Shunt Voltage Reference LM431

The LM431 wants a few mA to stabilize, at 100V you'd be dissipating a lot of power if you used that design. The old IC regulators, like LM305 and uA723, can be configured with external components to do this kind of thing. They need 10V or so minimum for their internal power, you might want to run 'em from low voltage power and just put the output device on the 100V.

Précis d'electronique wrote in news:37a921c4-ce51- snipped-for-privacy@q2g2000yqq.googlegroups.com:

Go to the NCP1200 product folder.

Look under Technical Information, click on Design & Development Tools.

You will find excel based spreadsheets for calculating all components for a basic flyback called "Flyback Current Mode Loop Compensation Design Worksheet". They also have spice models with transient and AC templates.

Why do you need a 100Vf refrence?

- a

Hi, thank you

My output is 90v and 0.100A, I already calculated R5,R6,R4 and I have the transformer but I get 5 volts in the output (in a squared signal) and I cant find a LM431 for my reference.

Précis d'electronique wrote in news: snipped-for-privacy@i31g2000yqm.googlegroups.com:

Where did you get the transformer? You know you cant use just any old transformer and that a flyback uses coupled inductors for energy transfer.

You can just use a zener/opto to feedback like this.

You do know that you need a rectifier on the secondary with a capacitor?

You could also regulate the Vcc winding and just have your output voltage regulated by the turns ratio.This woukdnt require an opto coupler but regulation wouldnt be tight depends on your requirments.

snipped-for-privacy@i31g2000yqm.googlegroups.com:

.

HI, Wurth Elektronik gave me the transformer (custom) for 90v

what rectifier? the D2 on the schematic?

Could you help me to calculate the values for the opto/zener feedback? that would be a 75 volt zener, a 1k resistor in parallel to the LED , or to the Rled too?

thank you

OA2 or VR-105? You can google for them.

--

Tauno Voipio, OH2UG

innews: snipped-for-privacy@i31g2000yqm.googlegroups.com:

Not sure, but I think you're trying to do this:

---[L2]-----+---> +90V | | \__|__ [R5] / \ \ | /___\--------+ | | | [R6] | | +----------+ | [1N5374] 75V, 5W Zener diode | Gnd ----+

You compute the R5 and R6 values as you did before, except that, with the zener added, you base the computation on Vout - Vzener. With Vzener = 75, then you base on 15 volts instead of 90 (Vout). Note that regulation will not be as crisp because of the addition of the 75V zener - but without it the TL431 max V is exceeded. The datasheet for a 5W 75V zener shows a range from 71.25 to 78.75 volts, at 20 mA. That means your output voltage could vary +/- 3.75 volts from nominal. A different current would result in a different (likely wider) range. The the 20 mA test current on the datasheet, the zener would be dissipating 1.5 watts.

It may be that the advantages of the 431 are swamped by the "mushyness" of the zener - depends on what your output voltage requirements are.

I don't know what you are referring to when you mention "a 1k resistor in parallel to the LED , or to the Rled too? "

Ed

snipped-for-privacy@i31g2000yqm.googlegroups.com:

er.

?

ge

thank you, I make the desgin in PIExpert for 100 output and gives me no error, just the same circuit, so I am more confused about that, take a look

and that is the resistor of 1k, for 1mA

innews: snipped-for-privacy@i31g2000yqm.googlegroups.com:

Ok. The 1K resistor is not a problem. The circuit as shown at the url you provided is for 5 or 12 volts and is not a problem at those voltages, but at 100 volts it is. The TL431 can't handle that much voltage, as you have said. Its max rating is 37 volts. I thought that was why you mentioned the 75V zener - to keep the voltage that the 431 sees well below its maximum rating.

In any event, the circuit as drawn at the site you mentioned won't work for your 100 volt requirement. There's nothing to protect the

431 from overvoltage - all you'll do is burn it out.

Ed

--
I've read through this thread a couple times and I'm completely
confused.

Do you have this?:

http://www.powerint.com/sites/default/files/product-docs/an43.pdf

Also, what more can you tell us about the transformer you have on
hand?

yes!, well I am sorry I cant explain my doubts

I have this transformer for my 90v output, WurthElektronik designed it for my specs. and I thought that a simple changes (resistors, caps) could make my 90v work, I have not made a smps before but with all the application notes around there I was just scary about the layout.

The problem so far is the TL431 (programmable zener by definition), that can not work above 36v, the solution is, another winding for AUX (ex 12v) or Zener-resistor as feedback, or zener for reduce voltage and tl431, but I have not seen this configuration on ONsemi and PI

PIExpert gives me this

Précis d'electronique wrote in news: snipped-for-privacy@m1g2000yqo.googlegroups.com:

[snip]

Sorry for the late follow-up I've been busy.

Here is an application note from PI with a 48V output showing how to use a TL431 at higher voltages With pretty well component by component explanation.

Who designed your custom transformer?

Like I said earlier you could also just regulate the primary side ie the Vcc winding if your transformer has one the output would just be regulated by the turns ratio.

You could also just use a zener/opto for regulating the output.Where the output is approximatly regulated to Vzener + Vf of the optos LED. I'm sure PI and Onsemi have examples of this I've seen them before just never had the need to use it when you can get a 431 for the same price as a zener.

You could also look at no opto coupler isolated controller like the LT3748.

You would need a Vcc winding on your transformer.

What controller are you useing? What are the transformers specs? What are the output tolerance requirments ?

You may get more help if you provide more details.You've gone from an unknown Onsemi controller to now an unknown PI controller.

Is this your first SMPS?

Hammy wrote in news:Xns9DE7AA8753CFDHammyhamsterca@69.16.185.247:

[snip

If you go to the onsemi folder for the NCP1200 they have a pspice model for it with a template using a simple zener/opto feedback. Just so you know regulation isn't going to be tight probably +/- 10% or so when you factor in all tolerances.

Here's a screen shot of the spice template using the zener/opto. I can't fit the whole schematic in but the relevant portions are visible.

This is just the provided template with the default components.I just added R7 for Izener.

6.185.247:

or

ed =A0

I have not orcad or microcap but I have multisim and I am trying to make it work, so far I have this but it wokrs even without the NCP1200 so It is wrong

snipped-for-privacy@m1g2000yqo.googlegroups.com:

a TL431 at higher voltages With pretty well component by

Vcc winding if your transformer has one the output would just

output is approximatly regulated to Vzener + Vf of the optos

ust never had the need to use it when you can get a 431 for the

2. >

.

the output tolerance requirments ?

nown Onsemi controller to now an unknown PI controller.

Hi, thank you so much

Yes it is my first smps, thank you for that one der97.pdf, I think it is just what I ask for, just compute resistors and a zener of 82volts

my transformer is this one

(please note that I have 2 outputs, it is just for production for use either 90v or

12, not the 2 in the same circuit)

6.185.247:

or

ed =A0

I forgot to say that I am using the NCP1200, PI is just for the AN's

6.185.247:

or

ed =A0

do you use orcad or microcap?

I think he needs to stack the LM431's to achieve Vf = 100v eg: 4 lm431's stacked and set for 25v each.

Now I have no idea if he's rectifying the transformer voltage or just clipping the output at 100v.

Cheers

Do you mean that I can put in series (with Rupper and Rlower) four

431?

That is just possible?