[HomeBrew] Small 100Khz IF transformers

[Sandy and Kees Talen]

Continuing saga of making 100Khz IFs...........

Just got through testing a few JW Miller 15-H, K-TRAN, 262Khz
IF transformers. These are the small 3/4" square units (same 
physical size as used on the HBR-11/12/13).   

Coil inductance actually goes from about 2mH to 6mH which really 
surprises me that it will go that high. They are set at approx 3.5mH 
for the 262Khz frequency. The resonating capacitor integtated into
the base ranges from 80-95pf.  
 
The coils are very densely wound (something I can't do yet with the 
coil winder) using approx 37 feet of #40 wire (not bifilar .....which
also surprises me a little). There is no room to add windings to the 
diameter of the coil (will contact the powdered iron cup), but there is 
a possibility of adding additional turns to the ends of the coil. The 
cup provides higher permeability and self shielding, something the 
JW Miller 100Khz #1710 and #1709 don't have but they do have a 
flexible powdered iron material inside the can.

At 95% of the maximum inductance (should be lower to allow tuning 
room) the coil "Q" is "74" which is very close to what the 100Khz 
JW miller IFs measure.

Calculations show that a 430pf padder added to the 80-95pf internal
capacitor will bring it to 100Khz. This is more capacitance than 
I would normally add as additional padder, however the starting 
point is 80-95pf. This brings the total to approx 520pf.....that's high. 
The JW Miller 100Khz IFs use 333pf. Need to make some more 
"Q" compares but my JW Miller #1710 is about useless (patched 
broken wires at the coil 3 times) due to the multiple times removing 
it from the can, unsoldering wires, etc.

All the above said, the Spectrum Analyzer does not lie (the real circuit
application would be better) and at -6dB, the BW is 3.8Khz and the 
-20dB BW is 19.2Khz ....really pretty good. The JW Miller #1710 is 
measured at 3.5Khz and 15Khz. You can add another IF stage and 
put two IFs in series (HBR-13). The overall insertion loss is greater 
than the #1710 but .....close enough. 

Since the #1710 and #1709 transformers are "unobtanium", these 
JW Miller 15-H IFs should be more readily available (used in many
car radios) and ought to work ....and they have the cool JW Miller 
label just like the #1710 and #1709.  

After many attempts with the coil winder, I've been able to wind
several 6-8mH coils (1-#40, 2-#40, etc) needed for the 100Khz 
transformers. The coils are 1:1 and 2:1 spindle to cam ratios 
and really look pretty good. But I am unable the get the one turn 
adjacent to the next where only the layer to layer capacitance is 
reduced. The other dramatic effect of doing this is to reduce the 
physical size of the coils which you need to fit inside the 3/4" can. 
If you get too close to the can, capacitive effects reduce the 
inductance, more inductance would drive you closer yet and 
changing the aspect ratio reduces inductance.

Another technique would be to just take an old 455Khz slug tuned 
form, rip off all the wiring and wind a single high "Q" 8-9mH coil 
with a small resonating capacitor inside the 3/4" can ....fits easily.
(I can wind those with 2-#40). Then capacitively couple these 
between stages. There has got to be a drawback of doing that 
but I haven't found it yet ?? Another possibility is to take these 
retuned 262Khz IFs and put two in series which really narrows up 
the passband as shown with the BC-453 IFs (HBR web site). But 
it will require another IF stage because of insertion loss.

The only thing a little iffy is that the 262Khz units use 1-#40 which 
is about 14ma DC (but was designed for "6BE6 and 6BA6 tubes 
with 240V B+" according to the sheet). This is also what allows
the winding density. The 100Khz #1710s are wound with 2-#40 
....28ma.  I was going to try 3-#44 but don't have any of that.

Sure has been a fun learning experience (albeit reinventing the 
wheel). I know a lot more about coils. Also solving the right
angle drive problem on the coil winder was fun too (on the web 
site under coil winder).

73s  Kees K5BCQ
Radio homepage  http://www.qsl.net/k5bcq/