DM

Power Supply (8+) Group Test, LPS and SMPS

Rate this topic

18 posts in this topic

Original from here:

 

Part 1

Part 2

 

I asked the author for the permission to translate into English, so here we are. Beware that most of the translation was done by google, I didn’t fix the grammar (which I wasn’t that good neither), I just amend it so some totally meaningless phases now become readable.

 

---- Translation Begin ----

 

Is linear power supply (LPS) better than switching power supply(SMPS) ? The 8 models PSU group test.

 

 

Preface:

I have been an audiophile for several years, in recent years found that the role of power in the device is very important, good power will take the equipment to another level, the background is darker, higher contrast; like petrol in cars, regular unleaded vs. super unleaded.

 

Recently there is a friend posted that there is a SMPS power supply quitter than the LPS he used for many years, I directly commented his silliness, his ears are tuning into wood.

 

He didn’t like my comment and challenged me to his house on the next day, also as an occasion for eat and drink.

 

After an afternoon of AB testing, AB and kept drinking mountain tea, my mind has been changed. With the M2TECH EVO DAC, iFi iPower was better than the TeraDak LPS, darkerk background, the details are more clear, less background noise, why is this so? we all say that SMPS can’t be any good?

 

After an afternoon of AB testing, AB and kept drinking mountain tea, my mind has been changed. With the M2TECH EVO DAC, iFi iPower was better than the TeraDak LPS, darkerk background, the details are more clear, less background noise, why is this so? we all say that SMPS can’t be any good?

 

I sighed for a long time in my heart, this is really switching power supply? Picked it up, look closer, although shell is plastic, but the workmanship is good, replaceable plug, also the weight reminded me that this is really a type of SMPS. Well, this SMPS can has this kind of effect … A small idea started to grow in my heart, I wanted to take the iPower to my cousin’s lab with Audio Precision test gear to test how much the performance gap actually is .

 

First some basic knowledge:

(from wiki + some of my own)

 

Power is the heart of audio equipment; its importance is self-evident. Providing good power to the audio equipment is the same as providing a good working environment for the audio equipment. under the same conditions, a good working environment, audio equipment will perform better.

 

In our ordinary use of audio power, can be divided into two categories, namely, linear power supply (LPS) and switching power supply(SMPS). Both have their own strengths and weaknesses. The following is the advantages and disadvantages of these two types of power supply.

 

Its working principle can be compared as follows:

When removing a smaller flow from a larger diameter water pipe, two methods can be used:

 

Linear power supply (LPS):

Install a faucet, and open the valve in a smaller position, so you can get a smaller flow, and this is the working principle of linear power supply. (The valve can be seen as a transistor) linear power supply voltage regulator transistor to withstand a lot of "pressure" (the specific performance is converted to the form of heat dissipation);

 

Advantages: linear power as long as the water pipe in the water pressure is varying too much, the output is very stable, that is, we said the linear power supply has small output ripple;

 

Disadvantages: the valve need to withstand high pressure and this pressure will turn into heat and needs large heatsink, "faucet" role are generally pure copper coil, the cost of expensive materials;

 

Switching power supply(SMPS):

The other is to let the water pipe to a large "bucket", the small water pipe connected to the bucket to take water, need to do is rhythmic on / off the water pipe on the valve to ensure that the water inside the bucket Not completely, and will not overflow because of too much - the basic principle of switching power supply is the case.

 

Advantages: the self-regulation output are not sensitive to external voltage fluctuations, low cost of materials, the size can be very small, high efficiency conversion rate.

 

Disadvantages: frequent switching outputs, poor design will lead to the large output ripple.

 

So, in the sound system, everyone's cognitive LPS better than the SMPS.

 

There is another reason, but also let a lot of people to accept the LPS, is the circuit is simple. For the LPS, for anyone with little electric background can also go to ebay and get some parts back and assemble their own LPS, and the results are not bad. And the maintenance for LPS is also simple, the requirements of the instrument is low, basically a multi-meter is all one needs.

 

But for the switching power supply, it is difficult, the circuit itself is quite complex, the circuit needs a very good foundation to make the finished product, and for the results to be good, it is quite difficult, maintenance is more difficult. Therefore, to do a SMPS will be more difficult than a LPS, if you want to do a good SMPS, will be even harder.

 

But in our lives come into contact with the power supply, switching power supply and accounted for a large part of power supply because the heat is low, you can make a small size products, so the vast majority of standard power supply is SMPS. But in the audio world, because the SMPS’s frequent switching will cause high-frequency interference and power ripple, hence not very popular.

 

Machine tested:

Finally, about time to go to my cousin's lab to test, but also at the expense of a Sunday time, the test process also encountered power outage for most of the day, but also took me a while to get all the power supplies, and finally to a group.

 

I would also like to thank my cousin taught me how to test these, had wanted to write my cousin’s name praise him, he personally asked me not to, stressed that he is only the behind the scenes worker for me ...

 

PSU01.png

 

* Price is not necessarily accurate, is from China Taobao (China’s amazon) search (was in RMB and converted to USD).

* 6 ~ 8 is an additional gift, to contrast.

 

 

To be continued...

1

Share this post


Link to post
Share on other sites

1. Breeze Audio DC-1 15V/1A LPS

 

image.jpg

Unrealistic specification ah, set the 1A load directly, the voltage dropped to 12.4V, error should fall within ± 5% to be considered normal.

 

image.jpg

Into 0.5A, the voltage is considered normal

 

Test ripple performance:

 

image.jpg

 

In the case of load 0.6A, the average ripple is 30.61 mV

 

image.jpg

 

In the case of load 0.3A, the average ripple is 303 uV

 

Audio band comparison

 

image.jpg

 

Comparison of the two load band, 0.6A cases are "flying" up, low power is ok.

According to 0.3A, this audio frequency band, the peak at 50uV, high frequency part of the more spikes, guess the spike are 50Hz AC is not handled well.

 

2. S.M.S.L Panda dual 9V 9V * 2 /?A LPS

 

image.jpg

S.M.S.L has two 9V outputs, no rated current, guess AMP is the output of the larger power, use used this output. The output line is relatively short, can only see the "backside", and finishing photos found no shot was done to the face, the power supply also been returned, only from Taobao find a photo for the front.

 

image.jpg

0.5A load, the voltage is 8.94V is ok.

 

image.jpg

0.7A load, the voltage is 8.7xV is ok.

 

 

Test ripple performance:

 

image.jpg

 

In the case of load 0.5A, the average ripple is 668uV

The results are pretty good to say.

 

image.jpg

 

Load 0.7A case, the average ripple is 235mV, this change is too big, when the load become large, performance degraded 300 times?

 

image.jpg

 

Audio band comparison

 

The white line is 0.7A, the blue line is 0.5A

Comparison of the two loaded band, two high-frequency processing are relatively good, but after the load, the low frequency are more, the average value of about 50uV it.

 

With their own Panda equipment, power consumption is not high, the performance may be ok.

Edited by DM
0

Share this post


Link to post
Share on other sites

3. TeraDak TeraLink X2 8.5V / 0.8A LPS

 

TereDak has quite a following, a lot of friends use it. Now this one in addition to a 8.5V ouput also has a USB output, the function is pretty good.

 

image.jpg

Rated is 0.8A, first at 0.4A, oh, there is a problem with the voltage ...

 

image.jpg

Into the subject of the rated current 0.8A, miserable, the output voltage dropped 3V!

 

image.jpg

Let’s try 0.1A, the voltage is higher than the standard 8.5V ± 5% standard, the power will be broken!!!

 

image.jpg

 

Test ripple performance:

In the case of load 0.186A, the average ripple is 286.1 uV

 

image.jpg

Load 0.36A case, the average ripple is 301.3uV

 

Audio band comparison

 

image.jpg

 

Red is 0.186A, purple is 0.36A, compared to the two loaded band, low frequency noise separation is large, the interference of 50Hz is still strong.

 

If this power supply can fix it’s low end noise issue (or problem with the load), it is not a bad power supply. May be the power supply need to be fixed (damn, it’s out of warranty currently).

 

PS: debugging Shao Shuai this power, try a variety of exclusion line, take the test board, and the city voltage is normal and so on, and finally the above results, speculated that this power is not bad is a serious design flaw.

Friends A reaction that is no problem in the use of power, and later check the EVO DAC power, only 0.34A, Fortunately, this goods to eat less.

0

Share this post


Link to post
Share on other sites

4. ZERO-ZONE SUPER-PSU 12V/6.5A LPS

This compared to the others is significantly larger and heavier, power rating is also the largest one, of course, the price is the most expensive.

 

image.jpg

 

When the load is 1A, the voltage is normal.

 

image.jpg

 

With 3A, the voltage is normal (less than 5% error).

 

image.jpg

 

Well, with the standard 6.5A, 11.426 voltage or in the normal range (± 5% 12.6V ~ 11.4V)

Is already 74w, strong performance and honest marketing.

 

Test ripple performance:

 

image.jpg

 

Load 1.0A, the average ripple is 16.09uV, very good.

 

image.jpg

 

Load 2.5A, the average ripple is 20.39uV, is also very good.

 

Audio band comparison

 

image.jpg

 

Red is 2.5A, yellow is 1.0A

 

Comparison of the two loaded band coincidence is also very high, be quite stable, the average ripple in the 250nV! The expensive and material, reflected the effect, and praise.

0

Share this post


Link to post
Share on other sites

5. iFi iPower 9V 9V/2.0A SMPS

 

This is the beginning of the article introduced a magic switching power supply, when I read the data, my cousin and I had the same idea, we want to open it up to see, is this thing any good, and also has a significant price advantage.

 

 

image.jpg

 

Loaded with 1A, the voltage actually 0.1V higher.

 

image.jpg

 

Load 2A, the voltage is still very stable.

Completely in the specification, product with a conscience.

 

 

Test ripple performance:

 

image.jpg

 

Load 1A, my god, the average ripple is 10.31uV!

This is so low, so low ah.

 

image.jpg

 

Load 2A, the average ripple is 12.22uV! The

This data is unbelievable, full load is also so low ripple.

 

 

Audio band

 

image.jpg

 

Yellow is 2A, red is 1A.

 

Comparison of the two loaded bands, two bands coincide highly, very stable, even the peak of the highest point is not more than 1uV, the overall average of 150nV, measured data lower than iFi official data of 1uV, really low-key marketing.

 

It can achieve this kind of performance with maximum loading, no wonder in the listening test it was much better than the LPS, my friends and I have heard it. Now the mystery has been solved.

 

And later looked at the iFi's official website, may be this is the ANC technology at work, seems a very good “dark-art” technology, indeed should receive some praise.

 

bb5.png

 

Also from iFi’s wesite, found that there are many USB, DC purifier, one day if have the opportunity must get some over and then test it.

 

Saw that micro iUSB3.0 actually has a ripple of 0.1uV, really want to test it in the future.

 

(PS: 0.1uV, this data is the same as the 9V battery!)

 

 

The following are some usually stuff we throw into the mix for fun:

 

image.jpg

 

6. Xiaomi 10040mAh battery bank 5V/2A SMPS + Battery

 

Test ripple performance:

 

image.jpg

 

Load 0.5A, the average ripple is 2.377mV

 

Audio band

 

image.jpg

 

Yellow line is the load 0.5A, white line is the load 1.0A, two lines of the average ripple between 50uV ~ 100uV bar.

There is a gap between the different loads, other battery banks also use SMPS inside, and perhaps this is what we call the SMPS is generally not good.

 

7. Apple 5S charger 5V/1A SMPS

 

Test ripple performance:

 

image.jpg

 

Load 0.5A, the average ripple is 3.459mV, slightly higher noise than the Xiaomi battery bank.

 

Audio band

 

image.jpg

 

Blue line is the load 0.5A, white line is loaded 1.0A, two lines of the average ripple around 20uV it.

 

There is a strange phenomenon here, when the load is heavy, the noise is lower, especially with high frequency noise.

This is better than the Xiaomi battery bank, Apple's quality is still a little higher after all.

 

8. Nanfu 9V alkaline battery 9V/? Battery

 

Test ripple performance:

 

image.jpg

 

Load 0.4A, the average ripple is 15.99uV

 

 

Audio band

 

image.jpg

 

The cheapest is the 9V battery, but the most linear, the fluctuations are particularly small, an average of 100nV, and can be used as a benchmark for sustainable power.

 

The test is finished.

0

Share this post


Link to post
Share on other sites

Test Summary:

 

PSU03.png

 

Finished statistics data, the picture is more clear.

 

Do not know whether it is test instrument issue or a human issue. Most specifications on power supply seems unrealistic. We later added some of the daily used items, battery banks, phone charger the first, battery for additional reference.

 

iFi iPower 9V HiFi SMPS 100% win over HIFI LPS, from the power measured, and the ripple test, with eye dropping new technology, traditional idea on SMPS to be reversed.

 

SMPS vs LPS is no longer the issue, now really depends on the actual product.

 

Before the posting I shared the article with other friends, they pointed out that most LPSs are not done by regular manufacturers, but by a personal studios.

 

As SMPS and LPS compared, SMPS is much more complex to design , not simply stacking with audiophile components. LPS is relatively simple. But iFi in addition to the power also has DAC, amplifier, etc., especially the IDSD DAC, and it’s a British brand and so on. It is not difficult to see that the technical reserves are not comparable to the personal studio ah.

 

Thanks to the cousin to accompany the whole day, from day to night (with power outage in the middle), the next opportunity to see if he dare to show his face, thank you for reading.

 

------- Translation Ended-------

 

As we are not too far apart, a few hours travel, the guy actually said if we have any other power supply would like him to test, he can test them for us, so let me know if you guys want to test something :D

Edited by DM
0

Share this post


Link to post
Share on other sites

My Conclusions

 

I hope the translation was not too difficult to read :)

 

To help the readers here, I have put together a scoring system (our of 100, IMHO) and a few conclusions:

 

1. 95 Nanfu 9V alkaline battery

 

2. 93 iFi iPower 9V SMPS

 

3. 90 ZERO-ZONE SUPER-PSU

 

4. 80 TeraDak TeraLink X2

 

5. 65 Breeze Audio DC-1 / S.M.S.L Panda dual 9V (both flawed)

 

6. 55 Apple charger

 

7. 50 Xiaomi 10040mAh battery bank

 

 

 

 

  • The best is the 9V battery (not surprised).
     
  • The SMPS based iFi iPower is so close to the battery was a surprise.
     
  • Many LPSs didn’t do well at all was another surprise.
     
  • LPS is NOT guaranteed to have good performance, need to pick very carefully too, e.g. Zero Zone will be a good one.
     
  • SMPS is generally not as good as LPS except the iFi iPower, which has better performance than the LPS.

 

:D

1

Share this post


Link to post
Share on other sites

The best thread of the year ! dam man, wish I had the equipment to measure myself. But it is damn interesting to read that the Zero Zone super PSU which is a R-core score that high.

 

ZEROZONE SUPER-PSU 30VA DC9V 2.7A Linear Power supply LPS for Audio Sources | eBay

 

I would love to also get cheap chinese R-core supplies and linear ones with capacitor storage tested too.. they are not expencive.. could be interesting to be honest.

 

Also the ZeroZone doesnt seem to have EMI filter (Berkeley USB to SPDIF/AES has a EMI filter on its AC input). I believe we could get good performance with R-core and a EMI filter on AC input, and then a EMI filter on the DC output of the R-Core with capacitors !

0

Share this post


Link to post
Share on other sites

Interesting, although it should be noted that there is much more to PSU performance than just noise figures: low output impedance over a very wide bandwidth, high slew rate, fast rise-time and fast settling time to control the regulator output voltage during fast varying load current transients.

 

Battery may be low noise champ, but it can't compete with a properly designed PSU in the areas of performance I mentioned.

Edited by Elberoth
1

Share this post


Link to post
Share on other sites
Interesting, although it should be noted that there is much more to PSU performance than just noise figures: low output impedance over a very wide bandwidth, high slew rate, fast rise-time and fast settling time to control the regulator output voltage during fast varying load current transients.

 

Battery may be low noise champ, but it can't compete with a properly designed PSU in the areas of performance I mentioned.

 

That is true, but it all starts with very clean and filtered current. My ADC has bad artifacts due to the 50hz switching issue. So Im looking to make it better. I already changed PSU once for it, now Im doing a normal linear one, and going to add EMI filter at AC side, and then another EMI filter at DC side before the capacitors and voltage regulator.

0

Share this post


Link to post
Share on other sites

I sent him a few power supplies to test, he told me new tests will be up next few weeks , stay tuned

Edited by DM
0

Share this post


Link to post
Share on other sites

better than most reviews on the net by the pros. argee with op most interesting thread here. CA is richer with this knowledge. really deatiled and properly measured. i showed this to my non audio freind (ok he likes listening to audio but doesnt spend crazy sums) and he told me your friend knows what he is measuring, he couldnt really offer more input. fwiw my ee friend has poked holes in some magazine tech measurments and criticized them. the aquavox is cheap enouugh i used one before when i had a squeezbiz.

0

Share this post


Link to post
Share on other sites

I am a little confused here.  Why does the Xiaomi battery pack need an SMPS?  Is the voltage of the batteries a lot higher than 5v and the SMPS brings it down to 5V?  I am using an Inland 15,600 mWh battery pack to power an Audioengine D1 DAC.  (Inland is the house brand of Microcenter.)  My impression is it sounds better than  powering the DAC than using one of the Apple battery chargers I have around here.  Is there an explanation for that, or is it my imagination?  

 

Perhaps this $20 battery pack isn't worth the trouble, although it would be useful for charging my phone if the power goes out around here, and that happens a lot

 

http:// http://www.microcenter.com/product/453791/15,600mAh_Power_Bank_Battery_Charger_-_LED_Flashlight_for_Mobile_Devices_-_Silver.

0

Share this post


Link to post
Share on other sites

I tried powering my uptone regen with a 9 volt alkaline battery and the improvements are substantial comparing powering the uptone regen with it's standard power supply

 

Very thanks for the article and translating it for us

1

Share this post


Link to post
Share on other sites

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