Topping DX1 op-amp used?

[mike7877]

I've looked high and low, and can't seem to find anyone, anywhere, mentioning which op amp(s) is(are) used in the DX1.

Does it have to be one of the top of the line models (something like the OPA1612, OPA604), or can op-amps, too, be made to perform significantly better than their datasheets suggest with a few strategically placed passive components on their pins?

Obviously I'm asking because until Topping, only stuff in the tens-of-thousands came close to measuring like the $100USD DX1. I'm beginning to wonder if it's always been this simple and these "high end" DACs and interfaces just feature a bunch of non-terminated surface mount components placed plausibly around the actual circuits, to throw us off and get us to dig deep into our wallets for their 'genius'

 

[Dunring]

I've had a couple of those, and the amplifier portion sounded good but not the best. RCA out as a DAC it was really good, I'd use it every day in DAC mode. I think the tech has hit a wall for performance and affordability.
I have a Topping D90 and D30 Pro, and there's nothing magical you can't hear using a DX1 in DAC mode.

 

[DVDdoug]

I don't know anything about the Topping...

It's not that hard to make a DAC that's better than human hearing. (There's always a possibility of better measurements...)

If the op-amp is driving headphones it needs more current output capability than a typical op-amp.

Otherwise op-amps used in DACs are not that critical. The are working at line-level (about 1V) with little or no (voltage) gain so signal-to-noise isn't a big problem.

or can op-amps, too, be made to perform significantly better than their datasheets suggest with a few strategically placed passive components on their pins?

The datasheet is probably "ideal". ...No power supply noise or ground loop noise, very-short wires/connections, etc.

Some companies may select op-amps, testing them and then putting the best ones in their higher end products or re-selling or scrapping the "rejects". But that's often a nightmare... You can end-up rejecting 90% of 100% of an order, and the manufacturer/seller considers the parts perfectly good. I worked for a company once that was selecting parts (not op-amps) and they eventually had to make a design modification to keep in-production.

But DACs can be "improved". Where I work (non-audio) we have some ADCs & DACs and they are individually calibrated-corrected in software for better accuracy & linearity than we get with the raw chip. In some cases we only correct for offset & slope, and in once case there are 10 correction ranges (slope corrections) between 0 and 100% . They are operating at slower than audio speeds and they are only 8 or 12-bits. That probably won't work(or won't work as well) at 16-bits or higher because audio DACs are already pretty good and there may be as much noise & drift as constant error.

 

[mike7877]

I've had a couple of those, and the amplifier portion sounded good but not the best. RCA out as a DAC it was really good, I'd use it every day in DAC mode. I think the tech has hit a wall for performance and affordability.
I have a Topping D90 and D30 Pro, and there's nothing magical you can't hear using a DX1 in DAC mode.

I'm pretty sure I saw the THD+n measurement of the DX1 tested to be within 1-2dB of the RCA outs - the headphone section is remarkably good, part of the reason I made this thread

 

[mike7877]

I don't know anything about the Topping...

It's not that hard to make a DAC that's better than human hearing. (There's always a possibility of better measurements...)

If the op-amp is driving headphones it needs more current output capability than a typical op-amp.

Otherwise op-amps used in DACs are not that critical. The are working at line-level (about 1V) with little or no (voltage) gain so signal-to-noise isn't a big problem.


The datasheet is probably "ideal". ...No power supply noise or ground loop noise, very-short wires/connections, etc.

Some companies may select op-amps, testing them and then putting the best ones in their higher end products or re-selling or scrapping the "rejects". But that's often a nightmare... You can end-up rejecting 90% of 100% of an order, and the manufacturer/seller considers the parts perfectly good. I worked for a company once that was selecting parts (not op-amps) and they eventually had to make a design modification to keep in-production.

But DACs can be "improved". Where I work (non-audio) we have some ADCs & DACs and they are individually calibrated-corrected in software for better accuracy & linearity than we get with the raw chip. In some cases we only correct for offset & slope, and in once case there are 10 correction ranges (slope corrections) between 0 and 100% . They are operating at slower than audio speeds and they are only 8 or 12-bits. That probably won't work(or won't work as well) at 16-bits or higher because audio DACs are already pretty good and there may be as much noise & drift as constant error.

Do you work at a place that makes oscilloscopes?


I think I read that the AK4493S has an output impedance of ~450 ohms. If driving 10-20k ohm inputs, that's almost enough! I think my RME Babyface Pro is 300 ohm output impedance on the balanced outs.

Sometimes I wonder why there are any op-amps in DACs at all... Kinda like how I wondered why preamps had to be anything more than a good passive pot with some gold plated RCA/XLR in's/outs. I found one which I now use as my preamp lol (one thing less to break/break down. Been a long time since I had a source that needed more gain than line level!)

Has a good alps pot with 4 conductors for balanced. Not their best but it's up there!
Actually it's not. But it's pretty good. The only reason it's not "up there" is because their best 4 conductor is a ridiculous $2,000 !!!!!!!!!!! (yes, it's for audio)
I'm betting they don't sell a lot of them, but the materials and time spent can't seriously justify two whole thousand dollars from a bulk supplier... It must go on preamps that sell for $200,000 (little do they know, with just a couple gold plated female RCA / XLR jacks and that pot alone, they could have an even better sounding preamp, AND a ZR1!)
(and insurance and fuel for 5 years and and and and and)

edit: think about it.. for $2000 you can get an entire computer. Pretty high end, too. And they charge that for a pot

 

[voodooless]

Sometimes I wonder why there are any op-amps in DACs at all...

Because some actually are not voltage output. This AKM DAC is though. Some even have 2V line drivers or even headphone amps. It all depends on the use case and budget.

Kinda like how I wondered why preamps had to be anything more than a good passive pot with some gold plated RCA/XLR in's/outs. I found one which I now use as my preamp lol (one thing less to break/break down. Been a long time since I had a source that needed more gain than line level!)
View attachment 318536
Has a good alps pot with 4 conductors for balanced.

That is actually not the best idea. Balanced connection work because the - and + sides track each other perfectly (only inverted). A multi-gang pot is never perfect. You’ll always have differences in the channels. This will introduce inaccuracies in the balanced signal, leading to additional distortion or a DC offset.

 

[mike7877]

Because some actually are not voltage output. This AKM DAC is though. Some even have 2V line drivers or even headphone amps. It all depends on the use case and budget.

Yes there are reasons. That one wasn't in the forefront of my mind. My point though, is all the voltage output DACs I've looked at the datasheets of, seem to be pretty capable of functioning without op-amps, but every design I've come across has them

That is actually not the best idea. Balanced connection work because the - and + sides track each other perfectly (only inverted). A multi-gang pot is never perfect. You’ll always have differences in the channels. This will introduce inaccuracies in the balanced signal, leading to additional distortion or a DC offset.

I know, I know... The inaccuracy will be amplitude though, and very likely less than a quarter dB if the pot is any good. Pretty much same thing if you're not doing balanced. Since there are two (per channel for balanced), I think one will take care of the error of the other. Well, fluctuations would be double, but the average error would be the same (as unbalanced). I think....

You can't hear a quarter dB, but it will affect imaging if the system/room is good enough


edit: the DC offset on the amplifier side shouldn't be too much of a problem. Most speakers have crossovers, and if you have an active system the pot preamp wouldn't be used this way. Even still, with active crossovers, all drivers should have a high pass cap a couple octaves below crossover point. In the voltage likely to be seen, only tweeters may be in danger of overheating without a cap (it should still be there in case spikes at power on develop unnoticed because 1" domes don't thump)

edit2: personally I have two sources, one balanced, one unbalanced, and I always use RCA out (taken from the correct polarity of the balanced input)

 

[voodooless]

Yes there are reasons. That one wasn't in the forefront of my mind. My point though, is all the voltage output DACs I've looked at the datasheets of, seem to be pretty capable of functioning without op-amps

Not if you want to drive a headphone.

I know, I know... The inaccuracy will be amplitude though, and very likely less than a quarter dB if the pot is any good. Pretty much same thing if you're not doing balanced. Since there are two (per channel for balanced), I think one will take care of the error of the other. Well, fluctuations would be double, but the average error would be the same (as balanced). I think....

No it’s not, and it’s not just the absolute level. You’ll get a DC offset. If DC coupled equipment is used,, that will be an issue.

 

[mike7877]

Not if you want to drive a headphone.

Of course... I'm talking about DACs in things like CD players and such. Should've been clearer

No it’s not, and it’s not just the absolute level. You’ll get a DC offset. If DC coupled equipment is used,, that will be an issue.

I didn't mention the DC offset not to take away from its validity, but because I have nothing to add. You're right...

edit: and yes some equipment may not be happy passing DC, but at the levels likely to be seen with a pretty good pot, I think it'd have to be some pretty testy equipment

 

[voodooless]

Of course... I'm talking about DACs in things like CD players and such. Should've been clearer

Most DACs are still balanced, most CD players are not. And even if so, the opamp after the DAC also acts as a low pass filter to filter out high frequency modulator noise. Just look at the AK4493 datasheet, section “Analog Output”.

 

[mike7877]

Most DACs are still balanced, most CD players are not. And even if so, the opamp after the DAC also acts as a low pass filter to filter out high frequency modulator noise. Just look at the AK4493 datasheet, section “Analog Output”.

I found this under Analog Out:
"The internal switched capacitor filters attenuate the noise generated by the delta sigma modulator beyond the audio passband"

But if it doesn't, or doesn't well enough - whatever the case - would there not be a way to use one (or two or three or four) passive components on the DAC's voltage output (arranged like a speaker crossover) instead of an entire amplifier on IC (with less bandwidth than the noise being output) to do the job?

My intuition says passive would be better than active regarding distortion. Unless you needed gain, then of course, active would be the only option

edit: I see the problem the input impedance of the amplifier wouldn't be a constant. Still, it could be designed for >5k ohms, which would cover 99% of everything. It'd have to be first order then. I guess it all depends on how high the noise from the switching is, too. It makes sense why they'd do it now - to simplify things at the cost of a little performance. (that is, if there is, in fact, too much noise beyond the passband on the analog DAC output)

 

[voodooless]

My intuition says passive would be better than active regarding distortion. Unless you needed gain, then of course, active would be the only option

edit: I see the problem the input impedance of the amplifier wouldn't be a constant. Still, it could be designed for >5k ohms, which would cover 99% of everything. It'd have to be first order then. I guess it all depends on how high the noise from the switching is, too. It makes sense why they'd do it now - to simplify things at the cost of a little performance. (that is, if there is, in fact, too much noise beyond the passband on the analog DAC output)

The capacitance is the main issue here:

Don’t you think that if companies could save a few $€£ by omitting an opamps, that they would do that?

 

[mike7877]

The capacitance is the main issue here:

Don’t you think that if companies could save a few $€£ by omitting an opamps, that they would do that?

Yeah, I guess they would lol.

Which is one of the reasons why it baffles me that active preamps exist at all - especially with every single digital and analog device (except phono) being amplified with op-amps to line level already! Most power amps, if fed line level voltage, are at or past clipping - why amplify it further?! (especially when the only thing that can be added is distortion...)

Of course there is "what if the source is too quiet?"
If it's too quiet, it probably doesn't sound good either, so instead of an active preamp that'll add distortion to everything you have, why not just replace that crappy thing?

And tone controls... But how often do you see bass, treble, and loudness on anything high end past 1985

edit for clarity: I'm talking about home equipment above, single ended

 

[MaxwellsEq]

Which is one of the reasons why it baffles me that active preamps exist at all - especially with every single digital and analog device (except phono) being amplified with op-amps to line level already! Most power amps, if fed line level voltage, are at or past clipping - why amplify it further?! (especially when the only thing that can be added is distortion...)

There's history here. Go back several decades and "line level" for domestic gear - effectively tuners, cassette decks and reel-to-reel tape machines were all about a volt. When CD came out it was about 2V. I had an old active preamp feeding a power amplifier. For the power amplifier to achieve maximum output, the preamplifier needed to have positive gain. But, the CD player was the first device which had enough signal voltage to drive the power amplifier directly. This meant that "passive preamplifiers" were unrealistic before CD arrived.

BUT more importantly, the best domestic source was a turntable, which needed tons of positive gain as well as RIAA equalisation. This meant ALL preamplifiers needed to have active gain and CD players were "too loud", needing attenuation.

Fast forward to today where a turntable is a "foible" and you don't need loads of positive gain. "Passive* is a viable choice.

But you also may need a buffer (perhaps with a gain of 1), to handle output or input impedance. In practice, to guarantee that devices play well together buffers are worth having and can be designed with minimal noise and distortion.

 

[mike7877]

There's history here. Go back several decades and "line level" for domestic gear - effectively tuners, cassette decks and reel-to-reel tape machines were all about a volt. When CD came out it was about 2V. I had an old active preamp feeding a power amplifier. For the power amplifier to achieve maximum output, the preamplifier needed to have positive gain. But, the CD player was the first device which had enough signal voltage to drive the power amplifier directly. This meant that "passive preamplifiers" were unrealistic before CD arrived.

BUT more importantly, the best domestic source was a turntable, which needed tons of positive gain as well as RIAA equalisation. This meant ALL preamplifiers needed to have active gain and CD players were "too loud", needing attenuation.

Fast forward to today where a turntable is a "foible" and you don't need loads of positive gain. "Passive* is a viable choice.

But you also may need a buffer (perhaps with a gain of 1), to handle output or input impedance. In practice, to guarantee that devices play well together buffers are worth having and can be designed with minimal noise and distortion.

Definitely some truth. I think my view of older stuff may be a bit biased from not being experienced with a wide variety of devices. My oldest working (most of the time lol) device is a Pioneer RT-707, and its output is pretty hot. Well, it can be - there are adjustable pots on the back to adjust the output level, and if you used CrO2 tape or good Type I (I always have), you could record (like I did) with peaks even past +3dB without saturation! (at least on the semi-accurate VU meter). Like I said, I don't have a lot of experience with older gear though, and I may have been spoiled with the RT-707's. If its gain is higher than most RTRs, it was probably to be able to work optimally with a wide variety of equipment - it was rack mountable and used by the military a lot, if what I've read over the years has been correct. Mine is probably a little more than overdue for a recap by now, for safety reasons at least, on the power supply. Lol it's 45 years old now!

Also, the tape deck I spent most of my time with, a lower-mid level later model Technics from the early 90s with excellent sound quality on most measures for tape (except for probably wow/flutter because it was new enough that its electronics were transparent enough for 1/8" tape at 1 7/8 IPS, but the cost saving measures (even with mechanical advancements of the late 80s/early 90s) made it obvious enough that the deck was not high end through and through, I also used quality tape in it whenever I could, which allowed me to record with peaks between +3 and +5dB, depending on the tape type and where the energy was (on the spectrum).

 

[Njoy]

I've looked high and low, and can't seem to find anyone, anywhere, mentioning which op amp(s) is(are) used in the DX1.

Does it have to be one of the top of the line models (something like the OPA1612, OPA604), or can op-amps, too, be made to perform significantly better than their datasheets suggest with a few strategically placed passive components on their pins?

Obviously I'm asking because until Topping, only stuff in the tens-of-thousands came close to measuring like the $100USD DX1. I'm beginning to wonder if it's always been this simple and these "high end" DACs and interfaces just feature a bunch of non-terminated surface mount components placed plausibly around the actual circuits, to throw us off and get us to dig deep into our wallets for their 'genius'

The headphone section uses RT6863 op amps in parallel, as they are two of them. I had it for a week and sold it.