Voltage References

"Jim Thompson" wrote in message news: snipped-for-privacy@4ax.com...

I just spotted a resistor that looks bad. Measured it's value in circuit and it always gave OL... its suppose to be about 470... so even if in parallel with other things I'd expect a reading. It has a litle bround spot(looks like dirt) on it though and is the first resistor in the path of the tach. Hopefully thats it and it will work. (although I can't see why a resistor would go bad like that in this case?)

I don' t know to much about programmable voltage regulators but it seems they are used to "cook" up your own zener? If so I don't see the benefit unless it has some type of hysteresis in it?

ok, I just replaced the resistor and it's working!! HOORA!

Can't believe it was the resistor ;/

I'm wondering why it occurred? and could I make it less susceptable? It looks like its a 470:2000 voltage divider... maybe increasing the values to prevent as much current?

"Jim Thompson" wrote in message news: snipped-for-privacy@4ax.com...

Don't know but it was bad ;)

I'm still trying to understand how it works. Are you saying that it is basically like a driver? An "active zener"?

So whats the difference between a programmable reference and an op amp?

why not? surely the it doesn't need that much currnet to work?

After all

1000:2000 is same as 2000:4000 as far voltage is concerned... and surely I could increase them a bit and the device will still work but reduce current?

I'm not sure how much input the programmable reference needs but increasing the resistance can't hurt unless it's just not enough to drive the reference? (but can help a lot if they used to low of values in the first place)

Does anything got to the REF pin besides those two resistors? Like maybe a capacitor? Changing impedances changes dynamic response. If I deviated from the original I'd stick with just upping the wattage, though 470:2000, if it's Ohms, doesn't imply a problem with dissipation... make sure of the values.

...Jim Thompson

--
| James E.Thompson, P.E.                           |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
 I love to cook with wine     Sometimes I even put it in the food

Programmable references are programmed with resistor ratios, just like programmable gain blocks (opamps) are programmed with resistor ratios. They do, in effect produce programmable zeners, except that they are more temperature stable, have lower initial tolerances, lower impedance (wider operating current ranges) and lower noise than actual zener diodes.

--
Regards,

John Popelish

but I can't see how they will produce a stable voltage? Your input voltage is not stable so how can the voltage out of it be stable?

V | R1 |

• --- VRef --- VOut | R2 | gnd

so, your voltage into Vref is V*R2/(R1 + R2)

So if Vout is proportional to V in then whats the point? it's not really a programmable voltage "reference" but a voltage follower?

Even suppose that V is fixed(which in this circuit it is not), Any fluxation in V will be passed to Vout?

So I don't see why you just don't use the voltage directly except of course you need a driver... so maybe that is the point? One could use an op amp just as easily? (a buffer)

No. Just a voltage divider.

But it didn't fix it. At least not completely. Now it seems to give the error code randomly. It at least does do about half the spin cycle now it seems after it gives the error and you restart it immediately it won't spin up... but if you wait several minutes or remove power it will then work fine for several minutes at least.

My guess is that whatever cause teh resistor to stop working is malfunctioning from time to time. It could be the voltage reference? (it is thing right after resistor)

There are many sot23 type transistors in there that could be bad too but they are all after the vref while the resistor that went bad is before it ;/

Unfortunately I probably don't have any programmable vrefs to replace it but two questions:

1. Will virtually any vref work assuming similar operating specs. (mainly voltage in/power and current out?)
2. Any way to rig it to work?

see my reply to Popelish about programmable vrefs... still unsure what they are for.

Programmable zener is just one common application (where th3e voltage divider is connected as feedback from the output back to the reference input). Your circuit uses the reference as a comparator with a built in offset voltage that acts as the threshold voltage for the comparator.

When the divider output voltage exceeds the reference voltage, the output pulls down.

--
Regards,

John Popelish

Think of it as an NPN bipolar transistor with a 2.5V Vbe and a huge gain.

Another way to think of it is as an open-collector comparator, with the '-' lead connected to an internal 2.5V reference. Then, the r lead corresponds to the '+' input to the comparator, the 'k' lead corresponds to the output of the comparator, and the 'a' lead is the ground lead of the comparator. There is actually a picture of this on page 10 of this doc:

That datasheet also has a schematic of the device, which may be helpful.

Regards, Bob Monsen

Ok, I can see that from the functional equivalent(The op amp) but then I guess my confusion is the "programmable" part. It's not really programmable? Sure we can set the output voltage based on the input voltage but we do the same with op amps all the time and we don't say we are "programming" an op amp.

A programmable zener would be a zener, that say, has a little switch on it that lets you select different voltages but functions exactly as a normal zener. If the switch is "continuous" letting us select any voltage then that is better... but still functions discretely and independently.

If now the switch is voltage controlled then we have a problem... if the input changes it will follow it exactly(or close) so any "errors" in our input to the switch will modify the output... now it's nothing like the first case I mentioned.

Hence I don't see the programmable part(well, I can see how you might say it is programmable) and the precision part. What's precise about it?

If you change two op amp buffers together and one really sucks then doesn't matter how precise the other is.

Now if it just functions as an on/off then thats different... but "programmable precision voltage reference" seems to imply so much more.

I can see what it means when it acts as essentially a switch:

Programmable means we can set the trip point at any point and precision means that it trips very precisely.

But if there are other more appropriate uses for such things other than switches then I'm confused.

You're confusing "programmable" with its use in your machine. Think of it more like the inverting input of an OpAmp with a 2.5V built-in offset voltage; or, as someone recently pointed out, an NPN with Vbe =

2.5V

...Jim Thompson

--
| James E.Thompson, P.E.                           |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
 I love to cook with wine     Sometimes I even put it in the food

It helps but I'm still a bit confused as to the general purpose of such a device. If it acts as a programmable threshold switch then it makes sense... if thats just one of the many uses(I'm talking about its specific use) then I'm a bit confused. (I'm also talking about ideal case)

E.g., an ideal diode acts as an ideal voltage controlled switch s.t., if the voltage >= 0 it is on and if V < 0 it is off.

A "programmable" diode is the same as a diode except the "trigger point" can be changed: if the voltage >= a it is on and if V < a it is off.

The problem is that the "trigger point" or "threshold" a follows whatever we hook up to it... if it's a sine wave then the threshold is changing sinusoidally... If there is a lot of noise then a will be noisy... not very precise ;/

Now maybe it, a, "precisely" follows whatever is attached too it and that is what it means?

I guess that depends on how broad your concept of programmable is. There are analog programming methods and digital programming methods. I have certainly seen text books that refer to opamps as universal amplifiers that are programmed to perform many specific tasks with passive components connected to them. In the days of analog computers, this was just about all that was meant by programming.

Its internal reference voltage. That makes the operation very predictable and dependent mostly on the precision of the resistors used to program it. This is similar to how a very accurate zero input offset voltage makes opamps so predictably programmable for DC functions. This device just has a precise non zero input offset.

It is programmable because it includes both a precision reference and a high gain block with a controlled phase phase situation that makes a wide range of feedback configurations possible.

You obviously have not programmed an analog computer.

--
Regards,

John Popelish

Late at night, by candle light, "Jon Slaughter" penned this immortal opus:

^^^^^^

Is that a brown round spot? (Sorry, couldn't resist)

- YD.

--
Remove HAT if replying by mail.