Numpty question on NPN WS 78L05

[Another Dave]

Please treat me like a child and assume I am stupid also when answering 

I have a lathe with a digital LED readout that failed, I had to strip it to change a bearing and though
about sorting it at the same time.

The power board is not supplying 5v to the digital panel.  Simple circuit but still over my head

Bridge rectifier 3 caps one diode one inductor and 2 resistors and the WS 78L05.

Bridge rectifier checks out OK doing the diode test. Will swap the caps for Panasonic and Rubycon
items instead of the unknown Chinese capacitors.

I have 2 78L05's arrive and putting them on the component tester one shows as 2 diodes and the other
is backwards to my original item.

I can post up diagrams  of the testers findings, but the question is should one some up as 2 diodes
when tested?

And the other one thats backwards, can i simply fit it the opposite way around?

Original item.. 1 =C,    2 =B, 3 =E  hFE = 5,   Uf = 1.35v

Replacement they sent,  1 =E,  2 =B,  3 =C,   hFE = 566,   Uf = 669mV

The other one says 2 Diodes  Uf 815mV   Uf1.38V

Wrong items?   Many Thanks 

[Performa01]

78L05 is an integrated +5V voltage regulator, not a transistor. These ICs aren't exactly identical, depending on the manufacturer, but their pin-out is.

If the original item has died, it just cannot give the same readings as a working unit.

It's just suspicious that one of your replacement parts actually appears to behave like a transistor.
Hopefully it actually is a 78L05!

[rstofer]

The 78L05 is not a transistor, it is a 3 terminal 5V voltage regulator.  Grab the datasheet:
https://www.fairchildsemi.com/datasheets/LM/LM78L05A.pdf

For a bit of time, I was wondering what all the measurements meant.  Then I realized you had stuffed the VR in a transistor checker.  I have no idea how that should work out.

So, your bridge rectifier gives some value of DC which is filtered by some of the capacitors, the 78L05 drops the voltage to 5V which if filtered by the rest of the capacitors and the 5V goes on to the display.

[wraper]

They are IC's. How are you supposed to test them in component tester? 78L05 from different manufacturers have different circuit inside, but nonetheless they all are 5V regulators which are compatible. You don't test ICs like transistors or diodes.

[Another Dave]

Ahh yes apologies, its a component tester,  The original shows a PNP transistor with that configuration.

I was puzzled when the original and new items gave different results.

In the simple voltage drop circuit is that likely to be my problem because im only getting approx 0.7v
from memory.

Original one is a WS 78L05 933SB

replacement.  ST L78L05 A GE 520

The otherone shows 78L05 on the package but printed on it is something totally different.
BC 547C C39

The ST one is the one thatcomes up as 2 diodes.

Thanks for humouring me

[wraper]

BC547C is NPN transistor, same package as 78L05

[Another Dave]

Safe to use the ST L78L05 ?  Thanks.

[rstofer]

Safe to use the ST L78L05 ?  Thanks.

Sure, you can probably get away with installing the 78L05 but I wouldn't do it.  Once it went through the component tester, I would consider it abused and rather than risk my display, I would just replace it.  One thing these VRs don't like is reverse voltage between the input and output.  Many designs will have a diode in reverse across the VR to limit this voltage during power-downs.  I have no idea what the component tester applied to the '05 or what the polarity was.  The fact that the original even needs replacement is worrisome.

Here's an article that touches on the problem:
http://www.infineon.com/cms/en/product/promopages/aim-mc/Selecting-a-linear-voltage-regulator/Voltage-Regulator-Reverse-Polarity.html

[Brutte]

I'd say there are only two methods to kill 78L05:
1. With a hammer
2. Applying reverse voltage

It is protected against most mishaps: over-current, over-temperature, thermal run-out, it is even protected against unconnected reference pin. If you managed to kill your IC then there must be something seriously wrong in the design. Reverse-biassed diode won't hurt. Once it is brought back to life I'd measure the power dissipation (that is P=U*I, namely voltage across IC times current that runs through it). Keep it below 0.5W

[Another Dave]

So the original item is unlikely to fail?   Best guess what would be the cause of the output voltage
to be 0.7 and not 5 volts?

Down to one of 3 of the unknown brand capacitors?

I tested the bridge rectifier a 2W10 and that appeared to be fine.

Thanks.

[wraper]

I'd say there are only two methods to kill 78L05:
1. With a hammer
2. Applying reverse voltage

It is protected against most mishaps: over-current, over-temperature, thermal run-out, it is even protected against unconnected reference pin. If you managed to kill your IC then there must be something seriously wrong in the design. Reverse-biassed diode won't hurt. Once it is brought back to life I'd measure the power dissipation (that is P=U*I, namely voltage across IC times current that runs through it). Keep it below 0.5W

Or it just dies itself if runs hot enough for prolonged time, or just fails without too much stress. Faulty Vregs (vast majority of which have similar protections) is not something rare at all, among the parts which fail most frequently. Anyway those protections kick in when the Absolute maximum ratings are already exceeded.

[Another Dave]

Do i replace the capacitors and cross my fingers? or replace this also?  Anything in a simple circuit like
this to provide a 0.7v output?

Fairly common for these boards to fail, it seems they do not like living in sheds in the UK. Its a bit
of a pig to remove and refit so i would rather replace the bits as a just in case rather than miss
an item and have it not working again.

Thanks.

[rstofer]

Yes, replace the capacitors!

The 78L05 is rated for a load current < 100 mA and there's a reason that the specs talk about loads in the 40 mA range and often less.  It is a very low current kind of device.

The 7805, OTOH, can deliver an amp or so.  It is much more robust!  Alas, it is also bigger as it comes in a TO-220 package.  Nevertheless, I would see if it would fit.  Watch the pinout!

The 0.7V number is one of those 'magic numbers'.  It is the forward voltage drop across a diode or base-emitter junction of a transistor (more or less).  If you are measuring 0.7V with the load connected to the regulator, one could suspect that the VR has gone into current limit as a result of something wrong in the load.  In other words, the poor regulator may not be the only problem.  If the output voltage is 0.7V with no load then, yes, the VR is toast and there may be nothing wrong with the load.

We can suspect that the VR is on the small side without knowing much about the load.  It's easy to just replace the VR with a 7805 but we still don't know anything about the load.

Pictures would be good, a schematic would be better.

[Another Dave]

It runs a tachometer with a 4 segment LED display. The only thing thats marked is the +5v and Gnd and the 3
signal wires.

When i removed the display i tested it with 5v and it came on fine. Then tested the voltage on the 5v pin and
there was just that 0.7 volts.

I will try and upload some pics later.

Thanks.


So far i have learned to check for codes before using a component tester

[Brutte]

Nevertheless, I would see if it would fit.

You know, 78L05 is not only a voltage regulator. If that was the case then a raw zener would do and OP's post would start from "I had a fire in my shed, please help".

This is a protecting device that should save other sensitive components in case of a mishap.

I would discourage swapping a 100mA limit into a 1500mA limit. Or doing such tricks with any circuit breaker.

If the L failed then there is something wrong with the circuit. With a 15x higher limit the next weakest point is going to be a transformer 

Check the current. If the device draws more than 100mA at 5V then the regulator goes into shut-down (this happens around 120mA - 200mA, depending on the manufaturer and tongue angle).

Similar thing happens when you heat it up above 120 degC. I'd suggest you should test this by touching the package with a hot soldering iron - the protection has to trigger then, otherwise the IC is broken and should not be installed. However, it should get back to life once when it cools down. I have never heard about such IC fail when subjected to prolonged heat - these are essentially electric heaters and are designed to operate in the range of up to 125*C (some up to 150*C).

[Another Dave]

Thanks, what about cold and damp?  The lathe sits in an unheated shed. Initially it was intermittent
and would work briefly then fail.

It would do that randomly for a few minutes to half an hour and then fail totally.

The weather didnt make any difference whether it was below 5 degrees or over 15 degrees C.

I really do not want to risk damage to the rest of the power circuit. Tempted to run it off a 5v USb charger. 

[rstofer]

I didn't randomly suggest upgrading the VR without at least a little thought.  First, there was mention that these boards frequently fail.

Fairly common for these boards to fail, it seems they do not like living in sheds in the UK. Its a bit
of a pig to remove and refit so i would rather replace the bits as a just in case rather than miss
an item and have it not working again.

Second, there is an LED display.  I don't know how many digits but 6 seems about right.  I don't know if leading zeroes are blanked but it probably doesn't matter.  When I think about average segment current of say 10 mA times 4 segments lit per digit (just a number I picked) times 6 digits, I come up with 240 mA.  Yes, I have the wrong answer, the load can't be that high.  But it isn't going to be a lot less than 100 mA either.  Third, every spec in the datasheet is at some low load current like 10 mA or 40 mA.  The device MIGHT be capable of delivering 100 mA but I wouldn't bet on it.  The fact that the device shuts down periodically, yet recovers, make me wonder if it isn't going into current limiting.

Perhaps the display isn't even fed via the VR.  There's no reason that is must be but it usually is.

Given the proper test equipment, it would be interesting to know just how loaded that 78L05 really is.

Dampness could certainly be a problem if there is condensation.  That's why they invented conformal coating for PCBs

Given that there a history of failure, simply continuing to do the same thing won't suddenly change things.  I like the idea of using a wall wart as long as it is exactly 5V and has low ripple.  Use a DMM on VAC to measure the ripple and again on VDC to measure the actual DC voltage.  I checked one of mine and it delivers 5.17V with 1.7 mV of ripple.  Most wall warts will supply on the order of 1000 mA - at least 10 times more current than that little 78L05 can deliver.

If one board fails, it's random perversity.  If two boards fail, it is a trend.

Is there any access to the schematic?

[Another Dave]

Its just a 4 digit display, generally the max is approx 3000rpm in high gear.

Nothing else except the controller and the pickup which i assume draws a tiny amount of power.

Well i went with the correctly marked ST 78L05 and  IT WORKS....    I now
have my rpm display back.

Thanks..  A lesson there not to stick everything in the component tester though.

Google the part number and see what it is first.

I changed the 3 capacitors and the 78L05 so it could have been any of those.


Thanks guys.   

PS,,,  I have learned another lesson. Don't leave the desk with the soldering iron
powered on.. oops...  Been on for about 4 hours.

[rstofer]

A good outcome!

[jitter]

Good that you fixed it.

Thanks, what about cold and damp?  The lathe sits in an unheated shed. Initially it was intermittent
and would work briefly then fail.

I didn't see you write anything like "the traces are turning green". This means that probably no condensation formed on that board.
Maybe connectors would oxidize faster in damp environments, which could result in intermittent failures at first.

It would do that randomly for a few minutes to half an hour and then fail totally.

The weather didnt make any difference whether it was below 5 degrees or over 15 degrees C.

I really do not want to risk damage to the rest of the power circuit. Tempted to run it off a 5v USb charger. 

Well, now that it seems to work, why not build a little circuit to see if the 78L05 that you replaced was the actual cause?

Normally these are almost bullet proof, but maybe if it was a "Shenzen market special" not that much...  Might the markings "WS" mean "Wun Hung Lo Semiconductor"? 

[Another Dave]

Nope no traces of green and no water ingress directly,  but its open to the damp moisture from
an unheated shed.  Its a chinese item rebranded for several sellers. Although the spec seems
to vary slightly.  Some mention a US designed or built circuit board. But im guessing that US
is a factory in China 

A few forums have mentioned that the boards do not like the cold damp conditions of a shed.

Good idea about building the circuit back up.  Use the original parts and see whether is was the
78L05 or one of the capacitors.

Mirror the original circuit,   voltage in with one diode and 2 capacitors  a 16v 220uF and a 22uF?
on the input side.  Maybe a 10 not a 22 but im sure the 22uF is after.  Must check that.

So Voltage in (7v - 30v) through a diode and 2 capacitors, then the regulator and one capacitor on the
output which im pretty sure was the 10uF and i should have a nice stable 5v?


Why a 220uF and a 22uF capacitor in parallel?  Because they do not make a 242uF capacitor?

Is it possible to explain in dead simple terms why a 220 or 330 would not work in this
instance?   

Many Thanks...




I need to start another thread now about a caravan battery charger that appears to be
lacking in components 

[rolycat]

Why a 220uF and a 22uF capacitor in parallel?  Because they do not make a 242uF capacitor?

Is it possible to explain in dead simple terms why a 220 or 330 would not work in this
instance?   

Electrolytics typically have a tolerance of +/- 20%, so a capacitor marked 220uF might well have an actual value of 242uF.

Dave has made a video on exactly your question:

 https://www.eevblog.com/2015/05/09/eevblog-742-why-electrolytic-capacitors-are-connected-in-parallel/.

[Another Dave]

Just watched that again, I have watched it before. So it seems in my case it wont be for long life
or for cost so it must be for the capacitance 

No issues with space the board is quite large for the small number of components.

Lets see if i can test the old parts.  Missing an inductor though...

Thanks.

[Another Dave]

Seems CPC sent me a wrong item.  I ordered this...
http://cpc.farnell.com/fairchild-semiconductor/lm78l05aczx/ic-v-reg-5v/dp/SC10569

And they sent me this...
http://cpc.farnell.com/on-semiconductor/bc547c/si-n-50v-0-2a-0-5w/dp/SC06669

Although the bag does state the SC10569 code.

Just tinkering with the old part and tried with a 9v battery, silly me 10v minimum 

[rolycat]

Seems CPC sent me a wrong item.  I ordered this...
http://cpc.farnell.com/fairchild-semiconductor/lm78l05aczx/ic-v-reg-5v/dp/SC10569

And they sent me this...
http://cpc.farnell.com/on-semiconductor/bc547c/si-n-50v-0-2a-0-5w/dp/SC06669

Although the bag does state the SC10569 code.

Just tinkering with the old part and tried with a 9v battery, silly me 10v minimum 

A 9V battery should be fine - ignore the "10V min to 30V max" description on the CPC website and check the datasheet. The dropout voltage for the 78L05 is about 2V, so any input between 7V and 20V or so should work. 8V or more gives slightly better regulation, but as the voltage increases you pay an increasing price in power wasted as heat.