Purifi class-D amplifiers

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • ergo
    Senior Member
    • Mar 2005
    • 676

    I have a Soekris DAC / Bruno Butzey's Balanced Pre project that I've also listened with the Purifi. That Balanced Pre uses the same LM4562 opamp in all stages including the output buffer. Power to the BPBP module is provided by SilentSwitcher. Both me and my friend found that sounding inferior to the WHAMMY buffer doing the same function. Interestingly it was not only low end control but also the depth of the sound stage and sort of 'air' that was missing in BPBP. At the time of listening the power supply for both cases was the SilentSwitcher module.

    The Neurochrome is a different circuit of course with only the opamp being the same (and there is only 1 in signal path instead of 3 in BPBP). So it is sure possible that it achieves a different result. It would be kind of interesting to get two modules and try them out in apples to apples comparison, but for me being in EU it'll probably cost quite a bit more due to +20% VAT on modules + shipping.

    ***
    I am making some progress on finalizing the WHAMMY based solution too. I had bought some Jung/Didden Super Regulator PCB-s times ago for ... I don't even remember what for. Anyhow, this project was a motivation enough to finish those. So the plan is now to have a MEAN WELL IRM-20-24 AC/DC unit supplying each positive and negative power rail for the buffers and each channels gets their own. So basically a dual mono build from buffer power supply perspective. Waiting on MeanWells to com in to put it all in action.

    Click image for larger version

Name:	Jung Didden SuperRegs.jpg
Views:	1
Size:	134.2 KB
ID:	865327

    Comment

    • mocenigo
      Junior Member
      • Feb 2020
      • 6

      Originally posted by ergo
      I have a Soekris DAC / Bruno Butzey's Balanced Pre project that I've also listened with the Purifi. That Balanced Pre uses the same LM4562 opamp in all stages including the output buffer. Power to the BPBP module is provided by SilentSwitcher. Both me and my friend found that sounding inferior to the WHAMMY buffer doing the same function. Interestingly it was not only low end control but also the depth of the sound stage and sort of 'air' that was missing in BPBP. At the time of listening the power supply for both cases was the SilentSwitcher module.

      The Neurochrome is a different circuit of course with only the opamp being the same (and there is only 1 in signal path instead of 3 in BPBP). So it is sure possible that it achieves a different result. It would be kind of interesting to get two modules and try them out in apples to apples comparison, but for me being in EU it'll probably cost quite a bit more due to +20% VAT on modules + shipping.

      ***
      I am making some progress on finalizing the WHAMMY based solution too. I had bought some Jung/Didden Super Regulator PCB-s times ago for ... I don't even remember what for. Anyhow, this project was a motivation enough to finish those. So the plan is now to have a MEAN WELL IRM-20-24 AC/DC unit supplying each positive and negative power rail for the buffers and each channels gets their own. So basically a dual mono build from buffer power supply perspective. Waiting on MeanWells to com in to put it all in action.

      [ATTACH=CONFIG]30129[/ATTACH]
      ergo, I do not know the BPBP, and I am VERY intrigued by what you are doing. I can tell you this: the Neurochrome Universal Buffer sounds incredible. You really have the feeling that "something opaque" has been removed from the systemm. The Neurochrome Universal Buffer into Neurochrome Input Boards with buffer bypassed and Hypex regs on board into 1ET400A is by far the best amplifier I have listened so far - in particular among those with my source and speakers. At home I have auditioned Abrahamsen (in other words, Electrocompaniet), Lavardin IS, a Naim NAC 552 + NAP 500 DR from a neighbour that went back home saddled, NAP 150, and Apollon NC500-based amps with Sonic Imagery opamps. Source is a Soekris 1541 (I am sure you have a good idea how it sounds) and the speakers you can see here. Your DIY skills surpass mine, but I would like to initiate an exchange of information with you. I am sure I can learn a lot.

      Best, and stay safe, sane, healthy.

      (reminder that my build can be seen here)

      Comment

      • ergo
        Senior Member
        • Mar 2005
        • 676

        Aha, your comments confirm what I was thinking I saw from the pictures - you essentially bypass the LM4562 opamp on the "PURIFI 1ET400A / HYPEX NC500 INPUT BUFFER" and instead the driving of the Purifi inputs is on the shoulders of LM49724 on "UNIVERSAL BUFFER". That chip has a much more capable output with +/-80mA instead of the +/-26mA specified for LM4562. So it is possible we get to similar outcome with your scenario versus WHAMMY.

        What I was suspicious of was a use case where one has only the "PURIFI 1ET400A / HYPEX NC500 INPUT BUFFER" - as is, we do not yet have user feedback on that scenario it seems.

        Anyhow, the Neurochrome buffer sure is nice small and easy to set up compared to all the complexity I'm trying to go through. Unless one has dual power supplies there is still a bit a question how exactly is the best way to split the power between the two modules. I see that you did it about mid way the cables between the modules and PS. And for Buffer lower voltage supply the ground split at the regulator out in your case.
        Without measuring it it will be hard to tell if that can create some small ground loops or hum. But if you don't hear any than you were probably lucky and it is good as is.

        Click image for larger version

Name:	LME49724.jpg
Views:	1
Size:	202.8 KB
ID:	865328

        Comment

        • mocenigo
          Junior Member
          • Feb 2020
          • 6

          Originally posted by ergo
          What I was suspicious of was a use case where one has only the "PURIFI 1ET400A / HYPEX NC500 INPUT BUFFER" - as is, we do not yet have user feedback on that scenario it seems.
          [ATTACH=CONFIG]30130[/ATTACH]
          The sound with just the input buffer of the Neurochrome adapter board is a bit laid back and less dynamic than with the Neurochrome Universal Buffer, but still with an edge over the Purifi FE02 board in the EVAL1 kit in clarity. I do not like the sound of the EVAL1 kit that much, relatively speaking. Absolutely speaking it was great, and better than my previous NC500 based monoblocks with Sonic Imagery discrete opamps - the main difference being in the highs, so smooth that it was closer to vacuum tube ideal. But with the Neurochrome Buffer the bass becomes even tighter and the plasticity of the soundstage (how well the sound sources are modeled and the layering is rendered) just blossoms. The sound now, with the best recordings, is even eerie, like watching a 60fps movie –*nearly disturbing, because we are used to the defects of a lower frame rate.

          Most people would probably not even notice the difference between the three buffers with decently sensitive speakers. But in my case, with horn loaded compression drivers and highly sensitive woofers downstream, and a Soekris DAC1541 upstream, all balanced connections, I can say I am very happy with the outcome.

          Comment

          • ergo
            Senior Member
            • Mar 2005
            • 676

            mocenigo, your description of a change in sound does sound similar to what we have experienced with EVAL1 vs WHAMMY.

            My build is still not done, but some progress is made...

            On Monday I should get the Mean Well AC/DC adapters and also a kit from Khozmo in Poland to solve the volume control.
            It was hard to decide what to choose. I did want a R2R relay based volume. Originally I did build the back panel holes for BPBP preamp which would have that. But as noted before the listening experiments showed it to be a degradation even when compered to lowering the volume from PC and thus potentially losing some bits. I then e-mailed Khozmo and it turns out they have more options available compared to what is up on the website. So below is an image of what I'm expecting now.

            At first I planned to put the two WHAMMY modules on top of each other... that then makes it hard to swap out opamps etc. So I printed two stands for them and listed them up. That makes enough room for both of them side by side and the coming volume control boards can hopefully fit below these.

            When I bought the case I chose the deepest one and in beginning there seemed to be loads of room left over. Seems I'm now able to fill it easily and parts even on multiple layers.

            Click image for larger version

Name:	Khozmo.jpg
Views:	1
Size:	124.4 KB
ID:	865346Click image for larger version

Name:	3D_print.jpg
Views:	1
Size:	94.0 KB
ID:	865342Click image for larger version

Name:	stands.jpg
Views:	1
Size:	95.3 KB
ID:	865343Click image for larger version

Name:	WHAMMY on legs.jpg
Views:	1
Size:	109.4 KB
ID:	865344Click image for larger version

Name:	Purifi front angle.jpg
Views:	1
Size:	132.1 KB
ID:	865345

            Comment

            • ergo
              Senior Member
              • Mar 2005
              • 676

              I have not had time to post, but a lot of work has been done after last update.

              I received the Khomo R2R relay volume and I'm very happy with it. Very nice engineering for affordable price.

              Click image for larger version

Name:	Khomo_volume.jpg
Views:	1
Size:	90.1 KB
ID:	865377

              So next step was to mill a suitable cavity and window into the Purifi enclosure front. As I do have a 3D printer also it seemed easier to go the easy route and make 3D printed holders attached with hot glue instead of over engineering and threading the aluminium etc.

              Click image for larger version

Name:	Front display.jpg
Views:	1
Size:	47.5 KB
ID:	865378

              I did use a bit of over engineering for mounting the relay board. I had some old rotating hard drive "decoupler rubber grommits". So the relay boards got attached with those to avoid the relay click transferring into the enclosure.

              Click image for larger version

Name:	relay board buffering.jpg
Views:	1
Size:	55.7 KB
ID:	865379

              Khomo board inplace before WHAMMY

              Click image for larger version

Name:	Khomo in place.jpg
Views:	1
Size:	125.9 KB
ID:	865380

              And the final view..... the case that originally seemed very big indeed is now full of 'stuff'

              Click image for larger version

Name:	Purifi inside final.jpg
Views:	1
Size:	135.6 KB
ID:	865381

              Comment

              • ergo
                Senior Member
                • Mar 2005
                • 676

                The front panel looks very nice and the dot matrix display is super cool through the tinted glass...

                Click image for larger version

Name:	Purifi front ready.jpg
Views:	1
Size:	57.5 KB
ID:	865382

                Comment

                • ergo
                  Senior Member
                  • Mar 2005
                  • 676

                  On electronics side - I originally had the Khomo volume and WHAMMY boards isolated from the ground with non conductive spacers. The XLR cable ground was connected to enclosure at the point of input XLR sockets, then went to Khomo relay volume and from there to WHAMMY and from WHAMMY to Purifi input that is again connected to enclosure ground firmly.

                  The result did not look bad when volume was at Max, but if I decreased it to 30 (which is moderate listening level) it looked problematic.

                  Click image for larger version

Name:	Old-ground.jpg
Views:	1
Size:	33.5 KB
ID:	865383

                  So then onto reading Bruno Butzeys article G word.

                  So onto making my amp from left image to right image concept

                  Click image for larger version

Name:	Grounding.png
Views:	1
Size:	21.9 KB
ID:	865384


                  Click image for larger version

Name:	Good ground.jpg
Views:	1
Size:	96.4 KB
ID:	865385

                  And not the spectrum is super clean.... and remains as clean down to lowest volume levels

                  Click image for larger version

Name:	1W_4ohm.jpg
Views:	1
Size:	98.9 KB
ID:	865386

                  I've never had a power amp that measures like this

                  Click image for larger version

Name:	100W 4ohm.jpg
Views:	1
Size:	100.2 KB
ID:	865387

                  Comment

                  • ergo
                    Senior Member
                    • Mar 2005
                    • 676

                    And last two

                    Click image for larger version

Name:	5W_4ohm multitone.jpg
Views:	1
Size:	106.6 KB
ID:	865388

                    and a plot from about 60mW to 100W

                    Click image for larger version

Name:	THD vs power.png
Views:	1
Size:	21.6 KB
ID:	865389

                    Comment

                    • Evil Twin
                      Super Senior Member
                      • Nov 2004
                      • 1531

                      LLC SMPS Development platform PCBs

                      This is for Sven, because I could not attach images to a PM message. Discussion related to PuriFi.


                      Click image for larger version

Name:	PCB-3D 4-01-2020 Top.jpg
Views:	1
Size:	64.4 KB
ID:	865433



                      Click image for larger version

Name:	PCB-3D 4-01-2020 Orthagonal.jpg
Views:	1
Size:	58.1 KB
ID:	865434
                      DFAL
                      Dark Force Acoustic Labs

                      A wholly owned subsidiary of Palpatine Heavy Industries

                      Comment

                      • jaytor
                        Junior Member
                        • Oct 2020
                        • 2

                        Hi folks. This is my first post in this forum. I've been an audiophile for about 45 years and a home theater enthusiast for 30. After a 35 year hiatus from DIY, I've gotten back into building some of my own gear. So far in the past year I've built a pair of GR-Research NX-Otica speakers, a pair of triple stack open baffle servo subwoofers, a pair of First Watt F5 Turbo monoblocks, and a pair of Neurochrome Modulus-286 monoblocks.

                        I'm almost done with a new preamp that uses a Khozmo balanced relay-based stepped attenuator with OLED display and some boards from AMB Labs.

                        And the reason I'm posting here - I've started working on a pair of Purifi module monoblocks. I'm building a linear power supply and have a couple extra features like audio-sense power-on and a peak power meter. I'm working on figuring out what to do for the input buffer. I think I'm going to start with the Neurochrome input board as a baseline, but plan to build a couple of additional buffers using custom PCBs. I am planning to build a tube based design, but am also considering something similar to the WHAMMY design that @ergo has described above.

                        @ergo - do you have any more thoughts on using the WHAMMY after a few more months of listening? What opamp did you settle on to drive the MOSFETS?

                        Thanks,
                        Jay

                        Comment

                        • Jonasz
                          Senior Member
                          • Nov 2004
                          • 852

                          Can't answer your question but do you prefer the FW or the Neurochrome and why?

                          Comment

                          • jaytor
                            Junior Member
                            • Oct 2020
                            • 2

                            The Neurochrome amp works better in my system. It has considerably more detail and images a bit better. I cross over to an active servo sub at about 70Hz, so can't really comment on the low bass.

                            I also have a Parasound JC5 and I think the Neurochrome is pretty comparable (although obviously much less power), while the lack of detail in the F5T compared to the JC5 was very noticable.

                            I have long balanced connections between my preamp and monoblocks, so I built the F5T with a Jensen input transformer to convert balanced to single ended. This may be contributing to the reduced detail. At some point, I'm going to rebuild these amps. I may just replace the transformer with an op amp based input buffer, but I've also been thinking about trying to do my own design using a single-ended class A output stage at lower voltage. My speakers are fairly efficient (93db/w) so I don't need a ton of power.

                            Comment

                            • Jonasz
                              Senior Member
                              • Nov 2004
                              • 852

                              Thank you jaytor!
                              Valuable information when trying to decide what to build. I don't need a lot of power either so there's lot of stuff to choose from but hard to listen before building.

                              Comment

                              Working...
                              Searching...Please wait.
                              An unexpected error was returned: 'Your submission could not be processed because you have logged in since the previous page was loaded.

                              Please push the back button and reload the previous window.'
                              An unexpected error was returned: 'Your submission could not be processed because the token has expired.

                              Please push the back button and reload the previous window.'
                              An internal error has occurred and the module cannot be displayed.
                              There are no results that meet this criteria.
                              Search Result for "|||"