Is resistance of a silicon diode constant?

[iampaul]

In our experiment we measured the forward resistance of a diode using a multimeter. We used that value and the voltage across the diode to compute the current passing through the diode.
Is that value really constant??

 

[NascentOxygen]

In our experiment we measured the forward resistance of a diode using a multimeter. We used that value and the voltage across the diode to compute the current passing through the diode.
Is that value really constant??

Your calculated value will change with temperature, and will also be different for different values of current.

Only where a device's I vs. V graph is a good fit to a straight line that passes through the origin can it be said to have a constant resistance, for the measurement you performed. Does the usual diode I vs. V graph look like that to you?

 

[yungman]

You cannot measure resistance of the diode using a multimeter. The voltage on the multi meter turn on the diode and give you some reading. Some meter only measure some reading at say 2K full range setting, other setting will give you an open circuit.

Diode has a minimum turn on voltage something around 0.4V to 0.5V to conduct very very little current. When it starting to turn on, then the resistance is inverse to the current. Without going into detail, if you pass 1mA at 25 deg C, the resistance is about 25Ω, 10mA give you resistance about 2.5Ω. But at this level, parasitic resistance kick in also. This is only the ball part number. This is call

[tex] r_e≈\frac 1 {gm}=\frac {V_T}{I_c}\;\hbox { where }\;V_T=\frac {kT}{q}≈25mV[/tex]

 

[Redbelly98]

In our experiment we measured the forward resistance of a diode using a multimeter. We used that value and the voltage across the diode to compute the current passing through the diode.
Is that value really constant??

As others said, it's not constant. But the ohmmeter reading is telling you the ratio of voltage/current through the diode||voltmeter combination, at whatever voltage and current happen to be provided by the ohmmeter.

Your calculation of current might be correct, it depends whether these two conditions are being met:

1. You measured the diode voltage with a separate meter, at the same time you were measuring resistance with the first meter.

2. The resistance of the 2nd meter, when measuring voltage, is a lot higher than the resistance reading of the 1st meter. You can measure the resistance of the 2nd meter in voltage mode using the first meter, without the diode, and see what you get. 10 MΩ is fairly typical of a digital meter, while an analog meter could be in the ballpark of 100 kΩ.

(Even if condition #2 isn't met, it might be possible to get the diode current if you provide us with your actual readings.)

 

[jim hardy]

I take it you're working up the learning curve.
One needs to understand his test equipment lest he get fooled by it in situations like you described.

most multimeters measure ohms by applying a current then measuring the voltage drop.

The current applied depends on the range selected.

Take your DMM out of autorange and measure diode on several differnt ranges RX1, RX10, RX1k etc. What do you notice about result ?

It's more fun with an analog multimeter - try this trick
http://www.kpsec.freeuk.com/images/scale.gif
on ohms scale what number is mid-scale?
The analog meter places that many ohms in series with internal battery to make a current source.
Internal battery is usually a 1.5 volt flashlight cell.
When you complete the circuit through an unknown resistance, current flows.
The meter measures the voltage across that internal resistor and the scale is made so it reads in ohms instead of volts.
Full battery voltage is full deflection meaning there's no drop across unknown resistor so its value must be zero, that's why zero ohms is over there on right hand end of scale.
Conversely if no current flows the unknown must be infinite and sure enough look at the little infinity sign on left end of scale !
Draw it out on paper, it's a clever scheme.
Clearly at midscale the nattery voltage is dropped half across unknown resistor and half across the internal one, and it's simple algebra to figure out what ohms to write on the scale for any deflection. That's why the scale is nonlinear, it's a voltage divider.

That's why a diode forward biased reads about 2/3 scale on an analog meter regardless of scale - 1/3 of battery voltage (~1/2 volt) is dropped across the diode and the rest across internal resistor..

So on a DMM use the diode scale it applies a milliamp or two and reads voltage directly.
But you have to outsmart the analog meter.

Now take two DMMS, place one on milliamp scale and other on ohms. Connect them together in parallel. The one on milliamp scale shows how much current the ohmmeter is applying, try several ranges.
The one on ohms shows the resistance of the other one's internal current shunt, try several ranges.

If you are lucky enough to have a Simpson 260 analog meter* , observe that on ohms it can put out in excess of 100 milliamps. Thay's why you never check transistors with a Simpson on RX1 scale but intead use RX10. It can hurt a small signal transistor on RX1.

*** Simpson 260 was the workhorse of electronics industry for ~ five decades. Many of us old timers still prefer it for all but precision work.. Be aware it has a 9V battery for highest ohms scale RX10K so a diode will read nearly zero on RX10K.


old jim

 

[yungman]

Hey Jim,

That's good explanation. Never thought of the reason behind the Simpson meter. That's actually a little before my time. I started with Fluke. Nice to know one setting of the Simpson drive 100mA, it'll blow most of the transistors I worked with! You make me wanting to go out and buy a small needle meter. A good needle meter is better for finding shorts on the pcb.

There are some use of the good old needle meter. They are much faster to the transient stuff. Also they are more rugged. When I was working with HV stuff in spectrometers, we blew so many of those DVM because of arcing, we change to the cheap needle meter and they survive.

BTW, CH258 on Direct TV just show the movie "Blast from the pass" that you talked about. I recorded it on DVD, but I yet to watch it. I record a lot of tv shows to watch it later. I am still over one year behind in watching the shows, still on 2010-2011 season! I like to watch tv this way so I can roll the commercials and also if they have a two part series or season ending cliff hanger, I can just watch it as they are all recorded already! This is what I called efficiency! Also one good thing to put on DVDs, with my age, if I dig out the DVD from season 3 years or more ago, it's almost like a new show!

 

[jim hardy]

You make me wanting to go out and buy a small needle meter. A good needle meter is better for finding shorts on the pcb.

Ahhh Yungman i hope you do.

I wanted a Simpson for decades, but they're soo expensive ! Finally found a secondhand 260 in a pawn shop for $40. A little epoxy fixed the cracked case and it's good as new. Taut band suspension movement survived whatever blow broke the case , what a great invention that was.


Observe today's better digital meters try to give a simulated analog display via the line along bottom of numbers.

Once a fellow gets a feel for the analog meter it excels at finding resistive or intermittent contacts,
the "output" jack blocks DC so you can measure ripple on a power supply or across a filter capacitor(primitive ESR test);
with practice you can estimate capacitor values by how far the needle deflects on ohms when you switch polarity;
and with practice you can get a go-nogo estimate on inductance of a transformer by how much the needle delays its upswing on RX1.

Analog Rules !

old jim

ps - another movie for you would be "Space Cowboys" - old engineers sent up to repair an aging satellite that the younger guys can't figure out because it's from from the vacuum tube days. Reminded me of our magnetic amplifier voltage regulators.

 

[yungman]

Ahhh Yungman i hope you do.

I wanted a Simpson for decades, but they're soo expensive ! Finally found a secondhand 260 in a pawn shop for $40. A little epoxy fixed the cracked case and it's good as new. Taut band suspension movement survived whatever blow broke the case , what a great invention that was.


Observe today's better digital meters try to give a simulated analog display via the line along bottom of numbers.

Once a fellow gets a feel for the analog meter it excels at finding resistive or intermittent contacts,
the "output" jack blocks DC so you can measure ripple on a power supply or across a filter capacitor(primitive ESR test);
with practice you can estimate capacitor values by how far the needle deflects on ohms when you switch polarity;
and with practice you can get a go-nogo estimate on inductance of a transformer by how much the needle delays its upswing on RX1.

Analog Rules !

old jim

ps - another movie for you would be "Space Cowboys" - old engineers sent up to repair an aging satellite that the younger guys can't figure out because it's from from the vacuum tube days. Reminded me of our magnetic amplifier voltage regulators.

Ha! I have to try the other tricks if I ever get one, I just know it is superior in finding the origin of the short.

I saw the Space Cowboys many times, that's a very good movie.