My Take on Threshold NS10

[apassgear]

Steen,

You may also try FET insted of the of the BT, say IRF 9610 and 610, who knows, it may help.

 

[Zen Mod]

Steen,

You may also try FET insted of the of the BT, say IRF 9610 and 610, who knows, it may help.

yup-but with recalc of base (grid ) resistor to maintain 10mA through zener chain and,also, minor change in zener values
who knows-thinking about greater transconductance ,in case of mosfets.....😀

 

[apassgear]

Steen,

If you are going to test the FET’s you should also add a grid stopper resistor at the gate just to ensure stability, and would suggest living the zener bypass cap also.😎

 

[Zen Mod]

first- 2200uF on output
sec-if needed-darlingtons plus one more 1N400x in zenner string
third-I'll power my brain.........

 

[Zen Mod]

just showing full page of smilies to my lil' one

 

[steenoe]

first- 2200uF on output

Thanks for the suggestions, I will try that first, and take it from there.
Here is a few more for the lil one🙂

Steen😎

 

[apassgear]

Choky,

I Have been looking to a lot of links (as you have) to shunt regs from the 12B4 line stage thread and seen a lot of interesting things there.

Me thinks that the problem lays on the ref voltage zeners which are causing the low level noise that is amplified on the NO common mode rejection shown by the NS10 circuit.

So instead of trying to overcome this by other means let’s try to better the voltage ref of the reg.

One way is using the TL431 as voltage ref, adding the said bypass (47u +0.1u film) plus a small grid stopper to the base, etc.

I’m sure you have other resources to test Choky 😎

 

[steenoe]

Me thinks that the problem lays on the ref voltage zeners which are causing the low level noise that is amplified on the NO common mode rejection shown by the NS10 circuit.

Tony, that sounds right to me, considering that the LM317/LM337 supply is very quiet. An adjustable precision voltage ref. might be the way to go.

Steen🙂

 

[Zen Mod]

hm....
just lookin' at tl431 datasheet...

for our shunt reg I can see now just two paths-one is to improve existing and other is to switch to solo tl431,even if I dislike more feedback in any sort of way.

first path-improving our reg ,in steps :
2200 or bigger cap on output is must
darlingtons-and one or two more 1N diodes in zenner string
test-is it better to decouple zenner string or base to gnd resistor?

also-more volts on input is better,but I think that we can settle with 10V difference,just for economy reasons (I will burn at least 20V and 50-70mA,anyway)

steen- you are my hope,just because you allready made working NS10 proto

all I can is to make psu proto with CRO testing,but I think that is easier and faster with ears 😉

let me know what are your plans,and mebbe I'll fire my solder at last

 

[Zen Mod]

just started breadbording-first look at just graetz with only 25VAC tells me that this option is nono
just too small margine for shunt type!!
take in account losses in bridge (even if shotky it is) etc

continuing with more volts on input.....

I'll let you know results in several hours

 

[Zen Mod]

ok,results:
start-we need stiff 35VDC MINIMUM at input

both series resistors are 100 ohms ,1W is OK

1000uF between them

zenner string -just two 12V/1W3 ,without 1N400x and WITHOUT caps across zenner string or base resistor

base resistor 100 ohms too, 1/4W OK

any darlington you have in drawer (2N6039 and cousin in mine )

2200uF on output
even better if you have place for some 100.000uF in box

current summ for 35V input is 45mA ,approx 10mA through zenners and 10mA through load (2K5); rest through shunt Xistor

better looking and measuring at 40V input,current is also bigger through Xistor -only;


all I can see on scope is just breadbord induced (read mess implementation) noise,but I can't see any residual hum or anything simillar

all caps can be bridged with some ceramics (22n good enough)




uh-yes-25V7 on output

read text looking at this picture

 

[Zen Mod]

and-steen-you don't need heatsinks on transistors at all ,at least to 45VDC input

 

[steenoe]

let me know what are your plans,and mebbe I'll fire my solder at last

I didnt have any plans, until you posted the next proposal🙂 I dont even have a darlington in my inventory (I think) I will try it, but it will propably be a few days before its finished, when I need to order parts🙂 Thanks for your work on this, Choky😉

Steen🙂

Edit, I still dont quite understand the other shunt reg has that slight buzz. Can it be a defective rectifier diode or something??
I will not use any of it, when I build a new one

 

[apassgear]

Yes, this happens Steen, some zeners are more noisy than others

 

[Zen Mod]

I didnt have any plans, until you posted the next proposal🙂 I dont even have a darlington in my inventory (I think) I will try it, but it will propably be a few days before its finished, when I need to order parts🙂 Thanks for your work on this, Choky😉

just try without darlingtons (plain BDs ) and with that 1n400x in place
have you 'scope?
[/B][/QUOTE]

Steen🙂

Edit, I still dont quite understand the other shunt reg has that slight buzz. Can it be a defective rectifier diode or something??
I will not use any of it, when I build a new one [/B][/QUOTE]

can be ,but you can only try to swap something,or see with 'scope....

I didn't tried that shunt reg on anything this evening,but I'll maybe tomorrow
hehe-mebbe I even make one quickie Choky Mono NS10 Clone just for testing of psu

 

[Zen Mod]

hehe I make a mess with bloody quoting......

 

[steenoe]

have you 'scope?

No, and this is not the first time I need one I better get me one and learn how to use it🙂 Just so you know, I am building a whole new shunt reg PSU, to see if we can rule out bad components. I would prefer to etch a couple of Russ's psu boards for the new ones🙂 But I dont think they will accomodate the 2200uF's at the output. We will get this thing right

Steen🙂

Edit, I just checked the psu board within the edit time😉 If the pads for the outputcaps are moved a bit away from the BD's, I think its good🙂 If you find the time to post a pdf Russ, I will give it a shot😎

 

[Zen Mod]

you can also put diodes on output-in series with + and -

try it first-then make pcbs

 

[steenoe]

Is this the darlington reg you had in mind, Choky? I fiddled around in Eagle to see if I could clean up the schematic a little bit🙂 This is without any 1n4xxx's.

Steen🙂

 

[Zen Mod]

Is this the darlington reg you had in mind, Choky? I fiddled around in Eagle to see if I could clean up the schematic a little bit🙂 This is without any 1n4xxx's.

Steen🙂

yup.
😀

put 22n in para with all 'lytics