Most, if not all amplifiers made today are not DC coupled from input to output.
I think Sansui made some in the 1980´s or so.
Indeed, but you also lose 100 + other components that the op-amp consists of in the signal path.
Add or lose bass - what is on the recording anyway, and what balance do you have in your setup?
Behringer makes some nice digital equalizers, I am using the SPDIF in and then out of a DEQ1024 if I feel I want to correct the frequency balance.
Last edited:
Looking at the AK4396 data sheet (http://www.akm.com/akm/en/file/datasheet/AK4396VF.pdf page 34), it shows a recommended output LPF circuit using 2.4K resistors, a 3.3nF capacitor between the DAC differential outputs, and 680pF capacitors in the feedback loop and the opamp's non-inverting input. It indicated that this provides a low pass filter down -0.8db at 80 KHz. However, when I run the circuit in SiMetrix (Free Demo, SIMetrix/SIMPLIS analog circuit simulation software), it comes up with an output of -1.6db at 80 KHz---TWICE the drop in frequency response that the datasheet claims.
Can anyone explain this to me? Is the datasheet wrong?
Can anyone explain this to me? Is the datasheet wrong?
Good that you are happy with your setup - that's all that matters. In mine, I found I lost bass weight rather than extension, and that wasn't a compromise I was willing to make.
Aha! My bad......if I run it with the "Ideal OpAmp" then it runs as the datasheet suggests. The SiMetrix simulator automatically put in a TL072 opamp, which shows the decreased response. Unfortunately, the database of opamps in SiMetrix is woefully out-of-date and doesn't include any of out newest favorites--the LM4562, OPA1612, OPA627, or even the NE5534. Does anyone know how to import data for these new chips to SiMetrix?
Tl072 is not suited as a IV... perhaps as filter... but even then the distortion will be too high.
For these kind of DAC boards you need an opamp that can perform the task of IV... so extreme low noise is not that important. Consider a LT1028 or AD826...
The board needs a dual opamp, so single opamp types like NE5534 or OPA627 will cause some problems...
I believe the AK4396's analog output is VOLTAGE not CURRENT. But, as you say, when dealing with a one-volt RMS source, the noise performance of the opamp is rather irrelevant. Oddly enough, noise performance is probably the strongest point of an LT1028. An AD826 has horrible THD and noise specs, although at 350V/usec, it IS faster than greased lightning!!
THD ain't bad at all... not true! And noise is hardly worse than the ne5532. And as a big plus: it sounds way better than the ne5532 or even an OPA2604.
check the data on https://by-rutgers.nl/IV-converter.html
and https://by-rutgers.nl/PDFiles/OpAmp lowpass AD826 Dist.pdf
https://by-rutgers.nl/IV-converter.html
From the Analog Devices datasheet:
"The 50 MHz bandwidth and 350 V/µs slew rate make the AD826 useful in many high speed applications including: video, CATV, copiers, LCDs, image scanners and fax machines."
Interestingly, they only quote THD (-76db) at 1 MHz---I guess that's because it's intended as a VIDEO amplifier. Even your 'rutgers' reference only quotes -86db THD (0.0046%). An LM4562 is on the order of -130db THD.
The voltage noise (15nV√Hz) of an AD826 is THREE times worse than an NE5532 (5nV√Hz); the current noise is about twice (1.5pA vs 0.7)
Your 'rutgers' references are talking about current-to-voltage converters; the AK4396 has a VOLTAGE output, unlike the CURRENT outputs of a PCM1792/PCM1795. So it's a bit different application, methinks.
Last edited:
-86db THD (0.0046%) is wrong.... get your math right for it's -104db!!! LM4562 sounds crappy... commonly known... Just try it... sound reproduction is more than thd and noisefloor...
Anybody tryed the THAT Corporation 1200 series op-amps?
It is a balanced input reciver, and was designed as a transformer substitute.
http://www.mouser.se/Search/m_ProductDetail.aspx?THAT/1200P08-U/&qs=sGAEpiMZZMv1cg40gLeF3ohbwM7uF68q
It is a balanced input reciver, and was designed as a transformer substitute.
http://www.mouser.se/Search/m_ProductDetail.aspx?THAT/1200P08-U/&qs=sGAEpiMZZMv1cg40gLeF3ohbwM7uF68q
Could I gently suggest you check your own maths first?
https://lightmachinery.com/optical-design-center/more-optical-design-tools/percentage-to-db/
You could.... but it doesn't mean the AD826 is a bad opamp for a lp...
Try it, it sounds great. Besides even that -86db is beyond the limits of you loudspeakers.
Besides... the THD figures (-130db) of the Lm4562 are not measured values in a specific lp circuit, so absolutely not comparable to the AD826's figures.
Last edited:
The TL072 THD spec is -90db, which you say is too high. The AD826 is -86db
Hmmmm......
What is the "specific lp circuit" that you refer to?
BTW, the formula to express THD is
db = 20 * [log (voltage ratio)]
0.0046% equals 0.000046. The log of this is -4.3372; db = -86.74484337
But, if you get tired of listening to your AD826, you could always watch cartoons through it...
Again... you're comparing datasheets with measured data, that's no comparisson at all! Apparently you miss the entire context of this thread... your bad.The TL072 THD spec is -90db, which you say is too high. The AD826 is -86db
Hmmmm......
What is the "specific lp circuit" that you refer to?
BTW, the formula to express THD is
db = 20 * [log (voltage ratio)]
0.0046% equals 0.000046. The log of this is -4.3372; db = -86.74484337
But, if you get tired of listening to your AD826, you could always watch cartoons through it...
And your ad hominem type of reactions are way out of line...
Yes, I wonder why nobody tried them as a transformer electrical substitute, just to see how it sounds.
They have to be assembled on a piggy-back small board though.
There's also the INA134 (single) and INA2134 (dual), both PDIP and affordable, and also a SOIC (single) that is worth trying too: INA826. Specs of the latter are better than those others and the THAT.
I was merely pointing out that the AD826 was designed as a VIDEO amplifier.
The content of this thread is NOT about current-to-voltage converters! The AK4393/96 DACs output VOLTAGE.
A large math error--TWICE!- and the chastising post is "get your math right"???????
?????????
Thank you for the info on the INA826, will order both the THAT and the INA, and have a go at them.
Must be something wrong with the outputstage, when the raw output from the DAC chip itself sounds so much more relaxed and clear.
Yes, you are absolutely right in pursuing the sound you want.
Just one thing: sometimes some types of distortion sound better or nicer (e.g.: tubes). In general it's not recommended to link yourself directly to the DAC, and some kind of buffering should protect it.
I believe someone may have tried it, but I found no info on it. You should also try the INA2134, which comes in PDIP version and it's considered an "audio" IC by TI.
Just one thing: sometimes some types of distortion sound better or nicer (e.g.: tubes). In general it's not recommended to link yourself directly to the DAC, and some kind of buffering should protect it.
I believe someone may have tried it, but I found no info on it. You should also try the INA2134, which comes in PDIP version and it's considered an "audio" IC by TI.