• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

SMSL D1 - ROHM DAC for everyone

I was not expecting this.
We were.

 
Changed buffer opamp in d1 to opa1612 and I can confirm that it has made substantial difference in sound quality. I was not expecting this.
Wondering why SMSL didn't use opa1612/1656. Maybe because this is a low price category product.
Measurements or its merely psychoacoustics ‍

But for real do you have any sort of data, measured anything about the mod? I found the idea of OP amp rolling quite fun, but virtually every test done with data to back it up found that rolling OP amps at best doesn't change sound in any meaningful way and at worst does actually degrade audio quality.
 
Changed buffer opamp in d1 to opa1612 and I can confirm that it has made substantial difference in sound quality. I was not expecting this.
Wondering why SMSL didn't use opa1612/1656. Maybe because this is a low price category product.
And you compared that to a second D1 in a real, externally controlled blind test and found out?
 
the rohm chip in d1 and d200 is $10(BD34352EKV) , unlike the rohm chip in d300 and dc-elite that is $100 (BD34301EKV), not sure how much of a difference it makes to the ear vs the second stage output amp that comes after the dac chip signal. so do you hear a significant improvement between d1 and d200 ?
We haven't done the blind test yet, as it requires a certain amount of effort. Perhaps I can manage a simple comparison before Christmas.
However, I don't expect a significant or clearly audible difference, if there even is one.

Most insiders assume that the Rohm DAC chips are identical, come from the same production line, and differ only due to selection based on measured values.
The datasheets and evaluation boards support this assumption.
Given the small differences, these differences are more likely to occur in the mind and the evaluation of the measured values than in the sound and the music heard.
 
Can anyone tell me what kind of chip this is?
Its marked A08QQAIx
1765205479051.png
 
Most insiders assume that the Rohm DAC chips are identical, come from the same production line, and differ only due to selection based on measured values.
Because its the exact same chip but failed the binning to become BD34301EKV.
 
Changed buffer opamp in d1 to opa1612 and I can confirm that it has made substantial difference in sound quality. I was not expecting this.
Wondering why SMSL didn't use opa1612/1656. Maybe because this is a low price category product.
I’m glad someone else tried swapping the OP amps and got results similar to mine. Earlier I mentioned that replacing the stock single‑ended, unlabeled OP with an OPA2134 worked out really well. one month Later I experimented with changing the two IV‑stage, unlabeled OPs to MUSES OPs — the sound signature shifted quite a bit, but honestly I didn’t like it. I quickly reverted to the original setup, and the sound went back to how it was before the mod. Of course, this could all just be placebo since I didn’t run any measurements, so take it with a grain of salt.

Nevertheless, I believe SMSL must have had solid technical reasons for replacing the unlabeled stock OP amp with an OPA1612, developing the D200, and positioning it as a high‑priced model. At the very least, it’s reasonable to assume they conducted proper measurements and validation. Haha.
 
I’m glad someone else tried swapping the OP amps and got results similar to mine. Earlier I mentioned that replacing the stock single‑ended, unlabeled OP with an OPA2134 worked out really well. one month Later I experimented with changing the two IV‑stage, unlabeled OPs to MUSES OPs — the sound signature shifted quite a bit, but honestly I didn’t like it. I quickly reverted to the original setup, and the sound went back to how it was before the mod. Of course, this could all just be placebo since I didn’t run any measurements, so take it with a grain of salt.

Nevertheless, I believe SMSL must have had solid technical reasons for replacing the unlabeled stock OP amp with an OPA1612, developing the D200, and positioning it as a high‑priced model. At the very least, it’s reasonable to assume they conducted proper measurements and validation. Haha.
As long as people don't use genuine blind tests with identical original devices and/or meaningful measurements to verify modifications, they're deceiving themselves in 99% of cases and wasting a lot of time on nonsense.
That's precisely why I started working with a second, identical device without any modifications for comparison 25 years ago. It's much cheaper than wasting so much time.

Disguised components, including op-amps, primarily serve to conceal the fact that they've found a well-functioning and inexpensive op-amp, sometimes not even from the audio sector, thus securing this competitive advantage. Standard NE5532 op-amps were never disguised in the past by SMSL, Topping, or any other manufacturer.
The D200 also features two unidentified op-amps in the I/V stage directly after the DAC chip, in addition to the six OPA1612 op-amps.

SMSL typically uses the circuit diagrams and evaluation boards of the chip manufacturers as templates for its devices and modifies them to achieve good measured values, as can be clearly seen in the D1 and D300. Unlike other manufacturers, SMSL is not very innovative in its circuitry and usually limits itself to specific components with which it has experience.

The D200 is quite different. This device appears to have been developed by an external designer, as it is built completely differently than what is typical for SMSL, with different components and circuit solutions, and is significantly more complex in many areas.
For example, it features a true analog balanced preamplifier function with one NJU72315 per channel for volume control. This makes the D200 a DAC with an analog preamplifier section after the DAC chip. It's a shame SMSL missed the opportunity to add analog inputs.
The D200 also contains two unidentified op-amps in the I/V conversion stage. Since there are no coupling capacitors, the subsequent OPA1612s are presumably part of a DC coupling circuit.
The use of a toroidal transformer for the power supply is also unusual these days. The elaborate generation of numerous dedicated voltages for the various sections is also noteworthy.
The digital section appears to be divided into several areas, and a total of five crystals are used.
 
Disguised components, including op-amps, primarily serve to conceal the fact that they've found a well-functioning and inexpensive op-amp, sometimes not even from the audio sector,
If i had to guess its either locally manufactured or some cheap JRC branded opamp.

The OPA1612 became the de-facto standard for dacs as its one of the best performing opamp without external components. Yes the NE5532 can do -140dB distortion but it requires careful frequency compensation and input low pass filtering.
 
When i get my new multimeter im gonna reverse engineer the output stage as it looks like the bog standard TI recommended design. If thats the case them i'm gonna have to make a separate buffered LPF filter for the balanced out mod.
 
When i get my new multimeter im gonna reverse engineer the output stage as it looks like the bog standard TI recommended design. If thats the case them i'm gonna have to make a separate buffered LPF filter for the balanced out mod.
Take a look at page 9/17 in the BD34302EKV Evaluation Board Users guide; Rohm seems to be suggesting exactly that there, as a kind of high-end version. It's a continuation of the circuit with the two I/V op-amps on page 8/17.
 
Take a look at page 9/17 in the BD34302EKV Evaluation Board Users guide
The schematic is so poorly drawn that its impressive. From the eval board the second half of the NE5532 is doing absolutely nothing.
This just showed that i will have to build the same buffer for the negative output or make my own buffer tapped after the LPF.
 
They finally fixed my DAC. They replaced the op amp.
But they couldn't replace the "A08QQAIx" chip.
 
They finally fixed my DAC. They replaced the op amp.
But they couldn't replace the "A08QQAIx" chip.
it seems the SMAL D1 originally used the NE5532. Replacing it with the OPA1612 is clearly an upgrade, as the specifications are superior and may enhance certain aspects of the sound. Thanks a lot for the information!
 
it seems the SMAL D1 originally used the NE5532. Replacing it with the OPA1612 is clearly an upgrade, as the specifications are superior and may enhance certain aspects of the sound. Thanks a lot for the information!
And where does it come from about such a replacement?..
 
Changed buffer opamp in d1 to opa1612 and I can confirm that it has made substantial difference in sound quality. I was not expecting this.
Wondering why SMSL didn't use opa1612/1656. Maybe because this is a low price category product.
Your changes are extremely unlikely to have made any audible difference. They also likely made the DAC perform and measure objectively worse than with the original op amps. See [1] or [2] for some additional info on this topic.

If you tested this by sighted testing like listening, then switching op amps and then listening again, you can't reliably judge the two configurations. Remember that echoic memory is only a couple of seconds long, but it takes minutes to switch the chips. In addition to numerous additional pitfalls, sighted testing is also strongly affected by bias - even when you are aware of the bias existing.
 
Your changes are extremely unlikely to have made any audible difference. They also likely made the DAC perform and measure objectively worse than with the original op amps. See [1] or [2] for some additional info on this topic.

If you tested this by sighted testing like listening, then switching op amps and then listening again, you can't reliably judge the two configurations. Remember that echoic memory is only a couple of seconds long, but it takes minutes to switch the chips. In addition to numerous additional pitfalls, sighted testing is also strongly affected by bias - even when you are aware of the bias existing.
But from the page[2] you provided, replacing the NE5532(-124dB+0.0032%) with OPA2134(-137dB+0.0028%) or OPA1612(-138dB+0.0028%) in the LPF slot resulted in positive improvements.doesn’t that prove there is evidence of measurable improvement?
 
Back
Top Bottom