Pass DIY Addict
Joined 2000
Paid Member
Hi Matt - your output curves look quite nice! If you really can't resist the urge to tweak, decrease the value of your output cap by a few pF and you wave form should look a little more square. Have a look at the "output test" section on my Aleph-x (sig link) build page for some examples of square wave output with different cap values.
My novice understanding is that the closer the leading edge is to a pure vertical line then the "faster" the amp responds to inputs, which determines the high frequency response. The slope of the top of the square wave somehow correlates to low frequency response, I wonder if it is similar to filtering in a power supply in that the the downward slope indicates the amp does not have the energy storage capacity to maintain the target voltage over time. If so it makes sense that that would effect lower frequencies because they have a longer wavelength, but just theorizing.
When I took the measurements I was lazy and left the probes in 1:1 mode because I wasn't dealing with high frequencies. Running off the scope's 0.5Vpp calibration signal the waveform is spot on.
For fun I just set them in 10:1 mode and checked the calibration, it was close but needed a hair of tweaking. Since the compensation capacitor is only used in 10:1 mode this should not effect my original measurements.
Thanks for the suggestion, next time I'm feeling like tweaking I might have to make some adjustments.
- Matt
Pass DIY Addict
Joined 2000
Paid Member
I believe you are correct, Matt. Larger amounts of capacitance on the output lead to high frequency roll off that is easy to spot in the output waveforms. Here is a comparison of output caps I made for my Aleph-X amps. With no cap at all, you can see overshoot on the leading edge of the wave form that many characterize as treble brightness or harshness that leads to listening fatigue. As capacitance increases, the leading edge of the wave form rounds over which is often described as treble smoothness.
An externally hosted image should be here but it was not working when we last tested it.
Eric,
That is a really interesting image, thanks for sharing it. I'm assuming you are currently running the 3pF cap in your Aleph-Xs.
Based on your testing it looks like I could see a slightly improved square wave response if I drop C4 down slightly from 10pF, say to 7pF. At this point I think I'll leave the 10pF in place, but in a future parts order I may throw in a smaller cap just to try it out. I wonder if the difference is even acoustically noticeable...
I am a little concerned that with a 10kHz sine wave I start to see waveform deformation at around 10Vpp on the output, this seems really low since it maps to only about 2w. My rails are about +/- 15v so I would think they could support a bit more output.
- Matt
That is a really interesting image, thanks for sharing it. I'm assuming you are currently running the 3pF cap in your Aleph-Xs.
Based on your testing it looks like I could see a slightly improved square wave response if I drop C4 down slightly from 10pF, say to 7pF. At this point I think I'll leave the 10pF in place, but in a future parts order I may throw in a smaller cap just to try it out. I wonder if the difference is even acoustically noticeable...
I am a little concerned that with a 10kHz sine wave I start to see waveform deformation at around 10Vpp on the output, this seems really low since it maps to only about 2w. My rails are about +/- 15v so I would think they could support a bit more output.
- Matt
Pass DIY Addict
Joined 2000
Paid Member
Hi Matt,
Yes, I'm currently using the 3pF cap. The tradeoff is using caps also creates HF rolloff which is apparent in bandwidth measurements of the amp. With no cap at all, I get bandwidth near 240kHz. With the 3pF cap, I have bandwidth of about 140kHz.
The extent to which this is noticeable is another topic. Many have argued that square wave testing is not valid because we don't spend our time listening to square wave... I just view it as a way to tweak things to make them as good as I know how.
I'm not sure what to tell you about the output of the mini-A. I do know that one of the characteristics of the balanced-X amps is the cancellation of distortion inherent in the output mosfets. The left side shows the output of the positive speaker terminal to ground and the negative speaker terminal to ground, each separately. The right side of the image shows the summed waveform across both speaker outputs - what the speaker itself sees. Does the output you are talking about look like either half of the curve on the left side of the image?
Yes, I'm currently using the 3pF cap. The tradeoff is using caps also creates HF rolloff which is apparent in bandwidth measurements of the amp. With no cap at all, I get bandwidth near 240kHz. With the 3pF cap, I have bandwidth of about 140kHz.
The extent to which this is noticeable is another topic. Many have argued that square wave testing is not valid because we don't spend our time listening to square wave... I just view it as a way to tweak things to make them as good as I know how.
I'm not sure what to tell you about the output of the mini-A. I do know that one of the characteristics of the balanced-X amps is the cancellation of distortion inherent in the output mosfets. The left side shows the output of the positive speaker terminal to ground and the negative speaker terminal to ground, each separately. The right side of the image shows the summed waveform across both speaker outputs - what the speaker itself sees. Does the output you are talking about look like either half of the curve on the left side of the image?
Last edited:
The output of the Mini-A looks pretty much the same as what you show on the right. With this design the negative speaker terminal is at ground potential.
I did some more testing this evening and with a 1 kHz sine wave as input and driving an 8 ohm load clipping starts around 24 Vpp, which is approximately 9w. This is pretty much spot on with the 8ohm rating according to the "aleph-power" spreadsheet using my current bias level of 1.145 A.
Since my main speakers are 4 ohm I may try to get the bias up to 1.4 A which works out to 16 W into 4 ohm and 9 W into 8 ohm. But for now I'm just going to listen to it. A while ago I used this amp to drive a set of VMPS Super Tower/Rs and was amazed at how well it performed with those huge speakers at reasonably high listening levels.
- Matt
I did some more testing this evening and with a 1 kHz sine wave as input and driving an 8 ohm load clipping starts around 24 Vpp, which is approximately 9w. This is pretty much spot on with the 8ohm rating according to the "aleph-power" spreadsheet using my current bias level of 1.145 A.
Since my main speakers are 4 ohm I may try to get the bias up to 1.4 A which works out to 16 W into 4 ohm and 9 W into 8 ohm. But for now I'm just going to listen to it.
A while ago I used this amp to drive a set of VMPS Super Tower/Rs and was amazed at how well it performed with those huge speakers at reasonably high listening levels. - Matt
Andersonix, that's a good suggestion. Right now it has to draw power over the umbilical during transients, it might be better if it has storage capacity right next to the amps. Space is tight, but I might be able to fit a some caps in there.
Thanks for the suggestion.
- Matt
Dear colleagues, could someone tell me what are the capacitors that could make a difference in the sound of the mini aleph? The capacitors 220uf would be critical of its quality? The audio goes through some electrolytic capacitor? I thank you for the information ..... Happy new year to all!!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.