Restoring and mod of a Collins 51S-I

LW1DSE

LW1DSE

Star Member
Hi all folks. I owned this vintage rig near the start of 2025. It was out of service, tubes and knobs missing. Lytics look good, mainly the PSU unit, some RCA jacks missing, the S-meter appears OK.

Some tubes were hard to reach and was substituted with newer with more gm (gain). The dial mechanism is complete but very hard to move. The device that move the cores into the RF coils was partially broken. Some resistors out of range, but not too far from proper value.

First work was to remove the 'lytic can and place under the chassis, fresh Jamicon 100uF 160V units. The original 4 diode rectifier was replaced by a molded 4 pin 1A 1KV unit. Once did it and using current limited power supply (aka a filament lamp 300W 220V in series in the live wire) I started the unit, and some noise come from the (external) speaker.

Once the devise looked more or less doing its job, I started modding it. The first target was to remove a npn germanium transistor acting as audio preamp in the CW and SSB modes. It and and all its circuitry, completely dismanrled. All the detectors and main audio channel is reworked. Secondary channel for remote operation, disassembled completely.

At first, the audio output tube (6BF5 designed with low gm for vertical deflection) replaced with a 6EH5, 3X the gm of the original. Socket non compatible, so except heaters, all other connections changed. The 12AX7 audio preamp is replaced with a triode-dual diode 6BN8. Cathode resistors of both removed and grounded directly. 6BN8 triode now uses 10M grid leak biasing and 0.001uF coupling cap to the audio pote.

All germanium diode detectors, completely deleted: the 4 of the product detector for SSB & CW are now a 6JU8A quad diode like 2u 6AL5 in a single bottle. Socket added to the chassis. Also diodes for AM and AGC were taken off, another 6JU8A acting as peak to peak detectors (voltage doubler) recovering twice the audio signal than a half wave detector.

1000046959.jpg
1000046436.jpg
1000046435.jpg
 
Last edited:
The mA (Smeter) is OK but removed from the panel during the work to prevent any damage during it.

The AGC subsystem is completely reworked. A system developed for Westinghouse TV's of the tube era, uses a high AGC negative voltage rectifier and trough high value resistors tied to +B. So the IF pentodes receive an attenuated AGC bias, excep the RF stage. It receives a bias that is a balance of the negative tension developed by peak to peak (voltage doubler) rectification from a IF dedicated exclusively to AGC service (that's original) and a positive bias from a voltage divisor from +B. The 6BA6 pentode that amplifies the signal for this function is replaced by a 6AM8 pentode plus diode with a 9 pin socket. The pentode continues being the AGC amplifier, and the diode has its cathode grounded and its plate wired to an intermediate point between the AGC rectifier and +B thanks a series of resistors of different values. If the AGC gives little or no bias, positive voltage dominates the point where the diode is placed, and conducts, maintaining the bias for the RF amplifier whose grid leak resistors starts, at zero bias giving its maximum gain. As the AGC obtains a better voltage, this joint at some signal level will become negative, the diode stops conducting and some part of the negative resultant acts on RF's grid. This method gives a better AGC delay tjan the original one and is known as "sinking diode"(Langford-Smith's Radiotron book page 1117 fig 27.38b and its text) and "clamping diode" by Whestinghouse (https://earlytelevision.org/pdf/westinghouse_t109_canadian.pdf) fig 4 and text.

The dial mechanism that was obstructed by dust, was cleaned with solvent and lubricated. Now it moves very easily.
 
As mentioned, the transistor at the audio preamp was extirpated. The double triode 5670 is removed and now a 6GH8 is placed there. The triode section continues acting as cathode follower for the AGC amplier (I don't know why is it there) and the pentode section with 10M grid leak, cathode earthed and 1M and 1uF 160V at G2 now is the preamp that replaces such germanium transistor. Thus, no more germanium devices at all and no sand in the direct signal path.

All the second audio path is removed. The audio 600R transformer, too. In its place there is now another IF can with the intention of making a narrow band FM detector, using the double diode at 6BN8 tube as Foster Seeley detector, but this part is not already done. The socket for the 6AK6 is actually free, but possibly there may be allocated a 6BN6 or 6EW6 for RF limiter and exiter of F-S detector.

The Smeter drive will be modified too. A 12AU7 tube as VTVM is added near it, and the 3 position switch is cleaned from all components. Thus, I'll use it at diferent fuction. In one position, it will show a fraction of the AGC signal, as any other Smeter does, but in other position it will be sourced from F-S output with zero centred Smeter to show FM (mis)tuning.

All RF and IF tubes' cathodes are now earted direcly, as all them receive a minimum bias from the RF GAIN pot and voltage leak at them for the Smeter, is no longuer needed.

To be continued soon.
 
Apart from the sections where you made a complete change the circuits, or even added something totally new, when you swap the tubes from one type/gm to another type/gm is it necessary to make many component changes (like lots of resistor values or capacitor changes), or do you find that in most cases it is possible to swap out a tube and maybe only need to change the pinout wiring?

73, Mark...
 
Yes, sockets were added, changed and rewired partial or completelly. Higher gm tubes mean usually more gain thus better s/n ratio. I listen ham station more than 1000km away from me at 40mts band using as antenna, only a 6 mts wire at the floor, inside my shack.

I forgot to say that I have two more sets partially dismantled, where I took original cables, wires and sockets. They came to me in catastrophic state.
 
1000047040.jpg


Originally it was a RCA phone jack. Such a foolery.
1000047045.jpg

14.5mS with same voltages and output traffo. No necesarily implies more power, but more voltage gain with same anode load.
 
Last edited:
You must log in or register to reply here.
Back
Top