Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Bullet Proof

Status
Not open for further replies.

spec

Well-Known Member
Most Helpful Member
"A common mistake that people make when trying to design something completely fool-proof is to underestimate the ingenuity of complete fools" : Mostly Harmless Douglas Adams


I was reading through the past posts on ETO yesterday, and very interesting they are too. A lot are beyond my field, but when I came across throbscottle's post, below, about power supplies blowing up, it stuck a chord. It got me thinking about all the lengths you need to go to in order to protect certain circuits, and also comply with the many specifications and standards: temperature, shock/ vibration, humidity, EMC, safety, through life costs...

In terms of protection, bench power supplies are some of the most difficult: quite simply, you never know what will be connected across the output terminals. Because of this, they are also difficult to stabilise in the frequency domain and at the same time give good fast regulation. This is not so difficult with the low gain Darlington output of this simple voltage regulator, but even emitter followers can make good oscillators.

So, out of interest, I took the most basic voltage regulator, the kind that was popular before the three terminal regulators, like the LM78xx/ LM79xx and LM317/LM337, swept the board. Then, I added most of the components needed to make the basic regulator reasonably robust, and also to function ok with worst-case components.

In the attached schematic below, PSU 1 is the original circuit and PSU2 illustrates what it turns into with protection. As is often the case, the additional components completely overwhelm the basic circuit. In voltage regulation terms, both PSUs perform the same. In fact, the original simple circuit performs better.

These circuits should work OK, but I haven't spent any time optimizing, so don't worry about the design details, just the principles.

https://www.electro-tech-online.com/threads/variable-psu-output-vulnerability.145895/

I'm in the last 1% of my hybrid psu design, and think I have found out a problem that must affect many variable bench psu's.

So, take an adjustable psu with an emitter follower output, connect a capacitor across the output, turn up the voltage, quickly turn down the voltage. Suddenly the emitter of the output transistor is now at a much higher voltage than it's base, possibly exceeding its vbe and killing it.

So I think that's why I keep blowing output transistors anyway. Does the theory hold water? Does it affect all variable PSU's? Or is it a nasty trick I have uncovered because the pre-regulator lowers the collector voltage when the output is turned down?


ETO_bullet-proof_PSU_Iss01.00_2015_11_30_Sh01~N.png


ERRATA

(1) As AnalogKid noted in post 11, 2D-4 should be replaced by a short circuit and D2-5 should show the correct symbol for a TrasZorb (Transit Voltage Suppressor) 2015_12_04
 
Last edited:
"A common mistake that people make when trying to design something completely fool-proof is to underestimate the ingenuity of complete fools" : Mostly Harmless Douglas Adams


I was reading through the past posts on ETO yesterday, and very interesting they are too. A lot are beyond my field, but when I came across throbscottle's post, below, about power supplies blowing up, it stuck a chord. It got me thinking about all the lengths you need to go to in order to protect certain circuits, and also comply with the many specifications and standards: temperature, shock/ vibration, humidity, EMC, safety, through life costs...

In terms of protection, bench power supplies are some of the most difficult: quite simply, you never know what will be connected across the output terminals. Because of this, they are also difficult to stabilise in the frequency domain and at the same time give good fast regulation. This is not so difficult with the low gain Darlington output of this simple voltage regulator, but even emitter followers can make good oscillators.

So, out of interest, I took the most basic voltage regulator, the kind that was popular before the three terminal regulators, like the LM78xx/ LM79xx and LM317/LM337, swept the board. Then, I added most of the components needed to make the basic regulator reasonably robust, and also to function ok with worst-case components.

In the attached schematic below, PSU 1 is the original circuit and PSU2 illustrates what it turns into with protection. As is often the case, the additional components completely overwhelm the basic circuit. In voltage regulation terms, both PSUs perform the same. In fact, the original simple circuit performs better.

These circuits should work OK, but I haven't spent any time optimizing, so don't worry about the design details, just the principles.

I would think that a diode connected between the base and emitter would clamp the voltage to a safe level. The Motorola MJ4030s I use have 2 built in shunting resistors to alleviate the problem.

You might want to study the circuit I used about 40 years ago when I built my first PS. It uses a series parallel arrange with a hefty power resistor. The power resistor takes all the hell from the heat while the transistors stay cool, calm and collected. The documentation and circuit is included for your edification. From the picture, can you see where I goofed in applying this circuit?
IMG_1356.JPG


IMG_1357.JPG

Ratch
 
I would think that a diode connected between the base and emitter would clamp the voltage to a safe level. The Motorola MJ4030s I use have 2 built in shunting resistors to alleviate the problem.

You might want to study the circuit I used about 40 years ago when I built my first PS. It uses a series parallel arrange with a hefty power resistor. ... The documentation and circuit is included for your edification. ... Ratch

Hello Ratch,

Neat approach; there are always differnt ways to do a circuit. Never used the series configuration for power supplies but I can see that approach would be especially good if you were dealing with higher voltages. I did use series transistors in my scope days when HF high voltage transistors were not availabe- used four trans in a long-tailed-pair arrangement to generate 200V differential for driving the x plates of the tube.

The power resistor takes all the hell from the heat while the transistors stay cool, calm and collected.
Beautifully put- you make me laugh personifying semiconductors- I think the same.:)

What was the application- was this a bench PSU?

From the picture, can you see where I goofed in applying this circuit?
That's not a goof- it's holes for extra cooling- done same thing many times. Why they made TO3 cans asymetrical I'll never know. The TO220 etc were so much better but not such a low thermal resistance. My favorite is the TO247 with a clamp on the top.

Post circuit of your PSU?

Be nice to know which part of the world you are.
 
Last edited:
The nylon bolts melted ?

Hi Grand

I never trusted nylon parts on power trans. For one thing you can't get enough pressure for good heatsinking. But OK in this application because, as Ratch says, power trans run cool.

Nice part of UK you live in. Have been many times.:)
 
The nylon bolts melted ?

Not so far. The heat sink gets warm, not hot. The nylon is a good electrical insulator, too. The nylon bolts hold the two heat sinks together. The transistor is fastened with metal bolts.

Ratch
 
Last edited:
Hello Ratch,

Neat approach; there are always differnt ways to do a circuit. Never used the series configuration for power supplies but I can see that approach would be especially good if you were dealing with higher voltages. I did use series transistors in my scope days when HF high voltage transistors were not availabe- used four trans in a long-tailed-pair arrangement to generate 200V differential for driving the x plates of the tube.

Beautifully put- you make me laugh personifying semiconductors- I think the same.:)

What was the application- was this a bench PSU?

Yes, a bench linear power supply. Had it for a long time now.

That's not a goof- it's holes for extra cooling- done same thing many times. Why they made TO3 cans asymetrical I'll never know. The TO220 etc were so much better but not such a low thermal resistance. My favorite is the TO247 with a clamp on the top.

I should have put the dissipation resistor on top of the heat sinks so the warm air from the resistor does not go through the heat sinks.

Post circuit of your PSU?

Too much trouble. I never did write down the circuit. Just did it from notes and scribblings.

Be nice to know which part of the world you are.

State of Minnesota. Currently closed for glacier repairs.

Ratch
 
'Clear Blue Water'
'Minnesota's standard of living index is among the highest in the United States, and the state is also among the best-educated and wealthiest in the nation' (Wiki)

Minnesota, land of blond hair and blue ears.

You are frozen in?

Currently in a snowstorm.

Ratch
 
Two things about the PS2 schematic.

1. What is the reason for two back-to-back Schottky diodes on the output (D2-4 (typo) and D2-5), especially when there already is one (D2-3)?

2. The current limiter R8 and Q2-3 introduce a voltage drop that varies with output current into the output voltage, significantly affecting the overall circuit regulation. If you move the current limiter to the collector of the output darlington (adding a PNP sense transistor), it will not introduce this error into the normal mode regulation.

ak
 
Hi Analog,

(1) The two back to back diodes are transorbs for transient voltage supression. Can't remember the part numbers but they are like zeners, but short once the threshold voltage is reached. They take hundres of amps for a short time. We used them on every outside word interfaces on some equipments. Our spec was to protect from the human body model for transient voltages which, if I rember correctly. was equivalent to a 100nF capacitor charged up to 4Kv.

(2) Fully agree about the voltage drop of the current sensing resistor but it is not my design it was the basic circuit arrangement for a simple power supply, way back. PNP transistors- you young fellows are spoilt these days. At the time pmp transistors were rare and frowned upon. But the other thing is that this starving current limiting approach is well behaved and doesn't get into a tange where the voltage and current servos fight themselves.

You dont like the circuit then. Give you a clue- I think it is pants.

The protection diode introduces another error. So in total the errpor in the voltage servo is (2* vbe )+ (Iout*Rsense) + (Dp Vf) : not good.

Not only that, but the voltage reference is hopeless: 18V zenners are poor: 4V7 to 7V5 better. so you would use a stack of zenners to give aroun 18V. The power supply ripple and dv rejection is awful. And the tempco couldn't be worse.

I never used the cicuit myself but it appeared all over the place at one time. My sole objective with this post was to demonstrate how a simple circuit is made much more complex when all the protection etc is added. Also to address throbscottle's problems with the output tran blowing with this very circuit.

Me, I'd use an LM317/337 with current booster trans if necessary, or perhaps one of the LDOs which are quite good.

Also, I made a precision lab PSU but that is another story.

I have read quite a bit of your stuff on ETO. I can tell you know how wiggely amps work. :)
 
Last edited:
1. I'm very familiar with tranzorbs. They have a semi-standard schematic symbol that is different from both the Schottky diode symbol on the schematic and a zener diode that they resemble. Also, tranzorbs are restructured zeners, optimized for large current peaks and transient thermal events; they do not short when the threshold is reached.

2. I was talking about a PNP signal transistor, not a PNP pass transistor. You're old - think 2N404 or CK722. And the other thing - I was not suggesting servoed current limiting, just the same starving approach reworked.

ak
 
ak, I get the impression that something went wrong here. I had forgotten that you were from Ohio, so let me translate:

PNP transistors- you young fellows are spoilt these days. English humour hard to translate. You would often say it to someone older than you. I have no idea how old you are.

I have read quite a bit of your stuff on ETO. I can tell you know how wiggely amps work
. I have read quite a few of your posts on ETO and they are very good. It is clear that you understand electronics.
_____________________________________________________________________________

About the circuit: as I said, it is only symbolic and its design/performance are not relevant here.

But as you are obviously interested, I would more of less do what you say but also add a long-tailed-pair to elliminate the out-of-loop errors, get more gain, and better temp performance.

About the Transzorbs- I realise now that I put two Schottky diodes in rather than two Zeners. I couldn't find the symbol for a Transzorb in the library and I was in a hurry. Transzorbs, as you say, are a form of Zener and do not fold back.
 
No need to translate, I got your humor just fine. I even figured out what "pants" means.

I have no idea how old you are.

Since you're from England, translate this: Need me, feed me.

ak
 
Last edited:
No need to translate, I got your humor just fine. I even figured out what "pants" means.

Since you're from England, translate this: Need me, feed me.

ak

I think it means: "If you want my services... pay"
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top