• 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

starfly

Active Member
Forum Donor
Joined
Jun 6, 2019
Messages
246
Likes
196
Well, I think my final parts (such as the case, not unimportant) should arrive this week and I can then finally get started on the build over the holiday weekend :)
 

kaka89

Active Member
Joined
Jun 24, 2018
Messages
257
Likes
197
The version you see has fixed gain, I may add one resistor and two jumpers for 3 gain settings (unity + 2 others) for production boards.
Please do allow gain settings!


AHB2's gain settings:

Low-Gain = 22 dBu (9.8 Vrms), Gain = 9.2 dB
Mid-Gain = 14.2 dBu (4 Vrms), Gain = 17.0 dB
High-Gain = 8.2 dBu (2 Vrms), Gain = 23 dB
 

boXem | audio

Major Contributor
Manufacturer
Forum Donor
Joined
Jun 19, 2019
Messages
1,208
Likes
1,743
Location
Europe
Please do allow gain settings!


AHB2's gain settings:

Low-Gain = 22 dBu (9.8 Vrms), Gain = 9.2 dB
Mid-Gain = 14.2 dBu (4 Vrms), Gain = 17.0 dB
High-Gain = 8.2 dBu (2 Vrms), Gain = 23 dB
These are sensible settings, but...
- 1ET400A gain is 12.8 dB, lowering it would mean adding an attenuation, not an option.
- The engineer in me likes both 17 and 23 dB, but the salesman knows that louder = better in people's ears.
So if implemented, it may end up with something like 12.8 - 17 - 26.5 dB.
 

barrows

Active Member
Joined
Jan 24, 2020
Messages
131
Likes
169
I would prefer any gain adjustment be done by replacing a resistor, rather than jumpers, this would allow for more exact adjustment/options, and would keep the signal paths short, not increasing any loop areas. Just make that resistor a bit larger for easy replacement, anyone with a soldering iron can replace a 1206 package resistor, it is easier to replace a 1206 than even a through hole part.
Remember, ~26 dB gain is needed for those with sources of 2 volt output to reach the amplifier's rated full output. So 26 dB should be an option.
 
Last edited:

kaka89

Active Member
Joined
Jun 24, 2018
Messages
257
Likes
197
Yes the default gain of 12.8 is low enough.
If I remember correctly 14.2 dBu (4 Vrms) is RME ADI-2's max output, there for 17.0 dB is a good start too.
It make sense to have 26dB as standard too.
 
OP
J

JimB

Addicted to Fun and Learning
Forum Donor
Joined
Sep 19, 2019
Messages
717
Likes
460
Location
California
These are sensible settings, but...
- 1ET400A gain is 12.8 dB, lowering it would mean adding an attenuation, not an option.
- The engineer in me likes both 17 and 23 dB, but the salesman knows that louder = better in people's ears.
So if implemented, it may end up with something like 12.8 - 17 - 26.5 dB.
The AHB2 is a 100W nominal amp, so for the same source drive, you do need to provide a few more dB gain to max out the 1ET400A. Remember too that the 'default' EVAL1 provides "~27dB" gain. So, as others have said, ~26 total gain is a good number.
 

boXem | audio

Major Contributor
Manufacturer
Forum Donor
Joined
Jun 19, 2019
Messages
1,208
Likes
1,743
Location
Europe
The AHB2 is a 100W nominal amp, so for the same source drive, you do need to provide a few more dB gain to max out the 1ET400A. Remember too that the 'default' EVAL1 provides "~27dB" gain. So, as others have said, ~26 total gain is a good number.
I should stop replying to threads while doing something else in parallel...
If implemented, the gains options would be 12.8 dB, something to get 4V RMS sensitivity into 4 ohms, and 26 dB.
 

boXem | audio

Major Contributor
Manufacturer
Forum Donor
Joined
Jun 19, 2019
Messages
1,208
Likes
1,743
Location
Europe
I would prefer any gain adjustment be done by replacing a resistor, rather than jumpers, this would allow for more exact adjustment/options, and would keep the signal paths short, not increasing any loop areas. Just make that resistor a bit larger for easy replacement, anyone with a soldering iron can replace a 1206 package resistor, it is easier to replace a 1206 than even a through hole part.
Remember, ~26 dB gain is needed for those with sources of 2 volt output to reach the amplifier's rated full output. So 26 dB should be an option.
That's exactly how it's implemented now (1206 + 26 dB) since I had the same reasoning as you.
It's also why I am not yet firmly decided to have the jumpers.
 

kaka89

Active Member
Joined
Jun 24, 2018
Messages
257
Likes
197
That's exactly how it's implemented now (1206 + 26 dB) since I had the same reasoning as you.
It's also why I am not yet firmly decided to have the jumpers.
Please make this process as easy as possible if not jumper. I think this is the main selling point of this buffer
 

starfly

Active Member
Forum Donor
Joined
Jun 6, 2019
Messages
246
Likes
196
Assembled a voltage regulator:

20200701_204110.jpg
 

tlag

Member
Joined
Jun 1, 2020
Messages
25
Likes
19
An update.

Did yet another visual verification of wire connections, then plugged in and tested for DC at outputs: -4mV on one channel, +1mV on the other.

Still no front power switch, leaving rear IEC switch on, switching using external power strip. Refreshingly quiet power up and down with barely a perceptible click out of 96dB/Wm speakers.

Kind of agree with the routing comments, but not critical, IME. Twisted up the AC lines some. Left the DC lines, as there really isn't enough length to do much else with. Anyway, cored coils, especially closed-cores, don't have nearly as much stray field as air-core ones.

With no input, super quiet with ear right in front of efficient speakers. Sound is great, just a slight bit "etched," maybe, in comparison to past amps, but also apparently clearer. Bass somewhat drier, not necessarily better. Sighted, so utterly suspect impressions. But also speakON vs. problematic binding posts. Kinda wonder what REW says, and if the SHD/Dirac Live will actually call for retuning. Thinking damping factor here, and possibly the HF response March Audio mentioned, though differences, if any, seem lower in frequency.

Chassis barely warmer than room after many minutes of loud playing, which admittedly might've been a mere watt or two, average.

All in all, clearly the best amp I've ever had, and one of the easiest, cleanest DIYs ever. Highly recommended: Purifi Eval1 with Hypex SMPS1200A400 in pre-milled Modushop 230x280mm chassis with Schurter IEC input module and Neutrik speakON jacks.
 

starfly

Active Member
Forum Donor
Joined
Jun 6, 2019
Messages
246
Likes
196
I'm using one of the modushop cases for my build, and the bottom plate has vents along one side. I was thinking of having the vents in the front where I'll mount the power supply, thinking the vents can maybe help keep the power supply slightly cooler.

Good idea to do it that way? Or better to keep vents toward the rear?
 

Matias

Major Contributor
Forum Donor
Joined
Jan 1, 2019
Messages
2,653
Likes
4,385
Location
São Paulo, Brazil
I'm using one of the modushop cases for my build, and the bottom plate has vents along one side. I was thinking of having the vents in the front where I'll mount the power supply, thinking the vents can maybe help keep the power supply slightly cooler.

Good idea to do it that way? Or better to keep vents toward the rear?
They have top plates all covered with ventilation holes if you want. Just search their website for the size of your chassis.
 

starfly

Active Member
Forum Donor
Joined
Jun 6, 2019
Messages
246
Likes
196
They have top plates all covered with ventilation holes if you want. Just search their website for the size of your chassis.

Yeah, and I do have that for the top plate. But the bottom plate has vents in the front or back depending on how you install it. And I'm wondering if better to mount power supply on top of those vents in the front, or just have vents in the back.
 

tlag

Member
Joined
Jun 1, 2020
Messages
25
Likes
19
I'm using one of the modushop cases for my build, and the bottom plate has vents along one side. I was thinking of having the vents in the front where I'll mount the power supply, thinking the vents can maybe help keep the power supply slightly cooler.

Good idea to do it that way? Or better to keep vents toward the rear?
Using standard top/bottom plates, I placed the vent holes under the Eval1 in back, and over the SMPS1200 in front. Thinking it might help with some convection if things got warm.

If truly concerned about thermals, though, the best, easiest thing I see is to drill the front panel to bolt the SMPS to it. That thermal path is really short and sweet, over an order of magnitude better than going down to the bottom. Hypex even gives you a handy template you can easily print out in actual size. Just pull the existing screws and replace; it's all insulated.

If you don't want to drill the front panel, next best thing would be to use some sort of thermal adhesive (pad).

Providing only casual SMPS contact with the front plate, I see absolutely no thermal problem in my application, but it's gotta be just loafing most of the time.
 

starfly

Active Member
Forum Donor
Joined
Jun 6, 2019
Messages
246
Likes
196
Well, it's a real pain to drill the holes just right in aluminum. Even when using a hole punch, which I am, the drill bit still drifts a little each time. This meant initially that the power supply was sticking out over the edge, so I wouldn't be able to mount the base panel flush with the front panel. So I moved the power supply back a little by drilling new holes, but then now it's at a slight angle. Ugh. Oh well, it'll do as once the case is closed I won't see it. Making some progress so far finally.
20200705_135829.jpg
 

tlag

Member
Joined
Jun 1, 2020
Messages
25
Likes
19
Well, it's a real pain to drill the holes just right in aluminum. Even when using a hole punch, which I am, the drill bit still drifts a little each time. This meant initially that the power supply was sticking out over the edge, so I wouldn't be able to mount the base panel flush with the front panel. So I moved the power supply back a little by drilling new holes, but then now it's at a slight angle. Ugh. Oh well, it'll do as once the case is closed I won't see it. Making some progress so far finally.
View attachment 72022
Wow, eventual four-channel?

I hear you on the drill bit dance. Do you start with a sharp smaller bit? I find that helps some. Would suggest it's fine for most of the holes to be a bit oversized. In fact, better for the bottom ones, especially, so you can be sure to pull the SMPS to the front panel. Aluminum isn't so bad, actually, especially compared to stainless or spring steel. It's really the hard anodization, which the hole punch should mitigate. Also, a drill press (or even cheesy drill guide attachment) can help keep bit plumb in thicker materials.
 

starfly

Active Member
Forum Donor
Joined
Jun 6, 2019
Messages
246
Likes
196
Wow, eventual four-channel?

I hear you on the drill bit dance. Do you start with a sharp smaller bit? I find that helps some. Would suggest it's fine for most of the holes to be a bit oversized. In fact, better for the bottom ones, especially, so you can be sure to pull the SMPS to the front panel. Aluminum isn't so bad, actually, especially compared to stainless or spring steel. It's really the hard anodization, which the hole punch should mitigate. Also, a drill press (or even cheesy drill guide attachment) can help keep bit plumb in thicker materials.

Eventual 3 channel and otherwise plenty of space to not have to pack things too tightly together. Will also add a 12V supply and a couple of boards and screw down terminals so I need some extra space.

And yeah I did start drilling with a small bit first which does help a little. When I'm ready to add the 3rd channel I might just do a cad drawing of the bottom plate with all holes cut out.
 
Top Bottom