Jump to content
IGNORED

Conversion of a Chord DAVE to DC power using two Uptone Audio JS-2s


Recommended Posts

8 hours ago, asdf1000 said:

Btw: I don't write all this just to argue.

 

I just think people in the past have taken some of what each has said and tried to make a generalisation. 

 

To write it a different way: has Rob Watts measured an LPS-1.2 or JS-2 ? Has John S measured Chord DAVE's power supply? I haven't read anything anywhere that says yes to either question.

 

I'm happy to be proven wrong though. Maybe John S will chime in.

 

Correct, I have never measured a Chord power supply.

 

There are two main aspects of power supply that I find important for digital audio:

1) leakage current

2) output impedance over frequency

 

#1: I have written MANY posts on this, in a nutshell there are two types of leakage current, high impedance, which only comes from SMPS and low impedance leakage which is found in both SMPS and LPS. Every SMPS I have ever tested has significant amounts of high impedance leakage. Fortunately grounding the output of the SMPS gets rid of almost all of this. Once you do the grounding SMPS and LPS are on a similar footing. Good SMPS have leas leakage then poor LPS and good LPS have less leakage than poor SMPS. So far with what I have tested the best LPSs have lower leakage than the best SMPS.

 

#2 output impedance is very important to digital. This is NOT a single number, it is a graph of output impedance VS frequency. MOST PS of either type have fairly high impedance at both low and high frequencies. Most SMPS to not have separate regulators AFTER the switching converter, and thus rely on how good the regulation is of the whole converter. Usually this is quite poor. To improve this you can make the switching converter system as a whole have a much higher bandwidth, which CAN be done but costs a lot of money, OR use a linear regulator after the Switching converter. Both approaches can radically improve the output impedance at the extremes, but increase cost and space.

 

Note that no test equipment manufacturer actually makes a power supply tester that measure output impedance VS frequency. If you want to measure this you have to design and build your own. Doing a good job of this is NOT easy or cheap. Thus VERY few power supplies have actually been tested for this. I have built several over the years and have tested some of both SMPS and LPS and the results are all over the map.

 

John S.

Link to comment
7 minutes ago, JohnSwenson said:

The important parameters there are PSRR (Power Supply Rejection Ratio), which is again a graph over frequency. High numbers mean what is on the input gets attenuated by that amount. But this can vary radically with frequency. The next is output impedance over frequency, this relates to how fast the supply output voltage can track changing load currents. A low impedance at high frequencies means it can keep the voltage steady even when the load changes are very quickly which happens a LOT with digital circuitry. The other end is also important. If the load triples its current draw and keeps it there for a long time can the supply keep it there without drooping, then going back up. Both are important for digital audio.

 

Then there is output noise, which gets a lot of attention since it is usually specified as one number (but again it should be a graph over frequency) is actually the least important parameter. Modern designs almost always have local regulators driving the circuitry and the noise of THOSE is usually the determining factor. For an external AC mains supply the first two are usually of primary importance. Of course practically nobody actually specifies these and they usually DO specify the output noise so that is what most people focus on.

 

John S.

I forgot leakage current. The leakage current is also important. Leakage current can cause low level voltage variations across the ground plane which can increase jitter. (see the ER white paper for details on how this works). Whether this is an issue is very system dependent since there needs to be a connections that forms the leakage loop and that depends on the system connections.

 

John S.

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...