True, but after you bust your behind fixing all those things, then you are back to hearing the dac distortion. Also, what I find and what I think Jam has been finding is that fixing all those other problems you know about improves sound in ways we don't know how to measure so easily. Where people go wrong with measurement is when they overdo it to produce perfect sine waves but don't listen to how music sounds. There is a path to proper optimization and I think we know to do it pretty well with AK4499 at this point. What's wrong with D90 sounds like its mostly in the clocking and in the output stage. Of course, if we peel back those layers of the onion we might find the next set of problems
I agree.
It's actually bloody intriguing.
Jam talked to the distributor we know. Word is this time around AKM is not field testing the chips the same way as they did for AK4499. They don't expect to sell a lot of the new chip set because the market will be small, only for a few very specialized dac products. So, the few people that want to go there can buy eval boards if they want to.
I don't know about the claimed reasoning, but I do know that there aren't a lot AK4499 dac products out yet, and most of them seem to be priced above what most consumers want to pay for a dac.
I don't know about the claimed reasoning, but I do know that there aren't a lot AK4499 dac products out yet, and most of them seem to be priced above what most consumers want to pay for a dac.
I would rather see AKM come out with a stand alone, switched resistor DAC chip, optimized for DSD 256/512 input. Basically just the DAC section from the 4499. then one could use it to make a no compromise DAC designed specifically for oversampling externally in a powerful computer using HQPlayer.
Now there are many software players that do oversampling to DSD, so this type of DAC chip would not be as limited as one might think at first. And there are already plenty of DAC manufacturers who do not use standard chip based oversampling internally in DACs anyway. And plenty who run their own oversampling algos in an FPGA or DSP chip.
I am not sure why AKM chose their switched capacitor style DAC for the stand alone chip? It seems like a step backwards in a way.
Now there are many software players that do oversampling to DSD, so this type of DAC chip would not be as limited as one might think at first. And there are already plenty of DAC manufacturers who do not use standard chip based oversampling internally in DACs anyway. And plenty who run their own oversampling algos in an FPGA or DSP chip.
I am not sure why AKM chose their switched capacitor style DAC for the stand alone chip? It seems like a step backwards in a way.
It's kind of disappointing that people seem ok spending $2000 or even more on an apple computer to basically just surf the internet that could be done on a raspberry pi device but they're not willing to spend the same amount of money on good digital audio equipment that actually improves their listening experience as opposed to surfing the internet on an expensive computer which adds nothing to the experience.
Its a psychological thing. How people think about the cost of different things has already been studied quite a bit. Just part of human nature.
Some of us wouldn't ever buy an apple computer! I need 5 channels of DAC output for my active speakers so it gets expensive at 1,000 per channel. Buys a lot of music!
What do you think about evaluation boards?
Do the manufacturers do a good job of putting together a decent evaluation board or do they follow the bare minimum approach?
Member
Joined 2009
Paid Member
I’ve been saying this for some time, since I figured out dsp phase manipulation makes such a difference. But it must be configured @ lp with music .....not 1 meter with test tones.
So many things in a system change phase relationships slightly.....add them up in the wrong direction it can certainly be audible.
There needs to be a high quality dac with dsp (the RME comes close but no phase manipulation) and 8 channels would be nice.
I’ve gotten a minidsp shd to play around with as it’s the highest quality Dac with dsp I can afford at the moment. 4490 is only 9 less than 4499!
Last edited:
Well, the AK4499 evaluation board does not include a USB to I2S interface, nor does it include an MCU to control the dac registers.
On the other hand, the board can work in pin control mode with only an external +-15v power supply and a SPDIF or TOSLINK input source. Analog outputs are balanced XLR only. No MCU or USB board needed in the 'SPDIF/TOSLINK and pin-control mode' case.
What the evaluation board by its design encourages designers to experiment with are several of the more critical power supplies.
Getting deeper into it requires addition of a USB board, clocks, MCU, some power supplies, and either a reproduction system that can accept balanced inputs or the addition of external differential summing stages or unbalancing transformers.
In other words, the evaluation board is the central piece of a diy experimenter project.
All the above having been said, once you know how to deal with USB the best way (mount a clean-power-modded Xing Accusilicon clock USB board on the solder side of the eval board, with the USB board ground plane facing the towards dac chip), and once you hook up an Arduino or other MCU and get used to working with that, then its just standard analog audio stuff: power supplies and differential to SE conversion if needed.
Also, eventually Some experiments will involve modding the evaluation board. Since it is mostly SMD, old guys like me do much better if equipped with a suitable microscope, such as: <removed> ...I find the analog microscopes better for soldering than their digital counterparts. IME hand-eye coordination is much better if looking through eyepieces, presumably because the head and eyes stay in a fixed location.
Last edited by a moderator:
The MCU thing is not hard to lean. Easy to do with an Arduino or many other common MCUs. The main thing is to use an MCU for which an I2C library exists. There are multiple library choices for Arduino.
Once an MCU is selected then some hardware interfacing can take place. Usually a good idea to use an IC I2C bus isolator or other isolator/translator device. The benefits can be preventing ground coupled noise incursion, and preventing potential damage to dac and or MCU if devices are powered on/off independently.
As far as programming goes, I have sample code I could make available. Its not hard to do once understood. However, detailed understanding of a programming language can take some time. Depends how deep one wants to get into it.
EDIT: Forgot to mention there is some low level RF embedded in the analog outputs of the dac chip. Have to filter that out (attenuate it sufficiently, to be more correct).
Once an MCU is selected then some hardware interfacing can take place. Usually a good idea to use an IC I2C bus isolator or other isolator/translator device. The benefits can be preventing ground coupled noise incursion, and preventing potential damage to dac and or MCU if devices are powered on/off independently.
As far as programming goes, I have sample code I could make available. Its not hard to do once understood. However, detailed understanding of a programming language can take some time. Depends how deep one wants to get into it.
EDIT: Forgot to mention there is some low level RF embedded in the analog outputs of the dac chip. Have to filter that out (attenuate it sufficiently, to be more correct).
Last edited:
