EEVblog Electronics Community Forum

Electronics => Beginners => Topic started by: the_raptor on December 03, 2010, 10:43:08 am

Title: How easy is it to kill a resistor?
Post by: the_raptor on December 03, 2010, 10:43:08 am
Hi guys,
I am currently assembling a Ten-Tec shortwave receiver kit (http://www.tentec.com/index.php?id=51) and have just struck a problem on the first stage of assembly (the audio amp). I hooked up everything to perform the test as directed in the manual which involves listening to the speaker output to make sure there is no "motor-boating" (a sign of unwanted circuit oscillation) and then touching a resistor with a finger or a test probe to produce AC hum. In doing this test I noticed a fairly faint oscillation which was NOT AC hum and seemed to be "motor-boating" so I went through the circuit with my multi-meter and found a 47k ohm resistor (yellow-violet-orange) that was measuring 10k ohm.  :'(

In the pre-assembly checking I am pretty sure I checked all components with my meter so I am pretty sure this resistor was fine before it got soldered in. So I am thinking I either damaged this resistor when I soldered it in (at worst it might have got a second or two more of heat then other components), touching a nearby resistor with the test probe of my meter somehow damaged a resistor that was one resistor and a cap away, or using 9v battery pack instead of 12v somehow didn't kill the audio amp but some random resistor (which is at the base of an NPN transistor on the other end of the circuit).

Is it really this easy to kill a resistor or was this one bad and snuck through testing?

P.S. The really annoying thing is that I am just getting into electronics again and have no 47k ohm resistor spare and don't really want to "spider-web" in a replacement, so this kills my weekend project.

P.P.S How do you even tell what watt rating a random resistor in a kit has? These ones look to me to be "standard" through hole resistor size that I remember from my youth.
Title: Re: How easy is it to kill a resistor?
Post by: Neilm on December 03, 2010, 11:51:32 am
Did you unsolder the resistor to measure it or did you measure it in circuit?

If you measured it in circuit then there might be a "sneak" path as you measured the 47k ohm resistor. This sneak path would act as a resistor in parallel with the 47k and reduce its resistance. To be sure that resistor is broken, unsolder it and remeasure it.

In answer to the question, generally resistor values change only if they are over powered or over voltaged, assuming that they are not blown up doing it. For instance I put 4 times the power on a resistor for 30 mins and the resistance increased by about 20%

Neil
Title: Re: How easy is it to kill a resistor?
Post by: saturation on December 03, 2010, 11:59:49 am
Hi

The smaller the watt rating, the higher the likelihood that overheating during soldering can damage it.  

The watt rating is unfortunately, to eyeball them.  Match 2 different sized resistors you actually have to this image but scale it to life size, then you have an estimated watt scale.

(http://www.bcae1.com/images/jpegs/resistor.jpg)

From left to right, the wattage ratings are:7 watt, 10 watt, 2 watt, 1 watt, 1/2 watt, 1/4 watt and a 1/8 watt.

Neilm is right, in circuit measurement of a resistor is likely to be off, some older DMM have a low volt ohms capacity but the most reliable way unfortunately is to desolder it.
Title: Re: How easy is it to kill a resistor?
Post by: the_raptor on December 03, 2010, 12:43:41 pm
Thanks for the info guys, I haven't had any trouble measuring resistors in circuit before but I will desolder this one tomorrow and check again. The other thing that makes me suspicious of it (apart from the wrong reading) is that the meter very slowly reaches the value of ~10k Ohms even when I set the range manually (I am using the "I won't call you a dickhead" Digitech meter from Dave's $50 round up). But if Neilm is right that could be another symptom of the "sneak" path (though I can't spot anything with the mk I eyeball).

I am really hoping my "motor-boating" problem is due this resistor (or just some interference) as I couldn't get a reliable reading on the large electrolytic caps this circuit is using before I installed them*, so it would be a major pita to find the problem if I eliminate this resistor.

* This is one of those areas where more expensive meters are worth the cost.
Title: Re: How easy is it to kill a resistor?
Post by: alm on December 03, 2010, 03:48:43 pm
Thanks for the info guys, I haven't had any trouble measuring resistors in circuit before but I will desolder this one tomorrow and check again. The other thing that makes me suspicious of it (apart from the wrong reading) is that the meter very slowly reaches the value of ~10k Ohms even when I set the range manually
This sounds like some significant capacitance in parallel with the resistor, the current through the capacitor will decrease until it's fully charged. Desoldering one leg of the resistor should clear this up.

Killing a resistor is usually fairly hard (unless you overload it with several times the rated power), I wouldn't expect a few extra seconds of soldering heat to do much damage. Resistors usually have signs (i.e. discoloration) when overloaded.
Title: Re: How easy is it to kill a resistor?
Post by: the_raptor on December 03, 2010, 04:13:13 pm
This sounds like some significant capacitance in parallel with the resistor, the current through the capacitor will decrease until it's fully charged. Desoldering one leg of the resistor should clear this up.

The resistor is connected to the base of an NPN transistor one of whose legs is connected to a 100 micro-f electrolytic cap which goes to ground. However I can measure the resistor on either side of the dodgy one and they give me the correct readings (4.7k and 10k).

I'd post pics but my camera has flat batteries.
Title: Re: How easy is it to kill a resistor?
Post by: Murphy on December 03, 2010, 10:07:53 pm
Measuring resistors you should see either the value or something less than the value due to other resistances in parallel with what you're measuring. If you measure more, you know its toast resistors tend to burn themselves open. If you measure less, the easiest thing to do is test the component out of circuit.

Many components can be verified good while in circuit but if that fails they most be removed to prove bad.
Title: Re: How easy is it to kill a resistor?
Post by: Zero999 on December 04, 2010, 01:08:28 pm
You only have to unsolder one of the leads to measure the resistor.

Measuring in circuit is a bad idea, although most modern DVMs use a lower voltage than 0.6V so diode junctions shouldn't influance the reading much.
Title: Re: How easy is it to kill a resistor?
Post by: the_raptor on December 05, 2010, 02:15:03 pm
I checked the resistor out of circuit it is fine ....

I did the test again this time moving it a bit further away from sources of EMI and it works fine. On the other hand I put together a crystal radio kit instead and couldn't get that working correctly (damn Chinglish instructions, I am just going to strip the thing for parts as the circuit design is crap anyway. I mean who makes an ungrounded crystal radio!)
Title: Re: How easy is it to kill a resistor?
Post by: mikeselectricstuff on December 05, 2010, 02:52:05 pm
Quote
Measuring resistors you should see either the value or something less than the value due to other resistances in parallel with what you're measuring. If you measure more, you know its toast resistors tend to burn themselves open. If you measure less, the easiest thing to do is test the component out of circuit.
Assuming there are no residual voltages, e.g. from charged capacitors. Or forgetting to turn it off.
Title: Re: How easy is it to kill a resistor?
Post by: williefleete on December 06, 2010, 03:14:26 am
when resistors fail their resistance usually increases even if they did, such high value resistors are pretty much indestructible and a 47k resistor would need at least 150 volts to double exceed the power rating of that particular resistor assuming its a 0.25 watt jobby
the kind of voltages in a kit is usually on the order of tens of volts at most and those resistors should be fine