Amanero boards cost from about $40 to $100. That is to say, the clone versions are the cheap ones and the real one which is made in Italy costs close to $100. You don't necessarily need an Amanero board if you use upsampling. Depending on which upsampler board used, you might want an Amanero, XMOS or maybe some other USB to SPDIF board.
Last edited:
Consider an R-core transformer instead of a toroid. They pass much less mains noise through the windings.
Also for op-amps. the ADA4898 is worth looking at. The dynamic specs are very good - noise is less than the LME49720 - and it has an ultra-linear input stage, which makes it especially suitable for I/V.
Op-amp requirements for a I/V converter?
I just looked at the ADA4898. Turned out not to be worth looking at. Distortion at audio frequencies is poor. The data sheet even says on the first page, "for use in 16-bit and 18-bit systems." You may be unaware that modern audio DACs are capable of performance well in excess of 18-bits. So, please be careful not to inadvertently give bad advice if you don't really know the subject matter.
R-core transformers, as normally wound, are of split-bobbin construction. So, it is true there is less capacitive coupling from primary to secondary as compared to the way torroidal transformers are normally wound.
However, either type of core can be wound in configurations other than as we typically see them. It may be worth looking to make sure you know what you will be getting.
Both R-core and torroidal transformers tend to have less stray magnetic field leakage than E-I core transformers do.
R-core transformers, as normally wound, are of split-bobbin construction. So, it is true there is less capacitive coupling from primary to secondary as compared to the way torroidal transformers are normally wound.
However, either type of core can be wound in configurations other than as we typically see them. It may be worth looking to make sure you know what you will be getting.
Both R-core and torroidal transformers tend to have less stray magnetic field leakage than E-I core transformers do.
Can you provide a link to the latest study on this subject. Earlier wisdom seems to say the opposite. R-core is cheaper, but not necessarily better, than Toroidal.
Pete Millett said:
More details at:Power Transformers for Audio Equipment | audioXpress
Another consideration is the practical aspect of availability. There are many manufacturer of Toroidal Transformer in the US, making it widely available and cheap to ship. The R-core transformers are often shipped across Pacific from China, making it more expensive than Toroid in the US, for the same capacity VA.
Last edited:
I already have toroids sitting in boxes collected and retrieved over the years. The thing is that if you filter before the toroid I see the toroid as just as good.but since I alrady got it....then its the best to me. To that end my DAC has a standard transformer yanked from a SHURE HTS2000, a C core and a Corcom EMI filter preceding them. These days, a filter before the transformer is mandatory for me. 20-30 years ago AC was cleaner. I tend to pull what might come in handy and store it just in case.
The distortion performance of the ADA part is very good, if not quite state of the art. The feature of interest for I/V stages is it's highly linear input stage.
This is where high feedback parts can fall down when facing HF content, such as that output by noise shaping converters, despite otherwise excellent performance. Many experienced designers advocate circuits that are comfortable facing large amounts of RF content even if their THD is somewhat worse. I'm not being dogmatic on the subject, I am offering this to people who may wish to explore alternatives. As such, I think this is fair advice to offer.
As far as transformers are concerned, R-cores typically have less capacitance than toroidals, Toroidals sometimes have an electrostatic screen, which will help, but still subjectively I've always found R-cores to sound cleaner, as have many other people. Again I think this is a fair recommendation to make.
Re. your comments not to give bad advice, I've tried to show why I've given the advise I have. People interested can read up on the subject matters described if they want to - part of the fun of all this is there is always so much to learn.
More generally speaking, you've helped a lot of people with their adventures, and generated a lot of interest with your considerable contribution. I admire that.
But I was also concerned that you were speaking with great confidence on matters you clearly did not understand, such as how SRCs work, or noise performance in regulators. If I came across as abrasive, I apologies, but you may want to consider being more open to alternative solutions.
R-cores are better in that they usually have less interwinding capacitance than toroidals, hence they couple less mains noise into a power supply. The piece from Audio Amateur doesn't mention this, but it's a consequence of how they are wound.
That all sounds very reasonable. It's just one aspect people may want to experiment with if they have the inclination.
@Spartacus, My concern regarding opamps is we have readers at all different technical levels interested in making better DACs. When someone says a particular opamp is worth taking a look at, that can lead to people popping them into sockets willy-nilly without much understanding. That sort of thing would probably not be conducive to forward progress in getting the best out of the DACs we are discussing here.
Maybe the opamps in question might have been worth taking a look at 6 years ago, but DAC resolution has improved a lot, and given that we don't have much in the way of test equipment to evaluate opamps from a time back when 18-bit performance was SOA it seemed like more of a distraction than something likely to be helpful. Thus, I have been encouraging people to follow ESS recommendations which I have shown can work to meet ESS specifications and also have compared SQ with Benchmark DAC-3 to verify that the result also sounds good and is getting pretty close to a well-known standard among modern audio DACs. If we can do that well with the LME49720 opamps ESS specifically and strongly recommends then it seems we would be getting pretty off-track to try something that doesn't look like it can do much better (or likely do as well) for this DAC chip.
If you had made the point that a particular opamp that is more typically used for lower resolution systems might be worth looking at for those that have some expertise as well as test equipment to verify results, then I would not take issue with it.
Maybe the opamps in question might have been worth taking a look at 6 years ago, but DAC resolution has improved a lot, and given that we don't have much in the way of test equipment to evaluate opamps from a time back when 18-bit performance was SOA it seemed like more of a distraction than something likely to be helpful. Thus, I have been encouraging people to follow ESS recommendations which I have shown can work to meet ESS specifications and also have compared SQ with Benchmark DAC-3 to verify that the result also sounds good and is getting pretty close to a well-known standard among modern audio DACs. If we can do that well with the LME49720 opamps ESS specifically and strongly recommends then it seems we would be getting pretty off-track to try something that doesn't look like it can do much better (or likely do as well) for this DAC chip.
If you had made the point that a particular opamp that is more typically used for lower resolution systems might be worth looking at for those that have some expertise as well as test equipment to verify results, then I would not take issue with it.
Hi,
Did you looked at this one even cheaper which seems to have good reviews?
SMSL M8A DAC DSD512/768kHz HIFI Audio Decoder Amplifier USB Receive XCroe200 Xu208+ES9038Q2M Coaxial/XMOS Asynchronous 2017 New-in Headphone Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group
Did you looked at this one even cheaper which seems to have good reviews?
SMSL M8A DAC DSD512/768kHz HIFI Audio Decoder Amplifier USB Receive XCroe200 Xu208+ES9038Q2M Coaxial/XMOS Asynchronous 2017 New-in Headphone Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group
Spartacus, at any time if you feel I am giving incorrect advice to people or explaining something in a way that is misleading or incorrect, then I would ask that you bring it up at the time you believe there is a problem. It is my wish to provide people with accurate information in a form that is hopefully as intuitive as possible. I am by no means perfect and I do make mistakes from time to time. No question about that.
However, to drop vague complaints sometime later when I have no idea what you are referring to is not helpful to me or other people reading your words. If you would like to go back and quote something I wrote that you believe to be incorrect that would fine. Maybe we can get it straightened out now so as to make sure that correct information is given out in this thread.
I can't see why someone would have any issue using the ADA part instead. Obviously some people just want to follow an easy path to good results. That's fine of course. Others like to try out alternatives, which may be better in some respect. Surely that's fine also.
You seem to be going by the assumption that THD says all that is worth saying about a part's performance. For I/V especially, this isn't always the case as I explained earlier. I've been looking for a white paper that explains why, and will post it if I find it.
Also, I'm not sure why you think the ADA's THD is so poor. Looks very good to me. THD is only specified down to 100KHz under various conditions. It will improve a lot as frequency drops and loop gain increases.
The fact that ESS recommends one particular part doesn't mean it's necessarily the best, just that it's a safe choice to get a specified technical performance. Manufacturers are heavily spec driven because of the nature of the business, but one spec never tells the whole story about either objective or subjective performance.
+1
Thanks for the link to the document, but I'm not sure what point you are making?
I, and the folks at this thread, are aware of the Pro-Ject PreBox S2, SMSL M8A and Topping D50 and the good reviews that they received. But they do not have the I/V balanced output stage and are not conductive to DIY style upgrade. The lowest priced name brand ES9028Q2M box with I/V output is the Topping NX7 at $500.
I am using a generic brand ES9028Q2M box which sounds quite good to me, at only half the SMSL M8A price and allows me to use an Apple Remote for the input and function control.
NEW ES9038 ES9038Q2M DAC HIFI Decoder AUDIO IIS DSD DOP 384KHz + Amanero USB | eBay
The ES9028PRO package is another step up in sound quality. To be honest, I do not know how audible the sound improvement is. I just trust Markw4 and Mikett that it is worthwhile. Then I can judge it myself when I get the ES9028PRO. And the other half of the fun is to do the mod.
Last edited:
