Why not use super high speed opamps?
Sep 4, 2007 at 4:28 AM Post #3 of 15
Quote:

Originally Posted by Pars /img/forum/go_quote.gif
What's so fast about those? How about these, which are used commercially in some of Corda's stuff, as well as in some PINTs, etc. Much higher slew rate that what you pointed to. Read Tangent's Working With Cranky Opamps for more info.


Ah, did not know these high speed opamps were used commercially, usually only fairly low speed ones are used.

That one has 6 times higher noise and 3x the settling rate though. I'm gonna see what else there is besides the standard "audio" opamps.
 
Sep 4, 2007 at 10:09 AM Post #6 of 15
Hey,

Lookie this.

http://sound.westhost.com/highspeed.htm

I know this one uses THS6012 which is a Current-Feedback OPAMP. But this article is very useful in learning how to use high-speed chips. I myself used it to build an amp using THS4022 Evaluation Board. (THS4022 is VFB opamp.)

Tomo

P.S. Lookie my THS4022 CMOY amp!

IMGP2063.jpg


I am pretty cheesy with this, but it sounds decent.
 
Sep 4, 2007 at 6:51 PM Post #9 of 15
Hey,

It is not good if you are looking for tube-sweetness or warmth. I suppose this amp is very very unbalanced and it would be wrong to call it "musical". But it sound accurate to the level of sterile. It is nice when you are looking for some particular instrument in ensemble or band. It is also good for monitoring. i.e. you can hear the musician tapping with his shoe.

I like being able to pick out one instrument. But note that accuracy does not necessarily mean good music. Plus, I like it because no one uses this opamp; makes me unique.
smily_headphones1.gif


Note that as the opamp gets faster, 1/f noise floor goes up. You need to pay attention to this. Also, opamp's main short coming is its low current capability. I chose THS4022 because of its low 1/f noise and high current capability (150mA max).

Tomo
 
Sep 4, 2007 at 7:08 PM Post #10 of 15
Quote:

Originally Posted by Tomo /img/forum/go_quote.gif
Hey,

It is not good if you are looking for tube-sweetness or warmth. I suppose this amp is very very unbalanced and it would be wrong to call it "musical". But it sound accurate to the level of sterile. It is nice when you are looking for some particular instrument in ensemble or band. It is also good for monitoring. i.e. you can hear the musician tapping with his shoe.

I like being able to pick out one instrument. But note that accuracy does not necessarily mean good music. Plus, I like it because no one uses this opamp; makes me unique.
smily_headphones1.gif


Note that as the opamp gets faster, 1/f noise floor goes up. You need to pay attention to this. Also, opamp's main short coming is its low current capability. I chose THS4022 because of its low 1/f noise and high current capability (150mA max).

Tomo


I think the amp I chose for replacing the AD8610's in my portaphile is a decent decision then.

http://focus.ti.com/docs/prod/folder...t/ths4211.html
 
Sep 4, 2007 at 8:14 PM Post #11 of 15
You might need to change some of the things surrounding the opamp, not a direct drop in without debugging. I know nothing of the portaphile, but the AD8610 is a fairly easy chip to work with. Look at the input biases and offsets on the ths4211. It's comparatively large.

You might want to pick up some additional resistors and caps as well, you may need to do some stabilization on it. Nothing big, but possibly replacing some resistors and adding coupling caps. The relatively high current draw for offset and bias, and higher offset voltage may cause high dc offset. Current over resistors has a voltage differential. If one end is tied to ground on the input, you'll have some sizeable voltage at the input terminal, which then gets amplified and shows up as DC offset.

It might be easiest to put it into low gain mode first, looks like 2 is what it goes down to.

You may want to read on the AD8397 chip on headwize. Read the background on the Mini3 and PINT, for example, and see what needed to be done to stabilize it. That might give you some ideas of what to look for and some potential prior ground for tackling issues.

Tangent has a guide as well, which was linked.

BTW, if you're only looking at slew rate and settling time, check out the OPA690. It's used in the Mini3 to drive ground.
 
Sep 5, 2007 at 1:38 AM Post #13 of 15
Hey,

I think it's a bit noisy amp. I told you about the 1/f noise right? Problem with the most high speed opamps are the fact that they are noisy at the low frequencies as they are only designed to be operated at high frequencies.

If you observe the page 11 figure 18 of the THS4211 specification sheet, you can see noise level rise as the frequency lowers. Compare this to the page 1 figure 1 of the THS4022 specification sheet. You see THS4022 seems about an order lower in appearance. But becareful about the frequency indexing (goes from 100kHz to 10Hz). This means THS4211 is considerably more noisy at audio frequencies. Though this may not be a problem if you stick with unity-gain or low-gain applications.

As Holland mentioned, it has fairly high input quiescent current. This will result in considerably high DC offset on CMOY amp if you simply swap the chip. THS4022 has the same problem. These opamps are designed to drive 50ohm cables and are meant to be operated at this impedance. (i.e. quiescent current though 50ohm resistor will develop negligible DC offset)

What I did was to increase the resistor that connects the positive input to ground to 1k ohm. This should allow problematic but "okay" dc-offset. However, this causes very serious problems for the volume control. Essentially you will need to replace it with one with much lower resistance.

I recalculated the volume control effecitiveness (Vout/Vin or Wout/Win) and realized I need to custom-make the volume control. So you see that Big Honking rotary switch with resistors clustered together like insects.

Listen to Holland about the stabilization and stuff. Essentially you will need to rebuild your CMOY amp from ground up. He didn't tell you but you will probably need to make sure you don't pick up RF interference because you might interfere with cellular phones. (I got THS4022 oscillate enough to interfere with TV set; naughty!
tongue.gif
)

I suppose it's a wonderful amp to "play" with because of all the variables. You can tinker all day and not get bored. If you are expecting to get an instant trial, you ought to look elsewhere. (probably couple weekends straight-though) I have been tinkering with it for nearly a half decade.

Try em out with an evaluation board. You need to replace some resistors, but you are pretty much free from worrying about RF circuit designs (people does this professionally so I won't bother competing until later date).

Tomo

P.S. I recommend using EVM if they have them. They are expensive but cheap compared to all that studying RF and designing RF circuit. I wonder why not many people do this while most people are happy with META board and all that. ... It's almost plug&play and straight trial. Who knows you might like it and keep it. No offense to ambs and other resident kit designers, but I kinda like having an amp that doesn't frequent Ebay.

P.P.S.

READ THIS LINK

http://sound.westhost.com/highspeed.htm

You have got to read it.
 
Sep 5, 2007 at 1:53 AM Post #14 of 15
I have in my PPA the THS4631 and it sound fabulous in all three channels...
 
Sep 5, 2007 at 2:22 PM Post #15 of 15
Crap!

I forgot to tell you check how the distortion changes over frequency. That changes depending on the load impedance. At high frequency, you may get distotion peaking on some load impedance.

Ciao,

T
 
You must log in or register to reply here.

Users who are viewing this thread

Total: 2 (members: 0, guests: 2)
Back
Top