OP
- Joined
- Mar 7, 2020
- Messages
- 1,126
- Likes
- 1,869
- Thread Starter
- #141
I do! Two of them actually.
-
Welcome to ASR. 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 amp testing (with AES17 filter)
- Thread starter mcdn
- Start date
OP
- Joined
- Mar 7, 2020
- Messages
- 1,126
- Likes
- 1,869
- Thread Starter
- #143
Yep, still two left
OP
- Joined
- Mar 7, 2020
- Messages
- 1,126
- Likes
- 1,869
- Thread Starter
- #145
Send me a PM and I will ship for cost of postage.
Hi mcdn, do you happen to have any of those pcbs left?
- Joined
- Mar 6, 2016
- Messages
- 106
- Likes
- 16
Why aren't anyone making an active filter?
Sure the voltage should be attenuated to around 7Vrms.
Then the inductors of 1mH/220uH can be ditched, as they have a resonant point of 1MHz, 2MHz in best circumstances.
Use low noise opamps with GPW of >30MHz.
Also capacitors of 1 - 5% is easy to get. Those inductors are 20% tolerance, which makes it impossible to have two identical filters.
Just a question.....
Sure the voltage should be attenuated to around 7Vrms.
Then the inductors of 1mH/220uH can be ditched, as they have a resonant point of 1MHz, 2MHz in best circumstances.
Use low noise opamps with GPW of >30MHz.
Also capacitors of 1 - 5% is easy to get. Those inductors are 20% tolerance, which makes it impossible to have two identical filters.
Just a question.....
Hipocrates
Active Member
This is a effort mainly from enthusiasts and diy'ers with an AES17 filter. Design an active filter is probably out of the capabilities of many here and the scope of the thread.
That said, if you as a industry expert have a schematic to share that could lead the way.
OP
- Joined
- Mar 7, 2020
- Messages
- 1,126
- Likes
- 1,869
- Thread Starter
- #149
Quick answer - the passive option gives very low noise and acceptable distortion, and if an active option is wanted it's more easily done digitally.
Longer version:
1. The passive option is inherently low noise. Measurements show -150dB (see https://www.audiosciencereview.com/...-testing-with-aes17-filter.46684/post-1886411)
2. The passive option has stable performance. Once built, the passive device will always have the same characteristics, making it possible to compare measurements easily. Active options may perform differently with different power supplies for example.
3. For power amp measurements this can all be done digitally anyway with 768/24 ADC sampling rates and digital filters.
On the other hand an active version could potentially have lower distortion, but since the application is measuring power amps rather than headphone amps it's not really worth it for a DIY device to go better than -120dB THD.
- Joined
- Mar 6, 2016
- Messages
- 106
- Likes
- 16
ok, yes the low noise could be difficult to match in a active filter solution.
It is just an idea. Adjust values by a factor 10. Meaning inductors would be 100uH. I will give it a shot.
A slightly higher noise floor will be the result of it.
Delevan has inductors with 1% tolerance and Self Resonant Frequency (SRF) of 8MHz ... that is a factor 4 to 8 higher SRF.
1812R-104F available from digikey.
for 4.7nF in factor 10 => 470pF Kemet C1206C471FBGACTU will give 1% tolerance and 630VDC.
It is just an idea. Adjust values by a factor 10. Meaning inductors would be 100uH. I will give it a shot.
A slightly higher noise floor will be the result of it.
Delevan has inductors with 1% tolerance and Self Resonant Frequency (SRF) of 8MHz ... that is a factor 4 to 8 higher SRF.
1812R-104F available from digikey.
for 4.7nF in factor 10 => 470pF Kemet C1206C471FBGACTU will give 1% tolerance and 630VDC.
- Joined
- Mar 6, 2016
- Messages
- 106
- Likes
- 16
Hi mcdn.
I would not say i get a better result, but -3dB is moved to 60KHz. and suck out hits bottom at -60dB at 500KHz. 255KHz is -16dB.
It makes sense as the low pass frequency is moved up to 60KHz.
I will post it when i have cleaned it up and checked it an extra time.
BR Sonny
I would not say i get a better result, but -3dB is moved to 60KHz. and suck out hits bottom at -60dB at 500KHz. 255KHz is -16dB.
It makes sense as the low pass frequency is moved up to 60KHz.
I will post it when i have cleaned it up and checked it an extra time.
BR Sonny
Audioidiot
New Member
- Joined
- Jan 26, 2020
- Messages
- 2
- Likes
- 1
I have ordered a PCB for the Class D measurement filter based on the Audio Precision AUX-0025.
The LT-spice simulation showed a perfect response equal to the original.
Filter curve
zoom in of filter curve
The overshoot/passband ripple can be tweaked with R1 and R2
Zoom of the R1 and R2 influence. I'll probably go for the red curve
The boards are 130x80mm using 1% MKP caps 630V and 4 1mH (thin wire) air inductors from Jantzen. and is fully balanced
I will get 10 boards as that was almost the same price as the minimum 5 pieces, and shipping to the Netherlands was 4 times the price of the boards.
The mounting holes are connected to ground and both sides have a ground plane.
I will let you know how it works when it is build, and maybe some readers are interested in 2 boards to make this
The LT-spice simulation showed a perfect response equal to the original.
Filter curve
zoom in of filter curve
The overshoot/passband ripple can be tweaked with R1 and R2
Zoom of the R1 and R2 influence. I'll probably go for the red curve
The boards are 130x80mm using 1% MKP caps 630V and 4 1mH (thin wire) air inductors from Jantzen. and is fully balanced
I will get 10 boards as that was almost the same price as the minimum 5 pieces, and shipping to the Netherlands was 4 times the price of the boards.
The mounting holes are connected to ground and both sides have a ground plane.
I will let you know how it works when it is build, and maybe some readers are interested in 2 boards to make this
Similar threads
