Rank: Member
Groups: Member
Joined: 7/26/2012(UTC)
Posts: 4
Thanks: 1 times
|
Hi, yesterday i finished to build a kit composed of LCD/B-PS, COD dac with single S/Pdif receiver and legato 3.1 as output stage. The legato has been modified changing the resistor values as suggested here. I will post detailed photos of the assembled board if it can help, but 'd like to describe the problem first. Subjectively the dac sounded fine, then i tried to make some measurements with RMAA and a pc-based spectrum analyzer, just to check if it was really working as intended. Tests were carried out with a Terratec DMX 6-fire usb. This is the RMAA spectrum with 1Khz test signal,coming from the SE outputs, isn't that harmonic distortion a bit too high ?  To see if there was something wrong with the measurement technique i tried connecting a Zero Dac and do the same test, results were still disappointing, as you can see there's a very high noise floor and still some kind of heavy distortion present, so maybe it' a problem with the soundcard, but in loopback measurements it does fine and i used it in the past to do similar tests on an acrive crossover filter i was configuring and it measured well. In addition to that i have also tried the same tests with an EMU-0404 and the COD and i still get that kind of distortion so i guess the card is fine and there's something wrong with (both?) dacs.  As a first attempt to find what's wrong i tried to repeat the test on the balanced outputs with Bal/Se "conversion" quickly done by putting a coupling cap in series to eliminate DC offset and connecting the soundcard SE inputs to the + and GND terminals on the legato. The signal coming out this way seems much cleaner, considering how it was adapted to SE   This is a summary of the measurements, in order the card's loopback baseline, the Zero DAC (even that doesn't seem to be working properly according to these tests...), and then the Cod/Legato with SE and "balanced" outputs. I believe all this suggests that the balanced input stage is working as designed and something might be wrong in the Bal/Se stage, so i should begin to check that area of the board for errors, am i right ? Or should i just disregard these measurements altogether ? I can provide any useful info or picture. Thanks in advance for any suggestion. Edited by user Thursday, July 26, 2012 3:49:27 PM(UTC)
| Reason: Not specified
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
You should always use a balanced to SE converter if you want to use or measure an SE signal. :) Otherwise its not going to be very good at all. It look like you did (Legato), but it may simply be that the operating point of the Legato is not exactly right just yet. I will give you some suggestions for things to try later. Edited by user Thursday, July 26, 2012 6:24:13 PM(UTC)
| Reason: Not specified
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
To start what voltage rails are you running the legato at?
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
Also please measure the voltage across R1-4 on the legato with the DAC connected and everything powered but no music playing. We may need to increase or decrease the CCS current based on your answer. :) Edited by user Thursday, July 26, 2012 6:45:29 PM(UTC)
| Reason: Not specified
|
 1 user thanked Russ White for this useful post.
|
|
|
|
Rank: Member
Groups: Member
Joined: 7/26/2012(UTC)
Posts: 4
Thanks: 1 times
|
I'm running the legato at +/- 15V, across R-1, R-4, which are 355 Ohm resistors i'm reading around 5.5V. I've checked the board for bad solder joints or short circuits but they seem good, another thing i noticed is that there seems to be a slight channel imbalance ( around 0.7 dB as measured by my soundcard ) between left and right channel, even using the proper SE output on the legato. The board gets quite hot, especially the opamp i can't touch it for more than a couple of seconds, is that normal ?
The setup of the dac and receiver is as recommended, the dip switches on the COD are all set to 0 (-).On a side note, i've just bought the PCB from TPA and sourced the parts myself but they are just the recommended ones except maybe for the output inductors, for which i found an equivalent ( at leas i believe so ) replacement.
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
Ok excellent. I believe I know what to do. We need a bit more current out of the CSS. Let change the 1.5K to 1K at the CSS. The other option is to increase R1-4 but that is less desirable. It has been a while since I tried this with COD and I did not use SE output. What I believe is happening is the common mode of the balanced signal is too close to the VCC rail of the BAL/SE opamp. If you where using the balanced output directly it would like be very very clean. Edited by user Thursday, July 26, 2012 10:24:43 PM(UTC)
| Reason: Not specified
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
Now after changing the resistors what you will be able to do is adjust the negative rail until you reach the optimum THD. Edited by user Thursday, July 26, 2012 11:52:44 PM(UTC)
| Reason: Not specified
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 7/26/2012(UTC)
Posts: 4
Thanks: 1 times
|
I have swapped the eight 1.5K resistors with 1K, i had available only 1/4W metal film resistors, not an ideal choiche as they will be operating close to their maximum power dissipation...however the change doesn't seem to have solved the problem. However, i've managed to find balanced connectors and did an RMAA measurement with the Emu0404 which has balanced stereo inputs and results seem very good, even if i don't know the card baseline because i don't have a proper loopback cable, i'm tempted to conclude that the balanced outputs are working well.  This is the THD graph for the above mentioned 32/192 measurement  could be better but i'm still not absolutely certain of the reliability of these measurements.But reliable or not, the same test with SE outputs doesn't really look good.  These tests were taken with +/- 12.3 V rails, tried experimenting with higher values with no appreciable reduction of the apparent distortion out of the SE output, i say apparent because it's not subjectively audible when listening to music and might as well be an artifact of these measument techniquest. Both outputs are simultaneosly on and with the EMU i can cycle between them and listen to musing using the direct monitor, i can't really say there's something terribly wrong with the SE output.... Maybe i should try to assemble a quick active balanced to single ended converter on a breadboard with a couple of opamps to see if there's something wrong with the onboard one ? Ah by the way now i read around 6.9V across R1-4 with 12.3V rails will do further test with higher supply voltages. No AC coupling caps added to the balanced outs. Forgot to add that yesterday i tried replacing the 4562 opamp with an opa2604, with no changes. The DC offset of the +/- balanced outs with respect to GND is 5.3V now, was around 9V yesterday with 15V supply and different resistors but i'm not 100% certain i'm remembering correctly. Thanks for your help !  Edited by user Friday, July 27, 2012 11:51:22 AM(UTC)
| Reason: Not specified
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
Leave the positive rail at 15V. Only adjust the negative rail. :)
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 7/26/2012(UTC)
Posts: 4
Thanks: 1 times
|
Hi, while experimenting with the legato something must have gone wrong and suddely the Dac stopped working. I noticed the analog VR on the COD module was getting ridiculously hot and has stopped to provide the correct voltage ( in fact it didn't provide any ). I have since then bought another COD module and an IVY-III, even if i'm not sure the legato has been damaged in the incident; all parts stock, except for R13-R16 replaced with 356 Ohm resistors as suggested on the manual ( actually i used the 355.55 Ohm resistors removed from the Legato ). What's really strange is that i'm getting almost the same result...here is the SE output  The balanced outs are marginally cleaner. How is it possible ? this is an entirely different circuit, it's very unlikely that i made again some random soldering mistake that yelded the same outcome as with the legato. Ma it be related to the digital part of the circuit ? My jumper configuration should be as recommended, but i might be wrong, i uploaded an image where it should be visible  On the cod module i just installed the jumpers and terminal blocks, nothing else.
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
Maybe try a different test rig?
|
|
|
|
|
|
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.
