I do not think it is to save cost, indeed, and I know about the debate about regulated vs unregulated power supply in linear power supplies for power ampifiers. But I am not 100% sure how this argument applies to SMPS since it cannot be the excess heat dissipation due to regulation, for instance. Reduction of dynamics perhaps?
Regulation is even less critical with class D amps, especially these ones. Because the PSRR is so high. All regulation does is drastically worsens the dynamic load response, along with gobble valuable current to be dissipated as heat.
Not always, because I know power amps with regulated power supplies that did not lack in dynamics. Also, because of the way SMPSs work, the wasted heat is much much less than with a linear regulation, but I hear your argument.
Several claims by who? So several people have compared the 2 modules in isolation, powered by the same supplies, without an input buffer in the signal path? This amp module is so transparent that when used with a colored buffer, all you're going to hear is the buffer. Which would explain their results. If you want to test the amp module itself, do so without a buffer, and a source with around 10 ohm output impedance max. Along with an output voltage of around 10V. With any other method you're not testing the module itself in isolation.
It only takes about 2 minutes of comparing the best Connex has to offer with the SMPS1200 to discover that Hypex did a better job. I've only compared the main rails because the other rails are junk on both supplies anyways.
Uneducated people, perhaps the time to modernize the word "EDUCATION" should become a major priority in our modern societies.
If regulating the main rails was a superior way to do things, Hypex would have done so. These guys understand how class D amps work. The reason these fly by night Chinese junk brands do it is to fool unknowledgeable audiophiles into thinking their supplies are superior. Well 1 look at the quality of caps Connex decided to use speaks volumes about their standards of quality. Spending $2 more on the BOM they could have used decent caps. But that extra $2 probably would have doubled their BOM. So it was a no go.
1 more thing is the data sheet specs of this module are based on being powered by the SMPS1200A400. So if any other supply should be considered as an alternative, the DIY community will need to see measured data of the module powered by the alternative supply. And the specs from the analog outputs of the module should at minimum match the specs with the SMPS1200A400. But still even when Purifi comes out with their supply, even if superior, it will be missing 1 aspect that only the SMPS1200 can claim. 10000+ in the field over 10 years, with 7 revisions. The only way this accomplishment can be achieved is with time.
Perhaps it's as simple as an SMPS, a Class D amplifier is an SMPS, represents a negative impedance to its supply. Even if the supply is passive you can run into all sorts of problems if its impedance in conjunction with that load results in gain, an 'amplifier'.
If the supply is an SMPS without regulation the problem still exists. If the supply is an SMPS with regulation then you really need to be looking at such issues. For the Purifi amplifiers with such high loop gain it is extremely likely that they are going to exacerbate the problem.
This in part is the 'dangle more output capacitance on my regulated SMPS' question written harder. Sure more capacitance is going to give you blah bass and burble transient with blibble mid transparency and lucid highs but it is also likely trash the loop stability in the supply itself assuming it had decent bandwidth to begin with.
Same applies to linear regulators. The datasheets don't suggest particular types of input and output capacitors, more often than not placing limits on the range of ESR, because the designer has shares in the company that supplies those particular capacitors. It's all down to loop stability.
In respect of the 1ET400A, picture, VP 1/2 VN 9/10 GND 3-8. I kind of wonder about the current carrying capability of those pins but no doubt the sums work out.
Of more interest is the pair of electrolytics close to the connector. The bank of electrolytics to the rear of the the board and, more importantly, the pair of rod inductors also to the rear of the board.
I would hazard a guess that this is a deliberate design decision to implement a CLC section that isolates the connected supply from the main amplifier and the likely horrible negative impedance it might present to that supply.
Forward numbers are possibly meaningless but 1uH and 2200uF would have a corner frequency of just above 3KHz and would be critically damped with an ESR of 21mR.
In many respects the inclusion of that filter should cause pause for thought when arguing about the nature of the supply itself. At least it seems that way from my armchair.
...
If the supply is an SMPS without regulation the problem still exists. If the supply is an SMPS with regulation then you really need to be looking at such issues. For the Purifi amplifiers with such high loop gain it is extremely likely that they are going to exacerbate the problem.
This in part is the 'dangle more output capacitance on my regulated SMPS' question written harder. Sure more capacitance is going to give you blah bass and burble transient with blibble mid transparency and lucid highs but it is also likely trash the loop stability in the supply itself assuming it had decent bandwidth to begin with.
Same applies to linear regulators. The datasheets don't suggest particular types of input and output capacitors, more often than not placing limits on the range of ESR, because the designer has shares in the company that supplies those particular capacitors. It's all down to loop stability.
In respect of the 1ET400A, picture, VP 1/2 VN 9/10 GND 3-8. I kind of wonder about the current carrying capability of those pins but no doubt the sums work out.
Of more interest is the pair of electrolytics close to the connector. The bank of electrolytics to the rear of the the board and, more importantly, the pair of rod inductors also to the rear of the board.
I would hazard a guess that this is a deliberate design decision to implement a CLC section that isolates the connected supply from the main amplifier and the likely horrible negative impedance it might present to that supply.
Forward numbers are possibly meaningless but 1uH and 2200uF would have a corner frequency of just above 3KHz and would be critically damped with an ESR of 21mR.
In many respects the inclusion of that filter should cause pause for thought when arguing about the nature of the supply itself. At least it seems that way from my armchair.
...
Attachments
Last edited:
I would like to mention that the Connex was still “approved” by Kim and Claus at Purifi. Furthermore, a “two minute” test to verify that a unregulated supply has more dynamics than a regulated one sounds to me like confirmation bias (just as mine could be). Whatever. Then there is the deleted post “I could sell [my SMPS1200A400] for a big discount” that, when inquired privately, resulted in a “just get it from the hypex site” - and the aforementioned deletion.
It is quite sad that a frustrated person is trolling. Maybe even with a commercial interest. I prefer my bar brawls when everybody (including myself) show their faces.
It is quite sad that a frustrated person is trolling. Maybe even with a commercial interest. I prefer my bar brawls when everybody (including myself) show their faces.
Please, Maty. Mike from Mivera is a skilled salesman, an assembler with no engineering background, and definitely not a trusted source. Actually, not even that skilled, since with his aggressive tactics he was forced to set up his own forum because it was the only way he would not be banned. I am not sure for what six different regulators are for. The Purifi board may need three (gate drive, and +VOP, -VOP) which by the way I may provide soon with 1uV noise regulators, from our friend Aleksandar. Add two for his balanced buffer, and what's the other one for? Control logic? Why should it have that level of filtering?
So if you believe what he says when he disses the hypex so much, you need to stop for a moment, take a deep breath, erase all related information from your brain and start over. Please, do it for yourself. Do not trust a salesman.
Last edited:
of course how could I have missed it. It is not that inconspicuousYou sure know a lot about that Hypex supply, and before writing the next sentence, stay assured that I agree with you - it takes time to develop a good product.
But then, there, there, go to the Hypex HR department, explain your HR contact you are burned out and take some medical leave. They will understand.
Yes the amp will work with that supply. But some people want more than mere functionality.
What are you talking about? I didn’t delete any post?? I was talking about selling the crappy Connex supply dirt cheap and the post is still there. You asked me via PM you’re interested in the Hypex supply, and I told you to get it from Hypex. I’m not a Hypex supply dealer.
Last edited: