Collins' classic "over the top" PTO
High sophistication, low practicality.

The key question is: Is it necessary to complicate things so much to design an oscillator with very good stability and precision for a frequency of only 500 Kc?

Collins initially gained a well-deserved reputation for building solid, reliable devices, especially during World War II and later years, but over time he began to develop a tendency to provide complicated solutions to not so complicated problems. In the early 1950s, the military's rejection of the R-390 (non-A) model due to its impracticality in maintenance and repair meant that Collins had to replace it with the R-390A. It must be said that Collins was not alone in this approach, during the same decade both HP and Tektronix also had this tendency; the result was to produce expensive and uncompetitive devices (similar to what happened with cars in the 70s/80s). When the army was the main customer everything was going well, but when the army sought competitiveness at a better price, problems began for the large electronics companies that emerged from the industrial effort of World War II.

The consequence of this sophisticated design is that if you have to work on these classic PTOs, the experience can be a real pain.

The R-388 and 51J type Collins PTO



The photo above is part of an article that clearly shows the lack of practicality of these PTOs when maintenance or repair is required (similarly to what happened with the R-390 non-A model).

Masterclass on the classic Collins PTO by W6SAI — William I. Orr on the 51J Collins PTO, 6 pages — (Ham Radio Mag, 12-1969)


Test result of the 51J-5's PTO, a classic PTO used professionally by the army for years.

Operation; What should happen, but it rarely happens:

    Band #2 example (single conversion, high-side mixer)
  • When the Megacycle dial reads 2.5 Mc and the Kilocycle dial reads zero, the PTO (VLO) should oscillate at 3 Mc.
  • When the Megacycle dial reads 2.4 Mc and the Kilocycle dial reads zero, the PTO (VLO) should oscillate at 2.9 Mc.
  • And so on, until...
  • When the megacycle dial reads 1.5 Mc and the Kilocycle dial reads zero, the PTO (VLO) should oscillate at 2 Mc.
    Band #3 example (single conversion, low-side mixer)
  • When the Megacycle dial reads 3.5 Mc and the Kilocycle dial reads zero, the PTO (VLO) should oscillate at 3 Mc.
  • When the Megacycle dial reads 3.4 Mc and the Kilocycle dial reads zero, the PTO (VLO) should oscillate at 2.9 Mc.
  • And so on, until...
  • When the megacycle dial reads 2.5 Mc and the Kilocycle dial reads zero, the PTO (VLO) should oscillate at 2 Mc.
  1. The PTO frequency can be accurately checked with a frequency counter on the control grid of the 2nd mixer, V5 pin 2.
  2. When the passband tuning is in the center, the BFO frequency should be 500,000 cs.
  3. And when setting CAL you should hear zero beat.
  4. The 51J-5 BFO frequency can be accurately checked by double reading on a frequency counter at the control grid of V11, pin7 (V11 section 2). Switching from even to odd bands should adjust PASSBAND TUNING to the same reading on both bands, this is the BFO center frequency point. This procedure is only valid for model 51J-5. 51J-4 does not have an even/odd band correction circuit.
Band #2 PTO-Dial test on the 1-Mc area, 1.5 to 2.5 Mc range (mixer is high-side: 2 Mc LO-1.5 in, 3 Mc LO-2.5 in)
Mc DIAL PTO fr at zero beat (±) Shift from zero beat to 0 mark (Kc dial)   Mc DIAL PTO fr at zero beat (±) Shift from zero beat to 0 mark (Kc dial)
1.5 2,000,100 0 (hairline on 0)
2.1 2,600,160 +4,260
1.6 2,100,160 +1,150 (beyond 0)
2.2 2,700,120 +4,850
1.7 2,200,130 +1,570
2.3 2,800,100 +5,800
1.8 2,300,140 +1,870
2.4 2,900,100 +6,400
1.9 2,400,100 +2,900
2.5 3,000,140 +6,720
2.0 2,500,120 +3,300
BFO at center '0' is 500,060 cs. See 'd' above.

Band #3 PTO-Dial test on the 1-Mc area, 2.5 to 3.5 Mc range (mixer is low-side: 2.5 in-2 Mc LO, 3.5 in-3 Mc LO)
Mc DIAL PTO fr at zero beat (±) Shift from zero beat to 0 mark (Kc dial)   Mc DIAL PTO fr at zero beat (±) Shift from zero beat to 0 mark (Kc dial)
2.5 2,000,150 0 (hairline on 0)
3.1 2,600,160 +3,640
2.6 2,100,060 +470 (beyond 0)
3.2 2,700,120 +4,300
2.7 2,200,080 +1,000
3.3 2,800,150 +5,250
2.8 2,300,120 +1,500
3.4 2,900,100 +5,970
2.9 2,400,060 +2,200
3.5 3,000,140 +6,180
3.0 2,500,170 +3,000
BFO at center '0' is 500,060 cs. See 'd' above.
    Procedure:
  1. PASSBAND TUNING at 0 (BFO at center), CW-SSB, AVC FAST, CALibrator ON.
  2. Frequency Counter on V5, pin 2.
  3. Dial at 1.5 Mc (2.5 on band #3)and set zero beat.
  4. Match zero beat with '0' mark and set it with the hairline (ZERO ADJ).
  5. Take FC readings on all subsequent 100 Kc.
  6. Theoretically the differences should be the same for both bands. A more precise test should be carried out (numbers on the frequency meter change rapidly with just a slight touch of the tune knob).
- Both tables show that the frequency sampling at the PTO is wider than the frequency sampling at the dial. The PTO frequency offset should be "compacted" to fit the Kc dial markings.
- Sign values mean that the zero beat on each 100 Kc does not reach (-) or exceed (+) the 0 Kc mark on the dial for the indicated number of cycles (FC in REL).
- There is a linearity fault in the synchronization of the PTO frequencies with Kc dial marks. PTO frequency offset needs to be tweaked, but Collins designed the adjustment of these PTOs with little practicality.
- Nonlinearity occurs when the total length of the PTO frequency range does not match the total length and frequency range of the Kc dial. The movement of the PTO from the start frequency to the end frequency is "longer" or "shorter" than the Kc dial marking. Thus, the PTO frequency area must be "narrowed" or "stretched" to match the Kc dial markings.

Conclusion:

  • A deviation of 6+ Kc is outside the correction by hairline positioning, it is excessive, and must be corrected by overhauling the PTO. W6SAI explains the procedure to achieve this (see above), but Collins' design does not make the job easier (that's why I say "high sophistication but low practicality").
  • A PTO used professionally for years by the military with only linearity issues is an excellent result, confirming Collins' reputation for making high quality devices; however, the complexity of resetting these PTOs overshadows the conclusion. Collins should have considered the ease of adjustment without having to disassemble the PTO from the chassis, and also not started from the assumption that if any problems occur, the device must be sent back to the factory for overhaul.