Building my first cmoy, couple of questions
Feb 27, 2006 at 4:57 AM Thread Starter Post #1 of 9

pacemkr

New Head-Fier
Joined
Feb 6, 2006
Posts
19
Likes
0
I used tangents guide and already put the thing on a pcb and its working on the first try. =)

Here a couple of questions arise.
1) I have some hum even when no source is connected, its pretty audible at full volume, quite audible at half volume, very audible with a source connected. Is that normal. I tried adding R5, didnt make any difference.

2) There is a very low buzz, that seems to be a result of grounding issues. When I move my hand closer to the amp, the buzzing goes away. How do I fix this?

3) Are 18V really needed? I hooked the amp up to one 9V battery (I'm using the OP2227, I also have OP2134, but I havent tried it) and it seems to be working fine. I'm asuming the opamp is more stable at 18V, but I'm already thinking of the next amp I'm going to build =) and I want to make it really portable. This one doesnt qualify since I'm adding a Meier Bass enhanced crossfeed along with all the switches, two batteries, and quite a bulky case. So I was thinking powering the next one off a single 9V battery, are there any sideffects?

4) I'm using low impedance phones (16 ohm Sony MDR-EX71) can this be why I'm hearing the "humming"? I should get some more serious headphones, but I really love the portability and sound quality taking size into account, so these will be my main phones no matter what. You might say that defeats the whole purpose of having an amp. Thats what I thought, thats why I didnt opt for a portable design right away. But for some reason sound going through the amp sounds better to my ears or I simply convinced myself that it does, either way something is working.
etysmile.gif
On that note, are there any modifications to the cmoy design that are needed for low impedance phones?

Just in case, here are the "modifications" I used that deviate from standard schematic in tangents guide. I used a TLE2426 for the VG. I did not connect the NR pin on the TLE, since I didnt hear any difference and the 1uF cap is bulky. I used 0.47 uF caps at C2.

Thanks.
eggosmile.gif
 
Feb 28, 2006 at 12:58 PM Post #2 of 9

Porksoda

500+ Head-Fier
Joined
Jan 7, 2004
Posts
571
Likes
0
Questions 1 and 2 are difficult to answer straight away. I can't really offer you any advice there other than that having 16 ohm cans doesn't help. Even still, it shouldn't be really loud.

3) 18V is not needed, but it ensures that the CMoy can drive headphones that need lots of voltage. For 16 ohm cans, you could probably do fine with 9V. With cans like the HD580s, you would probably need 18V.

4) Adding the TLE was a smart move first of all. That helps quite a bit with low imedence cans. Also, R5 helps. Try playing with this value a little bit. Also consider trying amps like the A47, which is a small step up in complexity from the CMoy and is specifically designed for low impedence cans.


Cheers!
orphsmile.gif
 
Feb 28, 2006 at 6:54 PM Post #3 of 9

Ikon

100+ Head-Fier
Joined
Nov 21, 2004
Posts
269
Likes
11
About any noise/hiss/hum:
Low-impedance phones don't really need the additional voltage gain of the amp and often exhibit all kinds of background noises/sounds with a too high gain -> keep the gain of the amp low (try 2 for example).

And yes, there are many "optimisations" of the CMOY for low-impedance phones. They usually consist of putting some kind of current booster in the feedback-loop of the op-amp, look at the PIMETA/PPA, PRR's TORI or Earl Eaton's design for example. A small patch is the optional R5 resistor which you already tried.
 
Feb 28, 2006 at 9:20 PM Post #4 of 9

pacemkr

New Head-Fier
Joined
Feb 6, 2006
Posts
19
Likes
0
Thanks for the info. From your suggestions I know that the next amp I'm building is deffinitely going to be powered by 9V. I'll check out the A47 design,thanks for the pointer.

The main reason I'm building this amp (and the next) is because I want to really understand how it works, so more questions arise in my mind. Why does the TLE help with low impedance phones? I understand it helps to keep the voltage division more equal, but why is the result more significant with low impedance load?

I tried a gain of 3.6 I believe and the hiss was still there. I'll try changing it again, just to make sure. But first I'll test drive the amp on my friends "proper" phones tonight.

Also why do low imedance phones require more current? I asume the voltage gain remains the same and if the impedance is low the current has to be high and if the amp cant provide that current, distortion. Am I at least half right?
 
Feb 28, 2006 at 9:48 PM Post #5 of 9

Ikon

100+ Head-Fier
Joined
Nov 21, 2004
Posts
269
Likes
11
Quote:

Originally Posted by pacemkr
The main reason I'm building this amp (and the next) is because I want to really understand how it works, so more questions arise in my mind. Why does the TLE help with low impedance phones? I understand it helps to keep the voltage division more equal, but why is the result more significant with low impedance load?


Lower impedance cans -> more current draw -> greater [potential] imbalance [in the virtual ground].

Quote:

Originally Posted by pacemkr
I tried a gain of 3.6 I believe and the hiss was still there. I'll try changing it again, just to make sure. But first I'll test drive the amp on my friends "proper" phones tonight.


Hissing starts at about 2.5 with my (rather quiet) source and 32-50Ohm cans. I go for 2 usually, just to be sure.

Quote:

Originally Posted by pacemkr
Also why do low imedance phones require more current? I asume the voltage gain remains the same and if the impedance is low the current has to be high and if the amp cant provide that current, distortion. Am I at least half right?


That's pretty much it. Higher impedance phones eat less current and thus keep a greater distance to the output current limit of the amp.
 
Feb 28, 2006 at 11:26 PM Post #6 of 9

Porksoda

500+ Head-Fier
Joined
Jan 7, 2004
Posts
571
Likes
0
Quote:

Originally Posted by pacemkr
Why does the TLE help with low impedance phones? I understand it helps to keep the voltage division more equal, but why is the result more significant with low impedance load?


The advantage of the TLE over just using two precision resistors is that it is active, whereas the resistors are passive. When you put a signal through the headphones, the current retuns to the virtual ground after it goes through the headphones. The virtual ground must then "absorb" this current. Much like your headphones, when you put current into the resistor divider based ground, its voltage varies. So as the signal goes through the headphones, the ground is also moving along with it. This is inherent in the passive nature of the resistor divider, and leads to crosstalk and all kinds of non-ideal behavior in the circuit.

The TLE, on the other hand, is active. Its internal circuitry can absorb much more current than a resistor divider, and its voltage will vary a lot less because it uses transistors to decrease its output impedence. This keeps the ground voltage more stable, decreasing crosstalk and actually making the op-amps not have to work quite as hard in addition to saving on battery life.
 
Mar 1, 2006 at 12:36 AM Post #7 of 9

The Monkey

Monkey See, Monkey DAC
A really sick dud
Joined
Oct 28, 2004
Posts
8,010
Likes
135
Quote:

Originally Posted by Porksoda

The TLE, on the other hand, is active. Its internal circuitry can absorb much more current than a resistor divider, and its voltage will vary a lot less because it uses transistors to decrease its output impedence. This keeps the ground voltage more stable, decreasing crosstalk and actually making the op-amps not have to work quite as hard in addition to saving on battery life.



Does the TLE have a noticeable impact on SQ in a cmoy?
 
Mar 1, 2006 at 1:00 PM Post #8 of 9

Porksoda

500+ Head-Fier
Joined
Jan 7, 2004
Posts
571
Likes
0
Quote:

Originally Posted by The Monkey
Does the TLE have a noticeable impact on SQ in a cmoy?


I've never A/B'ed the difference, so I won't make a claim either way as to how audibe the difference is. All I can tell you is that it is a better (more ideal) circuit with the TLE in, and that it should decrease crosstalk and, to a much lesser extent, distortion. The difference should be more pronounced with lower impedence cans. I have one with a TLE and I find it quite enjoyable. Never done resistor divider.
 

Users who are viewing this thread

Top