[Hallicrafters] SR-2000, SR-400, SR-400A and HA-20 Temperature compensating capacitors

[hallicrafterssr2000 at k9axn.com]

Good afternoon:
Found a great replacement for the difficult to find dog bone caps in the VFO of the SR-400, SR-400A, SR-2000, HA20, and other radios. These caps are a superb replacement for the dog bone type.
Search for Ceramic Plate Capacitors using advanced search with the International attribute on EBAY --- yes EBAY. If you have a better source let us know. Ceramic plate capacitors made by Philips, Mullard, or BCE were used for circuit trimming and were probably the most exquisite temperature compensating caps manufactured. 
Their attributes tell the story. They are rated at 100 volts but test by the manufacturer to 300 volts. I tested them to 600 volts with no damage or value warp. The N150 are PG, 2% caps and +/- 30 PPM. Before you disqualify the +/- 30 PPM, remember, even the C0G/NPO and all other commercially available temperature compensating caps are +/- 30 PPM whether they are 1%, 2%, 5%, etc. The +/- 30 PPM is universal regarding the temperature compensating value. The 2% represents actual value of the cap.
The original caps used in the SR-400, SR-400A, and SR-2000 VFO were 5% caps, these are 2%. 
Another attribute to keep in mind is the reaction time of these caps. The original dog bone caps came in at least three reaction speeds. The reaction times of the caps used in the Hal VFOs were considered medium to fast. Caveat, The late model VFOs switched to ceramic disc temp comp caps that are slow to react. Is this a problem? No, it will slightly extend the time to stability. You can expect these VFOs to reach stability in any part of the tuning range in about 1 hour. After that they will wander +/- 10 of so cycles until you return from Holiday. 
NOTE: The Ceramic disc caps used in the late model radios reach stability about 20 minutes later than the dog bone or Ceramic plate caps. 
What can you expect from the above VFO design?
Tune the VFO to 250, turn the radio on from a cold start, measure the frequency after 2 minutes then at 1 or more hours. If all is perfect, the frequency will be +/- 20 cycles and wandering 10 cycles from the 2 minute measurement. 
Tune the VFO to 0 Kc, the actual high frequency end, turn it on from a cold start, measure the frequency at 2 minutes then at 1 hour or more. You should measure approximately +325 cycles from the 2 minute measurement with the majority of the drift occurring in the first 1/2 hour.
Tune the VFO to 500, the low frequency end, turn it on from a cold start, measure the frequency at 2 minutes then at 1 hour or more. You should measure approximately –325 cycles from the 2 minute measurement with the majority of the drift occurring in the first 1/2 hour.
That is a perfectly compensated CTO VFO by design. In the real world, due to 5% components and +/- 30 PPM for compensation, you will likely find the neutral/no drift point other than at the 250 Kc point, but non the less somewhere in the 0 to 500 range. 
The +325 total drift at the low end and –325 at the high end is a result of the variable tuning capacitor having a temperature compensating value of +80 PPM. 
if it is fully meshed, the + temp comp drift is significant and if unmeshed very little + temp comp drift. The temp comp in this VFO is designed to be neutral at the 250 Kc point.
The Halicrafters engineers worked with some very forward thinking VFO compensating designs to negate the effects of the variable tuning cap. A bimetallic end plate and or series negative compensation was considered. What they decided was, the conventional use of wide spaced copper plates was more than adequate for even the most demanding customer.
Don’t believe you will find a CTO or PTO VFO that will outperform these instruments. 
Hope you found this of interest.
Kindest regards Jim K9AXN