New TK2050 board

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Administrator
Joined 2004
Paid Member
shorting the dc probably won't damage the dac especially since it is only loaded by 50k but it might sound better with a cap to block the load.

I was thinking about this and at first thought for sure he needed a cap. But running the DAC into a 50K pot the DC load is only going to be about 0.05mA. Not likely to hurt it, you're right. Tho just to keep things in balance, maybe a 50K resistor to ground on the unused leg of the DAC? The data sheet does not seem to say what the max DC load is.

Generally these are meant to run into a lower impedance load. The Cirrus engineer I spoke to said they spec them at a 1K load. He would not tell me how they perform on other loads. I have not found higher impedances to be a problem.

Since your amp has DC blocking caps after the pot, you are not worried there.
The only thing I'd worry about is the amp switching noise getting back up into the DAC. You might want to put a little cap (0.01-0.1uF) across the inputs to drain off some of that noise.

Sure is a simple way to go! :checked:
 
Be careful, I just worry about the DC output of DAC will lead to a DC output of amplifier.
So, the cancellation of the input caps will be a risk to try.
Don't worry about the noise you mentioned, it will not happen I think. May be considering more about the ground will be more important.
 
if you have an oscilloscope, you can make a simple measure.
put CH1 at Vcc, ch2 at output of amp on load. Amplifier drives with burst signal 60Hz or 100Hz 33mS duration and see voltage decrease on oscilloscope.:)

sadly no. Just a digi-multimeter. I really cannot imagine that a voltage decrease should happen and if it´s really the case - it must be responsible for the tight dry and quick bass I am listening out of the speakers. I love voltage decreases:D

So I have to believe that you are right because I am a greenhorn in this area. Maybe someone else with more technical understatement could throw in his/her opinion ??
 
DAC direct is awesome

shorting the dc probably won't damage the dac especially since it is only loaded by 50k but it might sound better with a cap to block the load. Try another of the same cap you are using for the amp. Any value between 1uf and 3uf usually works well and gives enough extension in the bass.

I added a cap between the cs4397 and the tk2050 amp, and the sound is so good! Everybody should try this!!
What is the downsides, why doesn't commercial DAC do this?
 
Sense

I added a cap between the cs4397 and the tk2050 amp, and the sound is so good! Everybody should try this!!
What is the downsides, why doesn't commercial DAC do this?

Running a dac chip directly out from the pins doesn't make sense to an engineer and they only want to do what makes sense even after we tell them something else sounds better.
.
http://www.diyaudio.com/forums/digi...5-4396-ultrasonic-noise-spectrum-graphic.html
.
http://www.diyaudio.com/forums/digital-line-level/137231-ak4395-vs-ak4396-listening-comparisons.html
.
 
Administrator
Joined 2004
Paid Member
Right, it's not good "engineering practice". In the real world, you never know what people will connect to your design, so you have to be careful. The most important thing is that is works. And with a wide range of equipment. You don't have that restriction, you know what you are using.

I've tried the cap only and like it. But to my ear the transformer output sounds better. A matter of taste, really. Scott likes the cap coupling better.

Don't worry about added DC offset in this configuration. Your amp has DC blocking caps at the inputs, so the DC from the DAC will not pass thru. No worries.
 
sadly no. Just a digi-multimeter. I really cannot imagine that a voltage decrease should happen and if it´s really the case - it must be responsible for the tight dry and quick bass I am listening out of the speakers. I love voltage decreases:D

So I have to believe that you are right because I am a greenhorn in this area. Maybe someone else with more technical understatement could throw in his/her opinion ??
SMPS are usually capable to mantain a very well constant regulated voltage till their current limit.
If you notice a voltage drop on a regulated smps this means that the smps is underpowered for the application (you need more watts).

A *very little* voltage drop (about 0.01 V @ 7A for a 200W AT smps) is caused by the voltage drop on the output cables (more is the current consumption, more is the voltage drop on cables), this voltage drop generates a little accuracy error on the feedback circuit.
If you take the feedback signal downstream the output cables you will not have this irrilevant voltage drop.

That's my experience. I love smps.
Maybe they could produce some noise on the output, but if the board is well filtered (capacitors...) you will not hear nothing bad.

*Theoretically* a linear amplifier should sound better than a class T, but practice teaches us that's not always true.
Quite the same for linear power supply vs. switching power supply.
My 2 cents ;)
 
space.php

Where can we find more information about the T3 board?

Thanks,
Robert
 
Right, it's not good "engineering practice". In the real world, you never know what people will connect to your design, so you have to be careful. The most important thing is that is works. And with a wide range of equipment. You don't have that restriction, you know what you are using.

I've tried the cap only and like it. But to my ear the transformer output sounds better. A matter of taste, really. Scott likes the cap coupling better.
Can you say more, in what circumstances would it not work very good?
I would guess that cap out or transformer out doesn't make that big difference, compared to the direct out vs through filters and opamps.. thats very the big difference is.

I got a PM asking for pictures on how i connected the dac for direct out, so here it is. I was luck that this dac was mounted on a bigger connector plate, so it was quite easy to solder in my case.

I'm thinking about if I could talk hifimediy into making this type of direct out DAC :spin:. I know at least it works very good with this amp !


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.
 
for Don,

The 2.2uF or 3.3uF will not make a difference at that impedance, but if you make the size smaller than 2.2uF the cap starts acting as a high pass , and it will reduce the bass that comes thru. so if you need less bass, 1uF is just fine, i know from my amps that 2.2uF is also perfect, larger is not needed.
I think that I will go with 2.2uF K73-16 PETP capacitors:
http://www.diyaudio.com/forums/parts/143675-petp-capacitors-one-best.html

I think that comparing the frequency response of a 2.2uF setup vs 3.3uF setup there should be no appreciable difference in the audible range (20-20.000Hz).
My 2 cents and thanks for your info.
 
Hi Hifimediy and Nick (great initiative of you making this board accessibe via ebay)!

I recieved my board for a couple of days and I have let it run day and night to burn it in.

I´ve made some listing comparison to my moderate modded 2x100 sureboard (valab MKP2000 4.7uF input caps, passive cooling, T106-2
coils, stock ouput and tank caps) driven by a 27v 350w meanwell SMPS turned up to 29v.

I think your card performs very well, hard to say which I like best, my modded sureboard have some more depth/3D-sound and
"loudness", your board on the other hand I find more neutral sounding (and perhaps more correct).
Anyway, your card is far better than a stock sureboard and would be the obvious choice for someone deciding between them.

Some other observerations:

* the heat dissipation are similiar (I use a Zalman ZM-NB47J sink), perhaps your board is slightly warmer.
* The gain are close to the surebords lower gain setting (I think the sureboard have 16 v/v gain and yours 15 v/v)

Now to my question: I don't need the volumepot so I'm think of removing it and solder the line in cables direct to the output
holes of the removed pot. Do I instead need resistors (lets say 20k) parallell between signal lines and signal ground to give my
source some load or can I skip that? According to Tripath's application/test diagrams it doesn't seems necessary.
 
Last edited:
Input termination

Now to my question: I don't need the volumepot so I'm think of removing it and solder the line in cables direct to the output

holes of the removed pot. Do I instead need resistors (lets say 20k) parallell between signal lines and signal ground to give my

source some load or can I skip that? According to Tripath's application/test diagrams it doesn't seems necessary.
You can try it with and with out an additional input termination resistor in place of the pot or even just connect the input cables to the output pins of the pot instead of going through the pot to hear what that sounds like and use the pot setting toward the quiet end of it's adjustment to give the termination load. The input gain setting resistors of the amp will form a virtual 20K load with the pot removed but my dac sounded a little better with a 20K resistor to ground right before the amp's input cap.
 
Hi, refer to high pass filter formula.
fc=1/ (2*π*Ri*Ci ), π=3.14. Ri=20k
if you use Ci=2.2uF, fc=1/(2*3.14*20000*0.0000022=3.6Hz
means your amplifier will allow down to 3.6Hz siglal input.
we use Ci=3.3uF now, and will get better bass.
But don't try to use larger caps, which will lead to power off pop noise( will not happen on our boards, but we still not suggest use unnecessary large caps:D).
thanks.
You're right, because larger caps are slower in filtering:
Ft is 2.4Hz or 3.6Hz
if Ft = 1 / (2*pi*Tau), then Tau = 1 / (2*pi*Ft) = 66ms (or 44ms)

So the caps charging time will be = 3.5 * Tau = 230ms (or 150ms)
So larger caps are not the preferable choice for filtering DC.
Bye!
 
After 10 minutes use the initial sound quality of the board in "stock" form is not good. A slight "ringing" in the upper treble and weak bass. At the same time it feels like "loudness" in the mid. The voice of Rickie Lee Jones feels even more intence. Hopefully it will become better after a while of break in.

I will make a SS psu later for the boards also.
 
You can try it with and with out an additional input termination resistor in place of the pot or even just connect the input cables to the output pins of the pot instead of going through the pot to hear what that sounds like and use the pot setting toward the quiet end of it's adjustment to give the termination load. The input gain setting resistors of the amp will form a virtual 20K load with the pot removed but my dac sounded a little better with a 20K resistor to ground right before the amp's input cap.

Thanks for your "input" Scott! I will try around with your suggestions.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.