My take on a discrete shunt voltage regulator

[dodo]

Originally Posted by ikoflexer

Hi Telstar,



I would like to clarify something. These three issues are all different:
1) test the design/circuit
2) test the pcb layout
3) test the manufactured pcb

1) This tries to establish whether the circuit itself works up to the desired specifications. I know it's too much to read, tons of pages, but we have been doing a lot towards this goal. Not only simulations, but built several actual prototypes. Measured noise and line rejection to the best of my ability. Compared with other regulators that I built (other designs). Stormsonic has also done some tests. Hifinutnut also built the design and has done some of his tests. I would have loved to send a built regulator to syn08 or Jan Didden to test, because these two guys have both the tools and the skills to run proper tests. syn08 has me on his ignore list. Jan Didden has shown absolutely no interest in this regulator. In any case, the design has been tested for stability by more than one person, has also been tested for normal functionality, and measured to some extent. The only thing we lack are state of the art measurements.

2) I have etched pcbs myself, manually, to test the layout design. Built several, they all worked.

3) The purpose of this test is only to check that the pcb manufacturer has done a good job. I do understand though that many members would feel much better to see the design tested again at this point. But I expect no design changes at this point.

If people have trust issues with the design then we should not do a group buy. I'm serious. We should stop the GB or delay it, and wait until people are happy with whatever measurements and tests.





No, sorry, you're not wrong. I misunderstood what keantoken said. At some point he said that his scope has internal noise of 10uV, and that would not compete with my 100uV/cm scope. However, his 3a9 diff amp module has 10uV/div as the most sensitive setting, and that is better than my 100uV/cm. This being said, I truly do not believe that many of us will achieve lower noise than 20-30uV peak-to-peak ripple. It's just my opinion after doing a lot of these and my own measurements.

Iko

I understand you point of view , but please take a note that everyone from us belive you made an excellent job.
So I would rather think that way

We (members ot this thread) support you with the statement , that the final regulator ( circiut and layout) shows excelent quality / price ratio,
( taking into the account lack of any exotic parts used )

Cosidering above we (and specially you iko as constructor) challenge some EE to show us ( with the support of the tests) that we could be wrong in our belive.
As always where there is a challenge , it should be some arward.
In this case the additional fee calculated per board would build some sort of arward budget

[merlin el mago]

Hi qusp, thks for info

[merlin el mago]

Hi Iko,

Please let me ask to a friend capable to do very good measurements, he is working all day with mechanical robots for medical labs, he has all we needs but first to offer I have to ask him if can do me the help.

[ikoflexer]

Thanks guys! Any help is appreciated.

I'll be gone now for a while, be back later.

[qusp]

no problem Merlin, as you can see its some pretty serious PCB action going on there; i'll let you know how the I/V stage board goes, will be doing it in 3-4 weeks

ikoflexer, no problem, just say the word and i'll PP you

[Telstar]

Originally Posted by dodo

I understand you point of view , but please take a note that everyone from us belive you made an excellent job.
So I would rather think that way

We (members ot this thread) support you with the statement , that the final regulator ( circiut and layout) shows excelent quality / price ratio,
( taking into the account lack of any exotic parts used )

I second this statement.

[dvb-projekt]

Would´t it be a good idea to integrate the mountings for a SK104 heatsink?

For those who want to place the pcb not at a side of a enclose, this makes it much much easier.

[syn08]

Originally Posted by merlin el mago

Thks for advice

Could you test the Iko regs. please?

Out of curiosity I already did, in simulation and practice. They are both a living disaster in terms of performance and stability.

But apparently they "sound good", so who am I to debate this "design".

[Bas Horneman]

Out of curiosity I already did, in simulation and practice.

Then why not show the results of your measurements?

[Thomas_2-love]

Originally Posted by syn08

Out of curiosity I already did, in simulation and practice.

Great! How did you setup the practice part?

Originally Posted by syn08

They are both a living disaster in terms of performance and stability.

Could you specify that? Specially the performance?

Originally Posted by syn08

But apparently they "sound good", so who am I to debate this "design".

Well, then don't. But could you give us some figures?