cmoy op amp virtual ground question

Status
Not open for further replies.
Jump to Latest
V

vrfood

Member
Joined 2010
#1
Hey guys.
So, i've build a couple of cmoy amps, following Tangent's instructions. Now i am looking to make a different layout and i decided to use an op-amp based virtual ground instead of a TLE2426 (because i can't get one locally and i don't really want to mail order one). I am going to use UA741 as Tangent suggests, but i am not entirely sure how to wire it. Here is a pic based on his schematics; is this the correct way to wire it?

Thanks in advance
 
#2
not too many headphone that would be OK with the 1 kOhm series R

and no one actually uses 741 except as a bad example


what headphones is the 1st headphone amp design question

headphone impedance and sensitivity vary by over an order of magnitude for Z, several orders for sensitivity


active supply splitting is an option with a single V supply

but headwise, head-fi crowds went off in a poor direction with "3-channel" amps becoming fashionable despite the technical arguments actually going against them if correctly analyzed, measured
 
#3
Well, tbh, i was really looking to just swap the TLE2426 for something available locally. As for headphones, i own a pair of ATH-M50s as well as rock-it-sounds r-50. I usually pair them with a topping tp-30 amp/dac, but they do sound good with the first 2 cmoy amps i built. They were made with 2 power capacitors and 4.7k resistors as virtual ground, just like in the Tangent example. I simply wanted to try out a somewhat different virtual ground, that's all.

SO, basically, you are saying to just use TLE2426, then, right? I mean, it's cheaper than the better op-amps for just rail splitting. I thought the generic 741 will do fine for just splitting the current. What would be better, but still kind of widely available (in Europe 🙂?

What do you mean by active supply splitting, please explain with a bit more detail. excuse my technical incompetence.
 
Last edited:
#4
the active ground needs to be able to sink 2x the current of the R,L driver amps when the audio is in phase between channels - most often the case for bass

so actually "cheaper" isn't the way to think about the active ground buffer op amp - its performance requirements are

the low Z of your headphones makes AC capacitor divider be physically big - especially if you're trying for portable

but both look like they could be fine from standard DAP which already have high current op amps and virtual grounds built in
 
#5
I meant cheaper in comparison to the TLE2426, which everyone agrees is the best/simplest option here. So, what would you advice me to do, then? Scrap the op amp buffered design, since i can only get 741 locally (without mail ordering from the UK or US)? Or go for it and just see how it goes, since it practically costs nothing.

thanks
 
#7
vrfood said:
I meant cheaper in comparison to the TLE2426, which everyone agrees is the best/simplest option here. So, what would you advice me to do, then? Scrap the op amp buffered design, since i can only get 741 locally (without mail ordering from the UK or US)? Or go for it and just see how it goes, since it practically costs nothing.

thanks
Click to expand...

Do you REALLY want a '741 in the signal path? (Hint: NO!) Aside from elevated distortion levels and severely restricted bandwidth, they have a much higher output impedance that limits their ability to drive loads of anything less than several Kohms. You're better off with the resistor/cap solution.

Mike
 
#8
Michael Bean said:
Do you REALLY want a '741 in the signal path? (Hint: NO!) Aside from elevated distortion levels and severely restricted bandwidth, they have a much higher output impedance that limits their ability to drive loads of anything less than several Kohms. You're better off with the resistor/cap solution.

Mike
Click to expand...

Thanks, Mike. That's exactly the kind of answer i was looking for.

So, let me summarize:

Either use TLE2426 or a better op amp instead of the 741, like the LMH6642 or AD817 Tangent has mentioned in his article about virtual grounds. He claims using a good buffered op-amp might be an upgrade over TLE2426, since the op-amp's output current is higher. What do you think about that? I suppose it depends on the type of headphones you wanna drive with the amplifier.
 
#9
jcx said:
TI, AD must have EU distributors?
NXP, ST are both big semi manufacturers in the EU
Operational Amplifiers (Op Amps) - STMicroelectronics
Click to expand...

They do, although i am in eastern europe, (Bulgaria), so it is pretty much like.. they either have it locally in the country, or i need to mail-order it myself, in which case i am better off ordering it from Mouser UK; that of course defeats the whole purpose of using a substitute to the TLE2426. Can you give me some guidelines as to what i am looking for in the op-amps? I know they are supposed to have high output current and be single channel, but that's about it. Since manufacturers don't list output current i have no idea which ones to look at. How about naming some models from ST which will do the job of splitting the current, without adding any noise/crosstalk etc to the signal.

thanks again
 
#10
vrfood said:
Since manufacturers don't list output current..
Click to expand...

You bet they do! 🙂 The photo below shows what to look for in a data sheet. This particular one is for an LME49710:

http://www.ti.com/lit/gpn/lme49710

Note the conditions given though for Iout, a 600 ohm load and with +/-17Vdc power supplies.
 

Attachments

  • LME49710_Iout.JPG
    LME49710_Iout.JPG
    67.7 KB · Views: 238
Last edited:
#11
reading datasheets takes some time to learn but the manufacturers are interested in engineers knowing which of their parts to buy

a rather opinionated look at op amps for headphones: NwAvGuy: Op Amp Measurements

discrete transistor amp design could a way to go if you really have supply problems but can still get basic small signal Q like 2n3904/6, BC556/7

discrete "diamond buffer" can be used for ground buffer in an active splitter, and as output buffers for the channel amp too
 
#13
jcx said:
reading datasheets takes some time to learn but the manufacturers are interested in engineers knowing which of their parts to buy

a rather opinionated look at op amps for headphones: NwAvGuy: Op Amp Measurements

discrete transistor amp design could a way to go if you really have supply problems but can still get basic small signal Q like 2n3904/6, BC556/7

discrete "diamond buffer" can be used for ground buffer in an active splitter, and as output buffers for the channel amp too
Click to expand...

I am going definitely to build NwAvGuy's o2 amp, but i thought i should get myself familiarized with DIY electronics a little bit first. That's how i got started with the cmoy, i read it was a good first project and stuff. Anyway, thanks for the help. Looks like i am going to either use the default resistor type of design or mail-order TLE2426 and use that, as i am pretty sure i won't be able to find any good op amps stocked locally.

cheers
 
#14
After searching the best (local) online store we've got over here, i am still not sure which op amps are suitable for this specific job. They do seem to have specific applications and i am confused. They stock NE5534P, can i use that? It says 1 channel, power supply ±3 ÷ ±20V and has output current of 38mA. Actually, can you guys check out that website, they have 216 op amps in stock. Could you do me a solid favor and post 2-3 viable options to split the cmoy's 9v batter power in two rails. I'd gladly do it myself, but i really have no idea what to look for, apart from what I already mentioned and there are quite a few possible options.

Thanks in advance.

p.s. they've got a pretty easy interface, i am sure it won't take you more than 2 minutes. cheers
 
#17
I know I saw a virtual ground circuit using either BUF634 or LME49600, and I think it was on the datasheet. It's worth a look; I think either one of these buffers would work well in a virtual ground circuit.

You could use an op amp based virtual ground circuit and put one of these buffers in the loop. Either one can source or sink 250 mA minimum and both buffers are really fast, so I think that this scheme would more than suffice.
 
#18
Fast Eddie D said:
I know I saw a virtual ground circuit using either BUF634 or LME49600, and I think it was on the datasheet. It's worth a look; I think either one of these buffers would work well in a virtual ground circuit.

You could use an op amp based virtual ground circuit and put one of these buffers in the loop. Either one can source or sink 250 mA minimum and both buffers are really fast, so I think that this scheme would more than suffice.
Click to expand...

Yes, these buffer chips are recommended for such use, but they cost much more than TLE2426 and according to Tangent's website, they would introduce higher output impedance, and higher battery draw. Not to mention the availability is a problem, again.
 
#19
counter culture said:
Why not just use 2 * 9V batteries?

You're only trying to get a bit of experience, and a split supply with 2 batteries has a lot to recommend it, to say nothing of twice the voltage.
Click to expand...

Thought about it, but it'd require a lot bigger enclosure. As i want it to be an aluminum box, it would also be heavier than it should for a portable device. I surely will make a cmoy with 2 9v batteries, but it will be a later project. Besides, i already got a Hammond 1455 case for it and it is 80mm x 53mm x 24mm aka small. Thanks tho 🙂
 
Last edited:
Status
Not open for further replies.
Top Bottom