The AN/URM-25s described by schematics

It's amazing the huge number of models that were manufactured under the URM-25 tag, even in the end (model F), the URM-25 production went back to the previous model (D). The evolution of electronic design was also important, the first URM-25 model has noticeable differences with the last.

Looking at the schematics, my thought is that there are two main groups of URM-25s, the early models, and the D and F models (which could be labeled as "the consolidated models"). But the most amazing is all the manuals edited for all models were issued from 1953 (URM-25) to 1956 (URM-25F), a lot of models in few years, having production, and some subsequent model D revisions, until 1966 (some of them not clearly documented). This peculiar production maybe explain why the early models are scarcer than the "consolidated" models, especially taking into account that the second group produced a lot of D sub-models (model D became the most popular model).

First group, "early" models:

• URM-25 to URM-25B (SG-44, SG-44B): Schematic shows a classic radio-service oscillator electronic design (e.g. Heathkit, Eico, etc.) + deluxe 'extras' (high-end modulator and meter circuits, a crystal calibrator was added later in model B). Units usually made by Federal Mfg & Eng Co.
• URM-25C (SG-44C): Same as above, but trying to match oscillator and buffer circuits to the level of 'extras': RF buffer amplifier is redesigned in order to improve output voltage to step attenuator. Units also usually made by Federal Mfg & Eng Co.
• Noteworthy peculiarities of the SG-44 models: RF range selection is done by changing a coil connected to a switch (very good and practical design as long as the switch contacts are accessible for cleaning). The first seven ranges have inductance adjustment (slug for the low-frequency end) and capacitance adjustment (trimmer for the high-frequency end); but the 8th range (18 → 50 Mc, IMO too long) has no precision adjustment; its adjustment is made by compressing or stretching the corresponding coil; (a service-level circuit design procedure: Heathkit, Eico, etc.).

Second group, "consolidated" models:

• URM-25D architecture (SG-85, SG-111): The "best selling" design. Main/standard model is SG-85/URM-25D (described here), sub-models are SG-85(and B)/URM-25H, SG-85(and C)/URM-25G, SG-85(and B,C)/URM-25J, and more clones. A fully revised model (circuit + architecture) is the SG-111/URM-25E (as indicates the SG number). Sub-models are popular but quite undocumented. Sub-models G & H add an additional 12.6 VAC secondary winding to the PS transformer to provide voltage to a load that has two 1N2970B zeners connected back-to-back to give regulated 6.3 VDC to the oscillator tube filament (voltage reduction is achieved mainly by a 7.5 ohm resistor in series with this secondary to ground which has a similar function to the series resistance of a VR tube). Since the PS unit contains the VR circuit, a dedicated output is needed for the oscillator tube filament, this implies some modification to the filament line in the SG unit. The new transformer runs noticeably hotter than the standard model transformer due to the inclusion of this secondary (12.6 VAC, 1.75 A) and because the original dimensions of the transformer have been respected. Sub-model J eliminates this secondary by adding a classic doubler circuit with electrolytic capacitors to the 6.3 VAC line to regulate voltage to the oscillator tube filament (6AH6) by means of 2 x 1N540 silicon diodes and a 1N3998A zener. In fact these improvements may be questionable, since frequency change due to temperature change in filament will only be evidenced by selecting high frequency ranges and simultaneously having significant changes in mains voltage, but these submodels were specific to be used on ships. The 1956 sub-model E, SG-111, does not have any filament voltage regulator, it looks like a revised main model unit with redesigned frames and some quality details (e.g. retainers in PS connectors). The main model was made by A LOT of manufacturers: Trad, Kings, Specialty, Keeney, etc. And sub-models by Wiewlex (G), Winslow (H), Measurement Control (J), Harvey Wells (E), etc.
• URM-25F (SG-103): The "best selling" after model D. It looks like the result of an architectural redesign + a cost reduction program. Panel is made of cast metal and the step attenuator is integrated into it (moltened), the entire circuit is compacted and adapted to the shape of the band selector rotary drum, remaining components are internally adapted to this cylindrical architecture (that's why front panel knobs are located in "odd" places), some to panel and others, like the power supply, inside back of case (surprising choice of a large, not very usual 0A3 octal tube, as voltage regulator). This is the only model with a different type of oscillator, model F has an Armstrong-type oscillator instead of a Hartley-type oscillator, so the band switch connections to the oscillator coils should be increased to 6 instead of 4 required by model D⁽¹⁾. No step attenuator as an independent component (but this optimizes shielding), no BNC protective caps, no rubber feets. Dial reading is vertical, not horizontal like the other versions, which is unusual and can be awkward, but model F has one more subrange than model D, this allows for more precise dial adjustment due to frequency expansion. The classic switch that adjusts RF response depending on range (10KC-300KC ↔ 300KC-50MC) is deleted, its functions (partially) are internally automated by the band switch shaft. Some people say this model is easier to repair, IMO this cylindrical design solves old discomforts but adds new ones, the only advantage is that everything is grouped together. All units made by New London Instrument Co.
• Noteworthy peculiarities of the SG-85, SG-111 & SG-103 models: RF range selection is done by changing a coil fitted to a rotary turret selector (bulky and cumbersome). All coils have inductance adjustment (slug for the low frequency end) and capacitance adjustment (trimmer for the high frequency end); (very good but it does not solve all the possible marking issues).

See below schematics of various URM-25s that are representative of the major versions, changes among models are evidenced by changes in schematics.