• 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

boXem

Major Contributor
Audio Company
Joined
Jun 19, 2019
Messages
2,014
Likes
4,852
Location
Europe
Upon further research I think the best option would be to add an inline attenuator, such as a fixed Rothwell attenuator. Pardon my ignorance, but which gain reduction level should I aim for (-10, -15 or -20 dB), to bring the EVAL1 back into balance with a typical balanced DAC that has a 4V output Voltage?
20log(1.5/4) = - 8.5 dB so -10 dB
 

TimoJ

Senior Member
Joined
Dec 9, 2019
Messages
412
Likes
448
Location
Finland
Upon further research I think the best option would be to add an inline attenuator, such as a fixed Rothwell attenuator. Pardon my ignorance, but which gain reduction level should I aim for (-10, -15 or -20 dB), to bring the EVAL1 back into balance with a typical balanced DAC that has a 4V output Voltage?
You could also change two SMD resistors (R17) on the Purifi preamp board. I think that would give better performance than using an external attenuator. (Theoretical performance that you won't hear.)
 
OP
J

JimB

Addicted to Fun and Learning
Forum Donor
Joined
Sep 19, 2019
Messages
731
Likes
493
Location
California
Dear audiophiles,

A quick question - given that the stock Purifi EVAL input buffer (as amirm mentioned) reaches max volume at just 1.5V input with the extra gain, how do you guys drive the amp unit? If I am not mistaken, most balanced (XLR) DACs max out around 4V. How do you control the volume?

I am trying to procure a completely new setup from scratch with the Purifi amp, likely a pair of Klipsch RP-8000F speakers and a new balanced DAC that I still need to decide on (currently leaning towards either the Gustard A18 or the SMSL SU-8). My concern is that if the maximum volume is reached at just 36.5% of the max volume of the DAC (1.5/4) that the incremental volume steps are actually quite large. Furthermore, what would happen if the DAC reverts to full output/volume (4V) for example because there was a power reset or something like that, that could mean the end of the speakers. Very keen to find out how to best tackle the volume issue or how you guys drive the amp!

Cheers
The nominal gain of the EVAL1 with buffer stage enabled is 27 dB, or ~22.4x. It takes 1.8V to reach the nominal rated 200W output (8 ohms), though the actual drive required to hit hard clipping will depend on your supply and mains voltage. Having up to 2V is not excessive.

And having a little extra gain upstream is not crazy. You might want to turn up some low level recording. So the concern then seems to be volume control resolution. What will you be using? If a simple linear stepped attenuator, then yes, you'll be maxing out at about half position, with crude steps down the range where you will listen most of the time. But, if not, then it won't behave that way. If, for example, you use the volume control on the Gustard (right?), you would have good control in reasonable dB steps, all the way up.

Maybe try your chosen volume control method first, and then, if the potential to overdrive the amp in some cases still concerns you, add a good in-line attenuator.
 
Last edited:
OP
J

JimB

Addicted to Fun and Learning
Forum Donor
Joined
Sep 19, 2019
Messages
731
Likes
493
Location
California

Aardappel

Member
Joined
Nov 19, 2020
Messages
62
Likes
56
Location
Melbourne
The nominal gain of the EVAL1 with buffer stage enabled is 27 dB, or ~22.4x. It takes 1.8V to reach the nominal rated 200W output (8 ohms), though the actual drive required to hit hard clipping will depend on your supply and mains voltage. Having up to 2V is not excessive.

And having a little extra gain upstream is not crazy. You might want to turn up some low level recording. So the concern then seems to be volume control resolution. What will you be using? If a simple linear stepped attenuator, then yes, you'll be maxing out at about half position, with crude steps down the range where you will listen most of the time. But, if not, then it won't behave that way. If, for example, you use the volume control on the Gustard (right?), you would have good control in reasonable dB steps, all the way up.

Maybe try your chosen volume control method first, and then, if the potential to overdrive the amp in some cases still concerns you, add a good in-line attenuator.

Thanks JimB, appreciate your response. Your suggestion to try a DAC first and add an attenuator later if required seems the most sensible way to go. Your post, if I understand you correctly, made me realise that DACs that have volume control in steps of 1dB are not linear stepped attenuators, so the actual difference in volume per step will not be too drastic.
 
OP
J

JimB

Addicted to Fun and Learning
Forum Donor
Joined
Sep 19, 2019
Messages
731
Likes
493
Location
California
... Your post, if I understand you correctly, made me realise that DACs that have volume control in steps of 1dB are not linear stepped attenuators, so the actual difference in volume per step will not be too drastic.
A dB of change is a dB of change, regardless of full scale limit.
 
D

Deleted member 65

Guest
Nice break from Purifi build, MTB on a sunny Sunday ...

1606047545710.jpeg
 

Murray A

Member
Forum Donor
Joined
Jun 16, 2019
Messages
36
Likes
109
Location
Canada
My Purifi build using the Eval1 components in the Ghent case. Under 2 hours assembly time. Added a fuse holder with 10 amp S.B. fusing to protect the onboard power supply fuse! Running with input buffer bypassed for best performance.

L1010317_00001_01.jpg
 

Doodski

Grand Contributor
Forum Donor
Joined
Dec 9, 2019
Messages
20,745
Likes
20,756
Location
Canada
My Purifi build using the Eval1 components in the Ghent case. Under 2 hours assembly time. Added a fuse holder with 10 amp S.B. fusing to protect the onboard power supply fuse! Running with input buffer bypassed for best performance.

View attachment 95182
Nice assembly. Very orderly and good layout.
 

Murray A

Member
Forum Donor
Joined
Jun 16, 2019
Messages
36
Likes
109
Location
Canada
Nice assembly. Very orderly and good layout.
Thanks for your kind comments.

I've been amp building for 55 years. I had the opportunity to test the Purifi Eval1 with my APX555 for one of the online webzines and was blown away by the eval1 technical and aural performance. This is truely a game changer and I believe the future of audio.

L1010323_01.JPG
 
Last edited:

Doodski

Grand Contributor
Forum Donor
Joined
Dec 9, 2019
Messages
20,745
Likes
20,756
Location
Canada
Thanks for your kind comments.

I've been amp building for 55 years. I had the opportunity to test the Purifi Eval1 with my APX555 for one of the online webzines and was blown away by the eval1 technical and aural performance. This is truely a game changer and I believe the future of audio.

View attachment 95189
Nice HP 11C too. :D I used a regular calc all my life and decided to pick-up a HP 11C for my electronics study many moons ago. Much easier for many operations. Those are collectors items now.
 

Murray A

Member
Forum Donor
Joined
Jun 16, 2019
Messages
36
Likes
109
Location
Canada
S1 is still there but has been bypassed by J3 pin 1 pulled low. The speakons allow some performance improvement due to 4 terminal connection that encompasses the feedback loop. I've been leaving the amp powered up continuously. It is never shut down.
 
OP
J

JimB

Addicted to Fun and Learning
Forum Donor
Joined
Sep 19, 2019
Messages
731
Likes
493
Location
California
My Purifi build using the Eval1 components in the Ghent case. Under 2 hours assembly time. Added a fuse holder with 10 amp S.B. fusing to protect the onboard power supply fuse! Running with input buffer bypassed for best performance.
Are you happy with the case? Do you recommend it?
 
OP
J

JimB

Addicted to Fun and Learning
Forum Donor
Joined
Sep 19, 2019
Messages
731
Likes
493
Location
California
S1 is still there but has been bypassed by J3 pin 1 pulled low.
You mean by the power supply?
The speakons allow some performance improvement due to 4 terminal connection that encompasses the feedback loop.
How so? The 4-wire sense point is in the DELTRON-style sockets. Did you make any other adaptions in your build?
 
Last edited:

Murray A

Member
Forum Donor
Joined
Jun 16, 2019
Messages
36
Likes
109
Location
Canada
You mean by the power supply?

How so? The 4-wire sense point is in the DELRON sockets. Did you make any other adaptions in your build?

The sense points are in the same vicinity as the delron sockets. I was was unable to measure the output Z difference between the two connections.
 
OP
J

JimB

Addicted to Fun and Learning
Forum Donor
Joined
Sep 19, 2019
Messages
731
Likes
493
Location
California
The sense points are in the same vicinity as the delron sockets. I was was unable to measure the output Z difference between the two connections.
You appear to have done a really solid job of connecting your speakONs.
 

Aardappel

Member
Joined
Nov 19, 2020
Messages
62
Likes
56
Location
Melbourne
What is the reason for the pretty sizeable difference in Purifi output power measurements between their own measurement (top chart) and the soundstagehifi measurement (bottom chart)? In 8 Ohm its about a 100W difference - is it because of the input buffer being bypassed or not? And if so, why would one want to bypass the buffer if power output drops by 100W..?

Screen Shot 2020-11-23 at 5.38.07 pm.png


Screen Shot 2020-11-23 at 5.38.27 pm.png
 
Top Bottom