Welcome Guest! To enable all features please Login or Register.

Notification

Icon
Error

4 Pages<1234>
Options
View
Go to last post Go to first unread
Russ White  
#41 Posted : Friday, February 26, 2016 10:34:15 PM(UTC)
Russ White

Rank: Administration

Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,978
Location: Nashville, TN

Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
??? Are you sure you have the ground clip attached correctly? Nothing is making any sense here. I can't see how it would be even remotely possible to have a higher frequency after the clock divider - there is nothing else between - just a the clocks/mux/divider. If you have good signal at the mck output of the cronus you should have mck/2 at the mck at the hermes mck out to the amanero. You need to be sure CS (clock select) is pulled either clearly high or clearly low - don't leave it float.

schultzsch  
#42 Posted : Saturday, February 27, 2016 10:22:08 AM(UTC)
schultzsch

Rank: Member

Groups: Member
Joined: 6/3/2010(UTC)
Posts: 53
Romania
Location: romania

Thanks: 2 times
Yes, I had attached the ground clips for every channel of the scope.
It`s true that when I connect the scope in I create some ground loops because of the system being battery powered but when I remove the scope the shoud not be any ground loops at all and everything should work fine so I removed the scope this morning from the setup.


Now if I set amanero for MCK/2 prescaler or cronus divider for MCK/4 I can stream files until 192khz at a normal pitch.It won`t stream at all 384khz files.
On the display of the controller I see 384khz for 192khz files.

If I leave amanero prescaler at MCK/1 and cronus divider at MCK/2 I have the same high pitch speed as in the other tests.
If I set amanero prescaler at MCK/4 and cronus divider at MCK/2 I get a slower pitch like the normal.

Edited by user Saturday, February 27, 2016 10:26:42 AM(UTC)  | Reason: Not specified

Russ White  
#43 Posted : Saturday, February 27, 2016 12:36:53 PM(UTC)
Russ White

Rank: Administration

Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,978
Location: Nashville, TN

Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
It is easy to check the divider for correct operation just run the cronus alone and check the clock signals at the 1:1,1:2,1:4 clock divider header. It may be that the divider is damaged or needs solder rework? Take a close look around it.
schultzsch  
#44 Posted : Saturday, February 27, 2016 3:31:16 PM(UTC)
schultzsch

Rank: Member

Groups: Member
Joined: 6/3/2010(UTC)
Posts: 53
Romania
Location: romania

Thanks: 2 times
On the header everything looks fine.
schultzsch  
#45 Posted : Saturday, February 27, 2016 3:33:02 PM(UTC)
schultzsch

Rank: Member

Groups: Member
Joined: 6/3/2010(UTC)
Posts: 53
Romania
Location: romania

Thanks: 2 times
Could it be the isolator ?
Russ White  
#46 Posted : Saturday, February 27, 2016 4:51:31 PM(UTC)
Russ White

Rank: Administration

Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,978
Location: Nashville, TN

Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
I suppose - double check all the solder joints on the hermes. The isolator chip itself is very robust.
schultzsch  
#47 Posted : Sunday, February 28, 2016 6:47:21 PM(UTC)
schultzsch

Rank: Member

Groups: Member
Joined: 6/3/2010(UTC)
Posts: 53
Romania
Location: romania

Thanks: 2 times
I have checked everything so many times that I am fed up. It`s all well soldered and in place. Can we please eleminate this presumtion for always?

Today I connected the dac with the ufl coax wires for all the signals except mck. For the mck I have used a single straight wire and it`s acting the same as when there is the coax wire with gnd connected on both sides. So it wasn`t the 4th ufl, the one with the mck that was causing ground loops when connected in the circuit.

I am thinking to disolder the isolator and solder in place some wires from one side to the other for gnd and the rest of the signals except for the power supply .

If I do this will the warranty be voided?

What do you say? it`s a smart ideea to proceed with?


Russ White  
#48 Posted : Sunday, February 28, 2016 11:44:09 PM(UTC)
Russ White

Rank: Administration

Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,978
Location: Nashville, TN

Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
I am as baffled as you are. :)

Why don't you simply check the signal at both ends of the isolator? you could do this without the amanero connected.
Russ White  
#49 Posted : Monday, February 29, 2016 12:05:23 AM(UTC)
Russ White

Rank: Administration

Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,978
Location: Nashville, TN

Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
Also something tells me you may need to simplify your power supply scheme.

Power up order of devices is important.

DAC must have a clock present in order to receive I2C - and should never have any input voltage if there is no supply voltage present.

You might try to simplify things so that your controller (I would stick with on-board until you get things up) only starts when that DAC actually ready to receive commands.

See if you can power DAC(with on-board uC) and cronus from same supply.

The trick debugging things it to try to eliminate variables - your setup initially had quite a few :) That's not a bad thing - as long as it works. :)

Edited by user Monday, February 29, 2016 12:06:58 AM(UTC)  | Reason: Not specified

schultzsch  
#50 Posted : Monday, February 29, 2016 12:42:32 PM(UTC)
schultzsch

Rank: Member

Groups: Member
Joined: 6/3/2010(UTC)
Posts: 53
Romania
Location: romania

Thanks: 2 times
Even if I use different psus to power everything up, they start together always.
I have an on/off relay board that help me with perfect powerup sync so this is out of question.

I thought checking the signal as you say but if I don't connect amanero half of isolator would be unpowered. It will pass the signal anyway?
Russ White  
#51 Posted : Monday, February 29, 2016 1:21:49 PM(UTC)
Russ White

Rank: Administration

Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,978
Location: Nashville, TN

Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
I wouldn't be so quick to rule anything out. Start with the simplest possible thing to test and move from there.

Just power it with the same supply on both ends - 2 short pieces of wire will do the trick. :) You are testing for good signal operation - not isolation. :)

Edited by user Monday, February 29, 2016 4:03:35 PM(UTC)  | Reason: Not specified

schultzsch  
#52 Posted : Monday, February 29, 2016 5:53:19 PM(UTC)
schultzsch

Rank: Member

Groups: Member
Joined: 6/3/2010(UTC)
Posts: 53
Romania
Location: romania

Thanks: 2 times
Okey.. I`ve soldered two bridges from one side to the other of the isolator for Vdd and Gnd.

For both photos I have put the scope on isolator mck/2 output that goes to amanero.

This is what I get without b3 mck connected to cronus

image2.JPG (102kb) downloaded 9 time(s).


When I attach b3 mck ufl to cronus things go wild Boo hoo!

image1.JPG (114kb) downloaded 10 time(s)..


In both the photos amanero was disconnected from the isolator.

Edited by user Monday, February 29, 2016 5:55:07 PM(UTC)  | Reason: Not specified

schultzsch  
#53 Posted : Monday, February 29, 2016 8:27:51 PM(UTC)
schultzsch

Rank: Member

Groups: Member
Joined: 6/3/2010(UTC)
Posts: 53
Romania
Location: romania

Thanks: 2 times
In the end I have disoldered the isolator to see if it really was him.
I have soldered in place straight wires.

It still acts the same as first when with the isolator in. After I have risoldered the isolator back but it seems it`s not working anymore. Maybe from overheating when disoldered.
Now I have finished with two problems: the first one and a broken isolator board.d'oh!


image1.JPG (93kb) downloaded 9 time(s).


It starts to become sad ...Silenced

Russ White  
#54 Posted : Monday, February 29, 2016 9:27:25 PM(UTC)
Russ White

Rank: Administration

Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,978
Location: Nashville, TN

Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
The isolator is easy to obtain - I would not worry much there.

You need to focus you effort around the mck to the cronus - which happens before the divider (which then goes out to the Amanero).

You can do that with nothing but the Cronus and the DAC.

In fact all the Cronus does for the master clock is apply a MUX to choose one or the other (44.1 base or 48)

You could check the clock on each side of the MUX for the selected time base, measure at the RHea output pin and it should be the same as on the other side of the MUX at the mck out header. Do this first with the Cronus alone (nothing else).

Once you do that power up the DAC (but don't connect it to anything) and measure the MCK - make sure there is no signal at - and no DC present.

Then connect GND and MCK to the DAC (which should share exactly the same supply for this test as the cronus) by straight hookup wire. There should be nothing else connected to the DAC. Now check MCK and it should be as clean as always.
schultzsch  
#55 Posted : Tuesday, March 1, 2016 6:40:04 PM(UTC)
schultzsch

Rank: Member

Groups: Member
Joined: 6/3/2010(UTC)
Posts: 53
Romania
Location: romania

Thanks: 2 times
Now I am powering cronus and b3 from the same regulator.


When powering b3 alone I get no ac/dc signal at his mck header so this is ok.

For cronus the same thing, when I power it alone it works correctly I get the right mck and mck/2 at the outputs.

Now when I connect b3 mck input at cronus mck header, on amanero side I get exactly double the frequency that has to be.

Here you can see the mck at b3 header where it is clean as always as you say.

image1.JPG (116kb) downloaded 9 time(s).

In this photo you can see what cronus sends to the isolator. This is when setup for mck/2.

image2.JPG (106kb) downloaded 8 time(s).


Maybe now you will ask why I didn`t use another channel to monitor both outputs at the same time. That is simple, to keep away any ground loops.

I did the test with the laptop powered from ac outlet and battery but the results are the same.


This time there were no ground loops and no bad soldering joints. It is only cronus which is not working properly.
Now there are months since I am trying to say this to you but you always put the blame on something else.

At the begining you said maybe it`s bad the firmware of the controller that you shipped with the b3 and to try another controller so I`ve built one with display (to see the sampling frequency) that showed better the problem which cronus has.

After this you still put the blame on other things but neither a time on cronus.
Ok so I went ahead and bought a scope to see for real what is going on. With the scope at first you still wouldn`t want to admit that there is something wrong with cronus. But now you can see clearly that it`s not serving me for the purpose I bought it. I hope it`s cronus and not the dac !

Now to find this out I spent some money and time Applause


I hope we can find a solution to this.

Thank you

Edited by user Tuesday, March 1, 2016 6:43:35 PM(UTC)  | Reason: Not specified

Russ White  
#56 Posted : Wednesday, March 2, 2016 12:49:31 PM(UTC)
Russ White

Rank: Administration

Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,978
Location: Nashville, TN

Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
I am curious, when did I say there was bad on-board firmware? I actually strongly suggest you use the on-board firmware unless you want to do something special. For one thing your sample rate calculator won't work correctly unless you run async - because it work by dividing by the master clock frequency, and since that is changed outside the knowledge of your firmware you will get a wrong answer in at best half the cases and at worst all cases.

In any case - I am not putting the blame on anything - I have no idea what is wrong - we are not talking about a pristine Cronus - this is a DIY project. All I can do is only suggest things to check so we can found out what the problem actually is.

That is what I have done. :) It is not as if what you are experiencing is common or normal - so I am interested to find out why it is happening to you.

Now I thought we already tested the clock divider - it was working fine last time? If I am going to suspect anything right now it would be that the clock divider is damaged and not creating the divided clock (as we can see it is the same freq and as the mck and not 1:2) There are a lot of reasons the divider could fail - most common would be an inadvertent short circuit at it's output.

The only other thing I can think of is that there is some solder bridge or short at the B3 end that is attenuating the clock to such a degree that the clock divider on the Cronus can no longer latch onto the signal.

If the Cronus was ever subjected to more than ~7VDC or ESD then component damage can easily occur and so the job of debugging DIY electronic projects is often finding what died. :)

Russ White  
#57 Posted : Wednesday, March 2, 2016 2:09:39 PM(UTC)
Russ White

Rank: Administration

Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,978
Location: Nashville, TN

Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
Make sure you check both clocks (bring CS high and low) as you may simply have an XO with low output drive.
schultzsch  
#58 Posted : Wednesday, March 2, 2016 2:58:20 PM(UTC)
schultzsch

Rank: Member

Groups: Member
Joined: 6/3/2010(UTC)
Posts: 53
Romania
Location: romania

Thanks: 2 times
When you mentioned that the onboard firmware could be damaged was on diyaudio or here on b3 section of the forum where I have started another thread in the past. If you like I could do a search for that but I am not here to do this things.

I apreciate what you did and still do for the diy comunity don't get me wrong.

Cronus works fine until I connect b3 mck at it's output.
When I say fine I am saying it puts out the right mck and mck/2.

Things change only if I load the output that goes to b3, load it with b3 mck input. When I do this the mck/2 that goes to amanero becomes actually mck.
If I load only the divider output when connecting amanero the mck/2 is actually what it has to be and remains like this.



schultzsch  
#59 Posted : Wednesday, March 2, 2016 2:59:15 PM(UTC)
schultzsch

Rank: Member

Groups: Member
Joined: 6/3/2010(UTC)
Posts: 53
Romania
Location: romania

Thanks: 2 times
I have never powered up cronus with more than 5.5v
Russ White  
#60 Posted : Wednesday, March 2, 2016 3:28:14 PM(UTC)
Russ White

Rank: Administration

Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,978
Location: Nashville, TN

Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
If I said anything like that on DIYAudio it was not meant to suggest that the B3SE firmware itself would not work - but possibly that you had a blank chip (not likely because Brian does an great job getting them out). But if you did have one that was blank it would be really easy to tell - you could just switch the mono switch and see if it had any effect - if it does then the controller is just fine and has been programmed correctly.

Can you please check both clock time bases? I am starting to wonder if you have an XO (45.1584Mhz in particular) with low output drive. It would not hurt to pull the Rhea modules off and touch up the solder joints on the pins and the pads of the clocks.

Also check the impedance at the clock in of the B3 - it should be very high.

Edited by user Wednesday, March 2, 2016 3:31:54 PM(UTC)  | Reason: Not specified

Rss Feed  Atom Feed
Users browsing this topic
4 Pages<1234>
Forum Jump  
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.