In my opinion it get’s even worse, as the dam1121 is specified to work in a synchronised Multichannel configuration, but if you set the dam1121 to accept a external clock, only the “master” board can accept an i2s input. The “slave” boards i2s input must be connected to the “master” boards i2s output. This is fine if meant to work as dual mono. But you can’t feed it with different i2s inputs. So there is also no synchronised Multichannel as it is stated on the website.
Quote: “ For a board to use its own input, it must be master. But a master will not disable its oscillator, which are required to use an external clock.
It don't sound that complicated to implement a master mode with oscillator disabled, the problem with all changes are testing and verifying....”
(Source of Information:
Building with the Soekris dam1121
And:
Building with the Soekris dam1121
)
Quote: “ For a board to use its own input, it must be master. But a master will not disable its oscillator, which are required to use an external clock.
It don't sound that complicated to implement a master mode with oscillator disabled, the problem with all changes are testing and verifying....”
(Source of Information:
Building with the Soekris dam1121
And:
Building with the Soekris dam1121
)
I personally listen to some music on Youtube, there are a lots great live recordings in high quality, like these:
Halestorm - "Here's To Us" captured in The Live Room - YouTube
Kelly Clarkson - Because of You (LIVE @ GRAMMY 2006) - YouTube
Adele - Set Fire To The Rain (Live at The Royal Albert Hall) - YouTube
And I want my DACs to just work great, whatever your feed it. Anybody else not happy with that are welcome to design a DAC to meet their own needs, or move on to something else.
And would people who keep repeating themselves and don't bring anything constructive to this thread, please disappear.... cab, you couldn't even figure out to build a DAC, you just keep bitching, and andrea_mori, I don't know if you have build anything, I know your focus on the clock, please stop repeating yourself, but stick to your own projects.
I'm sorry, but if you only listen to YouTube music, I suspect that you don't even know what the dam is capable of.
i have optimized the dac over and over again for 3 years, with different power supplies, input and output boards. Now i would also say that it can keep up with most dacs on the market or even outperform them.
But I’m still waiting for features still mentioned in the manual e.g. the I2S FSEL IN/ MCLK OUT and FPGA MCLK OUT/ SLV IN pins.
That was initially one of the reasons why i chose the dam1021.
Let's see what will become of andrea_moris dac. sounds very promising to me
Multichannel, It meant for digital crosssover use with multiple output channels, still with a single stereo source....
I'm not going to add new features to the dam1121, unless somebody comes by and whant a lot of them.....
I'm sorry, but if you only listen to YouTube music, I suspect that you don't even know what the dam is capable of.
i have optimized the dac over and over again for 3 years, with different power supplies, input and output boards. Now i would also say that it can keep up with most dacs on the market or even outperform them.
But I’m still waiting for features still mentioned in the manual e.g. the I2S FSEL IN/ MCLK OUT and FPGA MCLK OUT/ SLV IN pins.
That was initially one of the reasons why i chose the dam1021.
Let's see what will become of andrea_moris dac. sounds very promising to me
I listen to a lot of music, from multiple sources.... And my DAC have work with all of them....
Well, the intended way to implement a (crossover) multichannel setup is to use the filters of the DAM to separate the frequencies. So this multichannel setup is well doable.
You can not expect magic that a relative short FIFOs driven by ONE clock can track two independent I2S sources. This said, it should be no problem to feed different boards with different I2S sources as long they are derived from the same clock and have not to much delay between the different sources.
Thanks zfe and soekris.
I myself was not aware that crossover is also called Multichannel in this regard.
I like my dual mono built and will stick to it. The sound is great and I have never had problems with this configuration. The Multichannel would just have been an excuse to buy more dam1121 well, two more only, not that many.
I have one additional question. The audio enable pin on the dam1121 which is on the pin header with the analog outputs; as many others I have different power supplys for analog and digital part of the dam. Would one connect this to analog 3.3v or digital 3.3v? Thank you.
I myself was not aware that crossover is also called Multichannel in this regard.
I like my dual mono built and will stick to it. The sound is great and I have never had problems with this configuration. The Multichannel would just have been an excuse to buy more dam1121
well, two more only, not that many. I have one additional question. The audio enable pin on the dam1121 which is on the pin header with the analog outputs; as many others I have different power supplys for analog and digital part of the dam. Would one connect this to analog 3.3v or digital 3.3v? Thank you.
Thanks zfe and soekris.
I myself was not aware that crossover is also called Multichannel in this regard.
I like my dual mono built and will stick to it. The sound is great and I have never had problems with this configuration. The Multichannel would just have been an excuse to buy more dam1121
I have one additional question. The audio enable pin on the dam1121 which is on the pin header with the analog outputs; as many others I have different power supplys for analog and digital part of the dam. Would one connect this to analog 3.3v or digital 3.3v? Thank you.
I'm happy that you like your build.... The EN pin just need a pullup to something, the example schematics say VCC33, which is a digital power.... If you don't have 3.3V around you can also do a 4K7 to +5V, there are a little pulldown on the dam1121....
I also have a couple of DACs in my projects, and I don't think I have to learn from you how to build anything.
But it could be useful for your customers if you measured the clock performance of your DAC, you may have some surprises.
You are right, I couldn't figure out how to build a properly functioning multichannel dac with the 1021. Sadly, the limitation wasn't my own skill set, rather the shortcomings of your faulty, misrepresented, beta product. Too bad you didn't actually test and verify the proper performance when used in a multichannel set up before you advertised it as good to go in that configuration. What's it called when you make a claim when advertising your product which turns out to be false and then refuse to refund the customer's money when the product won't do what you said it would?
Last edited:
Where did I say anything about your ability to design and build ?
I focus on the sound, not some single clock measurement. I talked very early on the clock performance is not optimal on the dam1021 as the clock goes though the FPGA, and all never products have improved on that, driving the sample clock directly from the oscillator with just a LVC flipflop between that and the LVC shift registers....
I don't have equipment to measure the clock like you have, so you're welcome to measure it yourself, I, and most of my customers, focus on audio measurements and listening to music....
This thread is really a technical thread for people havng build a DAC around the dam1021, or people considering building one. You're just repeating yourself with the same noise, so please stop it.
I launched a DIY DAC module on a DIY forum, expecting people here to have just a little technical knowledge, and you're not one of those. Other people have used multiple dam1021 with a dsp based crossover, so it can be done.... I never promised it to be easy or simple....
So as I just said:
This thread is really a technical thread for people havng build a DAC around the dam1021, or people considering building one. You never build anything and have your modules up for sale, so please stop making noise.
Who are these "other people"? I haven't read here of one bona fide third party account of success building a multiple channel dac with the 1021. The clock issues on the other hand have been well documented here. No amount of technical knowledge will overcome your misrepresentation- the problem is your product and your false advertising.
I will be pleased to measure your DAC, please send me a finished board and I measure the phase noise with my equipment, then I send it back.
lol he did promise once that clock syncing would be implemented on the 1021, later said it was "not possible". It's not exactly a deal breaker for dual-mono users, but it seems your problem has been preventing you from getting anything to work all these years. My condolences...
I remember reading through the datasheet of the Spartan FPGA and the jitter specification is very ugly compared to the difference between clocks. If it's really a problem, the 1021 system is basically screwed. I tend to think that our hearing is not quite at that level. Also, never forget that the shift registers have a very large specified tolerance on the propagation delay. Every 8 bits could be multiple nanoseconds off from each other, if the actual delay isn't grouped more closely than specified in the datasheet. Only incompetent engineers (and marketing execs) worry about leaky windows when the whole house is flooded.
That's great to know. You've been saying for ages that those linearity measurements are flawed b/c Sign Magnitude should be linear all the way down. It made sense to me theoretically, but it's always nice to have an explanation for how they messed up.
Last edited:
Soren, maybe you should just take back cab's boards and offer him a 1941 or something... The cost of a 1941 wouldn't set you back that much, and you can use his board for experimentation. Think about how much Amazon spends a year on free returns. Cab is probably the only guy on this thread that has done nothing with his 1021s.
After purchasing 4 of these and the issues started popping up I asked him if I could return them for a credit towards the newer boards, which he refused to do. Finally, after complaining publicly, he offered to upgrade them to the next rev to fix the Vref issues. Then the clock synch issues were discovered which basically made the 4 boards I purchased useless to me. All the while he simply blamed me for not having the DIY ability to make these work as he claimed they would. Blame the victim, never taking responsibility for his misleading and false claims. I have bought a lot of diy products over the years from several different vendors and have never had this sort of treatment or felt cheated until this experience.
Last edited:
cab, to me, is one of those few idiots that you can never make happy, no matter what you do, so it would be a waste trying....
I used to have a warning on my computer board website:
"If you don't know what you are doing, please don't purchase anything"
Living sounds, any chance that 100ps level jitter from the FPGA could cause the clock drift that you measured? Have you tried 1121 which doesn't have this clocking issue? If it's the shift registers, it seems to me (an amateur), that maybe there could be a software fix if Soren develops a calibrator. But that wouldn't explain the drift, only some sort of distortion. Maybe 1021 is doomed... Too bad it has all the nice input output features and a compact design :-(