• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

DIY Purifi Amp builds

barrows

Active Member
Joined
Jan 24, 2020
Messages
150
Likes
219
Hello
for memory, I don't know how much power you need but I have speakers with 88db of sensitivity, my dac have 4.5V on XLR output
I bypassed the buffer stage of the Eval1 , I have enough power to tear my head off and shake the walls!

What level is that in room? I live in a rural area, with plenty of room around me, with no need to worry about bothering anyone, and I play music at close to concert levels sometimes. 85 dB average levels and higher are common here. My loudspeakers (Focus Audio FS-888 Signatures) are actually nominally rated at 90 dB, but I think that is overly optimistic (perhaps today I will measure their actual sensitivity at 2.83 V at 1 m just for a more detailed answer here). I regularly push my NC-400 DIY amp to close to max levels on peaks. As my speakers are 4 0hm nominal, they crave some current, and I am hoping the slight additional current availability of the Purifi will give me a little additional headroom. Also, my room is medium/large: 14' wide, opening up to 18' wide, and over 30 ' feet long (I listen at 10' from loudspeakers) so it takes some power to energize the space as well.
I am aware of how close I am getting to the 2 V input level (for full output) of the NC-400 as I measure the DAC output with a pink noise reference track to keep track of this.
I am waiting for some pins from Digikey to get going again on my Purifi amp build (backordered, pins for the wiring harness), so I cannot test yet with that. I would really like to get myself a DAC with enough output to drive the Purifs bypassing the input gain, but as I oversample everything to DSD in software (requiring a -5 dB adjustment before oversampling to avoid clipping) I would need a very high output DAC to do so.
 

tomchr

Addicted to Fun and Learning
Audio Company
Forum Donor
Joined
Nov 5, 2018
Messages
984
Likes
2,607
Location
Calgary, Canada
Hello
for memory, I don't know how much power you need but I have speakers with 88db of sensitivity, my dac have 4.5V on XLR output
I bypassed the buffer stage of the Eval1 , I have enough power to tear my head off and shake the walls!
I suspect that's the case for many. I have no issue reaching ear-splitting levels with the 130 W (8 Ω) my "safe-n-sane" Modulus-686 build offers.

That said, having the ability to drive the amp to clipping will be a selling point for many. Also, some claim the Purifi/Hypex sounds better with a differential input. Now, I don't know whether this is "one guy on the internet says..." or whether it's reproducible on a larger scale in a scientifically controlled experiment, but in audio science doesn't always win.

I decided to include the proper differential output because it's fundamentally a better design (even if it is a little more expensive). In addition, it gives me a unique selling point, and brownie points with the "need more power" crowd, and possibly with the subjectivists as well. Most importantly, including the proper differential output does NOT reduce performance.

Tom
 

Red_Baron

Member
Joined
Jan 19, 2020
Messages
12
Likes
13
My Purifi/Hypex Input Buffer is a differential in/differential out buffer. If you provide it with a single-ended input, it will provide a single-ended output - just as the Purifi EVAL1 or many of the other buffers on the market.


I decided to include the proper differential output because it's fundamentally a better design (even if it is a little more expensive). In addition, it gives me a unique selling point, and brownie points with the "need more power" crowd, and possibly with the subjectivists as well. Most importantly, including the proper differential output does NOT reduce performance.

The earlier description implies psuedo-differential (dual mono with 180 out of phase inputs).

Is the new revision a proper fully differential circuit with a common-mode feedback loop or something else?
I am curious how the performance changes under this new mode. (SNDR, phase and magnitude balance at the output) It is nice to see this option available.
 

tomchr

Addicted to Fun and Learning
Audio Company
Forum Donor
Joined
Nov 5, 2018
Messages
984
Likes
2,607
Location
Calgary, Canada
The earlier description implies psuedo-differential (dual mono with 180 out of phase inputs).

Is the new revision a proper fully differential circuit with a common-mode feedback loop or something else?
I am curious how the performance changes under this new mode. (SNDR, phase and magnitude balance at the output) It is nice to see this option available.
"Dual mono with 180º out-of-phase inputs" is a pretty good description. Revision 1.0 of the Purifi/Hypex Buffer is the front-end of an instrumentation amp. Just as the Purifi EVAL1 and equivalent Hypex buffer are. The (now) obvious drawback of that circuit topology is that if you provide a single-ended input, you get a single-ended output.

In Revision 2.0 of my Purifi/Hypex Buffer (which will be the production version) the Buffer provides a differential output regardless of whether you connect a differential source or a single-ended one. It will contain common-mode feedback to ensure that the output common-mode voltage of the differential output remains at 0 V.

Tom
 

wineds

Active Member
Joined
Sep 1, 2019
Messages
117
Likes
82
Location
Melbourne, Australia
Finished the simplified circuit, parts list is the same with the exception of the 4N35 optocouplers and sockets plus proto boards. This replaces the 4 channel relay board. Both methods work well so It’s just a matter of preference. I have not had any problems with either one but proceed at your own risk. Will show a picture of the new version which is a little less cluttered, I used two separate proto boards to split up the two functions SMPS standby and LED amp enable. So now need to set back and enjoy. Thanks again for all the help, at this time it is one of my favorite amps.
JimM

Thanks for posting this I am considering implementing your approach using one of these :

https://www.amazon.com/gp/product/B01EY7626K/ref=ppx_od_dt_b_asin_title_s00?ie=UTF8&th=1

Did you also connect EVAL1 /FATAL to SMPS DC ERROR?

The other alternative is to use a 1 channel opto board output in parallel with the front panel switch to implement a 12v trigger. In this case 2 wires would need to be added to the back of the EVAL1 board as per the demo app that was reviewed by @amirm.

Lastly do you know what the standby power consumption is when SMPS is in STANDBY mode?
 

Count Arthur

Major Contributor
Joined
Jan 10, 2020
Messages
2,197
Likes
4,888
it is written in the datasheet ;) too much for me that's why I cut the 230v...

You can use a Hypex's softstart modules with their SMPS: https://www.diyclassd.com/product/softstart-module/38, which will reduce standby power to 1.2w. It also allows you to use a pushbutton, has an LED output and the option to use a PTC temperature sensor:

Annotation 2020-03-22 101459.jpg
 

boXem

Major Contributor
Audio Company
Joined
Jun 19, 2019
Messages
2,014
Likes
4,853
Location
Europe
After discution with Purifi this wiring is correct

it is written in the datasheet ;) too much for me that's why I cut the 230v...
Idle losses are when the SMPS rails are active but without load. Not related to standby consumption, which is not specified directly. Standby pin consumes 10 mA, if one applies 12V, 0.12W.
 

tomchr

Addicted to Fun and Learning
Audio Company
Forum Donor
Joined
Nov 5, 2018
Messages
984
Likes
2,607
Location
Calgary, Canada
Did you also connect EVAL1 /FATAL to SMPS DC ERROR?
I handle the handshaking between the SMPS and the amplifier module to ensure that the SMPS shuts down if the amp gets unhappy, yes.

Tom
 

KarVi71

Member
Joined
Nov 14, 2019
Messages
63
Likes
68
Still waiting for some parts but my build is coming along nicely...

Looks very nice.

As the small add on board on the Purifi modules seems to be a hot spot, perhaps you could consider not routing the cables over the one on the left in the picture?

Its hard to see how much clearance that there is, so perhaps its not a problem.

I am waiting for the modules from tomchr to become available, then I'll begin to work on my dual mono builds.
 

wineds

Active Member
Joined
Sep 1, 2019
Messages
117
Likes
82
Location
Melbourne, Australia
So
Looks very nice.

As the small add on board on the Purifi modules seems to be a hot spot, perhaps you could consider not routing the cables over the one on the left in the picture?

Its hard to see how much clearance that there is, so perhaps its not a problem.

I am waiting for the modules from tomchr to become available, then I'll begin to work on my dual mono builds.

So far I am finding that nothing even gets warm. But my speakers are 94db. In any case there ample clearance between the cables and the board.
 

JimM

Member
Forum Donor
Joined
Jan 7, 2020
Messages
61
Likes
28
Location
New Mexico
Dbfr did you connect Eval1 /fatal J3-3 directly to SMPS dc error J4-10 with out any buffer or pull resistor?
JimM
 

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,274
Likes
1,034

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,274
Likes
1,034
Hope a kit with neurochrome buffer will be available soon.

It would be a matter of purchasing two 1ET400A boards (and two cable kits), two Neurochrome input buffers, a Hypex SMPS1200A400, and for instance a Ghent case. once the bottom is properly drilled, the rest should be piece of cake (a bit of crimping or of soldering).
 

JimM

Member
Forum Donor
Joined
Jan 7, 2020
Messages
61
Likes
28
Location
New Mexico
Question on using 1ET400 /fatal signal and Eval1 /fatal signal to control the DC Error on smps1200A400. The way I read the 1ET400 data sheet the /fatal signal will function as a open collector transistor output so the two are wire or'ed together on the FE02A and then pinned out to the various connectors, my question is the /FATAL a logic level i.e. high then low when activated or just a open collector output. or does it need to be pulled high with a resistor when connected to the SMPS DC Error? In reading the SMPS spec sheet it states that Hypex amplifiers use open collector output to drive this pin, which implies the SMPS has a pull up resistor of 1k. And requires either around 10 ma or 13 ma drive capability. Can any one verify if this is the case and all we need to do is connect EVAL1 /fatal to SMPS DC Error. Many thanks:
JimM
 

wineds

Active Member
Joined
Sep 1, 2019
Messages
117
Likes
82
Location
Melbourne, Australia
Question on using 1ET400 /fatal signal and Eval1 /fatal signal to control the DC Error on smps1200A400. The way I read the 1ET400 data sheet the /fatal signal will function as a open collector transistor output so the two are wire or'ed together on the FE02A and then pinned out to the various connectors, my question is the /FATAL a logic level i.e. high then low when activated or just a open collector output. or does it need to be pulled high with a resistor when connected to the SMPS DC Error? In reading the SMPS spec sheet it states that Hypex amplifiers use open collector output to drive this pin, which implies the SMPS has a pull up resistor of 1k. And requires either around 10 ma or 13 ma drive capability. Can any one verify if this is the case and all we need to do is connect EVAL1 /fatal to SMPS DC Error. Many thanks:
JimM

My Eval1 /fatal signal is connected to DC Error on smps1200A400. I just measured /fatal on EVAL1 J3 and its sitting at about 19.8V with music playing. So there must be a pullup on the Hypex end I think? I also momentarily pulled /fatal to GND with a test lead and the SMPS shuts down and goes into protect mode.
 
Last edited:

JimM

Member
Forum Donor
Joined
Jan 7, 2020
Messages
61
Likes
28
Location
New Mexico
Great, thanks I will open up my case and connect the signals up, also I am going to enable the buffer on EVAL1 and lower the gain on the Benchmark DAC, for two reasons, one to see if I can tell any difference in sound quality, and two the gain is a little to high with the DAC at 0 db rather than -10 or -20 db. Right now I am running the volume control around 10:00 to 12:00 o'clock rather than 2 to 3, and I have to be careful when switching sources to keep from over driving the speakers. Will report back if I can tell any difference in sound quality.
JimM
 

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,274
Likes
1,034
Great, thanks I will open up my case and connect the signals up, also I am going to enable the buffer on EVAL1 and lower the gain on the Benchmark DAC, for two reasons, one to see if I can tell any difference in sound quality, and two the gain is a little to high with the DAC at 0 db rather than -10 or -20 db. Right now I am running the volume control around 10:00 to 12:00 o'clock rather than 2 to 3, and I have to be careful when switching sources to keep from over driving the speakers. Will report back if I can tell any difference in sound quality.
JimM

I would really suggest to use the buffer unless the DAC (or your preamp) is proven capable of driving low impedance loads. Excellent DACs (I am talking of behaviour of noise and distortion relative to the current signal level, not of beating SINAD world records) should perform well also with a volume setting of -30 Db or so – they should still give you 16-18 bits of undistorted resolution and effectively inaudible noise even with 26Db of gain and sensitive speakers downstream. My Soekris DAC1541 does that and I am sure your Benchmark DAC does that at least as well. [Effectively inaudible means: if you hear hiss/crackle/hum whatever only with your ear glued to the transducers, and as soon as you are one of two feet away you hear nothing. in my car there is only a hiss and I must be within inches from the mid/tweeter horn.] OTOH, with the buffer the DAC's (or the preamp's) output stage should be able to drive the amplifier more easily, with better bass response and less distortion.

If the gain is too much and you want to avoid risking to overdrive the loudspeakers when switching sources, you can use inline attenuators. If they are properly done they won't affect sound quality. I used JTS MA-123, and apparently the SHURE are even better. Some people swear by the fixed Rothwells.
 
Last edited:
Top Bottom