[R-390] CU-872A/U antenna coupler FS near Boston

[Roger Ruszkowski]

Fellows, The CU 872A/U was the ASA popular version of the antenna coupler.The screw on N (50 ohm small pin) connector was preferred over the 1/4 twist lock C series connector in other early models.The amplifier circuits are the same in all models. Parts is tube stages changed values to meet the bias requirements for different tube sets in different models. The CU 872 uses the skinny pin going for higher frequency.  The CU 872 also had a low end 2 Mhz band pass filter (sub chassis box) on the input. Part of the modifications to address problems in previous production models. A barrel connector will bypass the low end filter for AM band use. Class room speculation in 1968 was all the way down with a R389 receiver. Other transformer changes were made to round off the high end way over 30 Mhz signals. Circuit layout supports the high frequency operation. CU872's, R390's and Harley's are alike in that they all meet the minimum specifications.All outputs are close but you can detect differences in performance between outputs. Nice units for the multi receiver shack on HF.The units are left bank and right bank of four outputs. You can put the best of the tubes down one side and from front to back and have the best four outputs and keep the other side lite up in balance and not used with the weaker tubes. You can cherry pick the tubes and where you install them into the unit. Signal levels will not come up. However you are keeping the CU872 noise floor down and preserving  the head room between signal and noise what ever its source.Pick up a full set of NOS Russian made tubes and enjoy the next 17,000 hours of power on time.The idea that you can hang an R390 direct on the rhombic antenna line or any coupler output and measure the same signal strength is good. Hang 8 receivers and all are equal to the receiver numbers when feed direct from the antenna. Once we cascaded the couplers three deep 8 x 8 x 8 to 512 receivers (a lot of 47 ohm resistors were used as dummy loads. small losses were measurable to the test equipment. Operators in performance test could not discern differences in signals routed through various cascades of antenna couplers. Direct, one coupler, two or three couplers fanned out it all sounded the same in the headsets. The N series connectors exhibit less loss at higher frequencies that the C connectors. My CU 872's only had to work up to 32 Mhz because that was the top end of the R390 receivers in the station. CU 872's will keep right on working right up across the FM band. Some thing to keep in mind as ones choice of spectrum shifts. Our operators listened to different setups looking for a hot amp or dud outputs split phones and best two receivers in the station and a check list that would make NASA proud.  It has a 70 ohm antenna matching input impedance. Ever work any armored antenna cable that ignored 80 MM HE mortar rounds? Had more steel standing in the field than the air field had in the runway? Had more copper hanging in the air than the station generators had in their windings? Had more antenna lights than the airfield had runway lights? Every child needs a field station is sight just for the night lights and sense of security it brings. The armored cable from the Rhombic antennas terminated in a steel plate bulk head the size of a doorway (shrapnel proof). Between the bulk head connector and the CU 872 input connector mounted in a rack, was a specific length of a specific cable type. Read the manuals on mid trick once in a while.  The cable characteristics at the cut length was the best antenna impedance match to antenna coupler impedance match achievable across the 30 Mhz I was interest in and some more spectrum we had no equipment to access. It was common to go three deep and then receivers. One in the head to the bays. One in the bay to seven receivers and another third coupler to 8 more receivers. Eight operators with 16 receivers all wanting to listen different stations in the same far away city. Implications are that if all these CU872's want 70 ohms in and are happy with the output from another CU872 then CU872's must output 70 ohms. R390's are advertised at 70 ohms and we use 50 ohm sources frequently. Elegant military robust solution in a world wide production and application environment between real receiving antennas and receivers. A length of transmission line with offset impedance to match a source and load of different impedances thus tuning the entire antenna system  to the optimum transfer of power with out respect to an exact Z at any point in the system in either time or space. Use 70 for the CU872 end of the matching segment when doing calculations.   Seconds, thirds, and receivers were all three cables and six connectors in the patches. CU 872 back side to rack bulk head connector. Bulk head jumper cable. Bulk head connector to receiver or another CU872 input. Mill spec and good enough near loss less. Rack to rack jumps between rooms added another cable and connector set in the set up. To do the cable calculations backwards into what works best, it was found that the number 70 worked well for Z. When the head room starts with 48 CU872's six high in the rack and 8 racks wide plus maintenance rack you can believe this was an engineered installation that worked very well. Two operators could monitor every connector on the bulkheads in the bay room racks in under a minute at trick change. And reported any thing that fell down in the last 8 hours even if we were not using it. Three times a day it was ready if we wanted to use it. Or it was found and fixed real time. As a maintenance man I copied a lot of AM voice back side from Australia and keep it in my headphones. Every one else in the bay was working. The ditty bays were cooler than the maintenance shop so we would set idle time in the bays. When a Rhombic is catching an inch of rain per hour even it gets washed out. Wire noise goes way up with the falling rain and signals strength is lost through the rain. The eye wall of a hurricane either side is not the antenna reflector it is hypothesized up to be. Practical science experience being assimilated faster than beer.  A transformer output can be rewired to offer a phase flip at its output connector. An external match between the balanced CU872 output connector pair would be a two to one match into the balanced input of the R390. All mote science as most of us fellows do not operate in faraday rooms and the noise floor of CU872's and R390's are well below antenna atmospheric limits. The hardware is not the limit in the system.  If you live in a metro area and get blasted by overload a vacuum tube CU872 offers significant protection for virtual HF receivers.   Roger AI4NI  33C4H  >
> This is an interesting multicoupler. One interesting thing I've wondered
> about...It has a 70 ohm antenna matching input impedance. And i always
> wondered if the antenna connector was a 75 ohm N connector or the standard
> 50 ohm. (The difference is that the male center pin is thicker on the 50
> ohm version, and will damage a 75 ohm female part)
> Just something to look out for.
>
> Wikipedia has a nice photo of both impedances and genders.
>
> 73 Eugene W2HX
>