This is how it is all setup - after making a drawing that has been sent to Lars.
230V AC ground is connected to the case. The power supplies and the NCD modules are mounted on aluminum plates that are NOT in electrical contact with the case. The XLR input contacts are in electrical contact with the case, the RCA input and the speaker outlets are NOT.
As far as I can see everything is connected as it should according to the cookbook, the power supply has 56,5V+ to Gnd, 56,5V+ to Gnd, and 37,5V from Vgate to Gnd.
The GND are all connected together on the RCA input L, XLR L, XLR R and RCA R, and are also connected to "chassi GND" on XLR L and XLR R as well as to input GND on both modules.
Measurements gives as follows on speaker outlets with low-medium listening volume from Mp3 AC=0,4V, from unbalanced PC-card=0,8V, balanced=0,8V. DC from Mp3=0.03V, unbalanced PC-card=0,04V, balanced=0,04V.
Measures on unbalanced input terminal=0,0002V DC, unmesaurable AC (lower than internal variations of instrument)
Thats about it - only one channel measured as the other one is dead....
230V AC ground is connected to the case. The power supplies and the NCD modules are mounted on aluminum plates that are NOT in electrical contact with the case. The XLR input contacts are in electrical contact with the case, the RCA input and the speaker outlets are NOT.
As far as I can see everything is connected as it should according to the cookbook, the power supply has 56,5V+ to Gnd, 56,5V+ to Gnd, and 37,5V from Vgate to Gnd.
The GND are all connected together on the RCA input L, XLR L, XLR R and RCA R, and are also connected to "chassi GND" on XLR L and XLR R as well as to input GND on both modules.
Measurements gives as follows on speaker outlets with low-medium listening volume from Mp3 AC=0,4V, from unbalanced PC-card=0,8V, balanced=0,8V. DC from Mp3=0.03V, unbalanced PC-card=0,04V, balanced=0,04V.
Measures on unbalanced input terminal=0,0002V DC, unmesaurable AC (lower than internal variations of instrument)
Thats about it - only one channel measured as the other one is dead....
I don´t know this will turn out, or which steps that will be involved but I´m not giving up. The sound from the amp is far to good not to try to repare and correct the problems to.
Maybe the NCD1 needs adjustement on the input side to be able to handle certain soundcards? This E-MU 1616M is a proffesional soundcard used everywhere in smaller studios etc so it should be fairly good electrical wise.
Or maybe the NCDX has a better chance of adapting to PCs etc????
Maybe the NCD1 needs adjustement on the input side to be able to handle certain soundcards? This E-MU 1616M is a proffesional soundcard used everywhere in smaller studios etc so it should be fairly good electrical wise.
Or maybe the NCDX has a better chance of adapting to PCs etc????
If the Vgate should be max 20VDC and I´m using the standard NewclassD power supply one wonder how that can be?
The colours on the four trafo wires are not the usual standard ones but I have measured:
42V between Yellow&Red,
42V between Grey&Blue
31V between Yellow&Blue
48V between Grey&Red
5V between Yellow&Grey
5V between Blue&Red
I also measure 0 Ohm between Yellow&Red, as well as between Blue&Grey when the power is off, and Full resistance between all the other combinations.
So I put Yellow at pin 1, Red on 2, Grey on 3 and Blue on 4 on the power supply. I hope this is the right way?!?!?
The colours on the four trafo wires are not the usual standard ones but I have measured:
42V between Yellow&Red,
42V between Grey&Blue
31V between Yellow&Blue
48V between Grey&Red
5V between Yellow&Grey
5V between Blue&Red
I also measure 0 Ohm between Yellow&Red, as well as between Blue&Grey when the power is off, and Full resistance between all the other combinations.
So I put Yellow at pin 1, Red on 2, Grey on 3 and Blue on 4 on the power supply. I hope this is the right way?!?!?
I know you won't believe this because each module goes through Lars' infamous testing procedure, but it sounds like you have a defective pair.
Try usingp liars to pull on the magnet wire of the filter coil hard enough to slightly deflect the PCB to ensure it's properly soldered. Do both ends and both modules.
You may find a dry socket. No, this MCU isn't smart enough to leave the amps off upon a critical fault so it will just keep hammering your speakers until they pop, but I guess you know that now. Didn't save the module either hm wow that's interesting wonder why.
Failing that perhaps you should try the 627 input stage as it seems to cure all other ailments.
Try usingp liars to pull on the magnet wire of the filter coil hard enough to slightly deflect the PCB to ensure it's properly soldered. Do both ends and both modules.
You may find a dry socket. No, this MCU isn't smart enough to leave the amps off upon a critical fault so it will just keep hammering your speakers until they pop, but I guess you know that now. Didn't save the module either hm wow that's interesting wonder why.
Failing that perhaps you should try the 627 input stage as it seems to cure all other ailments.
CDL:
I sent hom some pictures to Lars and he came back with some suggestions of possibly wrong connections but it was only a matter of problems with seing in the pictures exact how things are done. It at least gave me another reason to check AGAIN how everything is setup. Everything was connected in the right even the fift fime I checked it.
That was yesterday and havent heard from him today - He use to come back as soon as he are able regarding work&travel so I hope he will come up with something more.
I have now olso measured the DC from the soundcard on high volumes (using another amp) and it is still 0,004something V, lower than my instrument can recognise.
I have measured between chassi-gnd and output-gnd from modules (=minus/Gnd on the speaker connector) and get 2.5 Ohm. Disconneting the module the speaker connector has infinite resistans to chassi so that is right.
Teflon1, What do You mean "dry socket"?
And I think this is an actual problem somewhere, I have burned more than a pair at this moment. I don´t know if there could be something like incompatibility between a soundcard and an amp, but maybe this is a case of that?????
The soundcard is a proffesional one with setup for both proffesional use and "consumer" use (=lover levels of output).
I sent hom some pictures to Lars and he came back with some suggestions of possibly wrong connections but it was only a matter of problems with seing in the pictures exact how things are done. It at least gave me another reason to check AGAIN how everything is setup. Everything was connected in the right even the fift fime I checked it.
That was yesterday and havent heard from him today - He use to come back as soon as he are able regarding work&travel so I hope he will come up with something more.
I have now olso measured the DC from the soundcard on high volumes (using another amp) and it is still 0,004something V, lower than my instrument can recognise.
I have measured between chassi-gnd and output-gnd from modules (=minus/Gnd on the speaker connector) and get 2.5 Ohm. Disconneting the module the speaker connector has infinite resistans to chassi so that is right.
Teflon1, What do You mean "dry socket"?
And I think this is an actual problem somewhere, I have burned more than a pair at this moment. I don´t know if there could be something like incompatibility between a soundcard and an amp, but maybe this is a case of that?????
The soundcard is a proffesional one with setup for both proffesional use and "consumer" use (=lover levels of output).
I would try connecting everything as though you were going to play music - first with the amp turned off - and power up the computer, play some music (amp still OFF) and then pause music player. Then measure as close to the input of the amp as possible. If there is no DC (or the 0,004V you measured are close enough), go on and power up the amp, measure on the input again, and then finally on the output. As far as I can read you already did something similar?
As for balanced / single ended, let's stick with single ended testing for the moment...
Unfortunately, NCD does not take care of DC, and does not contain a servo. Therefore it is very important NOT to feed it with any DC. You might consider using a coupling cap - as a temporary solution of course. Try something similar to what the cookbook describes for tube preamps.
I agree with you, I still think your problem is DC related somehow, maybe at higher outputs only (since the modules power up well, but fail when you try to go a bit louder). Maybe you could try provoking the amp to power off (using big resistor or disposable speaker!) and monitor DC and supply rail voltage while you're doing it. I had a stupid problem with my DAC once, which made one power rail fail (only) when drawing current...
Next step could be to measure for DC on the outputs of the opamp (pin 1 and pin 7), also check that input voltage (pin 4 and pin 8) are correct... maybe you have a failed input opamp?
I assume you are not using external power supply for the opamp ;-)
Perfect!! Then there should practically no resistance between PC chassis and amp chassis, and maximum 15ohm between PC chassis and input Gnd - which is the result you got.
Sounds right. Can you measure the same on the screws of the plugs connecting into the modules?
Concerning your remark with no measureable DC on V-gate - I read the later post where you found the right voltage after all, but please double check your V-gate for loose connections - I think a loose V-gate turning randomly on and off could cause trouble.
I would - at any cost - avoid switching on and off (and even rebooting) the computer with the amp powered on - wierd stuff sometimes happens when powering up and down PCs :-(
I think your soundcard is fine for this purpose, as you say it's a pro card and should be good. However, it seems that connecting a PC to a power amp directly can often cause trouble :-( In theory, everything should be allrigth since you are all "grounded" correctly etc., but I would still consider an attenuator of a sort on the input (I am sure you still have lots of gain), and a DC blocking cap. Pro gear usually ISN'T DC-coupled, or is made to be DC resistant some other way.
Another thought... could the soundcard be clipping at higher volumes?? Maybe if you have an extra soundcard of a sort, you can loop the output of the EMU into the input of the extra card and check for clipping (just use some kind of recording utility to record a bit of the EMU output while playing loud music, and check for clipping).
Last chance of course is that you might have a defective module - however, it's unlikeley that you had two defective ones. I think Teflon1 means a loose soldering on the PCB, or one which doesn't contain enough solder to make a good contact.
I sure hope we end up figuring out what the problem is! Maybe Lars will have more good input, but it does seem as though you have done everything right.
/cdl
Last edited:
ank You CDL for Your engagement. I try to anser to Your numbers above:
1) I have now measured DC on different listening levels (with another amp connected) and there is still almost none. Measured inside on the connectors of the amp.
3) It could be a good way of doing it, to measure on the pins on the op amp. I think I do that. What should the measurements be here?
6) No loose connection(s)
8) The soundcard has protection circuits (tells the manual & EMU Homepage) to prevent problems when switching on and off.
The soundcard comes with an software mixer that has to be setup (You can do wathever You posibly can think of and twice as much with it) and there is possible to set in level meters both on incoming signal and outgoing signal to and from the card so as far as I can see the levels are right. No clipping or overload.
And also, I have played on the little Sentec 30W amp with higher output volume from the sound card than ever used with the EMU. Both because of the fact that the Sentec has the Old DIN-standard and are only 30W. No clipping or strange things.
Lars think I should send it to him for measurements and fixing and I think it would be nice to give the whole setup (PC + amp) to him for measurent together. I might go down to Holsterbo and hand them oveer to him Dec 2nd and 3rd as my wife works in Kalundborg a few days every two weeks.
1) I have now measured DC on different listening levels (with another amp connected) and there is still almost none. Measured inside on the connectors of the amp.
3) It could be a good way of doing it, to measure on the pins on the op amp. I think I do that. What should the measurements be here?
6) No loose connection(s)
8) The soundcard has protection circuits (tells the manual & EMU Homepage) to prevent problems when switching on and off.
The soundcard comes with an software mixer that has to be setup (You can do wathever You posibly can think of and twice as much with it) and there is possible to set in level meters both on incoming signal and outgoing signal to and from the card so as far as I can see the levels are right. No clipping or overload.
And also, I have played on the little Sentec 30W amp with higher output volume from the sound card than ever used with the EMU. Both because of the fact that the Sentec has the Old DIN-standard and are only 30W. No clipping or strange things.
Lars think I should send it to him for measurements and fixing and I think it would be nice to give the whole setup (PC + amp) to him for measurent together. I might go down to Holsterbo and hand them oveer to him Dec 2nd and 3rd as my wife works in Kalundborg a few days every two weeks.
Hi again,
I now played with my last NCD channel higher and higher continuously measuring the DC on the input. The result was – not measurable.
At a listening level (on one speaker) that could be named as medium-to-fairly-high-but-not-very-loud by me the amp shutted down for parts of seconds a few times and started again. I just turned the whole thing down, not wanting another module or speaker (not many left in this house) to break down. I did throw the probes from the instrument away just to as fast as I possibly could to turn the amp down before disaster.
Same thing as before, but now with an actual measurement – NO DC.
So here I am with a one functioning NCD module, one lousy Voight horn and one burned NCD module, two burned great Thiel speakers and one burned Voight horn.
Any suggestions anyone?
I now played with my last NCD channel higher and higher continuously measuring the DC on the input. The result was – not measurable.
At a listening level (on one speaker) that could be named as medium-to-fairly-high-but-not-very-loud by me the amp shutted down for parts of seconds a few times and started again. I just turned the whole thing down, not wanting another module or speaker (not many left in this house) to break down. I did throw the probes from the instrument away just to as fast as I possibly could to turn the amp down before disaster.
Same thing as before, but now with an actual measurement – NO DC.
So here I am with a one functioning NCD module, one lousy Voight horn and one burned NCD module, two burned great Thiel speakers and one burned Voight horn.
Any suggestions anyone?
Damn - sorry there are no results so far. But better to be safe than burn more speakers though!
On the opamp outputs (I am guessing) there should be 0V DC (or, if there is DC, it should be balanced, i.e. +2V on one and -2V on the other or similar).
The supply of the opamp should also be symmetric, somewhere around 4.5V or so.
Sounds like a good idea to bring Lars the whole thing so we can have a proper diagnosis.
Good luck!!! and keep us updated!
/cdl
On the opamp outputs (I am guessing) there should be 0V DC (or, if there is DC, it should be balanced, i.e. +2V on one and -2V on the other or similar).
The supply of the opamp should also be symmetric, somewhere around 4.5V or so.
Sounds like a good idea to bring Lars the whole thing so we can have a proper diagnosis.
Good luck!!! and keep us updated!
/cdl
Thank´s again!
Yes I will keep You posted.
Another question: Anyone heard of anyone using NCD1, 4 or X in a multichannel setup for hometheater?
I started up with the best X-fi 7 channel soundcard as I thought could give a nice sound. It has the same type of DAC components as the EMU1616M card has so I thought it could be good - But it wasn´t. Not so good at all, especially compared to the EMU1616M card. So there went the 5-7 channels.
I have also - more or less seriously tried to talk Lars into contruct a multichannel preamp with Video functions (with all that stuff that comes with that) so far no luck
Yes I will keep You posted.
Another question: Anyone heard of anyone using NCD1, 4 or X in a multichannel setup for hometheater?
I started up with the best X-fi 7 channel soundcard as I thought could give a nice sound. It has the same type of DAC components as the EMU1616M card has so I thought it could be good - But it wasn´t. Not so good at all, especially compared to the EMU1616M card. So there went the 5-7 channels.
I have also - more or less seriously tried to talk Lars into contruct a multichannel preamp with Video functions (with all that stuff that comes with that) so far no luck
What source file did you use for testing - CD audio rip? I believe the X-fi internally resamples 44.1khz to 48 or 96khz, like other Creative cards do, could that be the reason for bad music performance? For some Creative cards it is possible to disable the resampling in software, but not my Audigy2 ZS, so it isn't very good for 41khz audio, but for movies (48khz / 96khz) I feel that the sound is OK. The obvious over-emphasis on the bass area doesn't matter so much for movies either.
Apart from EMU cards, I only found a few M-audio and Turtle Beach that would do true 44.1khz playback AND multichannel. However, they were either total overkill (pricewise too), or didn't do 96khz / 24-bit playback.
BUT - just checked the specs of the EMU1616M (VERY nice soundcard, my comments on clicks/DC earlier were because I was thinking of the cheaper EMUs w/o breakout box - yours should be great). Anyway, the spec say that it has 6 analog outputs??? Won't it let you do 5.1...? A question of software, maybe?
If not, for multichannel I guess we're down to:
- using an inferior soundcard, or
- the easy solution - SPDIF to an external decoder in a reciever, or
- use the capability of the Creative cards to output 3 distinct 2-channel SPDIF streams, i.e. one for main speakers, one for surround, and one for center/sub - 3 DACs needed . For a while I considered building 2 extra DACs to play with this, until I found out about the 44khz issue. I don't know how it works for AC-3/DTS but if the resampling issue can be bypassed on the newer cards, that still might be an option.
/cdl
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.