Voltmeter usage with breadboard ?

[ibot]

Hello,

I am trying to validate my circuit [for $50 robot] .. by checking the voltage at multiple points.

1. In tutorial, admin used the multimeter and placed the probes on pins [back] and tested it. But, I am using a breadboard. wondering, how can I test the same?

2. also, can somebody help me with, at what pins should we test the voltage [any specific pins that we need to check for sure]?

3. I tried this,  [from multimeter]  placed my red probe on output of 5V regulator and black on ground wire .. is this correct? it was showing approx 7v

4. In $50 robot, does the power should first go to regulator then to capacitor? or the opposite? I am guessing the power probably should go to capacitor first and then go to regulator. so, the regulator will get a continuous flow. Then, the output from regulator will be used to power up the Micro controller. Is my understanding correct?

Greatly appreciate your help. Thanks !!

[cyberfish]

First of all, do you know how breadboards work?

You can leave black always on ground.

5V regulator should give you, well, 5V. You're probably measuring the wrong pins (or connected it up wrong).

Capacitor then regulator (on the input side of the regulator). We usually add a very small cap (ceramic) on the output side, too, but that's not nearly as important, and is probably not required for 7805.

so, the regulator will get a continuous flow. Then, the output from regulator will be used to power up the Micro controller. Is my understanding correct?

That's right.

[SmAsH]

1. Yes, the same pins and stuff need to be tested.

2. the vin and vout of the regulator and the 5v pin on the mcu are the best ones to check.

3. If you have the points right, that's bad. Are you use your not testing vin?

4. Yes, cap first. Sometimes a small cap on the other end too.

[Soeren]

Hi,

Just to correct...
The 7805 needs around 220nF (100nF to 470nF) on the input and about 25µF on the output - emphasize on needs - to stabilize the system.

And for those that now wanna tell me, that they have been running without one and/or the other without problems, take your pick from the following...
1) Several people have been driving cars at high speed with their eyes closed for a scary amount of time and survived - I'll still say it's a bad idea.
2) You probably never heard the terms, but line and load regulation, noise, feedback and oscillations do matter in a voltage regulator, even if you cannot see it with a DMM.
3) I have had this discussion too often, go study datasheet app.notes etc. and learn instead of trying to make a point of randomness.

[amando96]

if you're feeding the 7805 with a normal 9v batery check if the 7805 is hot 5 seconds after turning on, if it is, you have wired it wrong...

i'd consider upgrading your regulator, linear ones like the 7805 waste energy...

[ibot]

All, thanks for the replies.

I checked my circuit again and found that I wired it wrong

I fixed it now. Now, i finally got 5V reading as output from regulator. But, I found my regulator was getting hot in a minute if I left the circuit powered on.

I may need to learn things that Soeren mentioned in his reply. Thanks a lot for all your replies. Greatly appreciate it !!

[chelmi]

i'd consider upgrading your regulator, linear ones like the 7805 waste energy...

Yes they do, but for some application (like powering a micro-controller) it is negligible.
Buying a switching reg. for the 50$ robot would be a waste of money IMHO.

Chelmi.

[Soeren]

Hi,

I found my regulator was getting hot in a minute if I left the circuit powered on.

Add a small heat sink if it's too hot and remember; what's a perfectly allright temperature for the electronics will still be able to give you a nasty burn if you touch it!

7805 has got temperature shut down, so if it gets too hot, it will shut down to protect itself.
In that case, add a sink.
If you have 9V in and 5V out, you have to get rid of 4W.

Another aspect of it is, that if the 9V comes from a battery, it won't be using the battery all the way down A 9V (6 cell) alkaline is considered flat at 5.4V and with NiMH the range is 10.15V down to 6.3V (7 cells), or 11.6V down to 7.2V (8 cells).
The 7805 needs to drop 3.0V to 3.5V, which means that it needs a 8.0V to 8.5V input.

[ibot]

Soeren,

Thanks for the reply .. but, I have a question ..

Why are we using a 9v battery and trying to use a regulator to get the voltage down to 5v ? can we just give a 6v input and regulate it to 5v? [so that it might be very easy for regulator to handle. ] ..

Greatly appreciate your help !

Modified:
hmm .. just going through the tutorial again. seems, we can use a 6v battery pack too .. what I am currently doing is a 6v battery pack going to servos and an alkaline 9v going to power the MC. i guess, I will use another 6v battery pack instead of 9v and see if that will keep my regulator cool !!

[Soeren]

Hi,

I will use another 6v battery pack instead of 9v and see if that will keep my regulator cool !!

Please reread my previous post.
The 7805 needs 8.0V to 8.5V in to give a regulated 5V out.

If you use a Low DropOut voltage (LDO) regulator, it will work from around 5.5V and up.
Your 6V battery is a 5 cell battery I assume, so will be ~7.25 when just from the charger and ~4.5V when empty, so will not be used fully.
A 6 cell battery (7.2V battery) will work great with a LDO regulator with a range of ~8.7V down to ~5.4V.

[cyberfish]

7805 has a typical dropout of 2V at 1A according to the datasheet.

Also, I know how the datasheet says it has thermal shutdown, but I HAVE blown up a 7805 before, so I'm not too sure what's happening.

I ran it from a lab power supply at 9V/4A and probably shorted the output. It blew up literally (a black piece flew away) after a few seconds. TO-220 without heatsink.

I am using a 6V battery with a LDO. Works perfectly. It's a lot more picky about the output capacitor, though. Both ESR and capacitance need to be in a certain range to stop the output from oscillating. Apparently modern ones are a lot less picky, but still harder to use than 78xx.

[ibot]

Thanks for all your help.

Now, I added a heat sink and my regulator seems much happy. not that heat any more ..

Thanks !

[Soeren]

Hi,

7805 has a typical dropout of 2V at 1A according to the datasheet.

Some of the makes do say 2.0V (typ. and usually 2.5V max, then). If you know the exact make and model of a 7805, you may get away with under 3V IF you are a keen designer that know how to read a datasheet (you seem to scoot quite indifferent past stuff like eg. line and load regulation). AND you verify the sample you've got AND you bake a batch of them at around 110°C for 3..4 hours AND you retest them after that.
Short of a single one of them, you can either take my word and add 3.5V or design inferior supplies most of the time, but I don't base my designs on luck.

Even if you really don't wanna see the light, at least get datasheets from as many manufacturers as possible and base your advice to people that you don't know and hence have no idea what they have at hand, on worst case data, then nothing will go too wrong. Even with the best (at least on paper) 78xx regulators, 2.5V should be the worst case drop. Some regulators have a combined line and load regulation of up to around 0.25V Total so far 2.75V (for one of the good regulators).
Since this is a battery fed regulator, luckily, there is no AC to take into account, or it would have been worse.

Also, I know how the datasheet says it has thermal shutdown, but I HAVE blown up a 7805 before, so I'm not too sure what's happening.

I ran it from a lab power supply at 9V/4A and probably shorted the output. It blew up literally (a black piece flew away) after a few seconds. TO-220 without heatsink.

Apart from the possible bad component (if eg there is an error in the thermal shutdown part of the silicium, things will go wrong), I find it likely that oscillations is the real sinner. Thermal shutdown was never meant for short circuit protection, but for an overload that takes some time to reach a critical temperature, as the temp. needs time to reach the thermal detector circuit.
I wonder how anyone can mix up "thermal shutdown" with short circuit protection"? It's not even spelled the same way

Running it from a lab supply could mean LF ripple if conventional iron and HF ripple if a switcher was involved. Further, if you used long (i.e. >10 cm) leads from the supply, the induction of the leads may have helped in creating oscillations.
If the regulator wasn't properly capped, oscillation was surely involved (those regulators are quite high gain amplifiers).

I am using a 6V battery with a LDO. Works perfectly. It's a lot more picky about the output capacitor, though. Both ESR and capacitance need to be in a certain range to stop the output from oscillating. Apparently modern ones are a lot less picky, but still harder to use than 78xx.

Sure they're more picky, they have much less room for error and have much higher gain.
I have never had any problems with LDO's though, but I do study the datasheets of any "new" component I get a hold of.
My preferred LDO is Micrels MIC29752BWT, a variable output regulator which has a voltage drop of around 125mV at 1A (250mV at 4A and around 425mV at 7.5A), so I switch down from a 12V (motor voltage) to about 1V over what I need and the LDO takes care of the remaining ripple with minimal losses.

Going from a smaller battery, a SEPIC/CUCK/Buck-Boost topology is the right way to go  if you are wanna get every last drop of juice out of it without heavy losses when freshly charged.

[cyberfish]

(you seem to scoot quite indifferent past stuff like eg. line and load regulation)

WHAO! no wonder I felt something behind my back when I read the datasheet. And yes, contrarily to popular belief and despite being a lowly second year student, I do know what they mean.

AND you verify the sample you've got AND you bake a batch of them at around 110°C for 3..4 hours AND you retest them after that.
Short of a single one of them, you can either take my word and add 3.5V or design inferior supplies most of the time, but I don't base my designs on luck.

So you are basically saying, screw what the datasheet says, and test it yourself?

The one I am using, made by fairchild, says 2.0V typical at 1A and Tj = 25C. National Semiconductor's 7805 datasheet also says 2V.

I usually trust datasheets.

Do you seriously go test every parameter on the datasheet? Why do we even need datasheets then?

Why stop at 3V? Why not 4 or 5?

Even if you really don't wanna see the light, at least get datasheets from as many manufacturers as possible and base your advice to people that you don't know and hence have no idea what they have at hand, on worst case data, then nothing will go too wrong. Even with the best (at least on paper) 78xx regulators, 2.5V should be the worst case drop. Some regulators have a combined line and load regulation of up to around 0.25V Total so far 2.75V (for one of the good regulators).
Since this is a battery fed regulator, luckily, there is no AC to take into account, or it would have been worse.

Which 7805 datasheet exactly says 3V?

The fairchild datasheet says max load and line reg are 100mV, the national one says 50mV for both, but with a 10V input. Although it does say minimum 7.5V to maintain line regulation.

Thermal shutdown was never meant for short circuit protection, but for an overload that takes some time to reach a critical temperature, as the temp. needs time to reach the thermal detector circuit.
I wonder how anyone can mix up "thermal shutdown" with short circuit protection"? It's not even spelled the same way

Thank you very much. I do know perfectly well the difference between thermal shutdown and short circuit protection. As you noted, they are even spelled differently!

I was just under the impression that short circuit/oscillation should have triggered thermal shutdown before the case blows up.

Running it from a lab supply could mean LF ripple if conventional iron and HF ripple if a switcher was involved. Further, if you used long (i.e. >10 cm) leads from the supply, the induction of the leads may have helped in creating oscillations.
If the regulator wasn't properly capped, oscillation was surely involved (those regulators are quite high gain amplifiers).

That is possible. The input was capped, though (don't remember about output, it was quite some time ago).

[Soeren]

Hi,

So you are basically saying, screw what the datasheet says, and test it yourself?

You misread. I'm saying that if you want to design on the marginals and the so called typical values, you need to know the particular component, as the data sheet defines tolerances that you obviously don't care about.

The one I am using, made by fairchild, says 2.0V typical at 1A and Tj = 25C. National Semiconductor's 7805 datasheet also says 2V.

I usually trust datasheets.

When it suits you, but come to thermal shutdown...
You did notice the Tj=25°C?
That means: When the temperature in the semiconductor crystal is 25°C hot.
You'll never see this in a live circuit, unless perhaps if you cool it with liquid nitro. when the metal tab on the outside is a steady 40°C, Tj can well be 120°C, all depending on heat sink design and thermal characteristic of ambient.
In eg. a short circuit situation, the semiconductor may well have gone permanently off before you could register a change on the metal tab.

Do you seriously go test every parameter on the datasheet? Why do we even need datasheets then?

Why stop at 3V? Why not 4 or 5?

Read what I write please.
I test which of a batch is within the extreme specs, when and only when, I need to go for the marginals, which luckily isn't that often. Usually, I just go to 3.5V over the output - when I care to use 78xx's.
I think I have explained the reason for the 3.5V to you quite in detail, but once again for good measure, you need what will make the regulator stay in regulation under any condition, and not a bit more, as that would just be wasted as heat (and heat really changes the regulation properties btw.)

Which 7805 datasheet exactly says 3V?

Eg. this

The fairchild datasheet says max load and line reg are 100mV, the national one says 50mV for both, but with a 10V input. Although it does say minimum 7.5V to maintain line regulation.

So, at least you're up to 2.5V now, only 1V to go 
You should never assume ideal conditions. When you help a person somewhere else on the globe, you must try to anticipate each and every thing that person can possibly do to make it inoperative and that includes component tolerances and sometimes an inferior regulator or other component.

Thank you very much. I do know perfectly well the difference between thermal shutdown and short circuit protection. As you noted, they are even spelled differently!

I was just under the impression that short circuit/oscillation should have triggered thermal shutdown before the case blows up.

Why?
If the power going through some sections of the chip is rising suddenly, you have to account for thermal mass/inertia. The thermal shutdown will only work if it gets hot before the rest has gone *boom*.
In my youth, I managed to kill a series of LM317 before I found out how much of an issue oscillations could be in regulators. I learned from that. You can too, or you can go kill off your own components (but it's really not fair to kill other peoples components IMO).

That is possible. The input was capped, though (don't remember about output, it was quite some time ago).

Next time something like that happen, take notes and learn. As a matter of fact, you should take notes on each experiment you do, jot down measurements, formulas, random thoughts etc. and with time, you'll have a personalized and very usefull reference.
I'm sure that won't happen again though, as you know now some of the possible modes of error.
When running from long(ish) leads, you'd place a large cap on the input of the regulator board (say 1000µF, depending on lead length, current etc., but rather too big).
Most regulators are actually rather resilient and I have sometimes had them work harder than they should but have had very few fail on me (and I have used thousands of these beasts over the years).