How do I Fine tune a 78L05 ?

Dimbulb · Post by **Dimbulb** » Wed Jun 13, 2007 10:01 am

I agree the capacitors are necessary as the circuit grows and the battery gets low. I was thinking that a tantalum on the output for quiet. This because I of reading some audiophile designers. on a separate post I would like to measure the noise but this needs to be a whole new post agree. now regarding voltage accuracy.

I will drag out my calibrated HP3456A since we are talking about 4 wire ohm sense now. I would like to make a real accuracy claim.

To be honest I never used this function. @%! ? no instruction manual.
I made up 4 leads about about 6 inches in length by cutting a few aligator jumper wires in half and soldered on the bannna jacks.

Dave Dixon · Post by **Dave Dixon** » Wed Jun 13, 2007 10:56 am

Nice unit! I found the manual for it here....
http://bama.edebris.com/manuals/hp/3456a/
Looks like you need a precision voltage supply to use the 4 lead input on that one. Probably a lot more trouble than it is worth to gain a tad more accuracy. Best of luck,
Dave

Enzo · Post by **Enzo** » Wed Jun 13, 2007 10:45 pm

Unless I missed it, during this testing, is your load on the regulator? Is the load that will be on it stable or does it vary?

haklesup · Post by **haklesup** » Thu Jun 14, 2007 12:02 pm

4 wire sense measurements are primarily intended for low resistance measurements.

A simple voltage measurement is accurate with two wire so long as the volt meter has very high impedance

2 wire current measurements are also accurate because current is the same in all parts of the loop and the meter impedance is low (the meter and lead resistance can act as part of the load and reduce I for a given V but not much)

In a low resistance measurement, the lead resistance may get added to the load resistance. In a 4 wire (kelvin sense) setup, the main leads supply a fixed current source and the sense leads measure the voltage drop across the load (where ever you attach them). The meter does the math for you (V1-V2/I) and displays the resistance.

Dave Dixon · Post by **Dave Dixon** » Thu Jun 14, 2007 12:12 pm

Good point Hacklesup. As soon as I realized (hindsight) that I had been thinking ohms, I went ahead and pulled up that specific manual. At least they DO have the option for four wire ratio voltage testing! I kind of got the chance to cover my azz on that one

I thought I managed to dig myself out of that one pretty smooth - with my reply!! AND now he knows where to find the manual.
BTW, there are several other helpful files concerning that meter, besides the user manual, on that webpage. I hope they help!
Dave

Dimbulb · Post by **Dimbulb** » Thu Jun 14, 2007 6:22 pm

now that we're moving on from checking the meter.

I am now using the LM431 with much success.
National Semi's Data sheet example on adjustment using a voltage divider works nicely.

What I would like to accomplish is to use the LM431 reference and follow it with something to source more current.

on a personal note, today in silicon valley it was very hot, just got home from work.

Chris Smith · Post by **Chris Smith** » Thu Jun 14, 2007 6:54 pm

Unless your into scientific study, your numbers must be exact for a specific reason, or the facts must have an exact out put or expectation,..... your numbers that are required are totally useless.

Electronics in general has a huge lea way for specific reasons,.... its too hard to be so A-Retentive let alone achieve this perfect status.

Dimbulb · Post by **Dimbulb** » Thu Jun 14, 2007 7:14 pm

I am not a scientist or electronic engineer as a hobby and my goals are for improvement of this hobby for other hobbiest like myself.

The focus is how to progress into completion of this voltage regulator.
The LM7805 was unable to adjust upward. It is still a good regulator.

The LM431 adjusts nicely so now the next logical step is how to best supply more current ?

I have tryed connecting the base of a 2N4401 result was unwanted thermal drift.

the application 100mA 5.000 V from a 9 volt battery, it is a supply for hobbiest that need a little more precision. some future improvements include a pre-regulator. Also getting a little more life from a 9 Volt battery.

question being will the battery regulator circuit still be working at... let's say 6.2 V

Working in the other direction toward getting more current could an op amp deliver 100mA 5.000V if so which op amp ? or what op-amp transistor follower would work?

I don't mind including supercoolers or Josephson arrays as discussion however my mind remains on the task at hand. I will take hobbiest a step closer to precision.

Chris Smith · Post by **Chris Smith** » Thu Jun 14, 2007 9:05 pm

In real life we have the 78xx and the 317 types of regulators.

Neither are as accurate as you suggest?

There is a reason.

Yes, perfection can happen with all sort of other regulators but the generic off the shelf items regardless of the name or number were never designed to be as accurate as you require.

IF the absolute is needed, then start with the VR reference diodes that are out there and then build it up.

Your problem here is the same as always..., can you get other exact components like exact resistors, or what, and always,.. for what reason.

Point x is far more than needed for accuracy for 99.9 % of whatâ€™s out there, and then some, so why is this obsession needed?

rshayes · Post by **rshayes** » Thu Jun 14, 2007 9:24 pm

Normally, the output current of the TL431 is limited by the current available from the series resistor. The avaliable current can be increased by adding an emitter follower. The base of the transistor would be connected to the output of the regulator, the collector to your 9 volt supply, and the output current taken from the emitter.

This results in an additional voltage drop of about .7 volts with an added temberature coefficient due to the base emitter voltage. These effects can be avoided by moving the top of the feedback voltage divider to the emitter of the transistor (the new output terminal). This puts the base-emitter voltage within the feedback loop and greatly reduces its effect on the output.

If high current is required, a pair of transistors connected as a darlington pair would reduce the current that the TL431 is required to handlle.

Dimbulb · Post by **Dimbulb** » Thu Jun 14, 2007 10:11 pm

Please help me to understand the following:

"These effects can be avoided by moving the top of the feedback voltage divider to the emitter of the transistor (the new output terminal)."

I have two 10 K resistors

the output at the cathode of the LM431

introducing the new transistor we connect the base to the cathode.
The collector to the 9V+ and
the emitter to the intersection of the two resistors.

is this correct?

if so where are the outputs ?

Chris Smith · Post by **Chris Smith** » Thu Jun 14, 2007 10:23 pm

The Study of heat and its components are fine, but how do you intend to fight reality with the normal drift, if drift is perfectly normal?

IF you want to learn the secretes of physics, then the Peltier will be your start, other wise your numbers are meaningless.

Up one second and down the next?

So what can you learn from random?

Nothing.

rshayes · Post by **rshayes** » Thu Jun 14, 2007 10:43 pm

Hello Dimbulb,

Yes, I think you have the right idea.

rshayes

Dimbulb · Post by **Dimbulb** » Thu Jun 14, 2007 10:49 pm

I am getting 5.604
measuring voltage across the collector base
the voltage slowly drops as the battery drops.

a preregulator should improve this.

thankyou

rshayes · Post by **rshayes** » Thu Jun 14, 2007 10:53 pm

The voltage from the emitter to ground should be very close to 5.00 volts. The voltage from base to groound wuld be in the neighborhood of 5.6 to 5.7 volts, which is what you appear to be measuring.

T&L Publications

How do I Fine tune a 78L05 ?

Who is online