The Mississippi Historical Radio and Broadcasting Society
| Vol 3 #5 | September 8, 1993 |
News Letter
Our regular monthly meeting was held August 8th, 1993 at the usual place and time. Attendance at this meeting was poor to say the least. Several significant items of business were acted on. First, the time and day of our regular meetings were changed. Club meetings are now on the second Tuesday of each month at 7 PM. Location remains the same. Perhaps Sunday was just too busy for most folks. Monday is often a Holiday, Wednesday is a church night for many, Friday is date / party etc. night, as is Saturday. That left Tuesday and Thursday -- Tuesday was chosen. The next major order of business was setting the date for our Spring Meet. Saturday, March 12th. was chosen. This should be a nice weekend, no other shows are scheduled (yet) for this weekend, and the J. C.'s building is available -- and has been reserved for us. The last major bit of business was a decision about where to hold our fall "Show and Tell.". Two meetings ago there had been some discussion about holding it somewhere other than the Village Fair Mall so a swap meet, auction etc. could be held. However, after two months, no alternative sight was put forward, so the decision was made to see if we would be able to get an available date in November. The Mall has invited us on November 5th and 6th, and has noted that we will have the Mall to ourselves (i.e. no car, crafts, arts etc. shows). The Mall is booked from then until Christmas, as well as the two previous weekends (Antique car show on the 22nd & 23rd, and a Crafts show the 29th & 30th); so we are glad to get this date.
So here are some dates to mark on your calendar over the next six months:
Sept. 14:Regular Monthly Meeting
Oct. 12:Regular Monthly Meeting
Nov. 5&6:Show & Tell at the Mall
Nov. 9:Regular Monthly Meeting
Dec. 14:Regular Monthly Meeting
Jan. 11:Regular Monthly Meeting
Feb. 8:Regular Monthly Meeting
Mar. 8:Regular Monthly Meeting
Mar. 12:Annual Meet & Show
Radio Collecting Today
{Editor's Note: The meter that is the subject of this article is in the West Wing of the Museum - it is that 1936 Philco multimeter. The meter scale referred to in this article can be seen in that photo}
After an absence of a couple of months, this topic returns with a series on "reverse engineering" or how to figure out what the value of that mystery part is.
This particular project is based on the restoration of a 1936 Philco VOM., however the principles are the same whether you are working on something similar, or a radio or whatever. The key is knowing what clues to look for, where to find them, and applying a little theory and a lot of pencil and paper -- calculator optional!
This meter was designed a long time ago, and the chances of finding a schematic are probably the same as winning the $10,000,000 Publishers' Clearing house Grand Prize. So what do you do when about 1/3 of the resistors in a something are open -- and the rest are highly suspect?
Get a pencil and a pad of paper, sit down, take a deep breath and use your head. Here's what we have: a meter that has many high precision resistors (marked like 467K, 33.7K etc.) that measure way out of tolerance. So you figure the unmarked resistors are probably way out of tolerance as well. So measuring them to determine what value they are supposed to be isn't an option. You are going to have to figure the whole thing out, draw your own schematic, and determine the correct values for all of the resistors in the meter. Impossible? Not at all. First, the meter will tell you several important items: The system current (which gives you the system impedance), the value of the Ohms resistors and how they are connected (there is a slight trick to this one).
All of that from looking at the meter? Almost. All but the first item -- for that you need a 100 Ohm resistor, a 1K pot, a 9V battery and a milli-ammeter. Hooking the battery across the pot will give you 0 - 9 volts on the wiper. Make sure the pot is set so there is 0 volts between the center arm of the pot and battery minus. Hook the 100 Ohm resistor, the milli-ammeter and the meter under test all in series between the center arm of the pot and the battery negative terminal and SLOWLY adjust the pot until the meter under test reads exactly full scale. You now have the design current (and impedance) for the meter (almost -- it will be refined later). In this case the meter indicated full scale at 1.128 milli-amps. 1volt divided by .001128 is 886 ohms per volt. The mid scale reading on the ohms scale is 3.5. We'll come back to that. Looking at the precision divider resistors, it is noted that the value written on the highest divider, the one between the 1000V and the 300V terminals is 467K. Since we know that at full scale this resistor must drop 1000 - 300 or 700volts, then 700/467K = 1.5 ma. -- real close to what we measured on the meter - but is shows that the range is calibrated by a shunt across the meter to increase the current from .01128 to 1.5 milli-amps. There are two reasons for doing it this way 1) no two meters are alike: ESPECIALLY ones this early -- so the current for full-scale may (and usually does) vary widely. By using a system current (in this case 1.5 mils.) that is larger than any meter might need, the divider chain (i.e. the dividers for 1000, 300, 100, 30 & 10V) can be the same values for every tester built, while the one shunt resistor across the meter is chosen to calibrate the whole system. 2) AC has a different set of problems because of the diodes used to convert the AC into DC for the meter movement -- So the shunt for AC needs to be different for DC. The function switch hooks the correct source and the correct shunt to the meter. In this case, the divider chain resistors were all identified. What is they weren't? Well we found a current of 1.128 mills. Round it up a bit to say 1.25 mills. (to allow for calibrating shunts). Then lets guess at a system current of say 1.33 mills. 1V/.00133 is 750 ohms per volt. Our 1000V resistor (which drops 700V in the chain) would then be 700/.00133 for a value of 527K. We could design a new divider chain using this format all the way:
300V = (300-100=200V) 200/.00133=150K
100V=(100-30=70V) 70/.00133=52.7K
30V=(30-10=20) 20/.00133=15K
10V=10/.00133=(7.5K - meter resistance)
Oops -- slipped in an extra item here, didn't I? Measure the meter movement -- in this case 52 ohms. To protect the movement, we need something in series with the meter to limit current in case of overload -- a ratio of 3 to 1 is common, so use a resistor around 150 Ohm to make a total resistance of 200 Ohm. The shunts and inputs go across the meter and resistor combination. By using an adjustable resistor across the meter, the current of the system can be balanced for full scale reading with full scale input. Since a separate shunt is used for AC, it can also be calibrated. The shunt resistors used for current measurements can now be calculated as well:
.00133A*200 Ohm = .226 Volts across the meter (and it's series limiter) to give full scale reading. So:
10ma. = .226V/.01=22.6 Ohm
100ma=.226V/.1=2.26
10A= .226V/10=.023 Ohm
That leaves the ohmmeter parts. There are two types of circuits used in these types of testers, one is series, the other is shunt. The meter also tells us the answer to which this tester is: If the Ohms scale reads (reading left to right) 0 to infinity, the circuit is shunt. If it reads infinity to 0, it is series. This meter reads infinity to 0, so it is series. A ways back I said something about the center scale reading on the ohms scale meaning something. It is this: because we are actually reading a ratio of a known to the unknown (resistor being measured), the results are logarithmic -- that's why the scale is weird. To measure low values of resistance the meter measures the voltage across the known resistor. Since the voltage is a ratio, the center scale reading is the value when the known and unknown are the same -- times the multiplier, if any. So in this case, the Rx1 resistor is 3.5½ (less a fraction for the resistance of the wiring & connectors). The Rx100 resistor is 350½. The meter, as we said is connected to read the voltage across the known --so a resistor is placed in series with the meter to set full scale deflection with a low battery. This is usually done by trial and error; the calculation: V(battery low) /meter current
will give you a starting point: Example 1.35V/.001128A=1200 Ohm
This assumes a 10% low battery. A shunt resistor in the form of a variable resistor across the meter serves to shunt the current to allow the meter to be zeroed on fresh batteries. For high resistances the unknown is placed in series with the meter. An example: Rx10K we know that our resistor should be 35K (3.5*10,000) How many volt battery? E=IR so .001128*35000=40V. A 45Volt battery would be used (remember our 10% margin for weak batteries!). Our variable shunt across the meter again allows us to set a 0 Ohm reading with the leads shorted together. The variable shunt is switched in by a switch mounted on the control in this unit; modern testers use the function switch to switch in the Ohm's zero shunt. In use the unknown is inserted in this series loop: a resistor of 35,000½ would cause the current to drop by 1/2, indicating 3.5 on the meter (times our switch set multiplier of Rx10,000).
Next time we will continue on this project looking at the actual values and how they worked out -- as well as expanding these techniques for use in Radios and other equipments. Calculating that burnt up screen resistor isn't so hard after all!
Until next time -- Happy Collecting!
P.S. Don't forget Tuesday night at 7 PM!!!!!
The Mississippi Historical Radio and Broadcasting Society Newsletter is published monthly by:
The Mississippi Historical Radio and Broadcasting Society
2412 C Street
Meridian, MS 39301
601 693-5958
© 1993, The Mississippi Historical Radio and Broadcasting Society.
Back to Archive Page