Power Supply Freds, Hexfreds, Ultrasofts, Ultrafasts and Fast Recovery

Status
Not open for further replies.
Jump to Latest
#63
till said:
why can´t your camera make pictures of scope traces? mine el cheapo 50€ camera does. I would like to see them. What happens with a bridge made with schottky diodes?
Click to expand...

My digital camera is one of those matchbox-sized so called
spy cams and can't take close ups and require more light.
My Nikon SLR most probably could do the job, but then I
have to use ordinary film. Anyway, the traces don't look all
that interesting and are quite shaky since my scope doesn't
have such fancy features as dual time bases. I think the
figures I posted pretty much sum up the results. I will do
some experiments with other diodes, but probably not tonight
(but I just brought them out of the box, so ... 🙂 ).
 
#64
Magura said:
I may be mistaken, but for what I know, the undershoot is part of the excuse for implementing inductors in a PSU????

Magura🙂
Click to expand...

You mean LC or CLC filters? Yes, that can reduce both ripple
and undershoot and stuff, but it is not a general cure. It
depends on your application. I might work well for loads that
are more or less constant, like class A amps, but can cause
problems for more varying loads. As for undershoot and
ringing, better cure it at the source than by heavy, bulky,
$$$$$$$ inductors. 🙂
 
#65
I agree to some extend, but since the CLC filter is a must anyway, its an easy solution. A monster trafo would also reduce the problem to some extend i guess, so a combo of those would help a lot. I though think your experiments are a good approach to eliminate the problem totally. I for one have no interest in anything that isnt class A, so I havnt seen what kind of problems it causes for changing loads.
I made a few measurements on a 500VA trafo with a CLC filter of 68.000uF 2mH 68.000uF and ultrafst rectifiers and could not measure the effect of the undershoot. I know it must have been there, but the filter solved the problem. The measurements vere made at 25V and 3A.

Magura🙂
 
#66
I think the initial point was HF caused by the diodes cummutation. I think it comes from the diodes PN-capacitance. Schottkys do not or in much lower level show this, those soft recovery types show it reduced. The HF may cause IMD when an active component does act as a mixer with the audio signal.
 
#67
till said:
I think the initial point was HF caused by the diodes cummutation. I think it comes from the diodes PN-capacitance. Schottkys do not or in much lower level show this, those soft recovery types show it reduced. The HF may cause IMD when an active component does act as a mixer with the audio signal.
Click to expand...

Yes, that sounds about how I have understood it too. As far as I
remember Schottkys have no reverse current at all and
should be the most ideal diodes, but they usually are limited
in voltage and current handling capacity.
 
#68
Hi,

As far as I remember Schottkys have no reverse current at all and should be the most ideal diodes, but they usually are limited in voltage and current handling capacity.
Click to expand...

Not for much longer they aren't : haven't heard of Sic diodes by Cree? High voltage and high current Schottkys.

Meanwhile and much cheaper anyway, use a proper snubber to rid yourself of the diode blues.

Having a good scope, preferably into the 20MHZ range helps too.

Anyway this has been discussed on the forum extensively and is generally well documented.

Here's a good source of information:

JIM HAGERMAN

See: "Calculating Optimum Snubbers".

Ciao,😉
 
#70
What about MBR10100 and MBR20100, i use them in some power supplys now. You can get them at onsemi.com , not expensive. For smaller power supplys i used 1N5822 3A 40V or quad schottky array UC3611 3A 50V in DIP8.

For Ultrafast Recovery fans i have a load of MUR860 and others i would be willing to trade some.

My 25Mhz 7313 scope is not enough for this, as posted. excellent 100 or 200MHz equipment can be bought for small budget at ebay.
 
#72
hi Frank,

No, I have missed the Cree diodes. Anyway, Schottkys are
often sufficient for the applications that ought to be the
most sensitive ones, like phono and pre amps.

Yes, I know the Hagerman article, which is excellent reading.
The main purpose of my measurements was not so much to
empirically finnd the right snubber values, but rather to see
if I could measure the phenomenons at all with the equipment
I have. I agree with till, a better scope is probably needed
and that is not enough per se either. I had to use the HP
filter to get any sensible measurements at all. There has been
a lot of discussion about various diodes and their properties
but very little actual measurements reported.

Some people have complained that I mostly use theoretical
arguments and present Spice simulations. Are they to complain
that I do practical things too now? 🙂
 
#74
Some people have complained that I mostly use theoretical
arguments and present Spice simulations. Are they to complain
that I do practical things too now?
Click to expand...

some diyaudio rules:

- no matter what you do or post, there will be someone who does complain and tell you are wrong.

- If you ask a question you may get
a) no response posting and no answer
b) a lot response and no answer
c) multiple answers telling te opposite and telling the other one who answered is wrong
d) the response you are wrong, no answer
e) paint it green will sound better but pen must be 50$ minimum.

- If you ask for measurements, proofs, reason - or worse start to measure, people will tell you are a heretic and definitely no audiophile.

- the most expensive resistor, cable or cap sounds best.

- Do not diskuss aktive components influence on sound, audiophiles don´t do so. Better speak about laying stones under CDP, Photos into fridge or blue LEDs.
 
#77
Folks, try to look at the CURRENT waveform, rather than the voltage waveform. You need a current probe, current tranformer, or something special, BUT you should see more than what you are seeing. Hi speed, soft recovery diodes are necessary for any of my hi end products.
 
#78
Question about schottky

Here I can buy 45V schottky diode. Can't buy higher voltage than that.
What is the maximum symmetrical supply can be built safely with that 45V schottky? The arrangement is full bridge 4 diode, and the transformer is CT. Will it get to +/-45Vdc or it can only make +/-22.5Vdc?
What is the general rule of thumb for selecting proper diode voltage for certain needed rail voltage?
 
#79
hi,

Here I can buy 45V schottky diode. Can't buy higher voltage than that.
What is the maximum symmetrical supply can be built safely with that 45V schottky? The arrangement is full bridge 4 diode, and the transformer is CT. Will it get to +/-45Vdc or it can only make +/-22.5Vdc?
What is the general rule of thumb for selecting proper diode voltage for certain needed rail voltage?
Click to expand...

try mbr2020, it's rated for 200volts piv...

for a 45V schottky diode, i'd use no more than 18-o-18 volts ac or about +/-25vdc rails just for my peace of mind...

tony
 
Status
Not open for further replies.
Top Bottom