[Premium-Rx] Experiment: How to readjust a drifted carbon comp resistor / was: wj 205 / A10R14

[Peter Ratuschni]

Hello all and thank you for the tips.

>From the first question "how to estimate the nominal value of a burned carbon comp. resistor"
the more generell problem arised what happens when such a resistor is exposed to overload.

The last experiment showed, that the resistance of a upwards drifted carbon comp resistor decreases
under it's nominal value when it is that much overloaded that it is burned outside.

I asked, is it possible, to controll this process in a way that the resistance is brought back in
the range of the nominal tolerance with no visible damage occuring? And, if this works, is the
obtained value stable?

Disclaimer: I do not recommend these resistors for reuse, as I do not know enough about their 
properties after the described treatment. In most applications metal film resistors are 
a good if not better replacement.
 
 
Some refinements to the experiment
- First I swapped the 800VA variac against a smaller Phywe type, because it has better adjustablility
  at the needed voltage range. The Phywe is a combination of variac and a DC supply. So you have
  both, adjustable AC from about 0-25V and DC from 0-20V. For the small ohm and wattage values I am
  dealing with this is more than enough. 
- Voltage and ampere meters were added for monitoring how much power is dissipated by the resistor.
  Moreover the ohm value of the resistor under load can be calculated.
- The whole setup was moved outside because if smoke appears I do not want to breath that in.

First I repeated the former experiment, now with DC instead of AC.
The resistor was again a 47 ohm, 5%, 1/4w, probably Allen Bradley RCR07G as this type was widely
used by wj in the seventies. All these were unsoldered from the wj 205-3 receiver. The receiver
was built around 1979/80, so the parts are about 35 years old.
The 47 ohm resistor was measured 69.3 ohms at room temperature.  

This is the result table. The duration for each step was about 30 seconds to one minute (the time I
needed to calculate the two values and to write them all down). 

nominal 47 ohm carbon comp resistor, measured 69.3 ohms
 Vdc  ,  mA ,  watt, ohms , comment
4.3   ,  63 , 0.27 , 68   , slightly above rated power
8.0   , 108 , 0.86 , 74   , heat makes resistance rise
10.1  , 144 , 1.45 , 70   , resistance begins to fall
10.56 , 160 , 1.69 , 66   , smoke, little bubbles on the inner (black) ring
10.5  , 163 , 1.71 , 64.4 , 
10.5  , 165 , 1.73 , 63.6 , 
11.0  , 170 , 1.87 , 65   , rised voltage, rised resistance
11.0  , 173 , 1.9  , 63.6 ,     
11.8  , 178 , 1.96 , 66.3 , stopped smoking
12.3  , 190 , 2.33 , 64.6 , 
12.0  , 220 , 2.64 , 54.5 , starts smoking again
12.0  , 230 , 2.76 , 52.2 , 
11.9  , 240 , 2.85 , 49.6 , supply switched off

The three inner colour rings are a little burned but still readable.
Measuring the resistance (with no load) gives
after cooling down : 33.5 ohms
after ~12 hours    : 34.2 ohms
after ~24 hours    : 34.6 ohms
after ~36 hours    : 34.9 ohms
 
Conclusion: The treatment has to be stopped earlier to reach the rated resistance. 


The next candidate is a 100 ohm resistor, same type as above. It measured 132.1 ohms.
This time I avoided to make it fried - what means no smoke, no bubbles.

nominal 100 ohm carbon com resistor, measured 132.1 ohms
 Vdc  ,  mA ,  watt, ohms , comment
15.0  ,  99 , 1.49  , 152 , voltage carefully raised to this point
15.0  , 111 , 1.67  , 135 , 
15.0  , 116 , 1.74  , 129 , supply switch off

No smoke, no detoriated colour rings.
Measuring the resistance gives
after cooling down         : 104.2 ohms, This is within tolerance !

To see, if this is a stable value the resistor was set under load for 2 hours:
  V   ,  mA ,  watt, ohms , comment
4.6   ,  44 , 0.2  , 104.5, 
4.69  ,  45 , 0.21 , 104.2, 
4.8   ,  46 , 0.22 , 104.3, the resistance seems to be stable under normal load.

The voltage variation was not intended here. This could be a response of the supply to 
decreasing temperature in the evening (the setup was still outside).

after cooling down : 104.7 ohms 
after ~10 hours    : 108.0 ohms
after ~24 hours    : 109.7 ohms, It is drifting up again !
after ~36 hours    : 111.1 ohms


Conclusions:
------------
1. It is possible to reduce the resistance of an upwards drifted carbon comp resistor
   by applying overload without causing visible damage.
2. For 1/4 watt resistors the effect starts at approx. 1.4 watt.
3. For 1/4 watt resistors the power applied should not exceed 1.7 to 1.8 watt to avoid
   visible damage. The exact power dissipation where it starts to smoke may differs
   resistor to resistor. The duration how long the resistor can widthstand this
   power without burning was not investigated. 
4. If the nominal value should be achieved, the point to stop the overload treatment is
   when the resistance reaches approx. 120% of it's nominal value (calculated from voltage
   and current).

The second question, how stable is the obtained value, cannot satisfactory answered yet.
It was observed, that the resistance increases about 0.5% to 2% per day when lying in
stock with no load applied. How far this goes and if the former higher value will be
achieved is not know yet.

- 

Experiments of this kind often produce sideeffects. 
While I carefully took care that the second resistor did not go up in smoke, I forgot 
the bread in the oven - so this was almost burned.
Later, while sitting down with bread and wine I asked myself what the best woman of all
would say to such a funny job. I imagined her putting on a well known face and as a possible
reply the following words entered my mind:

Unanswered questions lead to strange activities.
This is the fate of nerds and scientists.
And it is an inherent quality of human beings.


best regards,
Peter