So....as a newbie, with no soldering experience do I risk ordering a kit and and tooling up to give it a go?! Or is it best left to the professionals? Sure it's a question that has been asked a number of times, but I am struggling to fight the need to own a DDDAC! Anyone care to share advice/tips on tackling a build based upon the latest boards?
thank you
Spencer
thank you
Spencer
With no soldering experience I'd recommend the fully built boards. You'll still have to connect some wires and plugs, so little bit of soldering will be in order. Maybe you can find someone who can coach you through your first solder joints? I'd say give it a go!
Having done a bit more research I tend to agree. It would also seem that the fully built options present good value, as a four-board kit comes in quite close when you add the cost of power supplies, case etc.
OK, I tested Ian's reclocker, buffer and isolator against the Kali and the Kali won hands down. Pretty good for something that costs 1/5th of the price if you upgrade the clocks as Ian suggests. The Kali is an amazing bargain at that price. I think the flaw with Ian's clock board is that the oscillators are mounted on long legs and even worse long legged SOIC adaptors if you upgrade them. I upgraded them to the top spec Crystek clocks and still found the Kali to be better. The Kali comes with NDK clocks which match the top spec Crystek clocks and are the ultra small version SMD mounted which is exactly what you want with 40odd mghz flying around!
The Kali reclocker and Wave IO USB board give total flexibility for a Mac or PC front end and meaning no RPI required.
Not surprisingly Ian's 'new product' is an integrated buffer /isolator/clock board to match the Kali and it will be interesting to see if he can get near the Kali's price. For those using the Wave IO board it has a built in isolator so you don't need an additional one. Again, testing the Wave isolator against Ian's I could detect no advantage either way.
David
The Kali reclocker and Wave IO USB board give total flexibility for a Mac or PC front end and meaning no RPI required.
Not surprisingly Ian's 'new product' is an integrated buffer /isolator/clock board to match the Kali and it will be interesting to see if he can get near the Kali's price. For those using the Wave IO board it has a built in isolator so you don't need an additional one. Again, testing the Wave isolator against Ian's I could detect no advantage either way.
David
I agree -- if you've got a Kali that works as it should. I have three Kalis, and only one works well in the WaveIO / DDDAC setup. With the other two Kalis, I get some noise and "static" at the DDDAC output. It would be interesting to understand the difference between the Kalis, but Allo was not very helpful with this.
We may not find out the differences, but have all the Kalis the same revision? Or is only one particular revision working with the DDDAC1794?
Acc. to the website the revision is printed on the back of the PCB:
https://www.allo.com/shop/523-thickbox/kali-reclocker.jpg
It would be nice to know if the working Kalis have the same revision and the ones which don't work have a different one.
https://www.allo.com/shop/523-thickbox/kali-reclocker.jpg
It would be nice to know if the working Kalis have the same revision and the ones which don't work have a different one.
See discussion in the Kali thread (starting at post 731). I ditched the Kalis and went with the Hifiberries, which are easy-peasy no-problemo.
This is a nice read on changing the clocks on the Kali...
link to article - use translate button top right on page
link to article - use translate button top right on page
I would just like to point out this article deals with changing clocks on the HiFI Berry board, not Kali. The clocks which were used in the 'upgrade' are the Crystek clocks I used in Ian's reclocker. Whilst these are among the best spec around, the clocks used in the Kali-the ultra miniature NDK clocks-match these on spec, which was what sparked my interest in the comparison. So, I don't think you wil currently find better clocks than come as standard in the Kali and as the Kali clocks are SMD mounted any upgrade would mean SMD work which few of us are equipped to do.
David
David
using the WaveIO isolator I was only able to get noise if you don't connect the earth from the isolator to the Kali
Well, you obviously need the GND as a return path for the I2S signals.
It seems your Kali version is okay, and I guess you can be happy
Well, you obviously need the GND as a return path for the I2S signals.
It seems your Kali version is okay, and I guess you can be happy
Not true. It works without the earth WITH some static, hence my question. There is no earth i2s return. Ian's clock works with no earth connection and no static.
Without a signal return, the I2S signals are missing their GND reference. Without a GND reference, the I2S signal voltages are meaningless and the the I2S signal transfer cannot not work.
If Ian's board works in your setup without a GND wire between the I2S terminals, this means that the GND reference is accomplished via some other GND connection between Ian's board and the DDDAC, possibly via the GNDs of the power supplies, or maybe the chassis. Inserting a GND wire between I2S headers of Ian's board and the DDDAC adds a second GND connection, which means you get a ground loop (not good). I guess that Ian's board should be connected to GND only via a short I2S GND wire to the DDDAC. You can test this by disconnecting the I2S wires and then measure the resistance of the I2S GND to the DDDAC GND using a DVM. There should be no connection.
The reason why the WaveIO does not work well (noise) without the GND wire between the I2S headers is due to the I2S isolation at the WaveIO output. The isolation decouples the GND of the I2S output from the onboard I2S, so the GND reference of the I2S signals is floating. Inserting the GND wire between the I2S headers of the WaveIO and the DDDAC established a well-defined GND reference, killing the noise.
I'd suggest to clean up your grounding scheme(s) and repeat your comparison tests. Make sure there are no ground loops affecting the performance of your test setup.
Your confusing things. My post refers to a ground connection between the Wave IO isolator and Kali board. As my post clearly says without this ground between the isolator and Kali I got the symptoms you describe. There is a ground connection between the DDDAC main board and isolated gnds on the Wave (and thus to the Kali) via the +/- power for the isolator. All power supplies are completely independent and not connected. As my system works perfectly with none of the issues you have encountered I'm not sure why you suggest I should look at the earthing arrangement.........