|
|
|
|
Collins R-390A Restoration
BackgroundThis one came from Fair Radio. It was a "checked" unit, and was in excellent shape. I opted for the two substitute meters. I got lucky in that all the modules are Collins with close serial numbers, with the exception of the IF deck which is a '67 EAC. The front panel is engraved (which means it's not a '55 Collins front panel), but the nametag is still Collins. Here is my Collins R-390A when I first received it.
After seeing pictures of others in a black finish, mine now looks like this...
It's in a CY-979 cabinet. I picked it up by asking one night on the Southeastern AM swap net, which meets every Tuesday at 7:30 PM on 3885 KC. Initial WorkOver the first two weeks, I made some minor modifications to the radio. I replaced the blocking capacitor which protects the mechanical filters (C-553) with a 0.01 uF Sprague Orange Drop rated at 600 volts. I also made Chuck Rippel's suggested audio mod which extends the low frequency response of the audio section. I also tried my hand at aligning the RF deck slugs, and found some improved performance. But a few days later, one of the slug racks was hanging up, making some the the band "dead." I opened the radio back up, centered all the slugs in their rack holders, realigned, and closed her back up for a few days. Meanwhile, I started having some problems with the gears getting hard to turn, so I knew it was time to get further into the RF deck and especially the gears. TopRF DeckRemoving the RF deck first requires removing the Oldham coupler to the PTO. You just simply have to remove the spring. The coupler falls out on the table later when the RF deck is removed. With the RF deck out, you can now see all the manufacturer's markings on the sides of each chassis.
The IF deck was an EAC from the '67 contract. I have since found a Collins IF deck from the '55 contract which is presently in my Collins R-390A, making it 100% Collins again. You can also clean the inside of the front panel, put DeOxit on all the switches behind the front chassis and underneath all the decks, and grease all the detents on the detent switches. WD-40 on a rag works pretty good for cleaning dirty metal surfaces. Let's clean this one up.
Here's the underside of the RF deck:
Now, it's time to dig in. I started with the paper capacitors underneath. Two of the three were cracked. Notice that C275 was 0.033 uF. This capacitor was originally 5000 pF, but was changed in an early production mod. The production mod lists C275 as 3300 pF, but all that I have seen that look original have a 0.33 uF (33,000 pF) capacitor there. I also replaced one that several folks say often needs replacing, which is C-327 (original shown below).
Now it's time to move to the mechanical side of the RF deck. I started by removing the IF section slug rack and lubricating the little bearings that ride against the camshafts. You just have to loosen the springs and lift out the slug rack. I cleaned all the parts with Q-tips, Mobil 10W30 synthetic oil, paper towels, and patience. Some take all the gears apart and soak them, but I was not that brave this time. After about an hour or so, the gears were beginning to come back to life.
On almost all gear trains today, I completely disassemble them. This includes separation and re-lubrication of the split gears and re-tensioning of the split gear springs. This method can make a really stiff gear train very smooth again. TopIF DeckIn the IF deck, I started by replacing several of the paper capacitors which are known to cause problems (C-553, C-531, C-549, and C-547). A few weeks later, I replaced all paper capacitors in the IF deck with 400V Sprague Orange Drop capacitors. I found that quite a few of the tubular paper capacitors had fine, hairline cracks. After replacing the capacitors and checking alignment, the IF gain was considerably higher (had to decrease the IF gain potentiometer). The BFO neutralization had to be readjusted. The limiter now works as it should, with previous problems (distortion at even the lowest level) now gone.
TopPower SupplyThe Power Supply has been upgraded to replace the rectifier tubes with 3A 1KV selenium rectifier diodes (Navy field change #6). This increases the B+ voltage just a bit, but also has the effect of putting full B+ voltage on the plates before the tubes warm up. I added an inrush current limiter (Keystone CL80) which eliminates high transients when power is switched on, keeps the ballast tube from lighting so "brightly" during the first few seconds of warmup, and also lowers the AC voltage to the transformer (lowered mine from 120 VAC to about 117 VAC). Details of the inrush current limiter can be found at Boatanchor Dreams. TopFilter CapacitorsThe rebuild of the filter capacitors is now described here. TopRestoration SummaryRestoration: Replacement or revitalization of original existing componentsCurrent restoration progress:
There are currently no plans for further restoration. TopModification SummaryModification: Addition of new components that are not part of the original designThe current modifications include:
TopCreditsI could not have completed this work without the help of Chuck Rippel and numerous others who have shared their knowledge of the R-390A through their web pages. Chuck also responded to many of my questions by email, and was always most helpful. But probably the best resource was the 7 hour R-390A Video from HiRes. TopSensitivity MeasurementsI currently use a URM-25F signal generator. I used the sensitivity measurement procedure as found in TM 11-856A (page 173). I've made a 10:1 attenuator (50 ohms to 120 ohms) that I use to get the output even lower (down to 0.03 uV). The attenuator seems to work really well, and the 50 ohm load on the URM-25F keeps it's output meter calibrated correctly. The attenuator is constructed as follows: 40.4 ohms 110 ohms |-----/\/\/----------------/\/\/---------- | | | | / | | \10 ohms | URM-25F / R-390A | \ | | | | |----------------------------------------- The above circuit gives 10:1 attenuation of the voltage delivered by the URM-25F. The URM-25F sees 50 ohms load, and the R-390A sees 119 ohms. The 40.4 ohm resistor is 47|330|2200 and the 110 ohm resistor is 220|220. The copper pipe gives great attenuation of any stray RF, which is necessary when working with such low signal levels. If you don't believe me, remove the cover from your signal generator and see the difference.
I was able to make some major improvements in receiver sensitivity (10 dB S+N/N), as many of the measurements were above 2 uV at the beginning. The greatest gain was made by reducing the IF gain potentiometer. The "magic" setting was found by adjusting the pot, and making sensitivity measurements. An improvement of almost 5 times was made by backing the IF gain way down. The following chart is after IF gain adjustments and a recent alignment. Some problems at the band edges. I have a weak and gassy 6DC6 RF amplifier tube, and am anxious to see what happens after I replace this tube. I may also try some new tubes and tube swaps in some of the RF mixers. Remember that the original specification called for sensitivity better than 5 uV. Almost any well tuned Collins R-390A can be restored to about 0.5 uV sensitivity, and numbers below 0.2 uV are often seen. Others have reported numbers as low as 0.07 uV. Keep in mind that these are AM sensitivity measurements (with 4KC bandwidth). The CW sensitivity is considerably better (2 to 3 times). Measurements in the 0.10 to 0.30 uV range are incredible for AM, even by today's standards. Note that the sensitivity of my Collins unit climbs at the band edges, and seems best in the middle of each RF tuned circuit: 0.5 to 1 MC, 1 to 2 MC, 2 to 4 MC, 4 to 8 MC, 8 to 16 MC, and 16 to 32 MC.
Last edited: 12/17/2002 | |||||
|
Checkout
The Quest Online to see where I now
spend much of my time. God Bless!
© 2005 by Walter Wilson, KK4DF. All information is offered for non-commercial use without warranty, expressed or implied.
|
