• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

DIY Purifi Amp builds

mocenigo

Active Member
Joined
Dec 8, 2018
Messages
184
Likes
73
And here's my build. Even more crude and sloppy than LondonLad 's because I only have a wood board for the moment. Apparently we use the same power outlet, but the similarities end here. Most of the text that follows is a copy from my own post at diyaudio: After a few email exchanges with Kim Nordtorp Madsen at Purifi, I decided to use a Connex SMPS800RE, which can deliver 800W continuously and up to 1000W burst. I picked it configured at +-54V, even though I could have asked for a customised 65V (more lead time), or even take the 72V one and use the trimmer to go down to 68.8V. In this configuration I have about 160W on 8ohm and 320 on 4ohm. With my very efficient speakers (95db!) I could even have bought a +-25V version as well.

Then I sourced (from Mouser) two little 18V switchers manufactured by RECOM. These seem to be the only little switchers at 18V for < 1A (in this case about 0.3A). 24V switchers are also ok, Meanwell also has them, but the RECOM ones are exceptionally low noise, low EMI, and precise. I mounted the two switchers on a little breadboard, as you can seem, with the -V of one tied to the +V of the others and to the ground, so I get -18,0,+18 - for the modulator. This is temporary since I will replace this with a little toroidal and a ATLhifi modified Sulzer regulator.

I use the Connex for +VP, -VP, VDR, VDR_RTN which is internal tied to -VP, and for GND. I need only one GND since the positive and negative rail share the ground on the Connex SMPS (not the case on the Hypex), so I have one GND connection "dangling" and will later remove it.

Note that the Connex is, unlike the hypex, regulated, and a lot of DIYers in France and Germany swear by the Connex power supplies, considering them much better than the Hypex ones.

What I can say is that this little concoction sounds to me (yes, this is a subjective perception, and it may be the so called "cockroach effect" [1]) better than the NC500+Hypex power supply. More pleasing and smooth highs, for instance.

Roberto

[1] The cockroach effect, in Neapolitan language "effetto scarrafone", from the proverb "Ogne scarrafone è bell' a mamma soja", i.e. "every cockroach is beautiful for his mom". This is the apex of confirmation bias for DIYers: if you have built it yourself, it sounds better than anything else.


And it keeps on growing :)

IMG_5631.jpg
 

hombre

New Member
Joined
Feb 9, 2020
Messages
1
Likes
0
Alternate power supply for Purifi EVAL1?

500W +/-55V Amplifier Dual-Voltage PSU Audio AMP Switching Power Supply Board (on Ebay about $45 total)

I am wondering if these PSU's could be used with the Purifi modules as an all in one, simple, more compact alternative to the HYPEX + other power supplies because:

1. in addition to +/- 55V (you can pick voltage) they also have;
2. All auxiliary voltage: +/- 15V, 1A (shared GND with the main voltage)
3. and + 12V, 0.5A (separate ground wire);

If I understand correctly an UNregulated +/- 16.4 - 25V is needed for the EVAL1 so perhaps one could simply remove the + and - 15 volt VOLTAGE REGULATORS on the above mentioned +/- 55V PSU and get at least +/- 16.4 volts UNregulated out. I think this should work?

One caveat, the voltage output shown in the ebay ad (i.e. +/- 50V, +/- 60V is for 110VAC input. I have purchased these and the line voltage at my place is 123 VAC so the corresponding output is higher by about the ratio of the VAC increase. So, if one has 120VAC and you order +/- 55 V PSU, expect to get about +/- 60 V output.

+/- 55 X (120/110) = +/-60V ouput
 

DaniP

New Member
Joined
Dec 29, 2019
Messages
1
Likes
0
Very interesting thread. Would be possible to gather in the first topic all informations, recommendations, configurations built and functioning, scattered in these 23 pages? I think it would be very helpful in the assembly process to avoid missing important information
 

JimB

Active Member
Forum Donor
Joined
Sep 19, 2019
Messages
282
Likes
174
Very interesting thread. Would be possible to gather in the first topic all informations, recommendations, configurations built and functioning, scattered in these 23 pages? I think it would be very helpful in the assembly process to avoid missing important information
There is not a single, DIY 'kit' here, to be summed up into a first post - no "assembly process". Rather, there are notes from several people about what interests them, and what some have done, to date, with comments from yet others. More options are evolving, and will continue to do so for at least many months.
 
Joined
Jan 24, 2020
Messages
70
Likes
51
I must say, I do not understand why some of these companies that implement the modules do not make better choices when it comes to the internal layout of their amplifiers. The way this amp is laid out requires long runs of internal wiring for both the line level inputs and the speaker outputs. They are designing these things, so they have plenty of options to do a better job at this.
 

mocenigo

Active Member
Joined
Dec 8, 2018
Messages
184
Likes
73
And finally it is mostly complete (some cleanup to be done). Modified MODU case, plexiglass cover, internal perforated iron baseplate spray painted matte cobalt blue, added some "mounting cubes". There re venting slits on the bottom plate, and as you can see I have venting slits also on the rear plate.

IMG_5685.jpg IMG_5684.jpg
 

mocenigo

Active Member
Joined
Dec 8, 2018
Messages
184
Likes
73
I must say, I do not understand why some of these companies that implement the modules do not make better choices when it comes to the internal layout of their amplifiers. The way this amp is laid out requires long runs of internal wiring for both the line level inputs and the speaker outputs. They are designing these things, so they have plenty of options to do a better job at this.
It is called "hacking something together to be [among the] first on the market" – a known "bestworst practice" in many fields.
 
Joined
Dec 26, 2019
Messages
18
Likes
13
Where can I find an input buffer instead of using the eval1. I don’t like how the plugs and xlr are soldered to the board it makes the cut out much more difficult. ATM makes one that’s visually similar but says it’s for the nc500.
 

Matias

Addicted to Fun and Learning
Forum Donor
Joined
Jan 1, 2019
Messages
588
Likes
702
Location
São Paulo, Brazil
Where can I find an input buffer instead of using the eval1. I don’t like how the plugs and xlr are soldered to the board it makes the cut out much more difficult. ATM makes one that’s visually similar but says it’s for the nc500.
NC500 and 1ET400A are pin compatible, buffer boards can be shared between them.
The modules use different power supply voltages though.
 
Joined
Jan 24, 2020
Messages
70
Likes
51
And finally it is mostly complete (some cleanup to be done). Modified MODU case, plexiglass cover, internal perforated iron baseplate spray painted matte cobalt blue, added some "mounting cubes". There re venting slits on the bottom plate, and as you can see I have venting slits also on the rear plate.

View attachment 50061 View attachment 50062
Thanks for sharing. Right now I am on the hunt for a suitable chassis, and am still considering whether or not to go with any ventilation. It is running so cool (including the Connex supply) that I think I will go with no ventilation in a full size chassis: the modules will lose heat through to the outside of the case anyway, and if I can get away with it, I like the shielding and lack of dust entry which one has with a tight chassis with no holes. at the same time I am building a linear +/- 18 VDC very low noise supply for the OPAs and modulator section. The layout with the Eval board is pretty tricky to do the way I prefer, but I think I have it all figured out. Will be a while before i have time to finish this all up, and maybe in that time period some different input board options will become available. The Nord boards, kitted out with Weiss OP2 discrete opamps might be an interesting (expensive) solution: the Weiss OP2s are damn good circuits, both for measured performance and subjectively (I have heard the Medus DAC).
 

jsy

Member
Joined
Feb 2, 2020
Messages
51
Likes
21
And finally it is mostly complete (some cleanup to be done). Modified MODU case, plexiglass cover, internal perforated iron baseplate spray painted matte cobalt blue, added some "mounting cubes". There re venting slits on the bottom plate, and as you can see I have venting slits also on the rear plate.

View attachment 50061 View attachment 50062
Sir, nice looking box!

I was looking at your plywood mockup and trying to follow what you're doing from your images. That's quite a lot of power you've got running around in there!

What's going on with the neurochrome buffer, for starters? It looks like you've set it up as the entry point, and are routing the output straight back into the eval1 xlr connectors to save the bother of ripping out the sockets. I can't quite follow the eval1 datasheet - do the b1 and b2 jumpers disable the pregain stage and leave you with a unity gain buffer, or do they let you bypass the buffer stage altogether? Are you just keeping the eval1 around to slave off the rest of the control functions, so you don't have to bother wiring up a big mess of pin socket connections?

Thanks for posting your results
 
Joined
Jan 24, 2020
Messages
70
Likes
51
Where can I find an input buffer instead of using the eval1. I don’t like how the plugs and xlr are soldered to the board it makes the cut out much more difficult. ATM makes one that’s visually similar but says it’s for the nc500.
The Eval board is really pretty damn good, but I like the 1612 OPA (I use a 1612-1632 based output stage in my DSC-2 DAC). It also has purpose designed discrete regulators onboard (pretty trick). It is going to be very hard (expensive) to find anything "better" in terms of at least measured performance. I am just going to remove the onboard XLR jacks and Banana sockets, and add wires to the connectors of my choice. the Eval board is also quite affordable. If you can wait a bit ir looks likely that Neurochrome is going to produce a dedicated input board which would be dual mono in nature, I am assuming it will use the LM 4562 OPAs, which perform nearly as well by spec as the 1612...
 

JimB

Active Member
Forum Donor
Joined
Sep 19, 2019
Messages
282
Likes
174
The Eval board is really pretty damn good, but I like the 1612 OPA (I use a 1612-1632 based output stage in my DSC-2 DAC). It also has purpose designed discrete regulators onboard (pretty trick). It is going to be very hard (expensive) to find anything "better" in terms of at least measured performance. I am just going to remove the onboard XLR jacks and Banana sockets, and add wires to the connectors of my choice. the Eval board is also quite affordable. If you can wait a bit ir looks likely that Neurochrome is going to produce a dedicated input board which would be dual mono in nature, I am assuming it will use the LM 4562 OPAs, which perform nearly as well by spec as the 1612...
This is good advice - I agree. It is not easy to better the EVAL1.

Alternatively, if one cannot do the suggested UNsoldering job (but can do a soldering job), just mount the EVAL1 slightly back from the panel and wire up from it to your chosen connectors and panel placements. If one cannot even do soldering on the EVAL1, then make internal jumpers with (for example) locking banana plugs from the 4mm DELRON sockets (you can plug in from the front or back), and use pins from an XLR cable (without its body) to make jumpers from the input sockets. This has been suggested earlier in this (getting long) thread.
 

mocenigo

Active Member
Joined
Dec 8, 2018
Messages
184
Likes
73
Sir, nice looking box!
Thank you!

What's going on with the neurochrome buffer, for starters? It looks like you've set it up as the entry point, and are routing the output straight back into the eval1 xlr connectors to save the bother of ripping out the sockets. I can't quite follow the eval1 datasheet - do the b1 and b2 jumpers disable the pregain stage and leave you with a unity gain buffer, or do they let you bypass the buffer stage altogether? Are you just keeping the eval1 around to slave off the rest of the control functions, so you don't have to bother wiring up a big mess of pin socket connections?
The buffer on the EVAL1 kit FE02 board is bypassed by jumper configuration, so it is not a unity gain stage — the input does directly to the 1ET400A. So indeed I keep is as a “breakout board” plus some control.

I did not want to solder on the FE02 board since I will be getting new connections boards from Neurochrkme to test and probably also their speaker protection boards
 
Joined
Dec 26, 2019
Messages
18
Likes
13
The Eval board is really pretty damn good, but I like the 1612 OPA (I use a 1612-1632 based output stage in my DSC-2 DAC). It also has purpose designed discrete regulators onboard (pretty trick). It is going to be very hard (expensive) to find anything "better" in terms of at least measured performance. I am just going to remove the onboard XLR jacks and Banana sockets, and add wires to the connectors of my choice. the Eval board is also quite affordable. If you can wait a bit ir looks likely that Neurochrome is going to produce a dedicated input board which would be dual mono in nature, I am assuming it will use the LM 4562 OPAs, which perform nearly as well by spec as the 1612...
When removing the banana plugs I read to add a resistor to r8 a1 r9 a1 and same as the other side is there anything that has to be done when removing the xlrs. Or is it as easy as wiring 3 wires to the pcb board.
 

Similar threads

Top Bottom