Hi Andrew,
I'm referrring to Mauro's 'Project My_Ref' document:
I condensatori C5 e C22 servono a smorzare i picchi di tensione generati dai diodi, in combinazione con l' induttanza parassita del trasformatore.
Il filtro che si viene a creare (a frequenze di circa 15-30 Khz a seconda del tipo di trasformatore) riduce notevolmente il rumore di rettificazione ad alta frequenza.
English translation:
C5 and C22 capacitors are used to smooth out voltage spikes generated by bridge rectifiers, combined with the parasitic inductance of the transformer.
The resulting filter (at frequencies of about 15-30 kHz, depending on the type of transformer) greatly reduces the high frequency noise.
Can you elaborate?
Maybe it's the term 'ringing' that isn't correct/precise?
Last edited:
Mauro's description would apply to R+C snubbers correctly applied to the impulsive signals around the rectifiers.
A capacitor alone is not usually suitable for damping ringing/resonances.
It appears from Forum posts and a bit of my own experience that a low loss capacitor is more likely to induce ringing and resonant behaviour. This is the opposite to what you claim (damping).
However, while it appears to me that you are revealing that removing C5 did not make the sound worse, I find your language very confusing. I find your reference to AC input caps and optional AC caps to be completely uninformative & misleading.
No wonder Soongsc had to seek clarification.
A capacitor alone is not usually suitable for damping ringing/resonances.
It appears from Forum posts and a bit of my own experience that a low loss capacitor is more likely to induce ringing and resonant behaviour. This is the opposite to what you claim (damping).
However, while it appears to me that you are revealing that removing C5 did not make the sound worse, I find your language very confusing. I find your reference to AC input caps and optional AC caps to be completely uninformative & misleading.
No wonder Soongsc had to seek clarification.
Last edited:
I do recall there being a paper on snubberizing the diodes. The paper also tried cap across the rectifying diodes, and the performance was not as good. It was not used across the rectifiers like here.
This is a similar paper.
http://www.hagtech.com/pdf/snubber.pdf
I have not put in snubbers for MyRef yet.
I would expect the cap C5 for some power factor adjustments.
This is a similar paper.
http://www.hagtech.com/pdf/snubber.pdf
I have not put in snubbers for MyRef yet.
I would expect the cap C5 for some power factor adjustments.
Last edited:
I hear quite a few members referring to Power Factor.
Can you explain the mechanism that C5 would play in our comment?
Actually I think that C5 removal is beneficial (in the My_Ref_FE context)
It's true that I could be more clear and I've understood that such terminology was not appropriate, in fact I clarified without problems.
But the wrong terminology doesn't change the facts, or the informative contest of such posts...
I can't understand why you had to point it out...Why in most of your posts I feel a sort of 'negative' energy?
I have not pursued this before, but if the transformer secondary has inductive characteristics, using C5 to tune a resonant circuit to the frequency of the mains power should give a better transfer of power where the current and voltage are in phase with each other.
Does this make sense? I'm really trying to apply what we learned in school over 30 years ago.
But since I have moved to existing switching power modules, I may not get a chance to play around with these. Having the value that came with the kit probably is just causing resonance at a wrong frequency effecting proper power transfer.because I read your posts and find my reaction is a shaking of the head. Does that equal -ve? That is not what you said. What you said wasActually I think that C5 removal is beneficial (in the My_Ref_FE context)...Why in most of your posts I feel a sort of 'negative' energy?
I have trouble understanding some of your conclusions. I have trouble understanding some of the logic leading to those conclusions. I have trouble just trying to understand what you are typing.
Sometimes I have trouble simply believing what you type.
All of this is probably down to me misunderstanding what you type, combined with me not understanding the science behind your experiments.
Yes, it leaves me feeling quite -ve. But, I keep reading hoping that some day I will reach the "I see" moment.
Last edited:
Well, I just changed the MyRef C channel so that the LM3886 pin 7 connects to power GND. This seems more like what I am used to. But the details are somewhat less, not as much emotion in the music. I may get some time to do some measurements.
Last edited:
Pity that you left out the most important part...
Andrew,
I respect your opinions and knowledge and I also think it would be more useful if you point out the faults that you see in my posts and argument your point of view.
What you do, instead, is a sterile constant polemic banter.
If I'm doing something wrong I'd be happy to learn something but a more positive attitude from you is necessary...
Did you remove also C5?
It could give you back the lost emotion.
If you manage to measure the difference it would be great.
C5 is on the board, but It's not connected to anything because my switching modules provide DC, and I don't have the rectifiers on board. I think the problem is the input cap C13. It sounds very typical like what a normal X-cap would be like even with the bypass cap I use. I just don't have room for larger cap. Would still prefer removing the cap, which I may try at some point in time.
Yes, it should be an Arcotronics/Kemet, a very bad cap.
A cheap Mundorf white MCAP is way better and mounts perfectly (see attachment).
BTW you can DC-couple for testing purposes...
In fact my first six months of tests has been with a DC-coupled My_Ref.
Apart the lack of input DC-protection the only audible caveat is that you can't judge the real bass performance.
Regarding the beta test if we remain 3 we could proceed anyway at the price of a slight increase (36€) ordering the PCBs from China.
Any thought on it?
Attachments
The measurement thing is a nice plus but it's not required.
So at the moment:
- Me
- Soongsc
- BmcBob
- .
- .
- .
One of the biggest difference will be soundstage, I think is hard to appreciate the change without two PCBs.
IMHO, the best thing is to buy the two PCBs since they will be manifactured with same specs as definitive ones:
Well, maybe the definitive ones will have chemical gold finish instead of HASL if the price is reasonable...
- 2mm FR4
- 2 Oz/70u copper
- Blue solder-mask
- Serigraphy on both sides
- ROHS HASL finish
Anyway with two boards, if everything works as intended, you'll have a complete working amp.
It will be between the last week of December and the first of January since the PCB factory is closed for Christmas Holidays till Juanuary the 8th.
Dipped in Sapho ?
Wish I could participate but I don't have a clean source yet; don't have a DAC. Don't have the technical knowledge to measure and assess. I'm just a fan of your work.
I hope you can post the BOM soon so I can get an idea of cost and sourcing.
that does not help me.
It is discussing power transfer and using this method for a transformer limits the maximum efficiency of said transformer to <50%.
Most transformer operate at upwards of 90% efficiency. Maximal power transfer does not apply here.
Power factor correction seems to me to be something else.
Why not?
Exact dimensions are 114.3 x 100.3 mm
2 mm FR4
0.7um copper
blue soldermask on both sides
silkscreen on both sides
with and without electric test
For 6 PCBs and 12 PCBs
Thanks
Sorry, I don't understand...
Hopefully it will be posted this night.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.