Are these LM35DZ parts bad?

Thread Starter

Roderick Young

Joined Feb 22, 2015
408
I just got quantity 5 of LM35DZ (temperature sensors) from China via eBay. Put one in a circuit, 5 volt power supply. The output is floating. I checked the other parts with an ohmmeter. There doesn't seem to be any conduction between any of the pins. I would think that when I measure across the power pins, there should at least be a tiny bit of current flowing? It's as if there are just 3 wires coming out of a TO-92 package, with nothing inside. The parts are marked

(N)36RA
LM35
DZ

where the (N) is the National Semiconductor logo.

I'm thinking the parts are all bad, but have no experience with how this component works. Am I doing anything wrong? The seller has an excellent 99.8% feedback rating.
 

JohnInTX

Joined Jun 26, 2012
3,913
These are simple, active output sensors. If you have them hooked up right, you get 10mv/C out. Just like that.
I think your main problem is from China via eBay.
Jameco is right up 101 from San Jose.
Mouser and Digikey are both franchised TI distributors that will send you real, non-counterfeit, working parts.
All of these outfits have excellent (and real) feedback ratings.
Good luck!
 

Lestraveled

Joined May 19, 2014
1,946
+1 on what John said.

Make sure you have them hooked up right. They are voltage output devices. They do not need a resistor as some other sensors do. If your hook up is right, then contact the seller, they will give you your money back before risking negative feedback.
 

Thread Starter

Roderick Young

Joined Feb 22, 2015
408
These are simple, active output sensors. If you have them hooked up right, you get 10mv/C out. Just like that.
I think your main problem is from China via eBay.
Jameco is right up 101 from San Jose.
Mouser and Digikey are both franchised TI distributors that will send you real, non-counterfeit, working parts.
All of these outfits have excellent (and real) feedback ratings.
Good luck!
Indeed. It's just a hobby project, but majority of the BOM was sourced in a ~$300 order from Mouser, except a couple parts which only Digi-Key had. As I was building the board, I found that I had neglected to order the LM34DZ. It was a non-critical subsystem, so I thought, I can afford to wait, I'll just order from China, and get 5 of them for the same price as just the shipping charge from Digi-Key.

+1 on what John said.

Make sure you have them hooked up right. They are voltage output devices. They do not need a resistor as some other sensors do. If your hook up is right, then contact the seller, they will give you your money back before risking negative feedback.
Thanks, my first thought was exactly that I had hooked it up wrong. But no, with the flat side facing me, left to right, +V, out, GND. At that point, rather than risk burning up another one in circuit, I got out the DVM and tried to see if there was conduction between ANY pairs of pins on the remaining 4 devices. Nothing. Which is when I came to you guys (in California, the term is neuter, meaning both male and female).

I've actually had very few issues ordering things from China, before, and as you suggest, on eBay, every seller seems to roll quickly on a refund, especially on a $5 order that was DOA.
 

OBW0549

Joined Mar 2, 2015
3,101
I wouldn't use an ohmmeter to check semiconductors; most digital multimeters run their ohms function on very low voltages, like 200-500 mV. Often this is too low to forward-bias the semiconductor junctions inside a chip, resulting in an apparent open-circuit indication. Best to just power up the chip as you would in normal operation and see what it does. For an LM35 at room temperature, something around 0.25V would indicate it's operating properly.
 

dl324

Joined Mar 30, 2015
9,332
I wouldn't use an ohmmeter to check semiconductors; most digital multimeters run their ohms function on very low voltages, like 200-500 mV.
My ohmmeters have a diode check/continuity range that will measure the forward voltage of diodes up to 1.999V. It provides enough current to light LEDs; even those that have a Vf higher than 2V, it gives an open indication but the LED turns on.
 

Thread Starter

Roderick Young

Joined Feb 22, 2015
408
I wouldn't use an ohmmeter to check semiconductors; most digital multimeters run their ohms function on very low voltages, like 200-500 mV. Often this is too low to forward-bias the semiconductor junctions inside a chip, resulting in an apparent open-circuit indication. Best to just power up the chip as you would in normal operation and see what it does. For an LM35 at room temperature, something around 0.25V would indicate it's operating properly.
Fair enough. The ohmmeter was the second diagnostic, though. I was hoping that on the high-ohm range, the DVM might use its 9-volt battery for measurement. Maybe not. The first test was to hook the thing up onto a 5-volt supply and measure the output voltage, which turned out to be floating (DAC input gave widely varying values). I put the DVM between OUT and GND, and measured 0 volts. I then put the meter between +5V and OUT, and again measured 0 volts. If I got even a small voltage, I might have thought that the parts were just poorly calibrated, but the output pin was tri-state, bolstering my belief that I was sold 5 molded pieces of epoxy with printing and leads coming out, but nothing inside.

@Roderick Young
Make sure you connected correctly. The pinout is showing image from bottom of part -looking at pins. For to-92 package.

You should measure 0.2v at room temp.
I was thinking that the flat side is the flat side - perhaps someone can double-check my reasoning?

BadLM35DZ.jpg
 

dl324

Joined Mar 30, 2015
9,332
bolstering my belief that I was sold 5 molded pieces of epoxy with printing and leads coming out, but nothing inside.
I'd pop one open to see...

They don't work, so there's nothing to lose at this point. The Sellers on eBay who are knowingly selling counterfeit products will eventually get around to giving you a full refund to keep under the radar. But not before giving you a sob story; I know, I've been a victim too.
 

Thread Starter

Roderick Young

Joined Feb 22, 2015
408
I'd pop one open to see...

They don't work, so there's nothing to lose at this point. The Sellers on eBay who are knowingly selling counterfeit products will eventually get around to giving you a full refund to keep under the radar. But not before giving you a sob story; I know, I've been a victim too.
Just so everyone knows, I requested a refund from the seller, and they gave me everything back, including the shipping, within hours. I did not have to return the parts. I advised them to check with their supplier, as they may have had a "bad lot". I could envision an unscrupulous business selling inferior parts, but there's no money to be made in selling DOA or dummy parts. The seller will lose the shipping every time.

At my previous job, which included doing forensic analysis, corporate had the facilities to de-cap a part, even section it and take electron micrographs if desired. Today, without the tools, I find that I have better things to do. If someone else in the US wants to try building a test circuit to figure out what's inside, or actually break one open somehow, please contact me, and I'll slip one of these TO-92's into a first class letter for your amusement. Just agree to share your findings with this forum, ok?
 

DickCappels

Joined Aug 21, 2008
6,009
It would be interesting to hear what you find if you decide to take a hammer to one or two of them, or crush them in a vise. Is there really a die in there or is it merely an encapsulated lead frame?
 

Thread Starter

Roderick Young

Joined Feb 22, 2015
408
It would be interesting to hear what you find if you decide to take a hammer to one or two of them, or crush them in a vise. Is there really a die in there or is it merely an encapsulated lead frame?
It's a regular plastic TO-92. The one in the picture in reply #8 is one of the actual parts. I'm just too lazy, and not curious enough to open one to find out.

@Roderick Young
Send one to me and I will mill the top off and post some photos.
Send me your address in a private message, and you can have all 5. The one with solder on the leads will be the one that was installed in-circuit. The other 4 are virgin. Well, not really, I think they all have been f***ed.
 

Thread Starter

Roderick Young

Joined Feb 22, 2015
408
Just so everyone knows, I ordered 5 more LM35's from eBay, but this time from a vendor in San Francisco (close enough so I could go over there and beat him up if something went wrong). The new parts are all perfect - register about 180 mV output, going up to about 200 mV when I pinch my fingers around them, and much less when I put an ice cube on them.

It just occurred to me that maybe the bad parts could have been MOSFETs (without a body diode). If I had one of those universal testers, I could have plugged them in to see.
 

Lestraveled

Joined May 19, 2014
1,946
@Roderick Young

OK, I got the parts.



I checked them with my ohm meter and they look to be NPN transistors. My curve tracer was ready to go so I plugged one in and Lo and Behold, they are NPN transistors.




So mystery solved, they are not counterfeit parts, they are fraudulent parts, NPN transistors that had the part number changed.

Oh, and Thank you Roderick for the Christmas card.
 

Attachments

Thread Starter

Roderick Young

Joined Feb 22, 2015
408
look to be NPN transistors
No way! The diode test setting on my multimeter should have found the base-emitter and base-collector junction when I tried all possible lead permutations. I must have been totally out of it. Could have sworn I tested every single one on diode test, and they all acted like bare leadframes. How embarrassing.
 
Top