yes The Mikett version looks more do-able for me.
At least it is more readable for my poor skills.
Reading things about Avcc implementation on the net, I found this at Twisted pear website.
AVCC-SR Dual 3.6V Series Regulator Module
Trident-SR
Do you think it is usable with our board?
At least it is more readable for my poor skills.
Reading things about Avcc implementation on the net, I found this at Twisted pear website.
AVCC-SR Dual 3.6V Series Regulator Module
Trident-SR
Do you think it is usable with our board?
@terry22, The bottom line for AVCC regulators is to build one and measure and listen to the effects on DAC performance. Looking at data sheets can give ideas for good candidates to try, but looking on a few graphs is not the same as measuring and testing performance in a very particular application.
Also, ESS came out with their own voltage regulators for AVCC. Probably because they thought there was a need for one better matched to the needs of their DACs, IMHO.
In addition, ES9038Q2M does not need anywhere near as much current from the AVCC supply as the PRO chips do.
I would stick with the opamp circuit and keep it as close as possible to the DAC AVCC pins. Same for the ground wire for that circuit because noise and resistance there can be just as much a problem as on the 3.3v output (current always flows in loops).
If you stick with the suggested circuit you will know you can expect excellent results. If you want to try something else, we can give you advice for hooking up to work as well as possible, but no guarantees. My recommendation would be to leave the experimenting to more experienced builders who also have distortion measurement equipment.
@system54, Don't know how you are measuring noise, but 3mv p-p AVCC noise would be huge and awful. It needs to be down at -120dB or anything lower than that you can get. Noise is not the only consideration either, but any noise there will get added to and intermodulated with DAC audio outputs. At the noise level that should be on AVCC one would need to take great care in measuring it. Normal DVM leads out in open air already are possibly going to act like antennas and add more noise to AVCC than there would be if you didn't touch it. A carefully grounded scope probe would probably disturb the circuit less, but vertical amplifier noise ought to swamp any AVCC power supply noise, so the measurement would be of very questionable accuracy. All this is assuming you are using a proper AVCC supply. If you are measuring a stock Chinese DAC board it might be pretty bad, don't know exactly what it usually is though.
Also, ESS came out with their own voltage regulators for AVCC. Probably because they thought there was a need for one better matched to the needs of their DACs, IMHO.
In addition, ES9038Q2M does not need anywhere near as much current from the AVCC supply as the PRO chips do.
I would stick with the opamp circuit and keep it as close as possible to the DAC AVCC pins. Same for the ground wire for that circuit because noise and resistance there can be just as much a problem as on the 3.3v output (current always flows in loops).
If you stick with the suggested circuit you will know you can expect excellent results. If you want to try something else, we can give you advice for hooking up to work as well as possible, but no guarantees. My recommendation would be to leave the experimenting to more experienced builders who also have distortion measurement equipment.
@system54, Don't know how you are measuring noise, but 3mv p-p AVCC noise would be huge and awful. It needs to be down at -120dB or anything lower than that you can get. Noise is not the only consideration either, but any noise there will get added to and intermodulated with DAC audio outputs. At the noise level that should be on AVCC one would need to take great care in measuring it. Normal DVM leads out in open air already are possibly going to act like antennas and add more noise to AVCC than there would be if you didn't touch it. A carefully grounded scope probe would probably disturb the circuit less, but vertical amplifier noise ought to swamp any AVCC power supply noise, so the measurement would be of very questionable accuracy. All this is assuming you are using a proper AVCC supply. If you are measuring a stock Chinese DAC board it might be pretty bad, don't know exactly what it usually is though.
Last edited:
Thanx. I understand.
I will read the ESS pdf to look at the avcc schematic and maybe give it a try
If there is anything I can do to make giving it a try easier, I would be happy to talk about it with you. Have you ever built anything before from scratch? Do you have a temperature controlled solder iron with a small point tip? Anything that would help me understand your situation better might be good. Maybe there are a few areas we can focus on where you feel like you might run into problems?
The LM317/337 board I ordered from ebay came today and I'm pleased to report it is of good quality, the big caps are Nichicon HE(M) and the smaller ones are Panasonic FC. However, there was a little surprise too - the board has four diodes on it and is also a rectifier, which is annoying as I only wanted a regulator.
LM317 LM337 AC-DC 5V 12V Adjustable Voltage Regulator Regulated Power Supply | eBay
LM317 LM337 AC-DC 5V 12V Adjustable Voltage Regulator Regulated Power Supply | eBay
Ian, you can remove the diode if you like, or you can run DC through them if you like. They will pass DC as well as AC although there will be a voltage drop across them if you do run current through them.
Also, there is always some risk with buying things from ebay. You can look at the pictures in the ads and sometimes some sellers will show more pictures of a product that others, so it can be useful to look around. But, pictures can still only tell you so much, so you have to be prepared to deal with what you end up with. Shipping it back may be an option, but some things are so cheap it wouldn't make any sense to pay shipping costs back to China.
Also, there is always some risk with buying things from ebay. You can look at the pictures in the ads and sometimes some sellers will show more pictures of a product that others, so it can be useful to look around. But, pictures can still only tell you so much, so you have to be prepared to deal with what you end up with. Shipping it back may be an option, but some things are so cheap it wouldn't make any sense to pay shipping costs back to China.
Hi Mark
Yeah, the ebay China 'lottery'. I've dealt with it many times. I'm actually happy with this board as it's good quality.
I didn't spot the diodes in the pics as they are right in the middle between the heatsinks for the regulators.
I'm not sure how I will use this board, I've radically rethought my DAC project so everything has changed.
I just mentioned the board to let people know it's good quality with good quality caps, so a bargain for the price it costs.
Yeah, the ebay China 'lottery'. I've dealt with it many times. I'm actually happy with this board as it's good quality.
I didn't spot the diodes in the pics as they are right in the middle between the heatsinks for the regulators.
I'm not sure how I will use this board, I've radically rethought my DAC project so everything has changed.
I just mentioned the board to let people know it's good quality with good quality caps, so a bargain for the price it costs.
Last edited:
Hi Ian,
Naturally, I am curious how your thinking has evolved regarding you DAC project. Also, I might point out that the Allo Katana DAC, while in a slight state of flux right now, could potentially end up being a good option at a good price.
The problem with DACs as I see it is that it probably costs maybe around $200 to make a good one in board form without a case and without distributer and retail markups, but as price goes up buyers want more bell and whistle features rather than wanting improved sound quality. If they have to choose between the two, they will rationalize that the sound quality of the one with the bells and whistles is probably good enough, which in reality it probably isn't but they have no idea what they will be missing.
If whatever issues there are with Katana get ironed out, it could be the first DAC that really goes for sound quality first, while doing so at the lowest cost. Whether or not it will succeed at that is something we will have to wait and see on.
In the meantime, I can advise and show people how to make what I know to be a very good DAC for the money, perhaps the best DAC for the money at the moment, but it remains a semi-advanced project to undertake, and one that nobody I know of has fully implemented (not counting me
). There have been some behind-the-scenes talk about making a board or a finished DAC board available based on work done here and whatever additional development might occur, but so far it has only been talk and not something I foresee happening in reality.
Naturally, I am curious how your thinking has evolved regarding you DAC project. Also, I might point out that the Allo Katana DAC, while in a slight state of flux right now, could potentially end up being a good option at a good price.
The problem with DACs as I see it is that it probably costs maybe around $200 to make a good one in board form without a case and without distributer and retail markups, but as price goes up buyers want more bell and whistle features rather than wanting improved sound quality. If they have to choose between the two, they will rationalize that the sound quality of the one with the bells and whistles is probably good enough, which in reality it probably isn't but they have no idea what they will be missing.
If whatever issues there are with Katana get ironed out, it could be the first DAC that really goes for sound quality first, while doing so at the lowest cost. Whether or not it will succeed at that is something we will have to wait and see on.
In the meantime, I can advise and show people how to make what I know to be a very good DAC for the money, perhaps the best DAC for the money at the moment, but it remains a semi-advanced project to undertake, and one that nobody I know of has fully implemented (not counting me
). There have been some behind-the-scenes talk about making a board or a finished DAC board available based on work done here and whatever additional development might occur, but so far it has only been talk and not something I foresee happening in reality.
Last edited:
Last edited:
Well, I never listened to my blue version of the board before I messed with it, but I really wish I had to know if it sounded so bloody awful in stock form, if it did, it would have been returned to China in disgust immediately.
I thought it was a case of 'this thing will sound quite good in stock form, and incremental improvements will come with upgrades/mods' but it's more a case of 'it's total crap until you heavily upgrade/modify it' or at least, that's the impression I get.
So when I spotted one of these for 60% of the price of the ESS board, I jumped on it, as it's a case of it already sounds great without any modifying or upgrading, which is more suitable to my current skills and experience:
http://www.diyaudio.com/forums/digi...io-851-series-upsampling-dac.html#post5493318
I thought it was a case of 'this thing will sound quite good in stock form, and incremental improvements will come with upgrades/mods' but it's more a case of 'it's total crap until you heavily upgrade/modify it' or at least, that's the impression I get.
So when I spotted one of these for 60% of the price of the ESS board, I jumped on it, as it's a case of it already sounds great without any modifying or upgrading, which is more suitable to my current skills and experience:
http://www.diyaudio.com/forums/digi...io-851-series-upsampling-dac.html#post5493318
Ian, Oh, man! Well, I will agree with you about the stock Chinese DAC sounding pretty bad.
Looks like you may have gotten yourself into worse trouble with your latest acquisition though. I would not have advised acquiring it if no schematic to study first. You could have asked, don't know why you didn't. PM would have been okay.
Looks like you may have gotten yourself into worse trouble with your latest acquisition though. I would not have advised acquiring it if no schematic to study first. You could have asked, don't know why you didn't. PM would have been okay.
I was asked to provide some info on Katana dac in another thread. That post can be found here: http://www.diyaudio.com/forums/vendor-s-bazaar/294940-fifo-buffer-rpi-sbcs-197.html#post5493686
Hopefully it will be a case of figuring out the connections of that 8 pin header the transformer connects to. Cambridge often silkscreen info about the connectors on the board next to them.
Progress will be slow as I'm going into hospital tomorrow for a few days.
Hi Mark
thanks for all the info , could you give the link for the board above by any chance?
Hi Mark,
I have pretty much no knowledge in electronics.
I do things with arduino, breadboard and components, but I didn't made any custom pcb myself or builded some electronics from scratch.
I can solder components. I can barely read schematics but I try to learn.
I have temperature controlled solder iron with small point tip, solder wick and basic multimeter and magnifiying glasses.
Next thing I will do is changing the clock.
Then I will try the Avcc Mod and if it is a success, Then I will try the I/V.
Do you think it will be easy for me if I have the pcb(s) made online from schematics, then only solder the necessary components on it. then do the necessary modifications on the 9038q2m board. then connect the boards together?
With some patient help from you guys, I may be achieve this.
The schematic I used was identically the same as posted as by ESS document that Markw4 points out. The easy route is to use two dual op amps. Two LME49720s. Three would be for the IV stage and the fourth would be for the AVCC. The AVCC op amp stage is easy as pie and it works and a good AVCC is absolutely necessary.
This post shows when I was in the middle of laying it out on perf. I turn the perf upside down to make it easier to figure out what is going on. When I am soldering, it is obvious would be turned over.
ES9038Q2M Board If you will be using perf board similar to what is shown, get a pattern that has four or five connections perpendicular to the two perpendicular parallel rails. This will make the layout very simple and more compact. The parallel rails are used for the +/- 15V rails.
Had I not messed with the single op amp solution, my 9038q2m would possibly be close to Mark's. But now that proto board is idle. However, I still have most of the parts ready so it is possibly just another $40 or so to build another DAC. I'll wait and see what comes out of Katana. My HTPC video setup needs a good DAC not necessarily the best or close but a decent one. Once my 9028DAC started singing, my video setup started to be a real laggard. It's contagious. Be warned.
But again look at the circuit and it does not take a lot to make it. One afternoon and you're done. BUT when interconnecting between it and the board reread when I was connecting my hookup to the 9038q2m dac. Mark has a small error in one of his photos is identifying the DAC out lines. You will need to look at that closely and follow it. It would be good to see if someone else was able to pretty much replicate the results mark obtained.
Last edited: