Hi all,
I soldered up a Cmoy headphone amp the other day, roughly based upon the Tangent layout. However, I need some help in fault finding, because it has some rather show stopping issues!
I'm using an OPA2134, and 18V as the power source, for reference.
The two obvious problems are:
a) There's no sound from the output unless the source is supplying a very strong signal, and even then it is just occasional transients. This appears to be due/related t to a -2V DC offset that appears on both channels.
b) This is most likely related, in some way, to the wonky behaviour of the rail splitter, which is splitting something like 15V / -3V. Which, while the resistive rail splitter is rather dodgy, is very, very bad.
Can anyone suggest further action to help with tracking this one down? Photo's can be supplied on request.
Thanks!
I soldered up a Cmoy headphone amp the other day, roughly based upon the Tangent layout. However, I need some help in fault finding, because it has some rather show stopping issues!
I'm using an OPA2134, and 18V as the power source, for reference.
The two obvious problems are:
a) There's no sound from the output unless the source is supplying a very strong signal, and even then it is just occasional transients. This appears to be due/related t to a -2V DC offset that appears on both channels.
b) This is most likely related, in some way, to the wonky behaviour of the rail splitter, which is splitting something like 15V / -3V. Which, while the resistive rail splitter is rather dodgy, is very, very bad.
Can anyone suggest further action to help with tracking this one down? Photo's can be supplied on request.
Thanks!
Hi,
Supply is a single 18V supply (2 9V batteries in series, and yes, they meter at 18V, I did check), connected to the "power supply" described in Tangents (and the original Chu Moy) instructions - that is, a resistive rail splitter with a tank cap for each rail (or at least, I think that's all they do.) The caps are inserted observing correct polarity. There's no problem with the resistor values, I tested those, so it seems it must be a wiring fault I've made, but I can't seem to track it down.
Supply is a single 18V supply (2 9V batteries in series, and yes, they meter at 18V, I did check), connected to the "power supply" described in Tangents (and the original Chu Moy) instructions - that is, a resistive rail splitter with a tank cap for each rail (or at least, I think that's all they do.) The caps are inserted observing correct polarity. There's no problem with the resistor values, I tested those, so it seems it must be a wiring fault I've made, but I can't seem to track it down.
If you are going to use 2x9v batteries anyways what not run it off of a dual supply (+/-9v) this will get rid of the Rail splitter and could get rid of the DC offset because the supplies will be more Symetrical and stable than a virtual earth supply.....
I built my CMoy with a Dual regulated 15v suppl;y and it works great and sounds very clear, it just doesn"t have enough output current to properly drive my headphones to really high Volume...
I built my CMoy with a Dual regulated 15v suppl;y and it works great and sounds very clear, it just doesn"t have enough output current to properly drive my headphones to really high Volume...
Sorry to jump in here guys but I too have recently built the Chu Moy. It works fine when running from a split battery supply (after reducing the gain) but when I use it with a split regulated supply there is a lot of noise in there and the sound quality is totally different, it is almost like I am listening to music in a tunnel.....Any ideas?????
Thanks
Gareth
Thanks
Gareth
gareth said:
Maybe you need more filtering on your supply lines...Did you put 0.1uF caps close to the IC pins?? are you useing noisy regulators?? I am useing noisy LM78/7915 regulators with 100uF before the regs and 100uF after the regs with 10uF and 0.1uF close to the IC Pins and it is not noisy at all... did you seperate your audio and Power grounds and use a Star grounding scheme??
Just a few suggestions...
Minion said:
Maybe you need more filtering on your supply lines...Did you put 0.1uF caps close to the IC pins?? are you useing noisy regulators?? I am useing noisy LM78/7915 regulators with 100uF before the regs and 100uF after the regs with 10uF and 0.1uF close to the IC Pins and it is not noisy at all... did you seperate your audio and Power grounds and use a Star grounding scheme??
Just a few suggestions...
Just like you I am using 78/79 regulators (I wanted cheapos as this is my first build). I have +-12v supplies, 1n4001 diode rectifier, 0.047uf RF blocking caps, Panasonic 3300uF FC's (all I had available at the time), wima 0.47uF bypass into the regulators. Perhaps I need to revise the design???
Also I did not use the 0.1uF caps close to the IC...
Gareth
The 0.1uf caps close to the IC are pretty much a must with most any opamp design were low noise is a priority, they help with high frequency oscilation which can show up as a hum or whine...You can solder them directly to the Copper side of the PCB directly to the IC pins and ground ...It is an easy fix so you should try that and see if it helps....
If you haven"t allready you could also put a small (100pf-220pf) cap in the feedback loop which will help filter out frequencies that are out of the audio band....
Generally poeple say that it is best to have the big cap before the regulator and a small cap after so if your 3300uF is before the reg then that is good but if the 0.47 is after the reg you should then also increase it"s Value.....
Cheers
If you haven"t allready you could also put a small (100pf-220pf) cap in the feedback loop which will help filter out frequencies that are out of the audio band....
Generally poeple say that it is best to have the big cap before the regulator and a small cap after so if your 3300uF is before the reg then that is good but if the 0.47 is after the reg you should then also increase it"s Value.....
Cheers
Minion said:
Thanks minion,
So I connect the 0.1uF's from supply pins to earth(ground)? Do these need to be polarised?
The feedback ceramic I can put in as it sounds as if it will do a lot of good.
Yes the 3300uF are before the regulators so I would need to increase the 0.47's to what value?
At the moment, both the amp and power supply are built on perfboards and through another forum on the site I am starting to learn Diptrace and I hope to produce some boards soon!!!!!!!
Gareth
Edit..the 0.47's I have in parallel with the 3300's and be honest I hadn't thought of putting anything after the reg's. Also the amp is in it's own case and the power supply will (in future) have it's own case too! I have braided the supply cables into the amp.
The 0.1uF caps are non polarized, I don"t think you can get Polarized caps in 0.1uF, preferabley Poly but ceramic will work , and yes connect them from the power pins to ground...
Yes you definately need some Caps after the regs, I believe anywere from 47uF -100uF should be good but if the PSU will be far from the Curcuit then the more the better, you might even consider that is the PSU will be far from the curcuit that you should maybe put some 10uf caps on the supply rails on the Headphe amp PCB it"s self.....
Cheers
Yes you definately need some Caps after the regs, I believe anywere from 47uF -100uF should be good but if the PSU will be far from the Curcuit then the more the better, you might even consider that is the PSU will be far from the curcuit that you should maybe put some 10uf caps on the supply rails on the Headphe amp PCB it"s self.....
Cheers
First off, its pretty hard to make opa2134 unstable....
I suggest disconnecting your batteries from the cmoy and check if the supply arrangement you have is doing an ok job at splitting the battery supply... I build mine with a single 9V, split...
Next make sure you have your pins in the right order, pin 1 is the one with the dimple...
Measure resitors especialy the ones used on the input and feedback, to see if they match your schematic...
I think 100nf is overkill for decoupling on these chips... mine runs dead silent on 10nf Wima caps...right against the chip body
I suggest disconnecting your batteries from the cmoy and check if the supply arrangement you have is doing an ok job at splitting the battery supply... I build mine with a single 9V, split...
Next make sure you have your pins in the right order, pin 1 is the one with the dimple...
Measure resitors especialy the ones used on the input and feedback, to see if they match your schematic...
I think 100nf is overkill for decoupling on these chips... mine runs dead silent on 10nf Wima caps...right against the chip body
I've rebuilt my cmoy on a bigger bit of board, for troubleshooting purposes....and it's exactly the same. So either the chip is shot (possible, I did accidentally forgot to complete the ground rail earlier in the build, and that may have caused a circuit which has damaged the chip), or I'm making the same mistake, again.
I read somewhere that it's worth trying to add caps to the output, to prevent a bias current being drawn and unbalancing the rails. Any thoughts?
I read somewhere that it's worth trying to add caps to the output, to prevent a bias current being drawn and unbalancing the rails. Any thoughts?
Hy. Im thinking to solder a cmoy. This shematic
http://tangentsoft.net/audio/cmoy-tutorial/assy.html
But i have q
In pics he used opa2227 but in the part list is OPA2132PA ?
which to use?
http://tangentsoft.net/audio/cmoy-tutorial/assy.html
But i have q
In pics he used opa2227 but in the part list is OPA2132PA ?
which to use?
Im little bit confused...
can opa2134 work in this shematic?
http://tangentsoft.net/audio/cmoy-tutorial/assy.html
thanks.
can opa2134 work in this shematic?
http://tangentsoft.net/audio/cmoy-tutorial/assy.html
thanks.
I've seen it done with other opamps, and as I understand it, all that's necessary is small resistors on the inputs, and and summing resistors on the outputs. I'm sure someone will correct me if I'm wrong.
Of course, if by "more power" you mean more gain, this won't achieve anything, and you should read the tangent tutorial again. If, however, you're using low impedance headphones, and the current supply of the opamp is insufficient, it's possible that parallel opamps could help.
Of course, it's a poor solution, and you'd do better to look into using a high current buffer stage instead. Look at the circuit diagrams of the PIMETA to see how this is achieved with BUF634.
Alternatively, it's possible to place a buffer outside the feedback loop, and by doing it this way, you can perhaps use less expensive parts for the buffer, like TI's OPA551. This is the solution that I'd recommend as the easiest (and probably the cheapest) to add on to an existing headphone amplifier.
EDIT: To make the last option above clearer: I'm suggesting that instead of using a dedicated buffer IC like BUF634 or LMH6321, you could use a unity-gain stable opamp with high output current capacity, such as OPA551, configured as a buffer by connecting its input pin to its output pin.
This will provide higher current capacity, but at the cost of potentially reduced sound quality, as the buffer is outside the opamps feedback loop.
Of course, if by "more power" you mean more gain, this won't achieve anything, and you should read the tangent tutorial again. If, however, you're using low impedance headphones, and the current supply of the opamp is insufficient, it's possible that parallel opamps could help.
Of course, it's a poor solution, and you'd do better to look into using a high current buffer stage instead. Look at the circuit diagrams of the PIMETA to see how this is achieved with BUF634.
Alternatively, it's possible to place a buffer outside the feedback loop, and by doing it this way, you can perhaps use less expensive parts for the buffer, like TI's OPA551. This is the solution that I'd recommend as the easiest (and probably the cheapest) to add on to an existing headphone amplifier.
EDIT: To make the last option above clearer: I'm suggesting that instead of using a dedicated buffer IC like BUF634 or LMH6321, you could use a unity-gain stable opamp with high output current capacity, such as OPA551, configured as a buffer by connecting its input pin to its output pin.
This will provide higher current capacity, but at the cost of potentially reduced sound quality, as the buffer is outside the opamps feedback loop.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.