steenoe said:
Sorry, you don't get my point. I just answered Jeff, please read that post above this one, no use taking up bandwidth to repeat myself.
Jan Didden
Kuei Yang Wang said:
Hi Thorsten,
Nice story, but has nothing to do with my worldview. You also didn't get my point. I never said cables make no difference, I never said snubbers make no difference. Why do you continue to read in other people's posts things they never said? May I be so bold to ask you to read my above answer to Jeff, not to take up etc.?
Jan Didden
carlosfm said:
Carlos,
My point is, and was, that because your mind was set up, you could only reach the conclusion you did reach, whatever the effect of your changes.
Jan Didden
CarlosFM,
On your regulated version, I think you ended up on the regulated with a 33uF cap at the amp chip after stopping there after reducing the size. I think you said it was a Philips cap of some sort. I believe you did not bypass this cap.
I will probably try that and the .47uF.
This discussion makes me think that something interesting I might try on the regulated (when I get my RIGC together enough - a few minutes of slow but steady progress every day!) will be a higher value cap such as 100-200uF but use a very low impedance series such as the Nichicon HE series (UHE). The idea would be to get increased "reservoir” near the chip but reduce the impedance below what the .33uF likely is.
If you did not find much benefit (I don't remember if it was discussed?) from bypassing the 33uF then perhaps the level of impedence on the 100-33uf is small enough anyway and the difference was really the reduction in capacitance. Still the low impedence 47uf-200uF cap seems interesting to try.
On your regulated version, I think you ended up on the regulated with a 33uF cap at the amp chip after stopping there after reducing the size. I think you said it was a Philips cap of some sort. I believe you did not bypass this cap.
I will probably try that and the .47uF.
This discussion makes me think that something interesting I might try on the regulated (when I get my RIGC together enough - a few minutes of slow but steady progress every day!) will be a higher value cap such as 100-200uF but use a very low impedance series such as the Nichicon HE series (UHE). The idea would be to get increased "reservoir” near the chip but reduce the impedance below what the .33uF likely is.
If you did not find much benefit (I don't remember if it was discussed?) from bypassing the 33uF then perhaps the level of impedence on the 100-33uf is small enough anyway and the difference was really the reduction in capacitance. Still the low impedence 47uf-200uF cap seems interesting to try.
Konnichiwa,
Who then wrote:
"That is exactly the way those multi-kilo dollar cable peddlers, brilliant pebbles, magic pens, whatever, etc make their killing. They insinuate that if you can't hear the enormous improvement with their stuff, you're one of these suckers with wooden ears. And who wants to be that, right?"
That was written clearly by someone who believes strongly that anyone who hears differences from "multi-kilo dollar cable, brilliant pebbles, magic pens, whatever, etc " is deluded, which inherently implies that there is no audible difference to start with.
Who wrote:
"- Audibility means that the movement of the speaker cone(s) changes so much in the middle of the complex movements resulting from the music reproduction that the change is audible;
- Thorsten's simulations show that the impedance of a power supply changes to a lower value upwards of 250kHz with those snubber things;
- Carlosfm maintains he can hear the differences introduced by those supply changes."
Again, while not saying directly "there is no audible difference" you imply there cannot be one such, as anything in the powersupplies at 250KHz could not possibly cause something to be audible with music.
You may not say "X is reliably inaudible" but what you write is aimed at insunating that it is and that anyone who nevertheless hears any difference is a deluded fool and/or Suckered Sucker...
Sayonara
janneman said:
janneman said:
Who then wrote:
"That is exactly the way those multi-kilo dollar cable peddlers, brilliant pebbles, magic pens, whatever, etc make their killing. They insinuate that if you can't hear the enormous improvement with their stuff, you're one of these suckers with wooden ears. And who wants to be that, right?"
That was written clearly by someone who believes strongly that anyone who hears differences from "multi-kilo dollar cable, brilliant pebbles, magic pens, whatever, etc " is deluded, which inherently implies that there is no audible difference to start with.
Who wrote:
"- Audibility means that the movement of the speaker cone(s) changes so much in the middle of the complex movements resulting from the music reproduction that the change is audible;
- Thorsten's simulations show that the impedance of a power supply changes to a lower value upwards of 250kHz with those snubber things;
- Carlosfm maintains he can hear the differences introduced by those supply changes."
Again, while not saying directly "there is no audible difference" you imply there cannot be one such, as anything in the powersupplies at 250KHz could not possibly cause something to be audible with music.
You may not say "X is reliably inaudible" but what you write is aimed at insunating that it is and that anyone who nevertheless hears any difference is a deluded fool and/or Suckered Sucker...
Sayonara
janneman said:
Jan, please...
I'd like to taste some of your wine.😀
You have something with listening tests, or is it your ears that are unreliable?
I told you: high capacitance sounded very bad to me, as much as I wanted it to sound good.
As much as I wanted, Dejan's recommendation of 1R + 560nf didn't sound good to me.
Am I biased to 1R + 100nf? Why? Can you explain?
Is this all in my head?
Are we all deaf?
The ones who tried it after me?
Am I such an inexperienced listener that I let myself fool so easily?
You don't know me.
Even then, I'd like to know you. I don't know why, but I respect you.🙄
Upupa Epops said:
I'm looking, and here's what I see:
1. Single bridge, for CT trafo.
2. 15~22,000uf 35~40v (strange for who recommends a "DC supply range from ± 20V DC to± 40V DC").
3. 3.15AF fuse (these won't allow parties, and forget 4 ohm speakers😀 )
4. 220 nf/100v caps
5. 100uf/100v + 1uf/63v caps
So... am I missing something here?
This is not enough to make the amp sound decent with these big caps.
janneman said:
How then can a DIYer ever make reliable judgements? A blind test doesn't change the fact that you are involved in the results.
Do you hire uninvolved strangers to select the gear in your system?
High, I am new to this forum, so please, be patient with me...
I just thought to join in to add some measurement results, which might
clarify the picture a bit better.
I really enjoyed Thorsten"s analysis, and I think, in these relatively simple situations the results can be good. Given the fact that I can have access to an old but good HP network analyzer, I tried to put together a small and simple test setup. (Actually, I have done similar tests already before, so I just used the same jig again)
So, the setup is:
I am using the HP network analyzer S-parameter test setup, and measuring S21, that is, the forward transmission coefficient.
I have two measurement ports: in case of S21, the Port 1 puts out a swept sinus signal with 50 ohm characteristic impedance, and on
Port 2, with 50 termination, the signal amplitude and phase is measured.
If one connects Port 1 to Port2 with a 50 ohm coax, on the screen there will appear a 0 dB attenuated straight line for the full bandwith tested. In this case, I chose 5kHz - 100MHz.
The test jig consist of two Lemo 50 ohm female connectors facing each other, soldered to a ground plane. A wire closes the circuit for signal "hot". That is, I am creating a breakout for the coax, so I can short the signal "hot" to ground, trying to attenuate it. Because
we have 50 char. impedance, in our case a 5 ohm short to ground will mean .. -14dB! (because, looking at the test point, it originally had 25 ohm impedance to ground..)
For a better catch, I'm including the pic of the test jig:
I just thought to join in to add some measurement results, which might
clarify the picture a bit better.
I really enjoyed Thorsten"s analysis, and I think, in these relatively simple situations the results can be good. Given the fact that I can have access to an old but good HP network analyzer, I tried to put together a small and simple test setup. (Actually, I have done similar tests already before, so I just used the same jig again)
So, the setup is:
I am using the HP network analyzer S-parameter test setup, and measuring S21, that is, the forward transmission coefficient.
I have two measurement ports: in case of S21, the Port 1 puts out a swept sinus signal with 50 ohm characteristic impedance, and on
Port 2, with 50 termination, the signal amplitude and phase is measured.
If one connects Port 1 to Port2 with a 50 ohm coax, on the screen there will appear a 0 dB attenuated straight line for the full bandwith tested. In this case, I chose 5kHz - 100MHz.
The test jig consist of two Lemo 50 ohm female connectors facing each other, soldered to a ground plane. A wire closes the circuit for signal "hot". That is, I am creating a breakout for the coax, so I can short the signal "hot" to ground, trying to attenuate it. Because
we have 50 char. impedance, in our case a 5 ohm short to ground will mean .. -14dB! (because, looking at the test point, it originally had 25 ohm impedance to ground..)
For a better catch, I'm including the pic of the test jig:
Attachments
ok, so I will try to continue with my results:
here is a pic of the calibration of the test jig - I've put 1 ohm to the ground, and this is the result on the screen:
Like in the example (5 ohm ) before, with reference to the original net impedance of 25 ohm, now we have - 28 dB attenuation.
The impedance is flat up to cca 30 Mhz, then starts to be inductive - no wonder, with such a simple jig..
here is a pic of the calibration of the test jig - I've put 1 ohm to the ground, and this is the result on the screen:
Like in the example (5 ohm ) before, with reference to the original net impedance of 25 ohm, now we have - 28 dB attenuation.
The impedance is flat up to cca 30 Mhz, then starts to be inductive - no wonder, with such a simple jig..
Attachments
And now, lets start with the real thing!
Thorsten was talking about cheap bad Jamicons, so let us have a look at a cheap bad Jamicon 10000 uF, 63v. (it's really big..)
There are two signal traces: magnitude of the attenuation, and the phase. ESR is ~ 20 mOhm, not so bad.
Given the fact, that we have a 3 dB corner frequency of ~ 130 - 140 kHz, with respect to the 20 mOhm Esr, the equivalent ESL seems to be around 24 nH. This coincides nicely with the findings of Jacka Racman.
Sorry, Thorsten!
Thorsten was talking about cheap bad Jamicons, so let us have a look at a cheap bad Jamicon 10000 uF, 63v. (it's really big..)
There are two signal traces: magnitude of the attenuation, and the phase. ESR is ~ 20 mOhm, not so bad.
Given the fact, that we have a 3 dB corner frequency of ~ 130 - 140 kHz, with respect to the 20 mOhm Esr, the equivalent ESL seems to be around 24 nH. This coincides nicely with the findings of Jacka Racman.
Sorry, Thorsten!
Attachments
here it is the previous arrangement, but now with a 680nF + 1ohm
added. sorry, I just can't call it "snubber"!
What is possible to see, is that it has not changed too much.
But! Don't forget, that up to now we measured an impossible bypass arrangement - 10000 uF directly on the pins!
What would happen, if we turn back to the real world, and have not much, lets say 2 inches of distance for the big cap from the pins, and the Wima right on the pins?
added. sorry, I just can't call it "snubber"!
What is possible to see, is that it has not changed too much.
But! Don't forget, that up to now we measured an impossible bypass arrangement - 10000 uF directly on the pins!
What would happen, if we turn back to the real world, and have not much, lets say 2 inches of distance for the big cap from the pins, and the Wima right on the pins?
Attachments
- Status
- Not open for further replies.