LPS for dac

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
The first linear power supply you linked to already has a transformer, it is the big blue square thing with fins in the corners (actually that is a plastic cover over the transformer).

The other thing you need to know is that besides plus and minus 15 volt supplies, you will also need a 5 volt linear supply for Katana. If you choose a power supply and or a transformer to work with all three linear power rails you need, then you will have enough.

--------------------------------------------
In addition, you will need one more 5v supply for RPi and for half of the Katana Isolator. That power supply could be a 5v SMPS. Or it could be linear if you prefer.

--------------------------------------------

martin155, If I may ask, how much do you know about electronics? I sort of get the impression that maybe a lot of this is new to you? Please understand, no offense is intended by asking those questions. I'm just trying to understand what level of information you are trying to find out.
 
Last edited:
I am trying to find out which supply I linked has better components. I am aware that the first one has the transformer and that these are only 15V supplies. I would just like advice which one to choose.

For 5v I intended to use s://www.audiophonics.fr/en/linear-regulated-psu/regulated-linear-power-supply-board-lm317-mj15025g-dc-33v-to-33v-5a-p-12220.html with r-core transformer.

The 5v supply is ok right?
 
Yes, the 5v regulator is fine, but it will need a transformer. That's why I am not clear on why you aren't looking for a transformer with two 15v or 18v windings, and one 5v winding. That way you only have to buy one transformer. There is no advantage to buying more transformers based on what I know so far.

As far as which one of the transformers is better that you asked about earlier, it probably doesn't matter. If anything you might want a transformer with a shield between primary and secondary to help reduce line noise capacitive coupling to the secondary.

If you read my last post in the 'ES9038Q2M board' thread, it turns out that using very modest linear power supplies and some film caps to clean up the 15v rails helps a whole lot. I wasn't that sure when I made the post, but I have listened some more, and I am very impressed with the power supply solution I used in terms of the sound quality I am now getting from Katana. You may not even need to spend as much money as you may be thinking.
 
Last edited:
Yes, the 5v regulator is fine, but it will need a transformer. That's why I am not clear on why you aren't looking for a transformer with two 15v or 18v windings, and one 5v winding. That way you only have to buy one transformer. There is no advantage to buying more transformers based on what I know so far.


I wanted to use more 5v supplies so I would need two windinds for 15V and 3 windings for 5v. Truth is, I am not sure about the specs of transformer. When the input to supply is 9/24AC do I need to have transformer with specific secondary circuit to be able to get 5v (or the 15 in dual rail)? Somewhere I read the transformer should have at least 1.5x voltage I need, is that right?


But you were using 4 LPS right? I wanted to have all supplies in one case but wasn't able to find such product that is why I wanted to get the power supply kits and maybe integrate the dac into a case also in the process. If I do so could there be a problem with EMC?
 
Last edited:
I am just asking what AC voltage should I have on secondary of transformer in relation to DC circuits voltages (5 and 15V), thank you very much.

AC voltages for power transformers are usually measured in RMS volts. Depending on particulars, an unregulated power supply might put out somewhere between the RMS voltage and the peak voltage, which is 1.414 times higher.

An unregulated supply is what we would use to feed a into regulator which would lower the voltage somewhat more in order to regulate it. In that sense there is some minimum drop across the regulator needed for it to work. For LDO regulators, that minimum drop is low.

At the other extreme, regulators have input voltage upper limits that may be given in data sheets for components used in making the regulator. Regulators also have power dissipation limits: If the input voltage to a regulator is 20v and the output is 5v, then the drop across the regulator is 15v. If the current going to the load is 1A, then 1A times 15v equals 15 watts being dissipated in the regulator. So, you would have to check to see if your regulator could safely dissipate 15W (probably need a heat sink for that).

Hopefully, you are starting to get the picture that designing your own linear power supplies isn't exactly cut and dried, rather it may require some iterative analysis to see if it will work as intended.

If you want a simple rule of thumb, you could probably get by with using transformers with the same RMS voltage rating as the DC output voltage you will need, or the next higher available transformer output voltage. However, it will always be up to you if you are designing your own power supplies at all, to make sure the parts you use will be able to work together okay.
 
AC voltages for power transformers are usually measured in RMS volts. Depending on particulars, an unregulated power supply might put out somewhere between the RMS voltage and the peak voltage, which is 1.414 times higher.

An unregulated supply is what we would use to feed a into regulator which would lower the voltage somewhat more in order to regulate it. In that sense there is some minimum drop across the regulator needed for it to work. For LDO regulators, that minimum drop is low.

At the other extreme, regulators have input voltage upper limits that may be given in data sheets for components used in making the regulator. Regulators also have power dissipation limits: If the input voltage to a regulator is 20v and the output is 5v, then the drop across the regulator is 15v. If the current going to the load is 1A, then 1A times 15v equals 15 watts being dissipated in the regulator. So, you would have to check to see if your regulator could safely dissipate 15W (probably need a heat sink for that).

Hopefully, you are starting to get the picture that designing your own linear power supplies isn't exactly cut and dried, rather it may require some iterative analysis to see if it will work as intended.

If you want a simple rule of thumb, you could probably get by with using transformers with the same RMS voltage rating as the DC output voltage you will need, or the next higher available transformer output voltage. However, it will always be up to you if you are designing your own power supplies at all, to make sure the parts you use will be able to work together okay.


Thank you very much. So if I have 9V AC on secondary (I will have aprox. 6.3V RMS) and output of regulator will be 5V DC 2A so the dissipation in circuit will be 1.3x2 = 2.6W?
 
Thank you very much. So if I have 9V AC on secondary (I will have aprox. 6.3V RMS)

How are you determining that you have 9v AC on the secondary? Voltmeter? RMS voltmeter? Oscilloscope? Manufacturer's rating?

The reason I ask is that low cost voltmeters often measure average voltage then assume the waveform is a sine wave, and display what the RMS voltage would be in that case. Better voltmeters and transformer manufacturer's would be expected to give RMS voltage numbers. An oscilloscope is often used to see peak-to-peak volts.
 
Normally, something called a 9v transformer secondary means 9v RMS, which would be about 12.7v peak or 25.4v peak-to-peak. If you use a full-wave bridge silicon rectifier, that will subtract 1.4v from the peak, so the unregulated output voltage of the rectifier would be 11.3v peak. That is the voltage your unregulated filter capacitor will charge up to each half-cycle. If the cap is big enough and the current draw from the load is small enough, we can approximate that between charging pulses from the rectifier the filter cap voltage might droop in a straight line according to I = C(dv/dt). If the AC line frequency is 60Hz, then the period is 16mS. Since there will be two charging pulses per cycle, the time between them will be 8mS. Knowing dt between pulses and the load I, we can estimate the voltage drop on the unregulated supply before the next pulse. This gives us an idea of the minimum input voltage the regulator will see. The regulator drop out voltage can be subtracted from the minimum expected input voltage to see if the regulator will be able to regulate between charging pulses. If not, maybe the input filter cap can be increased.

Now we are at a point where we can estimate the regulator dissipation probably using peak input voltage and expected load current to see if the power is within the data sheet ratings of the regulator and if it will need any heat sinking.

Whew!
 
Last edited:
Thank you Mark I will try to find out if the voltage on secondary is already in rms.

If the values are already in rms then I will need 6v - min reg. possible input for 5v on output



Also I was wondering when the regulator has output e.g. from 3 to 30 DC, how do I set the proper value of voltage I need.
 
Last edited:
...how do I set the proper value of voltage I need.

That information will be in the regulator data sheet. Google the part number for the regulator and probably the first pdf that it finds will be the data sheet. Usually, after all the specifications and graphical data, there will be an 'applications' section of the data sheet. Read it carefully and it should have what you need there.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.