THS4130
is one op-amp from Texas Instruments
It is very high performance.
I do not know how suitable for audio,
at least compared to one good hi-fi preamp by fotios 😉
http://focus.ti.com/docs/prod/folders/print/ths4130.html
Datasheet PDF:
http://focus.ti.com/lit/ds/symlink/ths4130.pdf
is one op-amp from Texas Instruments
It is very high performance.
I do not know how suitable for audio,
at least compared to one good hi-fi preamp by fotios 😉
Productpage:
http://focus.ti.com/docs/prod/folders/print/ths4130.html
Datasheet PDF:
http://focus.ti.com/lit/ds/symlink/ths4130.pdf
I use the THS4130/1 (and OPA1632) for single ended to balanced converters. They sound excellent to my cloth ears.
I still find it curious how hot these IC's get. What's going on inside that they need to consume so much idle current?
I still find it curious how hot these IC's get. What's going on inside that they need to consume so much idle current?
theAnonymous1 said:
Have you tried the THS413x with Vcc more than +/-15Vdc?
Because - if i understood well - in electrical characteristics refered Vcc up to +/-33Vdc.
As for the heat disipation it is refered also that the case of this IC is something special.
Fotios
Bonsai said:
Nelson Pass told me that his X0.2 preamplifiers have a signal path like these ICs of Texas. If i have understood well, it means that there is not any conversion of the signal inside the preamplifier from balanced to single ended and vice versa.
The main problem of course, it is the stereo volume control.
If you are using a dual gang potentiometer, then there is the need of convertion of balanced inputs to single before the pot and vice versa after it if you want also balanced outputs.
On the contrary, if you are using a four gang volume pot. there is no need for convertion. With the use of THS4130 you can implement a line preamplifier with only four THS4130 for stereo operation. Two before the 4 gang pot for buffering the stereo balanced inputs and two after the pot for buffering the stereo balanced outputs.
From the data sheet, if we take into account the transient response presented in fig. 6, then i think that this IC can has the same performance with discretes and much more!
Fotios 😎
Fotois, if you look at moxamp's site you will see a circuit for a balanced volume control pot that does not need matched quad gang pots - quite clever. his can be implemeted as he shows via relays (its an R2R topology). This probably the best solution for a descrete based design where you do not want to be fiddling with quad pots (however, see goldpoint - they have some quad and hex channel controls - expensive though).
The other way to do this is to use 2 PGA2310's per channel in balanced mode. The absolute gain error on these devices is very impressive, so this is indeed a viable option and what I have on my short list for my next pre-amp after I complete my current 'minimalist' design.
The other way to do this is to use 2 PGA2310's per channel in balanced mode. The absolute gain error on these devices is very impressive, so this is indeed a viable option and what I have on my short list for my next pre-amp after I complete my current 'minimalist' design.
Bonsai said:
I know the PGA2310s already, thanks for the information anyway.
Because they used in Pass Labs X0.2 pre. (Nelson keeps closed his mouth when asked for these) but it is obvious because they are also Texas, and it is commonly known that Nelson has a colaboration with Texas.
Thanks also for moxamp link.
Fotios
fotios said:
Hmmmm, I never noticed that absolute max voltage spec before. The "recommended" max voltage is +/-15v. I think with +/-33v it would melt itself right off the board.
I could try this and get back to you, but I don't think it will survive very long at that voltage.
theAnonymous1 said:I think with +/-33v it would melt itself right off the board.
I bet you are only supposed to attempt that if you have the PowerPAD packaged version with a hefty groundplane/heatsink. With the normal package, a copper heatsink thermal epoxied to the plastic should help but still might not be enough.
theAnonymous1 said:
look at the data sheet, see the big peak in the upper end of the frequency response curves? that's instability, and your chip might be running so hot because it's oscillating. 12mA of idle current is a bit high for an op amp, but not excessive, so if your device is getting really hot, it must be drawing a lot of current, and oscillation will cause that, and at between 90 and 100 Mhz, the oscillation might not show up on an oscope unless it's a 50Mhz or faster scope (it will show up on a 50Mhz scope, but at a reduced amplitude, unless it's a digital scope, and it won't show up at all because of the scope's Nyquist "brick wall").
QSerraTico_Tico said:
So what?
Maybe Anonymous1 has to fix a possible mounting problem, or to replace some caps or resistors to reduce the GBP, arround the circuitry of THS4130.
There is also the option of OPA1632 which is almost the same with THS4130 but with lower GBP.
In practice now, it is or not suitable these two ICs for audio balanced signals according to your oppinion?
If you have any objection about, say it as well the reason for it.
We need clear and reliable informations from your experience, before we proceed in unnecessary expense.
Fotios
fotios said:
While oscillating it cannot sound good.
It escapes me why using a very high-bandwith opamp for ...........audio.
Besides that the THS4131 has a peaking frequency response, not a good thing
Edit: lineup needs a lot more words saying the same
Op-Amps
like that .. with a Gain Bandwith of above 100 MHz and some times even higher
must be handled with care for use within Audio.
It is not as easy as using OPA2134, NE5534, OPA228 and OP27 and other good audio chips.
Already when we use such as AD797, OPPA2604, OPA627, LT1028 begins our issues.
This involves proper PCB and Supply decoupling
as well as be careful of what impedances/capacitances/circuitry the IC will see.
The issue here is, that people think they are smart
when using very fast amplifiers designed aminly for video and HF applications.
Without taking proper actions to ensure stability and fight oscillations.
There is no problems nowadays to find several GiGA hertz chips.
If we think we will have the ultimate Audio Amplifier, that way.
But this is in vain, people.
Because most of that 2 000 000 000 Hertz capacity is a vaste anyway.
For producing signals that are close to perfect at and around 1000 Hertz.
It would be like using one 1000000000 MHz crystal
for one 10000 Hertz application.
Plain daarn stupidity from sensible Audio Engineering point of view.
To produce 44100 Hz for our CD Players is often used ~4 MHz resonances.
This is well enough. 4 MHz / 40 kHz = 100:1
As a note,
the well respected old NE5534 has got Gain Bandwidth 13 MHz.
And the OPA2134 is something similar.
This gives very fast amplifiers for audio.
Fast enough at normal gain
😎 Lineup Audio .. not Radio 😎
😎 Lineup Audio .. not Radio 😎
😎 Lineup Audio .. not Radio 😎
like that .. with a Gain Bandwith of above 100 MHz and some times even higher
must be handled with care for use within Audio.
It is not as easy as using OPA2134, NE5534, OPA228 and OP27 and other good audio chips.
Already when we use such as AD797, OPPA2604, OPA627, LT1028 begins our issues.
This involves proper PCB and Supply decoupling
as well as be careful of what impedances/capacitances/circuitry the IC will see.
The issue here is, that people think they are smart
when using very fast amplifiers designed aminly for video and HF applications.
Without taking proper actions to ensure stability and fight oscillations.
There is no problems nowadays to find several GiGA hertz chips.
If we think we will have the ultimate Audio Amplifier, that way.
But this is in vain, people.
Because most of that 2 000 000 000 Hertz capacity is a vaste anyway.
For producing signals that are close to perfect at and around 1000 Hertz.
It would be like using one 1000000000 MHz crystal
for one 10000 Hertz application.
Plain daarn stupidity from sensible Audio Engineering point of view.
To produce 44100 Hz for our CD Players is often used ~4 MHz resonances.
This is well enough. 4 MHz / 40 kHz = 100:1
As a note,
the well respected old NE5534 has got Gain Bandwidth 13 MHz.
And the OPA2134 is something similar.
This gives very fast amplifiers for audio.
Fast enough at normal gain
😎 Lineup Audio .. not Radio 😎
😎 Lineup Audio .. not Radio 😎
😎 Lineup Audio .. not Radio 😎
Bonsai said:
There is also the AD5222 or AD5228 which does not need a MCU for control. One or two push buttons are enough to drive them.
What is your oppinion?
Fotios
QSerraTico_Tico said:
While oscillating it cannot sound good.
It escapes me why using a very high-bandwith opamp for ...........audio.
Besides that the THS4131 has a peaking frequency response, not a good thing
Edit: lineup needs a lot more words saying the same
The subject that cause our interest about THS4131 or OPA1632 it is their "Balanced in - Balanced out" property and not so much their GBP.
So, your reply it is again deficient.
On the contrary, the reply of Lineup it is complette and detailed.
You must be focused in this and not in brief explanations for issues that are commonly known to most of us.
We are full and tired from "commando-replies" which explodes like grenades in our head.
Regards
Fotios
I agree, monolithic opamps lack the x-factor. I started exploring discrete designs because of that.
There are IC's that are almost as good as the real thing. By combining opamps with different sound signature, you can get very neutral results, but still not that magic glow that draws you closer to the music.
E.g. AD8599 as L/R and AD825 as active ground is an utterly neutral combination.
There are IC's that are almost as good as the real thing. By combining opamps with different sound signature, you can get very neutral results, but still not that magic glow that draws you closer to the music.
E.g. AD8599 as L/R and AD825 as active ground is an utterly neutral combination.
lineup said:
I think there are very good reasons to use very high speed opamps like the THS4031 and THS4130 for audio. Not for amplification of MHz range signals; not for the slew rate either, since audio signals don't slew fast at all. The GBP still means something, even for signals at frequencies far below its value.
What is important is that these ICs have a lot of gain in the audio band. Assuming dominant pole compensation (pretty safe, except for the 5532), there are only two values that determine how much gain an opamp has at 20khz:
1) The GBP
2) The DC gain
If you follow the gain curve backwards, it rises 20db/decade from the GBP until it equals the DC gain, then stops. The THS4031 has about 80db gain at 20khz, since it is about 4 decades from the GBP on that chip. You have more available gain, so you have more negative feedback than another IC in the same closed-loop configuration. This combined with the good open-loop linearity of these parts makes them perform extremely well in the audio band. More NFB in an opamp circuit, all other things equal, improves just about every measurable performance metric.
- Status
- Not open for further replies.