hi there.
recently, i build a DAC based on 2x TDA1545A with CS8414 receiver.
i was upgraded the I/V converter from opamp to a 12AT7 tube (no buffer.tight budget
)
in the spdif input, i did lampizate (thanks to mr.Fikus for the idea) using...ECC83 or 84?..i forget that number again
.
my source is my laptop connected to a soundblaster surround pro 5.1 which i modded w/ a linear PSU.
actually i was pleased w/ the sound but i want more improvement.
what could i do without spending too much money?.
thanks in advance
recently, i build a DAC based on 2x TDA1545A with CS8414 receiver.
i was upgraded the I/V converter from opamp to a 12AT7 tube (no buffer.tight budget
)in the spdif input, i did lampizate (thanks to mr.Fikus for the idea) using...ECC83 or 84?..i forget that number again
.my source is my laptop connected to a soundblaster surround pro 5.1 which i modded w/ a linear PSU.
actually i was pleased w/ the sound but i want more improvement.
what could i do without spending too much money?.
thanks in advance
The link to my blog is on each of my posts - look to the left there's a number (78 now) just above the icons for Yahoo and Skype - if you click on that you'll get it.
To run differentially you just need an inverter on the Data line (pin 3 of your DAC chip). A 74HC04 chip has six such inverters, you just need one of them.
What voltage do you have on the Vref pin (pin7) ?
To run differentially you just need an inverter on the Data line (pin 3 of your DAC chip). A 74HC04 chip has six such inverters, you just need one of them.
What voltage do you have on the Vref pin (pin7) ?
Vref at 3V sounds rather high - normally for 1mA output current its 0.83V, can go up to 1.67V (for 2mA).
With differential you can use a balanced input (or a transformer) to reject PSU noise - as you have to reference your I/V resistor to some voltage. With your tubes, perhaps best to use the transformer. So you have two I/V resistors per channel, going to two passive filters, feeding one trafo to go bal-SE.
The particular link to a good sounding passive filter (there are others) is here : http://www.diyaudio.com/forums/blogs/abraxalito/1029-better-sounding-anti-imaging-filter.html
With differential you can use a balanced input (or a transformer) to reject PSU noise - as you have to reference your I/V resistor to some voltage. With your tubes, perhaps best to use the transformer. So you have two I/V resistors per channel, going to two passive filters, feeding one trafo to go bal-SE.
The particular link to a good sounding passive filter (there are others) is here : http://www.diyaudio.com/forums/blogs/abraxalito/1029-better-sounding-anti-imaging-filter.html
I haven't tried winding my own transformer yet between DAC and amp/buffer stage but I am planning to try it. I do have a (very cheap) pair of hand-made trafos between the DAC (which is balanced out) and my amp (which has SE input). They work very well. They are made from PQ ferrite cores and cost me uround $1 for the two of them - not including my time
The chokes in that filter are commercially available 10mH ones, unshielded. So they may pick up some hum. But since then I have moved over to TDK shielded chokes. 1mH is the largest value of these so the impedances need scaling down by a favtor of 10 to use them. Or you could go for one of my more recent filter designs.
The chokes in that filter are commercially available 10mH ones, unshielded. So they may pick up some hum. But since then I have moved over to TDK shielded chokes. 1mH is the largest value of these so the impedances need scaling down by a favtor of 10 to use them. Or you could go for one of my more recent filter designs.
I've used PQ32 cores so 32mm across and almost as tall.
Resistance of primary is (from memory) between 100 and 200 ohms, I think I used 0.1mm wire. But this trafo is after an active (voltage amp) stage, its stepping down the voltage. If you want one after the passive filter then it might be a bit different - the signal levels are lower and the driving impedance is higher. You'd want 1:1 ratio I think.
One key design aspect when driving a trafo from a high-ish impedance such as a passive filter is the LF roll-off is a function of the primary inductance. So you need to decide how low you want to go down at the LF - pick your -3dB point and that defines your minimum inductance. Inductance is an issue for ferrite trafos as they haven't such high mu as the more expensive metal-cored ones.
Resistance of primary is (from memory) between 100 and 200 ohms, I think I used 0.1mm wire. But this trafo is after an active (voltage amp) stage, its stepping down the voltage. If you want one after the passive filter then it might be a bit different - the signal levels are lower and the driving impedance is higher. You'd want 1:1 ratio I think.
One key design aspect when driving a trafo from a high-ish impedance such as a passive filter is the LF roll-off is a function of the primary inductance. So you need to decide how low you want to go down at the LF - pick your -3dB point and that defines your minimum inductance. Inductance is an issue for ferrite trafos as they haven't such high mu as the more expensive metal-cored ones.
If you're asking is jitter a problem - then I haven't found it to be, no. I haven't paid attention to clocking issues on my TDA1545s so far and have made significant strides in getting better SQ.
If trafos scare you then try running an SE filter and see how you get on. You'll need to match the filter impedance to your I/V resistor - what value are you using now? The filter also loses 6dB.
If trafos scare you then try running an SE filter and see how you get on. You'll need to match the filter impedance to your I/V resistor - what value are you using now? The filter also loses 6dB.
1k would be the maximum value if you've got Vref =0.83V and 2 chips in parallel. Then you'd get 2V total swing which is the compliance limit for the 1545 on 5V supply.
Using the passive filter I linked it needs 1k5 I/V resistor and the same load resistor - thus the DACs 'see' 750R. You can make up 2dB of the loss incurred in the filter by increasing your Vref by a third.
Using the passive filter I linked it needs 1k5 I/V resistor and the same load resistor - thus the DACs 'see' 750R. You can make up 2dB of the loss incurred in the filter by increasing your Vref by a third.
I'm not familiar with the Monica circuit, I'll go search for it.
Upper limit for compliance is VDD - 1V.
Not sure what pic you're wanting....
<later> Found a thread founded by our very own hollowman here : http://www.diyhifi.org/forums/viewtopic.php?f=2&t=1216
Upper limit for compliance is VDD - 1V.
Not sure what pic you're wanting....
<later> Found a thread founded by our very own hollowman here : http://www.diyhifi.org/forums/viewtopic.php?f=2&t=1216
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.