10k
Plastic
- Joined
- Nov 14, 2012
- Location
- Houston, Texas, USA
(This is an article Joe Williams and I wrote for the Home Metal Shop Club)
There was a need to obtain a replacement coil to restore a DC contactor for a Monarch 10EE lathe of 1943 vintage. The coil was part of a pre-built assembly of multiple contactors and relays made by Struthers Dunn (RPXA595A). The original coil must have failed at some point in the past, and a modern solenoid coil had been installed – and then burned up (it measured a dead short), along with the Reverse selector switch.
It would not be good to use another modern coil like the one that burned up. Although an identical coil could be found, it was likely the wrong coil in the first place. And forget about finding any repair parts.
Fortunately, the design employed two identical coils – one for Forward and one for Reverse. Only one burned up. The good coil was removed for testing. There was a very faint number fragment on the good coil, but it didn’t provide any useful information.
It was decided to make a replacement coil. First, the physical parameters of the coil were measured:
Hole in center of spool: 0.50" x 1.22” long
Wire on spool from: 0.625 ID to 1.42" OD (including tape around outside)
Length of space for wire: 1.08"
Then, the electrical values were measured after removing the iron core from the coil:
Value: 3.25 H (Using a Sencore capacitor/inductor analyzer)
R: 2183 Ohms
Wheeler's approximations from 1928 can be used to calculate coil parameters. His approximation for a multi layer air core coil with a rectangular cross section is:
L (uH) = 0.8 * a^2 * n^2 / (6*a + 9*b + 10*c )
where
a = average radius of windings
b = length of the coil
c = difference between the outer and inner radii of the coil
n = number of turns
all dimensions in inches.
Other formulas are necessary to calculate other parameters, but fortunately, there’s a web page that lets you calculate coil parameters easily.
Input information is the Inductance, the coil ID and length, and the wire gauge. The wire size was unknown, but several of the output parameters were known. It was possible to guess the wire size and see if it resulted in something reasonable.
The most important output parameter is Resistance. A low resistance would likely result in another burned up coil because of the relatively high actuation voltage.
Assuming 36 AWG wire gives:
Coil OD: 1.46"
R: 1539 Ohms
L: 3710'
This is probably not right, because R is too low and the coil OD is too high
Assuming 37 AWG wire gives:
Coil OD: 1.33"
R: 1882 Ohms
L: 3676', 68 layers of 212 turns
This looks good!
Assuming 38 AWG wire gives:
Coil OD: 1.2"
R: 2360 Ohms
L: 3642'
This looks like too big for R and too small of a coil
So, the best guess is that the original coil used 37 AWG wire.
The next thing to check is to see if the wire selected can withstand the expected current. The nominal voltage for a 10EE DC panel is 115 VDC. The current through the wire is calculated as follows:
I = V/R = 115/1882 = 61mA.
Various sources list the ampacity of 37 AWG wire as about 28mA, assuming 700 circular mils per Amp. The calculated current is more than twice this amount.
This results in a dilemma. 34 AWG is the smallest wire that can handle the current. But as seen above, larger wire won't give you the right parameters for coil size, inductance and resistance. This won't work. A series resistor could be used to reduce the current. It would likely be a multiwatt, ceramic resistor, and there was no evidence of this on the original design. So the decision was made to wind the coil as calculated, test it after installation, and hope for the best.
The new coil form was made on a lathe from a block of Nylon. Other insulators could have been used, but this was on hand. A 4,000’ reel of #37 AWG copper wire rated to 200C for the coil was obtained inexpensively on eBay.
No special equipment was available to wind a coil. Instead, a small lathe was used to turn the coil. A small motor could have been used, or even a hand drill if it was mounted in a vice and didn’t turn too fast.
For this tiny wire, it’s not necessary to wind the wire so that it’s precisely spaced, so no special mechanism is needed. Electrically, the coil will have the same value whether the wire is perfectly spaced or not. A side note about the wire – it’s about 5 mils in diameter. Human hair ranges from 1.5 to 5 mils in diameter! This is very small wire.
There was a need to obtain a replacement coil to restore a DC contactor for a Monarch 10EE lathe of 1943 vintage. The coil was part of a pre-built assembly of multiple contactors and relays made by Struthers Dunn (RPXA595A). The original coil must have failed at some point in the past, and a modern solenoid coil had been installed – and then burned up (it measured a dead short), along with the Reverse selector switch.
It would not be good to use another modern coil like the one that burned up. Although an identical coil could be found, it was likely the wrong coil in the first place. And forget about finding any repair parts.
Fortunately, the design employed two identical coils – one for Forward and one for Reverse. Only one burned up. The good coil was removed for testing. There was a very faint number fragment on the good coil, but it didn’t provide any useful information.
It was decided to make a replacement coil. First, the physical parameters of the coil were measured:
Hole in center of spool: 0.50" x 1.22” long
Wire on spool from: 0.625 ID to 1.42" OD (including tape around outside)
Length of space for wire: 1.08"
Then, the electrical values were measured after removing the iron core from the coil:
Value: 3.25 H (Using a Sencore capacitor/inductor analyzer)
R: 2183 Ohms
Wheeler's approximations from 1928 can be used to calculate coil parameters. His approximation for a multi layer air core coil with a rectangular cross section is:
L (uH) = 0.8 * a^2 * n^2 / (6*a + 9*b + 10*c )
where
a = average radius of windings
b = length of the coil
c = difference between the outer and inner radii of the coil
n = number of turns
all dimensions in inches.
Other formulas are necessary to calculate other parameters, but fortunately, there’s a web page that lets you calculate coil parameters easily.
Input information is the Inductance, the coil ID and length, and the wire gauge. The wire size was unknown, but several of the output parameters were known. It was possible to guess the wire size and see if it resulted in something reasonable.
The most important output parameter is Resistance. A low resistance would likely result in another burned up coil because of the relatively high actuation voltage.
Assuming 36 AWG wire gives:
Coil OD: 1.46"
R: 1539 Ohms
L: 3710'
This is probably not right, because R is too low and the coil OD is too high
Assuming 37 AWG wire gives:
Coil OD: 1.33"
R: 1882 Ohms
L: 3676', 68 layers of 212 turns
This looks good!
Assuming 38 AWG wire gives:
Coil OD: 1.2"
R: 2360 Ohms
L: 3642'
This looks like too big for R and too small of a coil
So, the best guess is that the original coil used 37 AWG wire.
The next thing to check is to see if the wire selected can withstand the expected current. The nominal voltage for a 10EE DC panel is 115 VDC. The current through the wire is calculated as follows:
I = V/R = 115/1882 = 61mA.
Various sources list the ampacity of 37 AWG wire as about 28mA, assuming 700 circular mils per Amp. The calculated current is more than twice this amount.
This results in a dilemma. 34 AWG is the smallest wire that can handle the current. But as seen above, larger wire won't give you the right parameters for coil size, inductance and resistance. This won't work. A series resistor could be used to reduce the current. It would likely be a multiwatt, ceramic resistor, and there was no evidence of this on the original design. So the decision was made to wind the coil as calculated, test it after installation, and hope for the best.
The new coil form was made on a lathe from a block of Nylon. Other insulators could have been used, but this was on hand. A 4,000’ reel of #37 AWG copper wire rated to 200C for the coil was obtained inexpensively on eBay.
No special equipment was available to wind a coil. Instead, a small lathe was used to turn the coil. A small motor could have been used, or even a hand drill if it was mounted in a vice and didn’t turn too fast.
For this tiny wire, it’s not necessary to wind the wire so that it’s precisely spaced, so no special mechanism is needed. Electrically, the coil will have the same value whether the wire is perfectly spaced or not. A side note about the wire – it’s about 5 mils in diameter. Human hair ranges from 1.5 to 5 mils in diameter! This is very small wire.