - grounded), then the input impedance absent any input resistor is zero (said resistor, of course, is necessary for reasons which I encourage you to discover on your own). Selecting an input resistor of 10k, which seems to be popular, gives a cutoff frequency of... (calculate it yourself using the above equation).

As you can see, quite a difference, and, hence, why I said the input DC blocking capacitor should be scaled along with the input impedance of the amplifier. This also goes for the output, where the impedance in question becomes that of the headphones. Here the problem is compound not only by their lower impedances in general, but the range of impedances among them (between 16 and 600 ohms for pretty much any pair of dynamic 'phones). Replacing the 50,000 ohm value with 600 and 16, keeping the 0.1uF value the same ought to be a real eye-opener

edited: grammatical mistake, formatting

 

[stereth]

My equipment:

Sony D-EJ756 PCDP:
Headphone amp:
L: -0.6 - 1.3 mV
R: -0.6 - 1.2 mV (flunctuates, even when paused at volume 0)
Line out:
L: 0.7mV
R: 0.7mV

M-Audio Sonica:
L: -40...-6...-9...0...-20... (slowly flunctuating up and down, don't turn your back on this thing)
R: 0 - 3 mV

Compaq Evo N800c integrated sound card (SoundMAX):
L: 0.1mV
R: 0.2mV

So I guess I can build one cmoy with no input caps, for the PCDP and sound card volume control. The Orange Drops stay for in front of the Sonica.

How high can the offset voltage be before you have to worry about your headphones?

 

[Rotareneg]

Denon DCM-370 = 1.6 millivolt on both channels.
Technics SH-GE50 equalizer, source turned off = 0.1 mv on both channels.
Corda HA-1 at 50% volume, source turned off = R 5.4 mv, L 13.0 mv.

Given the DC resistance of my various headphones, here's the amount of current and power the DC current should cause, based on the higher left channel voltage:

HD600, 300 ohms DC = 43.33 microamps, 0.563 microwatts
ER-4S, 105 ohms DC = 0.124 milliamps, 1.610 microwatts
SR-80, 31 ohms DC = 0.419 milliamps, 5.452 microwatts

This means, assuming the amp actualy puts out that current, the SR-80's would put out enough heat to warm one gram of water 0.04 degrees celsius in a year.