DIY Purifi Amp builds

[JimB]

Just giving us a place to share what we learn about these. Examples of what I hope we will have here are features of the front end board and how best to use them, case options, best practices, etc.. On the FE board: how to wire from the Hypex SMPS1200A400; how to use the standby control (remove the switch and connect to the SMPS?); change the buffer gain; etc.. I'll contribute specifics as I can.

 

[JimB]

From Ghent today, I have:
"Certainly, we are very interested in this project, and we are discussing the best way to develop it inside our cases inside our company.

Also, we contacted Purifi before. Purifi are very interested and recently has started the initial discussion with us about developing cases for their module, so please wait some times."

UPDATE Mar 5, 2020: https://www.audiosciencereview.com/forum/index.php?threads/diy-purifi-amp-builds.10478/post-344268

 

[BobPM]

I am considering the Purifi boards with the Nord Rev. C or D buffer boards so I can use Ghent cases in mono block configuration. Nord appears to use the buffer boards without modification on both its NC500 and Purifi based amps and has them listed on their website. I am not sure what is best for power supplies. What is the difference between using the SMPS600 vs the SMPS1200A400?

 

[JimB]

I am considering the Purifi boards with the Nord Rev. C or D buffer boards so I can use Ghent cases in mono block configuration. Nord appears to use the buffer boards without modification on both its NC500 and Purifi based amps and has them listed on their website. I am not sure what is best for power supplies. What is the difference between using the SMPS600 vs the SMPS1200A400?

Have you confirmed with them that those input boards do work with the Purifi module? The cost of the "D" board is as much as the amp module!

As for the Hypex SMPS600N400, it is likely to be fine for a single channel of the Purifi board, with the appropriate cable adaptations, but I have not confirmed the necessary details of that.

 

[JimB]

I am hoping for a Ghent case like the NCORE-SE(200W), which is for 200mm wide, not 200 Watts. It might want to be a hair longer than that case (currently 315mm), for effective venting holes (TBD). I'm hoping they'll offer two flavors:
1. For the stock EVAL1, as-is, at the back panel; and
2. One with binding posts* (and no holes for the Purifi banana sockets).

A third variant could be to provide panel mounted XLR sockets, but I don't see the point, other than to provide different back panel symmetry. In all cases, only the back panel would change.

* The Purifi FE board uses silver plated 4mm plug sockets. I'd be happy enough if they were gold plated, but I cannot find any to do replacement. Binding posts are a bit higher quality and more versatile. One might also prefer speakON.

I plan to remove the FE board Standby toggle switch and jumper that to always on, letting power determine when it is on (like the NC400). So I hope for the panels NOT to have that hole, or to provide a nice plug.

FWIW, here is how I did my Hypex NC400/SMPS1200 version of that case.

 

[JimB]

How the EVAL1 scales into this case. Not a real build.

 

[mocenigo]

Have you confirmed with them that those input boards do work with the Purifi module? The cost of the "D" board is as much as the amp module!

As for the Hypex SMPS600N400, it is likely to be fine for a single channel of the Purifi board, with the appropriate cable adaptations, but I have not confirmed the necessary details of that.

Try to contact Apollon. They may sense a business opportunity here.

Roberto

 

[Dudi]

Hello,

I'm a regular reader of ASR, but never post comments. Please let me to take exception to share my Purifi DIY implementation. Purifi eval board is retrofitted to my previous amplifier, so it has volume control, remote and input selector with RCA-XLR converter. It sounds amazing, but slightly worse than the pure eval board.

1ET400A generates close to zero heat but SMPS1200 gets warm

(top cover is drilled at SMPS).

 

[JimB]

Hello,

I'm a regular reader of ASR, but never post comments. Please let me to take exception to share my Purifi DIY implementation. Purifi eval board is retrofitted to my previous amplifier, so it has volume control, remote and input selector with RCA-XLR converter. It sounds amazing, but slightly worse than the pure eval board.

1ET400A generates close to zero heat but SMPS1200 gets warm (top cover is drilled at SMPS).

Nice! Maybe we should move this out of the Review thread to the DIY thread mentioned above.

 

[WHO23]

Hello,

I'm a regular reader of ASR, but never post comments. Please let me to take exception to share my Purifi DIY implementation. Purifi eval board is retrofitted to my previous amplifier, so it has volume control, remote and input selector with RCA-XLR converter. It sounds amazing, but slightly worse than the pure eval board.

1ET400A generates close to zero heat but SMPS1200 gets warmView attachment 43322 (top cover is drilled at SMPS).

Wow looks good! Is that a Hattor NCore Amp retrofitted with the Purifi 1ET400A?

 

[JimB]

A nice retrofit/upgrade with EVAL1 kit here:
https://www.audiosciencereview.com/...fi-1et400a-amplifier.7984/page-56#post-289977

 

[Dudi]

Wow looks good! Is that a Hattor NCore Amp retrofitted with the Purifi 1ET400A?

Exactly, I didn't want to sell that nice looking amplifier, the result is much better than NC250MP based original was.

Just wanted to point to that, old/previous amplifier box could be a good base.

JimB: Of course, please move the post to the DIY thread. I finished to write/answer into the measurement thread.

Cheers

 

[JimB]

...

JimB: Of course, please move the post to the DIY thread. I finished to write/answer into the measurement thread.

Cheers

IANAA (I am not an Admin), so I cannot move posts (@Thomas savage), but we can choose to start posting over there on anything new that is Purifi 1ET400A based DIY.

 

[JimB]

Received already! Excellent service!

Notes. Neither I2C header is fitted (fine for me). They included two stand-offs under the edge of the FE board, between the output sockets. And note the color coding of the buffer gain jumpers - red and blue - a nice touch, though I anticipate removing the headers and soldering jumper wires in place for my configuration. I also anticipate removing the toggle switch and hard wiring that as well. The output connectors are now all unbranded. They might go as well, depending on the final case options from Ghent. The XLRs are Neutrik brand and appear to be gold plated inside.

Unfortunately, I'll be way from now, unable to do anything with them for a couple weeks... Also need to get them SMPS1200's on order!

 

[pos]

Now that is exciting!

There is something I would love to sort out here: my understanding was that the jumpers would let you choose the gain of the buffer, from 0dB (13dB total) to 13dB (26dB total).
But now looking at your photo the markings on the PCB indicate "buffered/unbuffered", which is somewhat different from a simple gain change. The manual states this is a bypass option, which is confusing again: does it bypass the gain stage or the whole buffer, in which case the input impedance would be quite low...

 

[JimB]

Now that is exciting!

There is something I would love to sort out here: my understanding was that the jumpers would let you choose the gain of the buffer, from 0dB (13dB total) to 13dB (26dB total).
But now looking at your photo the markings on the PCB indicate "buffered/unbuffered", which is somewhat different from a simple gain change. The manual states this is a bypass option, which is confusing again: does it bypass the gain stage or the whole buffer, in which case the input impedance would be quite low...

Can't say. The EVAL1 document says this is a bypass of the FE gain stage, resulting in 0dB gain from the FE. However, there are no input impedance specs listed at all (let alone in both configurations) to confirm whether the 0dB setting gets you a unity gain buffer, or a true bypass. It might be possible to "ohm it out", but that is not something I can get into for a while.

Given the purpose of the board, to facilitate evaluation and development, I would expect that it does indeed bypass the buffer gain stage. Hopefully, others will find the answer without resorting to tracing the circuit.

Personally, I expect to be using the buffer gain stage, but I might want to dial down the gain to 6dB or so. TBD.

Since this simple photo already triggered new questions, perhaps I should take a better photo of each of the two production board types and post them here.
-----------------------------------------------------------------------------------------------
January 6, 2020. New information here:
https://www.audiosciencereview.com/...iy-purifi-amp-builds.10478/page-8#post-296652

 

[JimB]

You can zoom well in, here. I'll be away now until Jan 2.

 

[JimB]

... The EVAL1 document says this is a bypass of the FE gain stage, resulting in 0dB gain from the FE. ...

View attachment 43556

....

Purifi has confirmed that the FE bypass setting provides a true bypass, not unity gain buffering.

Also, that not populating the I2C headers is now standard.

 

[pos]

There are two potential off the shelf solutions for a variable gain buffer:

1. the original hypex nc500 evaluation buffer board which should be compatible (from what I gather from Bruno's interview) and has the classic hypex gain stage that can be set to unity. That would be the simplest solution as everything should just fit together (to be confirmed of course).
2. the upcoming neurochrome universal buffer that can work with a regulated +/-12V power from an SMPS1200 and provide an adequate output level with that power to match the 1ET400A sensitivity. The performance here is astonishing, but it would require an additional PCB to adapt everything.

 

[Science is fun]

Actually the NC-500 buffer board has a bypass jumper. It doesn't set the buffer at unity, it bypasses it completely if soldered.