AUDIO PC: THE USB CARD

[rickmcinnis]

Nigel,

My curiosity has the best of me so I want to try an NEC equipped USB 2.0 card and hear what happens.

I do not think my fingers nimble enough to use a clock the size you recommended on page four,

I saw these on EBAY and wonder if you think they are worth trying. I have never heard of the brand so I am a little wary of getting something not as advertised considering the source, China. I am attracted by their size more than the accuracy claim.

http://www.ebay.com/itm/1PCS-TCXO-24M-2 ... 339e949fb0

To help the afflicted would you re-assure me how to wire this thing in.

Should ground be attached to something on the board or can I ground it at my star ground near the batteries? I have an existing 5 volt cable that was used for the JCAT which has failed me. ALSO - the little spec sheet says 3.3 OR 5 volts - I wonder does one get a choice or can the device use anything from 3.3 to 5 volts? Will ask the seller.

Any opinion?

AS always, in your debt,

[nige2000]

tried one of these before wasnt too bad


http://www.ebay.com/itm/Vanguard-TCXO-0 ... 2a32b63014

also the usual mouser digikey etc will have dip xo's

usually the pcie gnd only works fine but you can try both

think its usually 3.3v or 5v not a range between
probably go with 3.3v
theres a reg under the tin hood
had one of these apart a while back there just smd xo's with added circuitry i didnt particularly understand

[rickmcinnis]

So you are saying the SMD sounds better to you than the BIG ones?

Have yet to hear from the fellow in China about voltage. I would prefer a five volts version since I have the cable already made and attached. Was used for the JCAT.

I would suspect there is all kinds of circuitry in either one of these things!

Thanks,

[rickmcinnis]

The Chinese seller of the big TXCO says it is 5 volts default but can be used with 3.3 v.

I think I will give it a try. At least, the freight is included.

Got the 2.0 card installed yesterday. On the newegg webiste the picture showed it with the NEC chip but what I have is unmarked so I have no idea what it is. I do know it worked without any drama.

I think it sounds good. Half of the goodness of this has to be the pcie aspect, I would guess.

Did you see the posts from alecm at the SOEKRIS DAC site? Very interesting fellow and posts.

[nige2000]

The Chinese seller of the big TXCO says it is 5 volts default but can be used with 3.3 v.

I think I will give it a try. At least, the freight is included.

Got the 2.0 card installed yesterday. On the newegg webiste the picture showed it with the NEC chip but what I have is unmarked so I have no idea what it is. I do know it worked without any drama.

I think it sounds good. Half of the goodness of this has to be the pcie aspect, I would guess.

Did you see the posts from alecm at the SOEKRIS DAC site? Very interesting fellow and posts.

Interesting stuff have you got a link for the diyaudio post and for the USB card you got
Time is at a bit of a premium here these days

[rickmcinnis]

http://www.newegg.com/Product/Product.a ... 6815158035 is the card.

http://www.diyaudio.com/forums/vendors- ... z-305.html for the alecm posts - work backwards from here. He has found the +- 4 volts regs. Recommended installing extra capacitance on the output of something and it does make a big difference. I used 1000 uF OSCONs I had on hand, more than he recommended but it works well.

Ordered the Chinese clock.

Do you think there could be any value to replacing the caps on the card with OSCONS?

Blew up the cheapie CPU - have no idea how. I think now there might have been something wrong with it since with the now G3258 I am not having the same boot difficulty. The 3258s are onsale at the local computer store and I paid less for it than the CELERON. Wish I had noticed this before ...

spzzzzkt has a new blog for filters and other things - http://www.moredamfilters.info/

I do not think you are the only one busy at TIR NA these days!

[nige2000]

http://www.newegg.com/Product/Product.a ... 6815158035 is the card.

I never found any evidence that the usb2 chips are better its a shame you won't have the jcat to compare too
it is possible to match or better the jcat with diy usb3 card so it would have been a good reference
Maybe you should try a usb3 diy card too

http://www.diyaudio.com/forums/vendors- ... z-305.html for the alecm posts - work backwards from here. He has found the +- 4 volts regs. Recommended installing extra capacitance on the output of something and it does make a big difference. I used 1000 uF OSCONs I had on hand, more than he recommended but it works well.

Right so I have not looked at this in a while is it possible to inject 4v from lifepo4 at this point or do we still need +- 5v
If we can feed direct from lifepo4 we could negate the need for caps

Has anyone tried less than 4v coz 3.6v would suit the battery cells better

Ordered the Chinese clock.

Do you think there could be any value to replacing the caps on the card with OSCONS?

Quite Possibly
Maybe if you're feeding 3.3v from battery these stock caps could be a bottle neck

Blew up the cheapie CPU - have no idea how. I think now there might have been something wrong with it since with the now G3258 I am not having the same boot difficulty. The 3258s are onsale at the local computer store and I paid less for it than the CELERON. Wish I had noticed this before ...

spzzzzkt has a new blog for filters and other things - http://www.moredamfilters.info/

I do not think you are the only one busy at TIR NA these days!

Good to know I'm not the only one taking the time and risk with experiments
Hopefully I'll get a little spare time soon

[rickmcinnis]

alecm has not got around to attempting to replace the +- 4 volts regs.

BUT he was adding a transistor to the output of the opamp which brought the voltage to ... 3.6 volts and he said it worked.

The transistor mod was not actually correct but the thing worked with 3.6 volts. Now do you think one could get the batteries to supply equal voltage to both sides? That is critical here, as you know.

I am sending the JCAT back for repair so I will be able to compare the two. At this point it the STARTECH does sound better than the onboard USB and that is all I can ask for.

All of this is intertwined so I hope I am not doing wrong by mixing topics within a thread.

Yesterday I attempted to power the memory on the ASROCK board with the replacement BELLESON reg. Verified that it was working properly, it was producing the right voltage, a little higher this time, 1.32 volts. Plugged it in and nothing happens other that the keyboard lights. Attempted to boot numerous times with no luck.

I then replaced the choke and tried to boot. Same thing. Nothing happens but the keyboard lights up.

Going to place the CPU on that board into the machine I am using to write and see if somehow the remote memory power is destroying the CPUs. Do you see any scenario where this could be possible?

I told you I was going to catch up with you and there is no other way than losing the fear of stuff blowing up. Luckily the stuff is not terribly expensive, if it was I would not be as determined!

This machine has a 4130T in it and I suspect that will be better for the music machine than the G3258. G3258 is HASWELL but cannot use AVX instructions which is surprising. In the BIOS it lists having 8 cores! I know the thing runs much hotter than the CELERON. When I installed it I thought my BIOS setting had reverted to DEFAULT instead of 800 mHz and 0.8 volts (and dozens of other changes but those are the main ones) my big brass block was HOT. I mean really HOT.

I t had a more spotlit sound than the CELERON - when I was listening to it I assumed the BIOS had reverted to default and it sounded like it had. The space was there but very thin in density. Hardly any of the hash I heard with the other CPUs running at DEFAULT but it was not running at default. Not terrible but one can do better. It will work much better in a non music computer.

The previous CPUs would make it warm at best. Maybe the settings I was shown had nothing to do with what it was really doing? Very strange. I hope it still works but when I bought this one I got the five dollars insurance policy.

But I do think the ASROCK is gone. Going to that same store and getting a GIGABYTE b85 board that is not ideal but is the best they have that I have not already used. Looks like it has a superior CPU power circuit to the MSI boards I got from them. Yes, two of those are dead, too.

I previously only used GIGABYTE boards and then had some funny ones. Hope I can keep this one running for longer than two weeks. At this point I am nervous about installing the memory regulator.

Audio is always waiting for you when you get some time. First things first ...

[nige2000]

alecm has not got around to attempting to replace the +- 4 volts regs.

BUT he was adding a transistor to the output of the opamp which brought the voltage to ... 3.6 volts and he said it worked.

cool

The transistor mod was not actually correct but the thing worked with 3.6 volts. Now do you think one could get the batteries to supply equal voltage to both sides? That is critical here, as you know.

well im confident the voltage coming out of the cell will be the same as the voltage coming from its floating supply
the cell will add stability
how accurate does the voltage matching need to be?
i think i could match to the nearest 0.01v with the regs i got in

I am sending the JCAT back for repair so I will be able to compare the two. At this point it the STARTECH does sound better than the onboard USB and that is all I can ask for.

well its good you can get the jcat repaired
i hope your powering your usb/i2s separate when doing comparisons

All of this is intertwined so I hope I am not doing wrong by mixing topics within a thread.

no probs
but if were doing further dac mods we better report results in the dac thread

Yesterday I attempted to power the memory on the ASROCK board with the replacement BELLESON reg. Verified that it was working properly, it was producing the right voltage, a little higher this time, 1.32 volts. Plugged it in and nothing happens other that the keyboard lights. Attempted to boot numerous times with no luck.

I then replaced the choke and tried to boot. Same thing. Nothing happens but the keyboard lights up.

Going to place the CPU on that board into the machine I am using to write and see if somehow the remote memory power is destroying the CPUs. Do you see any scenario where this could be possible?

I told you I was going to catch up with you and there is no other way than losing the fear of stuff blowing up. Luckily the stuff is not terribly expensive, if it was I would not be as determined!

This machine has a 4130T in it and I suspect that will be better for the music machine than the G3258. G3258 is HASWELL but cannot use AVX instructions which is surprising. In the BIOS it lists having 8 cores! I know the thing runs much hotter than the CELERON. When I installed it I thought my BIOS setting had reverted to DEFAULT instead of 800 mHz and 0.8 volts (and dozens of other changes but those are the main ones) my big brass block was HOT. I mean really HOT.

I t had a more spotlit sound than the CELERON - when I was listening to it I assumed the BIOS had reverted to default and it sounded like it had. The space was there but very thin in density. Hardly any of the hash I heard with the other CPUs running at DEFAULT but it was not running at default. Not terrible but one can do better. It will work much better in a non music computer.

The previous CPUs would make it warm at best. Maybe the settings I was shown had nothing to do with what it was really doing? Very strange. I hope it still works but when I bought this one I got the five dollars insurance policy.

But I do think the ASROCK is gone. Going to that same store and getting a GIGABYTE b85 board that is not ideal but is the best they have that I have not already used. Looks like it has a superior CPU power circuit to the MSI boards I got from them. Yes, two of those are dead, too.

I previously only used GIGABYTE boards and then had some funny ones. Hope I can keep this one running for longer than two weeks. At this point I am nervous about installing the memory regulator.

Audio is always waiting for you when you get some time. First things first ...

ive never managed to blow a cpu yet i doubt the memory power mod is doing the harm
maybe a short on the 12v to cpu or reverse polarity or voltage over 14v
cant really think of anything else

[rickmcinnis]

I get the feeling the voltages have to match pretty close but one has to wonder how well those chips match? They are not precision parts so that must be enough. I remember SOEKRIS saying something to the effect that his circuit compensates for small variations.

If you haven't already looked these things are near the balanced output jacks on the left side if you were holding the board with those jacks closest to you. There are two sets of +/-, obviously, and they are in a line across the board. There is a via where you can measure the voltage right next to them.

I have to believe my memory reg is doing something to destroy the CPU. The G3258 was definitely gone so the memory mod killed two cpus. I win on that count.

What are you using for your memory reg? Do you do anything for start up? With the MSI the memory reg was always on since it was connected directly to the 5 volts battery pack. Should this turn on with the rest of the voltages? Not sure how I could do that other than having a switch after the batteries and I do not like the idea of a switch after the batteries. I have been plugging in the P24 when I want to use the machine.

I wonder if this reg is not putting out the current the first one did and this starvation is doing something that destroys the cpu. I did nothing differently. Voltage reads correctly but I have no idea about the current. I will ask BELLESON if they will test the thing and see if something is wrong.

WAVE IO is powered by Doede's 5 volt shunt reg. Will eventually put some batteries in series.

No overvoltages. I monitor the voltages and they are slightly low. All are slightly low.

Ended up getting an H97m D3H. Sure has a bunch of stuff on it compared to the ASROCK.

Now to see if the regulator destroyed my memory, too.

Take care,

PS the computer store replaced the G3258 with no questions. Got lucky on that one.