Regarding regulators I would try before ON Semicondutor MC78xx/79xx since they have much lower noise and output impendance than others LM78xx/79xx and they cost only ca. 1.5US$...
I've just replaced in a Marantz CD6002 the 5V regulator of the DAC's analog supply (there was a KIA7805) with a MC7805.
It seems another player... fuller sound, deep bass and more refined
So, I presume you are looking at the output on an oscilloscope?
What is the amplitude of the signal when it clips?
I've assembled and running in my second MINI24/96 (the first one was a development board... ).
I've attached a picture and a modified BOM (Takman REX became PRP9372) since before mounting the Rikens to the new boad I've compared again PRP and REX, PRPs are a bit more 'dynamic' and full.
).I've attached a picture and a modified BOM (Takman REX became PRP9372) since before mounting the Rikens to the new boad I've compared again PRP and REX, PRPs are a bit more 'dynamic' and full.
Attachments
Jeff, it looks like you are using Audacity to generate your sine wave.
I looked closer at it and it looks like the sine wave is clipped a little.
Someone else on this thread seems to think that Audacity generates clipped files also http://www.diyaudio.com/forums/equipment-tools/90605-audacity-sine-good-enough.html#post1060647
When you generate a tone in audacity select amplitude of 0.98 and this appears to be not clipped....?
Thank you for your reply,
I have a last question which is not very thread related.
Imagine I have the possibility of shorting a coupling capactior (letting the DC pass), but let's say the DC is negligible. Will the audio quality with a coupling shorted or with a good audio cap (like simlic II placed)?
I have a last question which is not very thread related.
Imagine I have the possibility of shorting a coupling capactior (letting the DC pass), but let's say the DC is negligible. Will the audio quality with a coupling shorted or with a good audio cap (like simlic II placed)?
I've just tried a sample with the amplitude of 0.98 as well as the sample from the topic mentioned above (generated with GoldWave, 1kHz) but it still clips at maximum volume.
After lunch, I will measure some analogue outputs with the same files to determine whether the files are correct or not.
Is it possible that the clipping occurs because my DAC has a little amount of DC offset on the outputs?
After lunch, I will measure some analogue outputs with the same files to determine whether the files are correct or not.
Is it possible that the clipping occurs because my DAC has a little amount of DC offset on the outputs?
how much DC offset?
I don't think the DC offset should be an issue with clipping, if we are talking about 5mV or so.
If you check the output of the DAC with your oscilloscope on the 1K resistors before the op-amp, that may help to explain where the clipping is occurring. Either straight off the DAC or from the opamp.
There is a slight possibility that the DAC chip is damaged. I have seen AK4393 clipping before. Not on this kit, but on equipment at work that has come in for repair.
About 1,32 mV on the right channel and about 7,23 mV on the left channel in idle mode.
I've checked the output on the 1 kOhm resistors and the signal is already clipped when it comes out of the AK4396.
Is my AK4396 defective? It would cost me a fortune to replace it (I bought it on eBay about two months ago, so I don't think that the seller will replace it for free, and sending it back would cost more than a new AK4396 on eBay...)
At the moment I am using a 2x12V 20VA EI-core transformer, but I was going to buy a 2x15V + 9V R-core soon. However, I don't think that the low voltage is causing the problem, because I've checked the voltage on several points under load and they are perfect. (I've measured +-17,8V before the voltage regulators under load, which is well within the vreg's recommended values.)
I've checked the output on the 1 kOhm resistors and the signal is already clipped when it comes out of the AK4396.
Is my AK4396 defective? It would cost me a fortune to replace it
(I bought it on eBay about two months ago, so I don't think that the seller will replace it for free, and sending it back would cost more than a new AK4396 on eBay...)At the moment I am using a 2x12V 20VA EI-core transformer, but I was going to buy a 2x15V + 9V R-core soon. However, I don't think that the low voltage is causing the problem, because I've checked the voltage on several points under load and they are perfect. (I've measured +-17,8V before the voltage regulators under load, which is well within the vreg's recommended values.)
You do not need to populate the op amp side. you do not need the -12V regulator.
Last edited:
About 1,32 mV on the right channel and about 7,23 mV on the left channel in idle mode.
I've checked the output on the 1 kOhm resistors and the signal is already clipped when it comes out of the AK4396.
Is my AK4396 defective? It would cost me a fortune to replace it
At the moment I am using a 2x12V 20VA EI-core transformer, but I was going to buy a 2x15V + 9V R-core soon. However, I don't think that the low voltage is causing the problem, because I've checked the voltage on several points under load and they are perfect. (I've measured +-17,8V before the voltage regulators under load, which is well within the vreg's recommended values.)
Does it play a perfect unclipped sine wave at any amplitude?
Or is it always clipped?
Was it a pre-assembled kit, or you built it yourself?
Do you measure +5V on Vref High. Pin 17 of the DAC chip?
It is not always clipped. At a volume of 0.00 dB (100%) it starts to clip approximately at the amplitude of 0.45 working with samples created by Audacity. At an amplitude of 0.98 it starts to clip somewhere above the volume of -5.93 (~80%).
It is a pre-assembled kit, because it was cheaper this way. (auction, got lucky) I bought it from the seller 'along1986090' on eBay.
It is +5.006 V under load so it's totally okay. I'll check other voltage levels as well.
Then I have no other options, the AK4396 must be replaced. Thank you very much for your generous help through all these days. It would have taken a lot more time (possibly forever) to figure out the problem on my own.
If no one has any other ideas, I'll report the problem to the seller but I don't really think that there's warranty for these bought-online things.
In the end, I'll buy a new chip if I must. The cheapest available for me costs $8. (link) It's not that much but I'm not sure whether I'll be able to replace it. However, that'll be an other story.
Jeffzen, where are you located? There may be other DIY'ers in your area that may have experience with SMD that may be able to help you replace the AK4396. The best way would be to cut the legs of the existing chip and then unsolder them one at a time. Also, a Kester "Flux Pen" helps tremendously when doing SMD work (I've done an AMB Y2 and "The Wire") to install the new chip.
Btw, regarding eBay kits from seller "zoe_tsang": Are the AK dac and CS receiver chips on these boards already soldered? (with just the passive parts to solder yourself..)
Because I want to follow erin's and ClaveFreme's recommendations in selecting caps, etc. and am confident in soldering these, but SMD no skills yet
Because I want to follow erin's and ClaveFreme's recommendations in selecting caps, etc. and am confident in soldering these, but SMD no skills yet