Strikes me that in DIY terms Purifi is a bit of a dead end. No idea what the module costs but by the time someone else has put the standard module, presumably unmodified and unmodifiable, in a shiny box with a Hypex supply, because nothing else approved, regulatory, is available with a boutique input buffer for good measure then you're at £1,000 per channel or £5,000 per channel if you believe in yourself. There's really not much else to discuss. It's not DIY.
Huh?
I am not sure what you mean by this? are you saying one might just be better off to purchase a Purifi based amp complete, as the cost is so competitive with building it oneself? If so I understand the point, but what if I feel I can build a better amp than Nord, et al, and do it for less, with a layout in chassis which I prefer? I mean, little details, like where the binding posts are, really matter to me. I have not been very impressed by the internal layout of most of the builds i have seen from the amp makers out there...
I am going to try a first go with the Connex 800 W supply, with a better regulator (discrete) for the VDR supply, and a custom built +- 18 VDC supply for the analog stage supply. If that works out well, and sounds good exceptional with no problems, then I'll sort out the case, the internal layout and all. There is still a fair amount of DIY here. Heck, just figuring out a good way to do the layout and internal wiring in a chassis seems like it is a big challenge to commercial amp makers...
I am not sure what you mean by this? are you saying one might just be better off to purchase a Purifi based amp complete, as the cost is so competitive with building it oneself? If so I understand the point, but what if I feel I can build a better amp than Nord, et al, and do it for less, with a layout in chassis which I prefer? I mean, little details, like where the binding posts are, really matter to me. I have not been very impressed by the internal layout of most of the builds i have seen from the amp makers out there...
I am going to try a first go with the Connex 800 W supply, with a better regulator (discrete) for the VDR supply, and a custom built +- 18 VDC supply for the analog stage supply. If that works out well, and sounds good exceptional with no problems, then I'll sort out the case, the internal layout and all. There is still a fair amount of DIY here. Heck, just figuring out a good way to do the layout and internal wiring in a chassis seems like it is a big challenge to commercial amp makers...
Mike just answered an email I sent him regarding the quality of the stock Purifi input board. Here's what he said:
"The board already has exotic Bruno Putzeys designed discrete regulators that power the onboard TI OPA1612 opamps, as well as the amplifier modulator boards. These discrete regulators already exceed the PSRR and self noise of the best Belleson offers. But if that's not good enough there's provisions on the board already to add additional TO-220 style regulators as well. Why these DIY guys would use external garbage IC opamp based buffers, then power them with much worse regulation than they already have, is delusional. Even if they could equal the performance with external regulation and IC opamps, the extra connections will become a huge bottleneck. The onboard OPA1612's have a direct low impedance connection to the amp module via PCB traces. The only way to best that board is to do things right and make a new one from scratch. But even then not a trivial matter. It took me 13 prototypes over 5 years to get to where I'm at with the Puribuffers."
Hopefully that helps some people. Anyways over and out!
"The board already has exotic Bruno Putzeys designed discrete regulators that power the onboard TI OPA1612 opamps, as well as the amplifier modulator boards. These discrete regulators already exceed the PSRR and self noise of the best Belleson offers. But if that's not good enough there's provisions on the board already to add additional TO-220 style regulators as well. Why these DIY guys would use external garbage IC opamp based buffers, then power them with much worse regulation than they already have, is delusional. Even if they could equal the performance with external regulation and IC opamps, the extra connections will become a huge bottleneck. The onboard OPA1612's have a direct low impedance connection to the amp module via PCB traces. The only way to best that board is to do things right and make a new one from scratch. But even then not a trivial matter. It took me 13 prototypes over 5 years to get to where I'm at with the Puribuffers."
Hopefully that helps some people. Anyways over and out!
Last edited:
You've said that three times today in this thread, and at least twice in another. And yet you continue to post. Make up your mind.
Call it ROI. No doubt you will be in a position to do an A/B comparison. No, well at least the layout is worth your time. Honest feeling.
The VDR supply is for the Mosfet driver. Bruno says he has relaxed the dead time on those to make the oooh it gets hot people happy. Bruno says he can do this because of the high available loop bandwidth.
Outside of dead time we are probably still talking tens of nano-seconds transition times. Do you think your better regulator (discrete) located off board is going to make a gnats whisker worth of a difference?
Don't forget that the upper Mosfet is bootstrapped from VDR and flaps up and down at 500KHz, variable. Do you think your better regulator (discrete) is going to make a difference to the drive for the upper Mosfet?
So there we have it. Bruno got it wrong but Mike can sort it out by doing things right.
Cost of the Purifi modules, DIY? About $300 for a single amp board, or $700 for a pair of them with an interface board. Add an SMPS1200A400 ($200) and a nice case (e.g., Ghent for about $200) and you're at about $1,100 for a stereo rig. $1,200 with shipping and tax.
Nutube vs Original B1
I am under the impression that the Purifi needs some gain at the front end.
I am not familiar with the Nutube variant but know from experience the original 2sk370 B1 buffer is amazingly neutral but does not have gain.
Why not go that route? Will a B1 with solid power supply drive the Purifi boards effectively?
Currently assembling parts to find out for myself but until then...
Thanks,
DB in MN
0°c mid-morning January
I am under the impression that the Purifi needs some gain at the front end.
I am not familiar with the Nutube variant but know from experience the original 2sk370 B1 buffer is amazingly neutral but does not have gain.
Why not go that route? Will a B1 with solid power supply drive the Purifi boards effectively?
Currently assembling parts to find out for myself but until then...
Thanks,
DB in MN
0°c mid-morning January
Not necessarily. It depends what you have driving it. I don't think I'd add just a unity gain buffer, but these choices are all about individual system needs (or just wants).
https://www.passdiy.com/pdf/B1%20Buffer%20Preamp.pdf
https://purifi-audio.com/wp-content/uploads/2020/01/1ET400A-Data-Sheet-1.00.pdf
Input impedance is 4K4 differential, 2K2 single ended. Gain is 12.8dB, X4.37. 9.6V differential required for full output at 65V rails.
The limit appears to be the common mode input range of 5V which suggests that with your buffer being single ended you are going to be limited to an equivalent 5V single ended input which will not drive the thing to full output.
You might consider using a pair of such buffers driven differentially to provide a differential drive to the amplifier....
Just making things up.
https://purifi-audio.com/wp-content/uploads/2020/01/1ET400A-Data-Sheet-1.00.pdf
Input impedance is 4K4 differential, 2K2 single ended. Gain is 12.8dB, X4.37. 9.6V differential required for full output at 65V rails.
The limit appears to be the common mode input range of 5V which suggests that with your buffer being single ended you are going to be limited to an equivalent 5V single ended input which will not drive the thing to full output.
You might consider using a pair of such buffers driven differentially to provide a differential drive to the amplifier....
Just making things up.
Since the Purifi input boards are only $100 a piece. I’m not sure why anyone wanting to do monoblock’s couldn’t just use 1 per module. And leave the other channel not connected? This way you get the ultra low noise Bruno Putzeys discrete regulation, and OPA1612 opamp which sounds much better than any of the old National crap. If you think the best possible buffer would use IC opamps, you’re not going to do better than Purifi has already done. Unless anyone doubts the ability of Bruno and the Purifi team. In that case why trust the amp modules themselves?
If that's a traditional differential input around a differential input, single ended output amp then the resistor values don't give the input impedance for the inverting input when driven balanced. Assuming that it is the in-amp config with 4*2k2s then its 4k4 for the +ve input but only 1.47k for the -ve input.
One of the saddest trends we are seeing in audio today is as long as the analyzer says some of the specs are great, you can use that data to sell people on products. Since 98% of the science minded audiophiles only have a vague understanding of what the specs mean, it’s very easy to manipulate them into buying a product based on the premise it’s built to high objective standards. They do so by cherry picking only the objective highlights of the design. THD+N is a huge hit among the self proclaimed objectivists. High IMD is perfectly fine, as long as they don’t know about it.
But from a brighter angle. If all you need to do is build electronics that can wow people with cherry picked specs, 15 Euro an hour can buy you highly competent engineers out of Asia who can do this for you easily. Because the ability to listen is no longer required. These types of engineers don’t even need ears to make products that are perfect for their target market.
One of the things I admire about the Purifi team, is they measure then listen. Measure then listen. And they repeat this cycle until they reach perfection. If this same philosophy is applied throughout your system from source to speaker, the chances of enjoying the end result is much higher, than the 15 Euro an hour from Asia, or 120 Euro an hour engineers from the west for the “analyzer only” service.
But from a brighter angle. If all you need to do is build electronics that can wow people with cherry picked specs, 15 Euro an hour can buy you highly competent engineers out of Asia who can do this for you easily. Because the ability to listen is no longer required. These types of engineers don’t even need ears to make products that are perfect for their target market.
One of the things I admire about the Purifi team, is they measure then listen. Measure then listen. And they repeat this cycle until they reach perfection. If this same philosophy is applied throughout your system from source to speaker, the chances of enjoying the end result is much higher, than the 15 Euro an hour from Asia, or 120 Euro an hour engineers from the west for the “analyzer only” service.
Last edited:
And now with a linear power supply for the EVAL1 circuitry modulator, using a "lost and found" transformer and the DIYinHK 0.56uV Ultralow noise DAC power supply regulator at 15V.
I am going to use a +-18V instead when it arrives in the post, and then I will use the DIYinHK super reg +-15V to power the Neurochrome Buffer. Later, when I replace the FE02 connection board with Tom Christensen's adapter boards, I will use a second DIYinHK superreg at +-12V in place of the +-18V to drive the modulators, and then I will think at what to do for the gate driver.
Having fun!
I am going to use a +-18V instead when it arrives in the post, and then I will use the DIYinHK super reg +-15V to power the Neurochrome Buffer. Later, when I replace the FE02 connection board with Tom Christensen's adapter boards, I will use a second DIYinHK superreg at +-12V in place of the +-18V to drive the modulators, and then I will think at what to do for the gate driver.
Having fun!
Last edited: