DIY Purifi Amp builds

[boXem]

It was perceived as a crass, but it's late here and it's been a crappy day, so pretty bad mood on my side... Sorry
I also rely on the specs for design, but I also like to understand what's behind them. In this particular case I have a pretty good understanding.

 

[boXem]

After some sleep, I realize that I needed it...
To clarify, both NC500 and 1ET400A are internally single ended designs, but need to be connected to the speaker in a balanced way.

From 1ET400A datasheet:
1ET400A has a single-ended ground-centered speaker output.
The system integrator might notice that the OUT- terminal is connected to GND as part of the Design and be
tempted to route the negative speaker terminal to GND elsewhere in the system. This, however, should be
avoided as the internal feedback connections sense the voltage between OUT+ and OUT- terminals. If the
negative speaker terminal is connected elsewhere, The Design cannot correct for voltage differentials between
OUT- and the negative speaker terminal. Best performance is achieved by treating the speaker outputs as a
balanced pair when routing the PCB traces from the boundaries of 1ET400A Design to the location of the
speaker terminals at the system level.

Also from the 1ET400A datasheet:
1ET400A monitors all supplies with exception of the optional VD supply and prevents operation unless all supply
voltages are within preset safe thresholds.

 

[tomchr]

I agree with your 200% that the Purifi 1ET400A has OUTC/OUT- connected to ground. It even says so in the data sheet.

Where we disagree is on the Hypex NC500. The NC500 data sheet says, "Do not treat OUTC as a ground terminal". You can see the screen shot from the data sheet in Post #396.

The single output inductor would indeed indicate that both designs are single-ended (I suspect they are very, very similar), but when the manufacturer (Hypex in this case) specifically states, "do not treat OUTC as a ground terminal" then I will not connect OUTC to ground. Hence, my recommendation to use the Guardian-686 with the Hypex NC500 and the Guardian-86 (or half a Guardian-686) with the Purifi 1ET400A.

My plan is to make an input buffer board for the DIY market that supports both the Hypex NC500 and Purifi 1ET400A.

Tom

 

[Frankie]

The Guardian-686 is for use with amplifiers that have balanced/bridged/BTL outputs, such as the Hypex/Purifi amps.



Tom

A strange phenomenon can occur with Class D audio amplifiers that have single-ended output loads: the power-supply voltage increases as power is delivered to the load. The phenomenon is known as power-supply "pumping," and an engineer unfamiliar with Class D amplifiers may find it very puzzling. If the power-supply pumping is not remedied, the voltage increase in the power supply can damage the system.

https://www.maximintegrated.com/en/design/technical-documents/tutorials/4/4260.html

 

[tomchr]

Did you take the time to read my post just above yours? Post #403, that is? Or Post #396?

Speaker protection circuits have no impact on supply pumping, so I'm really not sure what your point was there.

Tom

 

[Frankie]

Did you take the time to read my post just above yours? Post #403, that is? Or Post #396?

Speaker protection circuits have no impact on supply pumping, so I'm really not sure what your point was there.

Tom

When you said:

"amplifiers that have balanced/bridged/BTL outputs, such as the Hypex/Purifi amps."

It appeared as though you were not aware the Purifi amplifier was single ended.

 

[tomchr]

When you said:
"amplifiers that have balanced/bridged/BTL outputs, such as the Hypex/Purifi amps."
It appeared as though you were not aware the Purifi amplifier was single ended.

I wasn't. Then I became aware (thanks Fred). As indicated in Post #396 and Post #403 I am now fully aware that the Purifi 1ET400A is single-ended, and the Hypex NC500 is not (or at least should not be considered as such based on the information provided by Hypex and quoted by me in Post #396).

It seems we're beating the stain where the dead horse used to be. Can we move on now?

Tom

 

[Frankie]

I wasn't. Then I became aware (thanks Fred). As indicated in Post #396 and Post #403 I am now fully aware that the Purifi 1ET400A is single-ended, and the Hypex NC500 is not (or at least should not be considered as such based on the information provided by Hypex and quoted by me in Post #396).

It seems we're beating the stain where the dead horse used to be. Can we move on now?

Tom

Okay it's good to know your mistake has been acknowledged. However I'm still confused on why one would want to put this extra circuity in the signal path after their amplifiers? All I can see happening is degrading the performance, as well as degrading the effectiveness of the feedback system.

A better alternative for anyone worried about blowing their speakers would be to limit the gain the amps receive. This can be done without any additional parts in the signal path.

 

[tomchr]

Okay it's good to know your mistake has been acknowledged.

Good lord! I acknowledged that 24 hours ago. Move on, please.

However I'm still confused on why one would want to put this extra circuity in the signal path after their amplifiers? All I can see happening is degrading the performance, as well as degrading the effectiveness of the feedback system.

The extent of the degradation depends on the implementation of the speaker protection circuit. Some relays will degrade performance. On top, they also won't provide good protection against DC on the output as they tend to arc over. You can see some spectacular videos of that here: https://www.halfgaar.net/dc-protection-with-relays That's why I use MOSFETs to do the job. The challenge there is to provide the gate drive. It's a solvable problem... As you can see from attached graph, the performance degradation is very small - well below audible. It wouldn't surprise me if the spade connections between typical speaker cables and amp/speaker binding posts causes more degradation than what I measure from the Guardian-686. The amplifier I used for this test was my Modulus-686.

A better alternative for anyone worried about blowing their speakers would be to limit the gain the amps receive. This can be done without any additional parts in the signal path.

Correct. But limiting the gain does not protect your speakers in the event of catastrophic amplifier failure. If one of the output devices shorts out, you can end up with the full rail voltage on the output of the amp. Some prefer to protect their speakers against that.
The Hypex/Purifi do have a "DC on the output" flag that can be used to shut down the power supply (if the supply supports it). Those who use a non-Hypex/Purifi power supply will need to either implement a shutdown function in their power supply or provide the protection by other means if they want speaker protection.

Tom

 

[Frankie]

Do you have comprehensive measurements of IMD, THD+N, linearity, etc with this board used in conjunction with the Purifi amp modules? Once you get your build done using this board as protection, please send the amp into Amir to compare directly against the Purifi demo set. Seeing we went from thinking the amps were BTL, to acknowledging they are indeed single ended within a couple posts. Your followup rhetoric is not confidence inspiring for me. When using a direct competitors thread as a sales platform for your wares, please at least include full comprehensive data to back up your claims.

In the event of an catastrophic amplifier failure, the FATAL pin of the Purifi module kills the supply install in a proper build. The system is already properly engineered without any add on parts.

 

[tomchr]

I'm not building anything. I'm just making my modules available to those who wish to add speaker protection to their amps. Nobody forces them to buy them or use them. As you point out, I suspect Hypex/Purifi put some thought into the shutdown functions in their products. Some prefer other solutions, hence, may need to add speaker protection.

Nobody forces anybody to buy from you either, Mike Davis of Mivera Audio.

Tom

 

[Frankie]

I'm not building anything. I'm just making my modules available to those who wish to add speaker protection to their amps. Nobody forces them to buy them or use them. As you point out, I suspect Hypex/Purifi put some thought into the shutdown functions in their products. Some prefer other solutions, hence, may need to add speaker protection.

Nobody forces anybody to buy from you either, Mike Davis of Mivera Audio.

Tom

The reason people come to ASR, is to seek data. And all you have provided here is misinformation, and rhetoric with no data backing it. Also plugging your own wares, on a direct competitors thread. There's several examples earlier in this thread when you tried to argue your amps are better than these for high efficiency speakers. Well the data certainly doesn't tell us this. The Purifi amps run rings around your amps objectively. The cost is less, the efficiency much higher, and the build process much easier.

Not everyone who challenges your misleading data is Mike Davis. He has 250 members on his forum who are also following all of the Purifi based threads on the internet.

 

[Poseidons Voice]

Although you have to pay $1000 to join Mikey’s exclusive club!

 

[tomchr]

Although you have to pay $1000 to join Mikey’s exclusive club!

Or to even see product information. Oooohhh....!!! So exclusive!

Tom

 

[Frankie]

He must offer some pretty good value to attract 250 members at that rate!

 

[Absolute]

Wow, what a beautiful drama I have missed!

 

[Frankie]

There's a new breed of charlatan infesting the internet with misinformation. What they do is use confusing numbers and data to sell laymen of their wares, under the guise of objective, scientific proof. When in reality the information is too incomplete to tell us anything of value. This new breed of charlatan is far worse than the subjectivist charlatan, because at least the subjectivist isn't being ripped off if they like what they hear.

When these types of charlatans can just claim "you're not a trained listener" if you don't like their products. Because their incomplete, misleading data is supposedly proof of its performance. Mike is very aware of this type of charlatan, and is arming his members with proper eduction to see through both subjectivist and objectivist charlatans. Not only does he have an arsenal of high end audio analyzers to verify objective performance, he also conducts proper blind listening tests with local audiophiles, and takes feedback from his clients to improve upon his data.

 

[amirm]

Member Mike Mivera, aka Frankie is banned. Members should be on notice that he makes a new login every couple of weeks to complain as he always does.

 

[goryu]

Any way you can delete all of his posts?

 

[tomchr]

Should you choose to delete his posts, feel free to clean up mine a well so they don't appear out of context.

Tom