Carlos' snubberized Gainclone Power supply

Status
Not open for further replies.
Jump to Latest
#581
Jan:

Point well taken. I enjoy the Dilbert cartoons in EDN as they relate to these engineering pecularities.

I recently dug out my 1964 ARRL Handbook and note that caps -- 0.01uF were placed across the diodes in the 811 - 200 watt RF amplifier I built sometime around 1967 -- old habits die hard.
 
#582
I knew that I will not get away with this!

While writing, right then it came to my mind that it's not completely true. But I did not edit the post, because I did not want to get OT even more..

I think I'm completely abusing that article here, but would like to put another illustration. It's the case when cap's across the diodes can actually improve the situation..
For italians, it's self explaining. For the rest, it says, that dividing that one Ca into two, [twice the value], and connecting them to circuit ground, will greatly enhance CMMR for the common mode noise introduced by the line.
And in fact, this is a trick that I learned hard at the age of 19, and since then respecting it..
But note, that this still differs from simply putting caps across all the diodes.
 

Attachments

  • img-6739r.gif
    img-6739r.gif
    13.7 KB · Views: 1,536
#584
Carlos, I supposed that.. When I was in your beautiful country [Lisbona] I could get away quite well with the things written.. but spoken.. :bigeyes:

So, basically it is about the inrush currents at startup. He is a tube guy, using 1N4007, [which is just a joke for a real man..:bawling: ].
He suggests to temporarily insert a small value current sense resistor at the signalled position, and monitor the voltage accross it during startup, that is, the peak of the inrush current developing. That's all.
Anyway, the figure shows well where to put that 0.1 - 0.05 ohm series resistor that I was talking about earlier on, so as to be able to monitor
the current pulses recharging the filter cap.

ciao, george
 
#585
Joseph K said:
...it says that dividing that one Ca into two, [twice the value], and connecting them to circuit ground, will greatly enhance CMMR for the common mode noise introduced by the line......But note, that this still differs from simply putting caps across all the diodes.
Click to expand...

Now this I can understand - great idea...the ringing is damped out and much of the remaining noise sunk to earth.

Well the next PSU that I built is going to be quite an experiment.

thanks

mike
 
#593
Implementation considerations

As the larger capacitors with snubber seem a good idea I decided to try myself.

I found nice BC components capacitors of 4700uF each and they would fit on the rev A amplifier boards. The idea would be to use 4 capacitors to replace the 1500uF panasonics and have the snubber and the 100nF cap as wel as the 100uF caps also on the ampliefier boards. With this setup the complete configuration would be close to the chips except for the bleeder resistors of 2.2k which would be mounted on the PSU board.

This means that I would have 4 snubbers (2 on each amp board)

I'm an ME an not an EE and connot oversee the exact consequences but this seems as a good idea.

Let me have your thoughts as whether this is a good or a bad idea. (I'm already soldering the stuff together but would like some blessing of the knowledgable people around here)
 
#596
if carlosfm is listening --

i ran some distortion tests with a regulated and non-regulated GC power supplies using the lm4780 in bridged configuration -- the low end distortion was about 0.001% better (in the range of 10 to 23 Hz, the high end (above 5600 Hz) was about 0.001% worse.
 
#599
jackinnj said:
if carlosfm is listening --

i ran some distortion tests with a regulated and non-regulated GC power supplies using the lm4780 in bridged configuration -- the low end distortion was about 0.001% better (in the range of 10 to 23 Hz, the high end (above 5600 Hz) was about 0.001% worse.
Click to expand...

So which was which? 😀
 
#600
jackinnj said:
if carlosfm is listening --
Click to expand...

No, I'm just reading.🙂

jackinnj said:
i ran some distortion tests with a regulated and non-regulated GC power supplies using the lm4780 in bridged configuration --
Click to expand...

This sentence is not very clear to me...
Can you go into more details, please?
What regulated PSU?
What unregulated PSU?

jackinnj said:
the low end distortion was about 0.001% better (in the range of 10 to 23 Hz, the high end (above 5600 Hz) was about 0.001% worse.
Click to expand...

Are you talking about the unregulated PSU?😕
 
Status
Not open for further replies.
Top Bottom