[R-390] Filter Repair

[quartz55]

First I’d like to thank Graham Baxter, G8OAD, for getting me started on this and he has a good write up on the procedure here http://www.delphelectronics.co.uk/filter390a/ I’d just like to add my take on the procedure.

What you need.

Fine needle nose pliers or tweezers, hemostats, the usual stuff.1/32" gasket material from the auto store, couple of feet will do, FelPro or Kerropak is what I used.

Propane torch or similar to remove end caps

Large soldering iron to re-solder end caps, I used Ungar #4036 47 W I tried a 42 W #1239 and it wasn't enough to solder the end caps on.

Vise

Solder of course

Aspirin to remove coating from enameled wire

#40 enameled wire

Solid clip holder to hold things while soldering

Hardline foam PTFE, enough, like Swedish meatballs or Pralines or Jameson’s

Drill and tap for 6-32 machine screws to seal the bottom end cap, #6 will accommodate the straw for the gas.

Can of 'Private Reserve Wine preserver' to put inert gas in the finished filter, thanks Tisha.

Non-reactive RTV, I used Permatex Ultra Blue sensor safe RTV silicone gasket maker

5 minute epoxy

6-32 machine screws cut to about 1/8"

Teflon tubing for #22 wire

#22 solid tinned wire

 

If you need to replace the magnetostrictive wire

#40 nickel wire available on line or send me a note, I have plenty.

Variable power supply and large capacitor, I used a 13,500 MFD 60WVDC with 10 volts on it.

Small angle iron, piece of single or double sided copper board, flea clip

Time and Patience

You can see all my pictures either here http://s251.photobucket.com/user/DogTi/library/R390A?sort=3&page=3 or here http://s251.photobucket.com/user/DogTi/library/R390A/R390A%20Filter%20Repair?sort=2&page=1

The pictures are self explanatory.

Notice there are two types of filters, the shiny ones and the matt ones.  They will come apart the same, but the shiny ones have a split metal tube that holds the filter assembly in rubber donuts and is ultimately easier to work on.  The shiny ones have holders for the end caps welded on the inside of the end caps.  Most of the rest of the article will be for the matt ones since there seem to be more of them.

First thing is to remove the filters from the IF deck. Obvious what needs to be done there, make copious notes where wires go, I am fortunate and used a second IF module as a pattern but I do have drawings and pictures but the bottom/input notes are suspicious.  

I have found it better just to cut the wires and remove components than stick a hot solder iron down in the bowels of these things.  Of course the components like capacitors should be removed with all the wires intact, but long jumpers can be cut, there is usually enough to put back on or you can get the Teflon spaghetti and tined #20 wire or add Teflon wire with small shrink tubing if things get too short.  You will burn the wire insulation taking them off anyhow, so just cut them.  I mark and tag parts with blue painters tape.

The filters in my EAC radio, both the 2 and 4K filters were dead, so there was nothing to lose.  The 8 and 16K worked, but the AGC action in this RX was bad or nonexistent.  The filter connections have to be removed to measure any leakage from the coils to ground since one side of the input/bottom end of all the coils are at ground and there is only about 50 ohms of DC resistance in the coils.  The output side has a common line for the AGC and the output is connected to S503.  There again one side of the coils is connected to the AGC and it’s impossible to measure out one filter.  The 8 and 16 K filters I had measured about 20 Meg to ground at the coils out of the circuit.  After repair, I could not measure the resistance, above 500Meg. I'm using a Fluke 189 and it seems capable of measuring these high resistances, you may need another method if you don't have a real good DVM.  I ruined the 2K filter by breaking the rods so I replaced it with an FA 2.1K plastic type inserted in the can.  The 4K filter had a broken magnetostrictive wire; it was laying in the goop when I took it apart. I couldn't measure it with my calipers.  I managed to spot weld a # 40 nickel wire back in its place and the thing is working so far. That magnetostrictive wire must be about #50 because it's impossible to spot weld back on, it just melts.  I'd like to know how they made these filters.

If you suspect your filters, I'd recommend removing them to test them; it's about the only way to make sure unless someone has a better idea.  Once the filters are out, I'd suggest only doing one filter at a time unless you are real good at keeping parts separate.  Each one you do will get easier.

The end caps are removed by clamping the filter at the mounting extrusions in a vice, maybe even with a right angle plate in the vice so you can catch things.  The ground tab on the end cap appears to be solidly attached to the cap so I use some needle nose pliers to grab that to pull it off when the solder melts. I try to avoid messing with the terminals as much as possible, DON'T PUT THE PROPANE FLAME RIGHT ON THE TERMINALS.  Wrap a piece of wet paper towel folded to the size of the label to protect it.  It makes the top end cap hard to remove, but it's worth it.  Heat the end cap more than the tube. Try not to jerk the cap out or you risk breaking the wires off the bobbins.  Position your pliers and your hand so it will pull the cap out just a bit when the solder melts, you don't want to jerk it way out possibly breaking the wires at the bobbins. Perhaps rig up a lever arrangement by the vice. Twisting across the tube axis helps.

When you get one end cap loose, cut the wires right at the end cap terminals and set the cap aside, do the other end.

Now remove all the goop you can with a small screwdriver, or whatever you feel appropriate.  You can get the goop out all the way to the fishpaper circle that is at the end of the copper cup, which is where the rods and disks start.  Object is to get out as much goop so it doesn’t get into the disks before or during removing the assembly.

Once most of the goop is out, pull the filter assembly out of the tube.  Grab the assembly by the copper cup end to pull it out, make sure the ground strap is loose from the tube.  Don't pull on the wires or the tape that's sticking out of the copper cup, you'll risk pulling the wires off.

Notice the ground strap is on the bottom end of the filter and is on the Collins name side of the label, as well as the ground lug, and the terminals on the end caps line up with the screw holes in the mounting tabs.  You will need to know this when replacing things.

The thing you want to do RIGHT NOW is put some RTV (non-reactive type) on the wires where that little yellow tape is sticking out of the copper cups if the magnetostrictive wire is good and you can measure ~50 ohms across the coil. Clean around the wires as needed, but as little as possible. Don't get it in the center hole or around the bobbin in case you have to remove it. The object here is to keep the wires from breaking at the bobbin which will be impossible to repair if they do.

Apparently the bobbins can be re-wound, but I haven't done that yet.  If any glue is loose, remove it and if needed use some new 5 minute epoxy to repair anything that needs it.  I've had the magnets come off so if you want, add glue to hold them on. DO NOT TOUCH THE RODS AND DISCS if at all possible. Handle the module by the copper cups with the yellow tape over them. The rods and disks are quite fragile and if you break/bend anything, that's it, it's over. Hopefully you won't need to clean the rods and discs, but look for crap in between them and if necessary clean it out with maybe a #32 wire or some air.

Now is a good time to clean up the end caps, remove the resistor at the top/input end, try cleaning it without alcohol or it will take the color off, measure it and make sure it's what it says it is, take a close up pic of it, the colors will be easier to see.  I found the 8K had a 100K, and the 16K had a 57K, I forget what the 2 and 4 had.  Consider replacing the resistors.  Drill and tap a hole in the bottom cap for a screw of your desire but the straw for the inert gas needs at least #6.  Pull the old fish paper off the end caps and from inside the tube, clean up all the residual solder so you can get the things back together easily.  While the RTV is setting up, use a dime to outline with a pencil and cut it with some scissors inside the line.  Gauge where the holes need to be for the terminals with calipers and make an X with an Exacto knife and push a knife through to make sure the terminals will go through.  After installing the disks back on the end caps, make sure to drill a hole for the inert air. Clean out the fish paper from the tube and clean out the tube and remove all the solder you can, I knocked it out with the propane torch and a cut it out with a utility knife. Cut a rectangular piece of the gasket 2.5 long x 2.6 Circumference (nothing critical, but not too long) and wrap it in a circle so it will form inside the tube.

Now clean all the loose goop off the filter, I used an acid brush with the bristles cut short and alcohol.  Don't soak the yellow Mylar tape on the copper cups on the ends, you don't want to remove that if possible, unless something else is fallen out or the magnet or magnetostrictive wire needs fixing.  You should be able to get everything pretty clean and no sticky goop around.  I left the circular fish paper spacers in place; it makes a gauge to fit the PTFE foam back on.  You'll need some PTFE foam insulation.  Only thing I had was some 1/2" hardline, I stripped the copper shielding off it, not easy, but with some side cutters and perseverance you can get the foam and center conductor out.  Take an Exacto knife and strip off a piece cutting it against the center conductor like you're stripping electrical wire, you will have a long strip about 1/8" thick and 3/16" wide.

Now cut pieces off the rest of the foam from the center conductor, you will need to start cutting the foam against the conductor and just peel it away with your fingernails and keep cutting against the conductor, finally you will get a piece off, maybe 1/2 to 3/4" long depending on how long you scored it.  I used the thicker pieces glued right on the copper cups/yellow tape, not against the magnets and the thin piece on top of that to bring the dimension up to the thickness of the inner diameter of the tube/gasket material.  I used a couple of small clamps to hold the foam on the copper cups while it was curing.  You should be able to put the whole assembly rolled up in the gasket material now and test fit it in the tube.

Some (all?) filters have resistors in the top/input side.  When you remove them, make a note of what they are and replace or reuse them.  At this point the end caps should be clean and the tube too so the end caps will easily slide onto the tube.  Wrap the new fish paper around the assembly with the insulation and slide it into the tube, make sure the ground strap is on the bottom end and is at about a 45 degree to the mounting tabs with the outside ground lug on the Collins name side of the label, check the pictures, a test run is advisable here, but it wants to be tight, but it can also be glued in place with epoxy. To replace the wires, I clean the ends from the bobbin with the aspirin and solder for maybe 1/8" or so, install 2 new #40 wires on the end cap terminals wrapping at least 3 turns around the terminal in case the solder melts while finishing the caps.  Make sure the fish paper is replaced, the resistor in place, the hole drilled in the bottom cap and then I make the wires from the terminals short enough to just reach the wires from the assembly.  Then holding the end cap with a clip, solder the wires back together.  I use RTV again to insulate the soldered areas and also on the terminals to prevent too much flexing. 

Let this set up for a while, have some Swedish meatballs or Pralines or Jameson’s and put the end cap back in the tube and now measure the terminals to ground, it should be well above 500Meg and the terminals should be in the vicinity of 50 ohms.  If all is good at this point, tack solder the end cap on and then do the other end same way.  Measure again to make sure there are no shorts or breaks.  I would test the filter at this point in a test fixture for bandwidth and IL.  Then solder the end caps back in place with a hot iron and the wet paper around the label again.  Clean it up with alcohol and make sure all the measurements are still good.  It should be good to go at this point. I found that grease and junk will contaminate the glass insulated terminals, so they need to be real clean before and after assembly or you will be measuring low resistance to ground (20Meg or so).

I had one bobbin that measured about 20 Meg to ground no matter what I did.  Apparently one of the #40 wires in the bobbin had some broken enamel and was showing low resistance. I had to remove the bobbin, wrap it in some Mylar tape (the clear shipping tape) and re-install it back in the copper cup.  Not easy to do, but possible, the hardest thing to do is get the Mylar tape tight against the bobbin, try rolling it together on a hard surface and holding the first end with tweezers.

If you need to remove the bobbin, you will need to break it free from the copper cup, use something like an Exacto knife and go around the edge inside the copper cup, break away any glue, pry against the bobbin and the copper cup and finally it will break free, then you need to hope you can pull the bobbin out with that little yellow tape.  But I've done it several times.  Then you need to hope when you put the bobbin back in you don't crush the magnetostrictive wire, and I've done that too.

At this point it would be suggested the filters be tested in some arrangement before replacement into the R390A.  I have a R2005D Moto service monitor that puts out +10 dBm or so, steps in 100 Hz, I made an LC matching circuit from 50 to 100K ohms to match the filter input and use a Motorola RF meter (Boonton) to measure the output with a 150pf variable capacitor and a 100K resistor.  I have the plans for that circuit if anyone needs it, it’s good to about 60 dB, but I’d rather have more.

Now finish those Swedish meatballs, Pralines or Jamesons.

 

Magnetostrictive wire

If you have a broken magnetostrictive wire or you break one loose, I haven’t had any luck welding the original back in; it seems to melt before it gets stuck on the end disk.  So I use #40 nickel, it seems to work.  Make up a sliding arrangement so you can grab the #40 wire with a small flea clip as close to the end as possible, make a bunch of test runs sliding the wire into the copper cup so nothing but the end of the wire touches the end disk.  Notice the copper cup has a hole in the back end of it and the last cone is attached to the copper cup, so you are trying to weld that wire onto the last cone.  After making sure nothing shorted and things were centered as best as I could arrange, I charged up the 13,000 MFD capacitor to 10V, attached clip leads to the assembly and the sled, slid the flea clip into the copper cup and you will get a spark.  Open the flea clip and pull the clip out and hopefully the wire will be welded to the end disk.  I’ve done 2 that have worked, so it’s not impossible.

 

Bobbins

I haven’t tried re-winding bobbins yet. G8OAD has, so see his article.  I figure that 48’ of #40 wire is 50 ohms, but Graham says it’s two 610 turns, alternately in the bobbin.  Graham states that there are 2 paths for leakage, one across the glass feed through insulators, cleaning is apparent here, and the other is leakage from the wires to the bobbins/copper cups.  I found on one bobbin I had luck by wrapping the bobbin in Mylar tape and re-installing.  Re-winding is another option. Of course if the coil is open it needs to be replaced if you have spare filters or re-winding.

 

Dave

N3DT