|
|
R 390A Performance Rebuild UPDATE ENGINEERED R390A(F) COMES WITH SPEAKER, ANTENNA ADAPTOR, LAB QUALITY ALIGNMENT - SERVICE, AND GUARANTEE All upgrades are hand wired with parts like original. --------------Upgrades------------- Audio...2 watts HI-FI, -3db 50cy to 10kc at 8 ohms, class A in the same space. Low level stages reworked for lowest distortion and noise. Three new audio filters selected from front panel: Flat...HI-FI for signals in the clear. Big Bands and Jazz sound great. Low Pass...-3db 3.8kc, -60db 5.00kc, which is the carrier spacing in the SWL bands. Voice Band Pass...-3db from 170cy to 3.2kc. Picks the voice out of a noisy signal, as in ham AM on 80 Meters. CW Band Pass...-3db 750cy to 1kc...250cy BW Gaussian shape. Enhanced sensitivity and large signal handling...Linear diode detector is retained and supplied with adequate injection, 50 volts. A new RF amplifier (6BZ6) provides 6db lower noise floor and much better large signal handling. Minimum Discernable Signal...At least -145dbm, CW at 350cy IF BW and 250cy audio BW. That's 0.012uV, 50 ohms at antenna. Undistorted Large Signal...For a 90% modulated AM signal of +10dbm minimum at the antenna jack, no audible or visible distortion. That's 3.8 Volt PEP at 50 ohms. All new AGC system...New detector, detector driver, relay with front panel selected mode and hang time. All new distribution time constants. Symmetrical rise and fall on AM, very fast rise with selected decay on CW/SSB. No over shoot, no pumping, no distortion. The CW and SSB performance is very smooth. For a 110db change in input level, 110dbm (0.7uV) to 0 dbm (.214v), that's 316,200:1... the audio output level changes 8db, that's 2.5:1... Audio Line Channel improved...It also drives the stereo front panel fone jack at about 100mW for use with low impedance stereo headphones. Headphone volume is controlled separately by Line Level. The Speaker was carefully chosen for it's high efficiency (99db SPL) and smooth response (-3db 50cy to 10kc). It is guaranteed for 5 years by Eminence. The enclosure is the smallest size that will provide good efficiency and damping on the bottom end. Default finish dark walnut... Full bodied, precisely defined music and voice... This system will fill your living room and Field Day site. Redesign of a good R390A to this level includes very serious speaker and antenna adaptor. For much better intermod numbers for SSB and CW, replace the 2kc mechanical filter with an 8 pole crystal filter. This gives IMD numbers of 92 to 93db at both 20kc and 2kc tone spacing. Numbers Alone are Misleading Receiver evaluation, as expressed by statistics of the current crop of digital wonder toys, occurs under pristine laboratory conditions. High quality low noise sources supply specific frequencies, specific products there-of are carefully measured....Life in the Ether, however, is some orders of magnitude more complex. Signals that appear side by side on a Spectrum Analyzer display are, to a certain extent, stacked on top of each other when processed simultaneously in broad band circuitry. Look at the 6mc SWL band at night on an analyzer connected to a resonant antenna. What you'll see is a forest of signals, many of them measured in 10's of millivolts at 50 ohms. Stack all those up and you're going to need something on the order of an 813 not to clip once in a while. This would make the receiver cumbersome. The solution is to place as much passive selectivity as possible before active devices like amplifiers and mixers. Filters and active control circuitry can generate IMD also. The R390A with 7 (8 - 32mc) to 10 (.5 - 8mc) tracking high Q tuned circuits before the first sharp filter is probably the best existing radio in this respect. The R390A also, being mechanically switched, tracked, and tuned generates no noise or distortion because of it's control functions. Solid state amplifiers, mixers, and switches saturate or hard limit when over-loaded. A tube circuit with a large grid leak resistor (1Meg), and some cathode degeneration at base band doesn't. If total input exceeds operating bias, the excess is rectified and the operating point immediately increases to include the entire signal and the device keeps on cooking. With the possible exception of a direct lightening strike, you can't kill a tube amp with a transient signal. The tuned front ends, wide dynamic range, and soft limiting of tubes make these older radios far more satisfactory for use in hostile electronic environments like Suburbia and Field Day multi sights. Now there's the phase noise thing... Solid state Bozo's blew it... and continue to, with progressively less insight; if that's possible. But, WOW... are there a lot of buttons on those plastic front panels! 250,000 possible bandwidths! I'm going into sensory overload!!! OK, ... "Easy Big Fella"... Phase noise, for this purpose, is: unavoidably generated side band energy (FM and PM) that is on all signals. Phase noise gives any carrier a measurable finite width in frequency. The further you look down from peak amplitude, the wider the signal becomes. High dynamic range considerations involve interacting with this wider part of an oscillator's signal. Think of a synthesized tuning LO as looking like a Christmas tree, when what you really need is just the trunk. It's the needles and branches on a synthesized oscillator that cause the intermod problems. As any carrier is generated at, or multiplied to a higher frequency, the noise pedestal gets wider at all points. The higher the frequency, the more phase noise on any signal. Heterodyne mixing, which is the fundamental signal processing tool used in most receiver designs, mixes desired LO frequency offset plus LO noise (phase, frequency, and amplitude) with any and all incoming signals. Each mixing step adds additional noise. The design goal here should be to generate minimum total noise. The lowest noise way possible to generate an RF carrier is with a well done crystal oscillator. Therefore using crystal oscillators for high frequency LO's, and placing the VFO (the noisy one) at a relatively low frequency (R390A) will give the lowest total conversion noise, and as a result, the best measured and experienced performance. The conclusion, if we're willing to make it....With a tuned front end tube rig, IMD spec.s amount to a close-in, worst case when the radio is hooked up to an antenna...With the broad band solid state front ends, those IMD spec.s amount to an unobtainable best case... when the radio is hooked up to an antenna. The KD0ZS R390A answers that nagging question in the back of your mind as you look through the ever larger and more glossy advertisements: "Are these things for real?" BEYOND MECHANICAL FILTERS...STEEP SKIRTS AREN'T EVERYTHING The R390A you get back from me is quite a bit better than the original mechanical filters. Filter stop band on the mechanical's is limited to 60db or so because the transformers on each end talk to each other. For the same reason, the noise floor consists of power supply hum, line harmonics, induced at the mechanical filters (roughly -45dbc for most power supplies). An additional problem with the mechanical filters is the severe phase shift with-in the pass-band. Effects of this phase shift are most noticeable with signals that need to be coherent, like AM. To obtain dynamic range numbers that the design is capable of, requires modern 8 pole crystal filters with matching circuitry in place of, at least, the 2kc mechanical... For smooth HI-FI AM, I've developed a 16kc lattice LC filter with minimal phase shift. The LC filter allows the upgraded R390A to deliver crystal clear highs... For the CW operator, the simple crystal 1st IF filter can be replaced with 8 pole cascaded filters which put stop band numbers beyond measurable. An AGC voltage operated manual RF gain control allows reducing receiver gain to suppress noise and interference, but normal AGC resumes at the set threshold. A carrier operated squelch for extended net operation, or tuning around on 10 meters AM is available. This is a sensitive, active squelch that allows completely normal radio sensitivity. The limiter tube is used for amplifier and switch, the limiter is changed to diodes... No new tubes. All the above features are available in the "R390F Redknob". AGC feeds the 3 mixers in addition to the RF amp in the RF deck. There are no buffers to isolate the oscillators from changes at the mixers. The result is some slight pulling over the AGC range. The manual AGC position can connect the RF gain control to mixers only to eliminate that pulling, but normally it's not noticeable. Chuck Felton KD0ZS Felton Electronic Design 1115 S. Greeley Hwy Cheyenne, Wy. 82007 307-634-5858 |