RCA Radiola 17

A Once Sad and Tired Looking Radio Comes Back to Life

Back to Eric's DIY Theater Projects

The Beginning of the Journey:

The Philco 40-180 radio that I found at the local flea market was a great find, but turned out to be an electronics-only project. The console's wooden cabinet was in GREAT condition for a seven decade old radio. Thus, half of the fun was missing - there was little to do beyond cleaning the wood work. Then I spied another old radio - the RCA Radiola 17. This radio looked like a mess on the outside. It had obviously been mistreated over the years and probably ended up in a garage, basement, or attic where it continued to get in the way. It had scratches and gouges into the wood on every side. Here was a great opportunity to learn something about refinishing antique wood. These radios are fairly common, nearly 200,000 of them were sold by RCA starting in 1927. The original price was $130 for the radio itself. You needed to spend an additional $27.50 for the tubes, and $35 more for the external speaker at a time when the averatge annual salary in the US was about $1500-2000! Wow - when was the last time you dropped more than one month's gross income on a single purchase? It's rather a surprise that so many of them were made and sold. This radio has the distinction of being the very first "light socket powered" radio - other radios of the time were referred to as "farm radios" that ran on A+ and B+ batteries. This necessitated a never-ending cycle of maintaining and charging several lead-acid batteries for the radio owner (see Phil's Old Radios for more detials). Thus, the Radiola 17 was a big step forward in convenience for consumers - provided that you lived in a major city that had electricity... Things were a little different in 1927. The Rural Electrification Act that brough electricity to the rest of the country didn't happen until 1936. As a side note, I recently picked up an old "Twin Tube Handy-Charger" made by the Interstate Electric Company of St. Louis, Mo sometime in 1924 or 1925. This was an early model charger intended to charge radio A+ and B+ batteries from a lamp socket so the owner could avoid the recharging battery fee charged by the local garage.

Since this radio was in such abused condition, I thought it would make a great "practice" radio to learn about lacquers and toners. Given how common this radio still is today, I figured that I wouldn't feel bad about stripping down a rare and valuable antique. Let the learning begin!

Initial Assessment:

This radio was quite a mess - The images below show some of the interesting "features" of time and abuse. These photographs don't truly reveal the extent of the damage that has been done - looking at the original radio in-person revealed far more of the gruesome details and sorry condition. Yep, it's also missing the dial lamp hood, just like 99.9% of all other radios like this... These show up on ebay from time to time, but tend to be expensive given how easily they came off and got lost.

If You Are New to Wood Finishing:

You might want to pick up a copy of Bob Flexner's EXCELLENT book "Understanding Wood Finishing." It is 322 pages of excellent advice and guidance and can be found at a number of online retailers. He describes different types of finishes and has entire chapters devoted to Staining, Grain Filling, Lacquer, Shellac, and more. It is definitely worth a read because he points out some of the common troubles people run into and how to avoid them. Another good resource is "Flexner on Finishing" and several other similar titles by Bob Flexner.

Removing The Old Finish:

This one looks like it spent time sliding around in the back of a pickup truck! Much of the abuse was confined to the lacquer finish, but there were a number of places where the cabinet was heavily gouged down into the wood. In the right hand image above, you can see parallel gouges on the right edge of the cabinet and also at the top left corner of the lid. The lid had sustained so much abuse that the three hinges that hold the lid to the cabinet were bent and no longer flat. What a shame for such a nice radio.

This cabinet needed more than just some clean up with GoJo and some Howards Restore-A-Finish. There are some very talented people that I'm sure could have restored this without totally stripping it, but I wanted to learn, so I stripped all of the way down to the bare wood. The easiest way to do this is to go to your local hardware store for a few items: 0000 steel wool, latex gloves, a paint roller pan, a can of Acetone, and a can of Lacquer Thinner. Take the radio apart, remove all of the hardware, all of the innards, and remove the lid from the rest of the box. Stand the box on end in the paint tray, mix a solution of 1:1 Acetone and Lacquer Thinner, dip the steel wool, and just start wiping. The tray is useful to catch your liquid so it can be reused. Just use a little at first. It will evaporate quickly, so replenish as necessary. Keep wiping with the steel wool and in a few minutes you will have a completely clean wooden cabinet. It might take a fresh mix of stripper to do the final wipe to get the color even and the surface clean.

After the first wipe with my solution, it became clearer how people can use lacquer thinner to help hide some of the scratches and damage to the finish. I was surprised to see all of the white spots disappear just before the entire finish melted away...

Filling the Gouges:

There are two major steps in preparing the now raw wood for new lacquer. The first is to apply wood filler as necessary to fill in the gouges that have resulted from mishandling and other human-caused impacts. Wood filler comes in a number of colors that basically correspond to lighter or darker woods. Since this cabinet is made from mahogany, I chose a darker color like Walnut. This Timber Mate product is one of the few that will actually accept lacquer on top of it without causing new problems. ALL of the wood filler products will CLAIM to be take stain, lacquer, etc, but many simply don't perform. Trust me on this one, Elmer's Wood Filler from Lowes just won't do. Get the right stuff for your job - pick up a can of Timber Mate and use it according to the instructions on the label.

Let it dry for a day or so and lightly sand it smooth. It does a great job at filling in the holes and gouges. I got mine from WoodCraft.com - just search on "Timbermate" and you'll find all of the colors that you need. The next step is to find some "grain filler." This is a very different product than the wood filler above. Wood filler takes care of deep scratches, gouges, and places where physical abuse has removed wood that shouldn't have been removed.

Filling the Grain:

"Grain filler" is a completely different animal than "wood filler." This stuff is for filling in the natural pores in wood that are often described as "open grain" that is characteristic in woods such as oak and mahogany. Below is a close up image of open grain in a raw piece of mahogany.

These woods have a naturally-occurring open grain to them and if you just start applying lacquer, they lacquer will coat the uneven wood surface, leaving the holes from the grain very visible in the surface texture. This is what gives many so-called "restored" radios an amateur appearance - people skipped the grain filler step. Filling the grain is the crucial step to getting that smooth, even, professional look to your work. If you are looking for that "smooth as glass" finish, this is the part you cannot skip.

I ordered my grain filler from Constantine's, though they call it "paste wood filler." Don't confuse this with the Timber Mate wood filler above - they are very different products intended for different uses. Right out of the can, it is a very thick and heavy substance - like a stiff peanut butter. You need to mix it with mineral spirits - just a few drops at a time in another clean container - in order to thin it out for use. When properly thinned, it should have the consistency of a thick paint. Use gloves while working and just slather some on your cabinet. Rub it into the surface, you want it to get into all of the tiny nooks and crannies that are the grain. Rub across the grain (this is the ONLY time that you rub wood across the grain!). The point is to fill the grain and by rubbing with the grain, you are likely to remove what you just placed there. Rub with your hands and in a few minutes, the mineral spirits will begin to evaporate, leaving the filler in the wood grain. Let it dry to a chalky appearance and then rub again - across the grain - with a cheese cloth or burlap. Some people use the edge of a credit card to scrape across the grain. This removes the heavy stuff that will clog your cloth. Hang on to what you scrape off, it can be reused by adding a few more drops of mineral spirits. Go gently, you want to remove only the filler that is above the surface of the wood and leave behind the stuff that fills in the tiny pores. If you rub enough (more gentle rubbing, not hard rubbing), the wood surface will get a bit of a shiny sheen to it and the grain filler will stay were it belongs - in the pores. When you've removed all of the excess, let it dry overnight and repeat the process. It is almost impossible to completely level the grain in one shot.

Below is an image of the lid from the Radiola 17 after being stripped, the gouges filled, and the first coat of grain filler nearly removed. A little more rubbing with a rough cloth made it nice and clean and smooth. I will have a bit of a sheen to it when you are finished. Let it dry overnight and repeat in the morning. After two applications, the wood should be relatively smooth and free from voids in the grain. At this point, it is important not to handle the wood with bare hands - oils from your skin will get into the wood and cause problems later on.


About Lacquers:

Most radios from the 1920's and 1930's time period were finished with lacquer. Some of the best lacquer is available from Mohawk and comes in either clear or with "toner" already in it. I ordered mine from WoodShopProducts. The basic varieties of toner are as follows:

Toner Lacquers:

Clear Lacquers:

Finally, do yourself a favor and order a can of Mohawk "Blender Flow Out" in gloss finish as well. This is to help with repairs as you spray the lacquer. Lacquer dries VERY quickly (it is "tacky" in mere seconds and begins to dry in just a minute or two) and there are a number of problems that can occur when you are not well practiced with toner. Every now and then, you might apply too much lacquer and get a run in the surface. Or an eyelash or bug lands in your freshly applied lacquer. Keep a pair of tweezers handy for picking out little hairs and bugs, but even if you are quick, it will leave a disturbance in your surface. Just a quick "spot shot" of blender flow out will "re-wet" the lacquer and allow it to flow and self-level so the offending area really just melts back into the surface and disappears. The same with runs. If you happened to get a run, turn the wood so that the run is flat and parallel to the ground, give it a spot shot (not too much, it takes a few seconds to work) and watch the run just melt back down and disappear. It works like magic, but use it SPARINGLY - it is essentially a lacquer thinner in a spray can so if you use too much, you'll create bigger problems for yourself.

Here are some comments from Jamie on the PhilcoPhorum:

"No matter how careful you are, you will still get runs! I hate it. Sometimes it just happens. There is no way to prevent it. As I said in the other thread, before you do a nice radio, get some Mohawk Blender Flow Out. It's like a spray can of lacquer thinner. It increases drying time and thins the lacquer. It is insurance in a can!

So, say you get a run.. grab the Blender Flow Out and spritz the run. The glob of lacquer magically spreads out into the surrounding lacquer and the run vanishes. If you use too much, the toner gets too thin and you have a "bald spot", so be careful. Another scenario - You spray your final coat of clear and a dog hair floats onto the wet lacquer!... ARRRggghh!!.. No problem - Pick the hair out with some tweezers! Now you have two little ugly marks in your finish. Spritz it with Blender Flow out and watch as the lacquer oozes back over the marks and they vanish. It's good stuff! Very handy! It saved me from having to re-strip a radio today."

Refinishing my Cabinet:

So, if you are working on a total strip and refinish, you need some Acetone and Lacquer Thinner for stripping, wood filler for fixing large dings, grain filler to fill the grain on all exposed surfaces, some Sanding Sealer as a first coat, some ultra classic toner lacquer for adding color, and some clear toner for the top coats. If you are just working on a single radio, you'll need one can of Sanding Sealer, one can of Ultra Classic Toner lacquer in the desired color, and likely several cans of clear for finishing. For the Radiola 17, Mohawk Dark Walnut/Oak is just about a perfect match - I used the back panel from the radio as my color reference because this panel was never stripped. Ultimately, my cabinet came out just a pinch darker than it was originally, but the difference is VERY subtle. At one point, I ordered about six different colors in order to find the best match to the set of radios that will one day be refinished.

There are many weather considerations to take into account when using lacquer. First, it CANNOT be humid. Humidity in the air gets trapped in lacquers (same happens with polyurethane) and turns the surface hazy and milky. Wait for a day with COOLER TEMPERATURES (50-70 degrees F seems best). Wait for a day with LOW HUMIDITY (lower than 50% is best). Wait for a day that is NOT WINDY- the wind carries dust, debris, and bugs that will get stuck in your wet toner. I work in the garage after moving the cars out. Spray the lacquer (use a breathing mask!), wait a few minutes for it to begin to set, and then open the doors to allow for some fresh air flow. This allows you to control the conditions as much as possible for best results. Finally, it is useful to WEAR GLOVES when working with raw wood or when appling lacquer. Oils from your hands will stain the raw wood and can prevent lacquer from adhering properly causing even more problems. Quickly wiping the wood down with some mineral spirits will remove most finger prints. After sanding down a layer of lacquer to apply more, give it another wipe of mineral spirits and let it dry COMPLETELY before applying lacquer. Have you seen the relatively new "crinkle finish" spray paints at the hardware store? Let's just say I know how they get that surface now - make sure your mineral spirits are COMPLETELY gone before applying new lacquer or you'll discover how they make crinkle finish, too...

Another imporant tip for using lacquers: NEVER USE A TACK RAG! Tack rags are impregnated with wax (this is what makes them sticky). The wax will rub off onto your wooden surface and will cause problems with lacquer adhesion - it will bead up and not adhere evenly. Use a non-lint rag, wipe with mineral spirits and let it dry, then apply your lacquer.

There are several great threads about creating beautiful lacquer finishes. One of the best is here: Philco 90 Cabinet Refinishing This thread is a complete start-to-finish tutorial on how to refinish a wooden cabinet properly. Read it several times. Then read it again.

Other threads on cabinet finishing are available here and here.

I tried my best to follow this advice, and still a few things went wrong as I was working.

Too Much Lacquer at Once:

It took a while for me to figure out why this happened, so here it is: If you use too much lacquer at once, you are virtually guaranteed to get an "orange peel" texture on your cabinet.

One solution is to wait for things to dry thoroughly (several hours on a warm, dry day - overnight otherwise) and do some gentle wet sanding with fine sandpaper (400 or 600 grit) and a sanding block. Get yourself a hand held sanding block (never sand without a block, you'll make the surface uneven), attach the paper, add some water to the surface and start sanding. Go gently, you don't want to go through the clear and begin removing the toner that you added, you just want to even up the surface. Sometimes it works best to apply two or three coats (separated by about an hour to dry) and then let it sit for a few hours more to dry and harden. Then sand it a little to smooth things out. You want the ENTIRE surface to be smooth and evenly textured from the sand paper. If you have any small round "dimples" in the surface that still look shiny after sanding (you can see one near the bottom of the image above), you need to even these out as well. Add a little extra lacquer to that spot (just do a spot shot with the spray can), let it dry for a few hours, sand it down, then add another few coats to the entire surface and even it out again.

The better solution is to apply many thin coats (apply lacquer as a "dusting" - move quickly with the spray can, keep 10-12" inches away from work surface) work better than fewer thick coats (you don't want a thoroughly wet and shiny looking surface after each coat). Trust me on this one- I spent lots of time "fixing" things because I either made the lacquer too thick, didn't wait long enough for it to dry, or both! Ugh! The good thing about lacquer, though, is that with some patience, you can correct nearly any screw up without having to start all over again by stripping things back down to bare wood. If you are in "fixing" mode because you got too impatient the first time through, it is even MORE important to lay down successive coats as a dusting and slowly build the surface back up. If you go light enough with the dusting spray on a warm dry day, you can apply successive layers about 20-30 minutes apart. After about 4-5 dusting coats, minor suface irregularities will have mostly disappeared, provided you started with a physically smoothly sanded surface. MANY THIN COATS OUTPERFORMS MULTIPLE WET LAYERS OF LACQUER!

In the image below, you can see the cabinet after I added a few coats of clear lacquer, but before I added any toner. Yes, things are being presented a little out of order here, but it makes a good illustration anyhow. Looking at the images at the top of the page, you can see that the Radiola 17 has a nice inlay of wood around the front perimeter. After filling in some dings that can be seen in the image below (slightly darker areas on the radius), I added some grain filler, and then added a few coats of clear lacquer. I then realized that the grain filler actually darkened the wood while it filled the pores in the wood. This made the lighter inlay blend in with the rest of the cabinet and nearly disappear. This isn't what I was looking for at all. Rather than stripping the entire cabinet, I used painters tape, masked off the inlay and used a soft cloth with some lacquer thinner to scrub off the lacquer and remove the grain filler, but only for the inlay stripe. I worked carefully so the lacquer thinner wouldn't drip and cause other problems, and with some time and effort, I managed to lighten the inlay again. It then took a little while to build this area back up with lacquer so the front surface was smooth again. In the image below, you can also see some of the dings that were filled with the TimberMate filler. These become much less obvious once the toner is applied.

After removing the grain filler from the inlay (note the difference in contrast between images above and below), I taped the cabinet off again and started to apply the toner. Everything was going great until I sprayed a little too much toner and started a run across the front rounded edge. Ha- I was prepared for this with some Blender Flow Out! But, I used too much and in addition to helping to thin out the run, the flow-out ran down under the piece of tape I was using to protect the inlay and hit the front of the cabinet!

Total disaster - a gaping hole in my nice clean and smooth finish! But all was not lost. A little gentle use of steel wool removed the dark wing-like edges of the run above the inlay to the point where you couldn't see them anymore. The spot below the inlay was sanded with fine paper to smooth things out again and then coated with more clear coat just to even out the surface. So then I had a slight depression that was lighter in color than the rest. I then sprayed more toner to restore the color. I used steel wool again to gently remove any excess color and blend in the color edges. Rub - GENTLY - with 0000 steel wool just a little then carry it out into the sun to see how close the colors are. Keep going until the colors match and the new area is blended in with the surrounding area. Just a little at a time, stop and check, and keep going until it the color matches in darkness and blends in well with the surrounding area. Then coat the entire area in multiple coats of clear. Next I used the sanding block again to even out the surface texture to build up the hole again. I added some lacquer, let it set and dry, sanded it, and repeated until the hole was filled in. This took a few hours, but kept me from redoing all of the work I had already done. In one case, I think I took a piece of paper, folded it in half, and cut a small oval about an inch in diameter. This was used as a "spray mask" that allowed me to add toner to a limited area on the cabinet. My wife held the paper a few inches above the wood and I held the toner spray can a few inches above the paper, spraying toner through the hole in the paper. This allowed me to apply the toner exactly where I wanted it and provide a more "diffused" edge to the toner. Had I put the paper directly ON the wood, the toner would have left a crisp edge that would have made more work to blend it in. If you need to apply lacquer to just one small spot (perhaps you made a ding because you got too close and bumped the surface with your spray can), you can apply lacquer to a spot with a Q-tip. Spray the lacquer into a bathroom paper cup, dip the Q-tip into the liquid, and then dab onto the spot you need to fix. Let it dry, sand it smooth, repeat as necessary, then finish off with several "dusting" coats of lacquer. If you do this, your repair will be invisible when you are done. If you keep applying "wet" coats of lacquer, any repairs that you have made will travel right through each new lacquer layer that you add and will be clearly visible after things dry. The solvents in the spray lacquer will soften up the lower layers as they dry and reveal all of your screw ups again... Like I said: many layers of light coats of lacquer - you don't want a soaked surface.

Final Touches to the Cabinet: There are a number of techniques for putting finishing touches on the wooden cabinet. If you don't want your new cabinet to be quite so shiny, you can wet-sand with very fine sandpaper (2000 grit), or some 0000 steel wool with some soap and water. If you would like a more shiny surface, use polishing/buffing compounds. Either way, the key is to let the lacquer cure/harden for 2-3 weeks first. The thread below provides some detail:

Getting a shiny smooth top coat

My Refinished Radiola 17 Cabinet:

Here are the results of lots of work, errors, and fixes along the way. Overall, it came out VERY nice and is ever-so-slightly darker than the original color. What an interesting learning experience this was! I purposefully left a few of the original dimples and dings behind - you can see on of them in the top left of the front of the cabinet as an indicator of its past. I wanted it to look nice, but not perfect.

What a beauty! At this point, I am willing to call the cabinet finished and turn my attention to electrical restoration.

Power Supply Restoration:

Before you do ANYTHING with your new antique radio, there are some safety protocols that you need to be aware of. Some of these protocols are intended to protect you and some are intended to protect your radio from further damage.

Safety Protocol for YOU:

First, there is a protocol to follow for testing tube radios/amps that needs to be followed BEFORE you plug the power cord into the wall. Failure to follow protocol risks destroying old equipment, starting a fire, or killing yourself. A tube radio power supply harbors LETHAL voltages, often in the range of 200-400v. This voltage can be present EVEN WHEN THE POWER CORD IS UNPLUGGED. This voltage CAN KILL YOU. There are a few basic safety rules that apply here:

1) REMOVE ALL jewelry. ALL of it! NO rings, NO necklaces, NO bracelets, NO danglies, NO nothing that will accidentally conduct electricity!

2) UNPLUG THE DEVICE AND DISCHARGE ALL CAPACITORS. This is typically achieved by using a specially made (by you) insulated wire with insulated alligator clips at either end and a 10k-100k power resistor rated at 1 to 3 watts in the middle. USING ONE HAND ONLY, clip one end of the wire to one side of the capacitor. Using that SAME HAND again, clip the other end of the wire to the other side of the capacitor. Let it sit there for a few seconds and then move on to the next cap. Do this for EVERY cap in the device. Use ONE HAND ONLY! Keep the other hand behind your back AT ALL TIMES. Exposing yourself to high voltage that spans both hands (and your HEART in the middle) is just asking for an ambulance ride! Tube amp power supplies make excellent defibrillators. A properly beating heart does not respond well to being defibrillated! You've been warned... Also, you'll find many people online just say "short the cap with a screwdriver." DO NOT DO THIS! This is potentially damaging to the screwdriver, to the equipment you are testing, AND TO YOU. This much power can literally blow chunks off the tip of your screwdriver. If this flying chunk of screwdriver tip hits you or someone else in the eye, you have a major problem. DON'T BE DUMB!

Safety Protocol for your RADIO:

DO NOT PLUG IN YOUR RADIO! No matter how tempted you are to "try things out" or "see if it still works" or whatever else, DON'T DO IT! You need to carefully and methodically inspect your tube gear FIRST so that you prevent doing even more damage than time alone has caused. Here is my short protocol for things you want to inspect first. A much more detailed and thorough protocol can be found here: Trouble Shooting Antique Equipment.

1) You will need to remove the metal chassis that contains the electronics from its wooden cabinet so you have access to the electronics. Once the chassis is freed, take lots of pictures - closeup pictures from different angles. Photograph each tube and each connector. Make notes on how things are wired and connected to each other. Note colors of wires, positions of wires. Label all of the tubes with masking tape. Draw a diagram of the chassis and label where each tube goes. Now remove the tubes, get yourself a old bath towel and lay it down on the table. Turn the chassis upside down on the towel and take lots of pictures of the underside before you touch/disturb anything. Use good lighting, take close ups that show details of where wires travel and what they connect to. Only then should you begin to poke further into the chassis. Your notes and images will become an invaluable resource when you start removing and replacing things. If you have an air compressor, this might be an excellent time to use it to clear out some of the old dust/dirt/debris. As always, use a bit of care not to dislodge weak/brittle wires or other structures. It might be best to blow some air across the chassis before your compressor hits maximum pressure to aviod unintended damage. 100 PSI is NOT your friend in this situation...

2) Inspect the power transformer with your ohm meter. Check the primary for resistance. It should not be open (mega-ohms of resistance), nor should it be a dead short (zero ohms). Check each of the secondaries for resistance. They, too, should not be open, nor should they be a dead short. These will typically measure tens, hundreds, or thousands of ohms. Check to make sure multiple secondaries are still isolated from one another. You should find mega-ohms of resistance (or infinite resistance) between secondary windings.

Alternatively, you can use a dim-bulb tester to check the health of your power transformer. This protocol comes from Leigh from the AntiqueRadios forum.

One test would be to remove all tubes, then power up with your dim bulb tester. Searching the web for "dim bulb tester" will reveal many examples. They all work the same way - a light bulb is placed in series with your radio to protect the radio's transformer from a shorted condition. The bulb might come up very briefly but should settle down to dark or almost dark. If it shows any significant brightness you have a shorted transformer or circuitry connected to the transformer. At that point you would need to disconnect all secondary wires, then try again. If the bulb still comes on, the transformer is bad. If not, trace the loose secondary wires to find the problem.

If everything looks good at this point, re-install all tubes except the rectifier. Bring it up again on the dim bulb bester. The bulb will light, hopefully not too bright. Measure the voltage at the transformer primary. Calculate the percentage (60v / 120v = 50%). Apply that factor to your nominal filament voltages. Then measure those filament voltages. If any are not close to your calculated value, that points to a possible problem.

3) This is probably a good time to make your first modification to your old piece of gear: add a polarized AC plug and add a safety fuse. Old AC power cords have plugs with two blades that are the same size so you can plug it into the AC wall outlet either way. This is a bit dangerous with old gear, so replacing the plug with a new polarized plug is a good first step. Then, follow the power cord back into the chassis and add a fuse to the "hot" wire. Do this somewhere AFTER the cord strain relief (where the cord physically enters the chassis) and BEFORE the power line hits the transformer. You can any type of fuse holder here, depending on how "original" you want to keep things looking under the hood. Typically, a 1A slow-blow fuse will be fine. When in doubt, start with a smaller fuse first, say 0.5A. I would not recommend going above a 1A fuse at all.

When attaching the power cord to the plug, you need to tie end ends of the wire with an "Underwriter's Knot" so that strain on the power cord does not pull the wires out of their screw mounts to the blades of the plug.

4) Check each of the tubes for resistance. Identify what type of tube you have, the numbers are usually stamped in silver on the glass envelope, either on the top or the side. You might need to wipe the tube with a damp paper towel to remove the dust and grime before you can read anything stamped on the outside. Look here at Frank's Electron Tube Database or Google the tube number and find a data sheet for it. This will help you identify the pins and their function. Start with the heater/filament pins - these are typially the pins that are connected in the tube diagram. Test them with your ohm meter for conductivity. Conductivity between the heater pins should reveal a small resistance - typically under 10 ohms. Make sure there are no other shorts across the remaining pins. Test remaining pins in pairs, all combinations, one pair at a time with your meter. If any other pair of pins besides the heater pins show connectivity, you probably have a shorted tube that needs to be replaced.

5) Inspect the power supply capacitors. You can perform one or two tests with your typical inexpensive Digital Multi Meter, and you can perform a few more if you have an expensive DMM. but NO digital multi meter (unless you've dropped a few grand on a bridge tester) is equipped to help you test what really needs to be tested in antique tube radios and amps: do any of your capacitors "leak"? This is a simple test: caps are supposed to allow AC to pass and are supposed to block DC from passing. If DC passes through your cap, you have a leaky cap. The ONLY way to test for this is to remove the cap from the circuit, apply a working voltage (200+ volts DC) to your cap and see if any of the current leaks through. See here for how to build a capacitor leakage tester.

In case you missed it the first time through, here is an EXCELLENT detailed writeup for how to thoroughly inspect a tube-based power supply BEFORE applying AC power. Be sure to read Max Robinson's write up about Trouble Shooting Antique Equipment. The first one-third of this page is very detailed and definitely worth a read to help keep you from screwing things up even worse than they are already likely to be from the simple passage of time. There are many, many things to go wrong with old electronics and they need careful and methodical attention. Simply plugging in an old radio can lead to any number of complication including fires! Be smart - do some reading and learning first!

If you get the power supply up and running, you are probably most of the way there to making your radio work...

Under The Hood of the Radiola Socket Power Unit:

A first inspection reveals that the radio itself needs some work as well. Removing the front bezel pieces from the cabinet revealed that the power on/off switch was a newer replacement part. Apparently, it also was much harder to operate than the original switch was because it bent the metal plate that held in in place. With some gentle work, I was able to straighten out the metal disk that holds the power switch in place and I eventually found an original RCA power switch on ebay to replace the part that came with the radio. The mounting holes for the power switch disk are a bit off center (from top to bottom) on the front of the cabinet. This requires that you orient the power switch vertically with the toggle pointed down (instead of up) when the radio is on - try to mount it in any other orientation and it won't fit into the mounting hole.

The two rings and the front escutcheon recieved a gentle scrub with some Gojo to remove the heavy dirt and grime. They are supposed to look "antiqued," so don't go scrubbing too hard. Don't use any harsh chemical cleaners, either. Something like Brasso will certainly clean this up, but it will look like shiny gold when you are done. This is not what the original radio looked like at all, so don't go there.

Click the image below for a high-resolution schematic of the Radiola 17 radio chassis. Also, note the screw-in plug on the end of the power cord. This radio was the "first of the breed" and came with a screw-in plug so it could be screwed into a light socket (hence the "socket power unit" name) back in the days before standardized wall outlets existed. All other radios prior to this one were referred to as "farm radios" and ran on batteries. I found Service Notes for this radio on the RadiolaVille website.

Below is the power supply module that contains the power supply transformer and various capacitor and inductor filter units (each in its own metal can in the image below), as well as the rectifier tube. You can see the remnants of the original power cord in the lower left corner of the image as well - clearly this needs to be replaced... The two wires flying off from the power supply terminal strip are for the pilot light located above the tuning dial.

A quick look at the underside of the power supply chassis reveals that some work was done on this radio long ago. The main Resistance Unit (the black rod-like structure at the bottom of the image below) that provides different voltages for the radio apparently stopped working some time ago and was bypassed with individual resistors. Even these replacement resistors look old... By the way, check out the original pride in workmanship displayed here. The wires are bundled together with string to keep thing neat and orderly under the chassis! You don't find that in too many products with this level of pride in construction today!

All of these resistors are rather old looking and each measures as an open circuit, thus they need to be replaced. The underside of the capacitor block for the power supply reveals the addition of a wire wound resistor (bottom right corner of image below) that does not belong there. A quick guess is that this is intended to compensate for some original resistor value that has gone bad or there is a more significant problem lurking about and waiting to be discovered.

The first order of business here is to clean up some of the apparent evils that have been committed and perform an electrical evaluation of the remaining parts. The power supply module here consists of three major components, each confined to one metal can on the top side of the PSU chassis: the power transformer, the filter chokes (labelled filter reactor in the service notes), and the filter condensers (oversized rolled paper-in-wax capacitors). Smaller components include the rectifier tube itself and the power resistor strip on the underside of the chassis. My DMM revealed each of the "new" resistors soldered across the power resistor to be significantly out of tolerance or just plain open. Thus, all of them were removed and the Resistance Unit was removed and cleaned up with a wire brush. All of the original internal connections of this strip are also open, which makes adding new resistors easy and straight forward since there is nothing to interfere with new parts. The original Resistance Unit values are shown in the image below and match the orientation of the Resistance Unit in the image above. A quick clue is that resistance measurement of each segment is a pretty direct physical correlation with the distance between the tabs of the Resistance Unit (the 205R segment is the shortest while the 3750R segment is the longest). If you get this backwards while re-wiring your radio, the voltages will be off - I learned this the hard way, luckily, it didn't cause any new problems... When you have things in proper order, the DC voltages you find on the Resistance Unit with respect to ground (terminal #11 - closest to rectifier tube when viewed from above) should be close to those indicated in red below.

The orientation of the Resistance Unit in the Socket Power Unit is shown on the schematic below. Terminal #11 (the Ground / 0v reference) is on the left and Terminal #1 (physically closest to the rectifier tube when viewed from above) is on the right.

Finding exact matches for these resistor values is pretty much impossible today since they don’t conform to standard values (evidently developed after 1927 – go figure). I selected values that match as closely as I could find – in each case my new resistor value was within 4% of the original value. Since most resistors of this time period were likely to be +/- 5% anyhow, these values are close enough to make things work without creating new problems. My replacement resistor values included 422R/5w, 3k6/5w, 2k2/3w, 200R/3w, 1k62/3w. The high voltage from the rectifier connects to the 410R resistor (not a bad idea to make the first two resistors 5w rated ones since all other voltages derive from the 410R resistor) and the coils that attach to the transformer center tap create a 0v reference point between the 205R and 1k69 resistors. Voltages between these points vary according to the individual resistors as shown above.

There is also a nice table in the service notes about connectivity measures you can make on the terminal strip of the power supply, though the notes are not terribly detailed or specific: they don't provide specific resistance (ohm meters were likely not very common in the 1920s) only indicate open or closed status that can be tested with a low voltage light bulb and a battery, presumably to make testing easier for the radio repair technician (this was a big business in the 1920's). This is about as far as you can go without either powering things up or further disassembly. Given how difficult it would be to replace components in this power supply, I opted to further disconnect the capacitors and chokes so more detailed measurements could be made. So, out came the soldering iron so I could remove some wires from the caps. Remember, make notes, draw diagrams, make detailed photographss of everything first so that you can be sure to get things back where they came from! My LCR meter is capable of making a fair variety of readings, so I measured everything I could. This revealed that all of the capacitors are reasonably close to their original values (as indicated in the service notes), with one exception (cap in the top right of the image below). One of the 1uF caps measures closer to 0.3uF and the dissipation factor is a bit higher than the other caps. I might parallel this cap with a smaller modern 0.68uF cap to make up the difference.

Measuring each of these caps for DC leakage current with my new capacitor leakage meter didn’t reveal any additional problems. The service notes recommend using no more than 200vDC for testing the caps, so I connected each one to my new leakage meter and ramped up the power to 200v. The leakage current started off at a high of about 20uA and fell over the course of about a minute to somewhere in the range of 10-13uA. For power supply caps, these measurements seem to be within allowable parameters. Essentially, you want to insure that leakage current for waxed paper power supply caps is below about 30-40uA.

Next up is the filter reactors (inductor coils). There are two coils stuffed into one of the metal cans on the Socket Power Unit. The challenge is that the service notes provide no measurements for what these values are supposed to be. One of these is easy to measure, the other is not. One end of both inductors attaches to the filter caps, so this is easy enough to disconnect. The other end of the first inductor connect to the resistor unit, so measuring resistance and inductance of this coil was easy. The problem is that my measurements reveal that this coil is open and not function properly. The second inductor is not as easy to measure, since the second wire from it goes directly into the power transformer, so there is no easy way to reach the end of the wire in order to measure it. Thus, I had to cut this wire in order to measure the inductor.

Darn my luck, this coil appears to be just fine and now I’ve broken a 90 year old wire that I didn’t need to break, so now this needs to be repaired. All in all the Socket Power Unit was in reasonably good shape: I needed to replace the power switch, the resistors in the Resistance Unit, and a bad inductor coil with a new resistor. It seems that no one knows what the appropriate value for this coil is. Not that this would make much of a difference as there is no physical space to just “add” in a new coil. Specifically, at this point there are two options. The first is a quick a dirty solution to simply replace this inductor with a 1k 10w power resistor (thanks to Norm at AntiqueRadios.com for this suggestion). This is likely to increase 120Hz hum in the speaker (which it did). The other option is to replace the entire power supply unit with a new one. This shouldn’t prove to be too difficult as there are quite a number of these floating around today. The trouble is the expense of shipping one as they weigh as much as the proverbial boat anchor. For now, I’ll substitute the coil with a power resistor and see how things go.

At this point, it was time to put the Socket Power Unit aside and turn my attention to the radio chassis and see how much work is necessary.

Radio Chassis Restoration:

A look at the radio chassis below seems to reveal fewer problems, but some cleanup is an obvious starting place. I removed the condenser tub from the top of the radio chassis so I could clean things up a bit. Go gently with cleaning here. If you use chemical based cleaning solutions, you can remove the paint from the tub and if you rub too hard on the tuning dial, you'll start to remove the numbering on it. Oops... Luckly, the pot metal that the tuning wheel is made from is still intact and the condensers turned easily. Many of these old tuners are made from pot metal that seems to have a habit of just self destructing over time. Finding one that is not bound up is a great starting place!

Click the image below for a high-resolution schematic of the Radiola 17 radio chassis.

The radio chassis was not without its own problems, though these were minor. The volume knob was missing from the front of the radio (replacement knobs are typically available on ebay - though you need one with a brass collar and set screw). Removing the radio from the cabinet revealed what had happened. Apparently, the front of the radio got hit hard with something and it cracked the knob to the point where it fell off and was likely thrown away. The impact also cracked the ceramic shell for the volume control. I attempted to glue it back together (super glue worked well), but a piece from the collar that holds the center post in place was missing which caused the center post to be too loose to function properly. Luckily, I found a working replacement volume knob on ebay for a few dollars and I happily donated the broken volume control to a fellow Radiola 17 owner who had a good volume control housing but damaged electrical components inside. That one worked out great all of the way around!

I spent some time with the schematic to trace each of the wires from the power supply unit to the capacitors and tube sockets. Each of these wires were good, none were broken and none seemed to have shorted to one another or the chassis. Hiding on the bottom of the radio chassis are a number of parts that need to be checked out as well. There are two grid resistors that are little more than wire wrapped around a phenolic rectangle. Each of these should measure 800R to 1000R - I've found both values on old schematics, so I'm guessing this value isn't critical. There is a larger "dogbone" style red resistor (grid leak resistor) that should measure close to 3M. Then there are three tiny, flat capacitors mounted to the vertical board. These are the grid condenser and two AF condensers. I have no idea what these are supposed to measure. My grid condenser measured 1nF, the AF condenser on the left measured 200pF, and the AF condenser on the right measured 33nF. At least they weren't open. Given that their construction appears to consist of physical layers of conductor and insulator, I presume they are relatively stable over time. The volume control should measure 2k from end to end, just as it should.

I moved on to the output transformer, which I was happy to see still had connectivity. The primary measured 565R and the secondary measured 770R. No idea what they are actually supposed to measure (originally a 1:1.5 ratio?), but I was pleased that it was intact. There is also a ganged capacitor unit made from three metal cans strapped together. This is the set of bypass condensers for the tubes. Without disconnecting any of these wires (some are hard to get to), I was able to make a few measurements. The block on the bottom (closest to the metal chassis) contains two caps that share a center connection to ground. These caps sit across the filaments of the UX-226 tubes and each measured 1.3uF (original specs were probably 1.0uF). I'll call that good enough. The middle cap has only two terminals and connects directly to the cap block above it, so it is not possible to make an independent measurement of this cap. The cap block on top contains two caps and a shares a connection to ground. I was able to measure 1.5uF across the two outer pins and 1.0uF from each of the other two pins to ground. I'll call these good enough as well.

The RF coils and the AF transformer are a bit more difficult to measure because they are so cross-connected with other part of the circuit. I don't have any measurements for these, nor do I have any values to compare my measurements to once I made them, so I stopped here.

The last detail was to replace the antenna and ground wires that go to the volume pot. The antenna wire is blue and the ground is black, so I replaced these with appropriately colored new wires. This radio features an interesting arrangement for volume control: the attenuator connects directly to the antenna wire. The amplifier always provides the same level of amplification, the only thing that varies is the strength of the signal from the antenna. Primitive but effective!

The Original Speaker:

I was fortunate that the original speaker came with my radio, but it's condition was just as bad as the radio cabinet. I found a set of Service Notes for this speaker some time ago, though I don't remember where they came from.

The grill cloth is torn and generally not repairable. This isn't too big of a deal since new cloth is available from Sonny. The bigger problem is the holes in the pot metal. Pot metal is perhaps the biggest scurge to the antique world. When mixed properly and according to formula, it is very stable over time and holds up quite well. When mixed improperly and/or with impurities (because someone picked up spilled chuncks from the ground and threw them back into the melting kettle) it seems to quite literally self destruct over time. Since the same types of holes are present on the inside of the metal cabinet as well, it seems that the damage seen on this speaker is the result of spontaneous self-destruction (see this account of pot metal cancer). So, I set out to see what could be done with the speaker to repair it. The most straight forward path is to fill the holes and cracks with something like JB Weld or Bondo (yep, the same stuff used for auto body repair) and repaint it. This is likely to involve a fair amount of labor and leaves open the future possibility that "new" chunks of the speaker will just decide to fly away one day. The good news is that the speaker itself is still in remarkably good shape. There are no holes or tears in it, the mechanism is free from rust, and the voice coil and attached transformer are all still in great condition.

The most obvious structural problem to tackle first is the broken corner. Of course, it is the front corner that's broken... JB Weld is an excellent candidate for this type of repair. My first effort was simply to put some JB Weld in and across the crack and clamp it, but when I released the clamps a few days later, the crack re-opened a bit.

As it turns out, this bond didn't have enough strength or surface area to hold the cracked corner in place for very long. For the next iteration, I took the same approach as using rebar to strengthen concrete: I mixed up some JB Weld and embedded within it a paperclip to span each crack. This added more strength to the crack and hopefully will prevent it from opening up again. If you look closely at the darker patches, you can see the partial outline of the paperclip on each side. I will add one more paperclip and some additional JB Weld at the apex of the crack that you can see in the lower middle of the image below. This should do a reasonable job of holding the corner in place as I start to work on filling the pot metal voids with some Bondo.


First Power Up:

OK, this is always the part where you hold your breath, back up a bit, and proceed with caution. Everything seemed to check out: I spent some time replacing parts in the power supply and measuring the remaining components. Measurements of the radio chassis revealed no obvious problems. I cleaned the tube sockets and the tube pins. Resistance measurements of each tube revealed that the heaters were all still intact and that there were no other shorts. None of the caps were shorted, none were open. A new power cord was in place and the hot lead from the wall socket now sported a 1A fuse. Pre-flight check list complete...

I made the first power up of the entire radio for the first time WITHOUT the rectifier tube in place. After a few seconds, each of the tube heaters lit up. Excellent! I let it sit this way for about about 30 minutes so I could monitor the temperature rise of the power transformer and choke cans and just keep an eye on things in general. There was a small temperature rise in the power transformer, which is to be expected, and everything else seemed fine. I could measure about 700v AC across the plate pins for the rectifier, so this looked promising as well. All that was left was to power up the radio with the rectifier in place. Here is where I tend to get a bit nervous. Up to this point, the highest voltage in the radio is 5v AC for the tube heaters... Now its time to get real.

Moments later, the rectifer was in place and the power was flowing to the entire radio. Time to make a few more voltage checks:

All in all, not too bad. While none of the voltages were spot on, they were all reasonably close 90 years after this radio was manufactured. Each of the voltages were just a pinch high. The first resistor in the Resistance Unit is specified as 410R and I replaced it with 422R. Perhaps a somewhat larger resistance (something in the 440R - 460R range?) would be a better fit to bring voltages closer to target values. I presume this is because my wall socket provides about 122-124v AC instead of 120v.

As expected given the higher voltages at the power supply terminals, plate voltages at the tubes were a bit high as well. The plates on the UX-226 tubes measured 150v DC, the 227 plate measured 55v and the 171 plate measured 158v. The four position terminal block in the radio chassis measured 168v, 0, 148, and 57v from the front of the chassis to the rear. The new 1k resistor that replaced the open choke had about 64v DC across it.

Next, I checked the voltages at the speaker terminal: 0v DC (excellent) and 12v AC (not so excellent). This was a bit concerning, so I shut things down and asked a few questions online before going much further. Turns out that the speaker NEEDS to be connected for the radio to function properly. Without the resistance offered by the speaker, the plate voltage of the 171 tube drives some high voltages through the transformers in the radio chassis. It caused a little bit of alarm because my first voltage measurements without the speaker connected caused some flashing to occur within the 171 tube and also made the plate pin spark just a bit against nearby cloth-covered wires that go to the AF Transformer. Connecting the speaker cured this behavior! Pfew!

By some coincidence, the radio was actually tuned to a local station when I first connected the speaker and high volume voices came pouring forth! Wow - I was quite surprised just how loud the volume was! Using an old antenna, I was able to tune in about 4 radio stations, two were very loud, the others were very faint. Success! What a thrill to bring such an old piece of electronics back to life again!

All Finished (almost):

Here is the working radio back in its re-finished cabinet. I need to find a dial hood and one more knob that matches the original one that came with the set and I'll call this one done. Nearly every one of these old radios is mising its dial hood. There is an excellent YouTube video about how to create a replacement dial hood while you scavange for a genuine replacement. The original knob has a brass insert and a set screw so it firmly locks onto the shaft for the tuning mechanism. The replacement knob (on the left) that I purchased on e-bay has a plastic collar with a brass leaf spring that doesn't hold very tightly at all. It sits there, but falls right off when you touch it. Overall, not too bad for a radio that was manufactured in 1927!


After having this radio apart for so long and working on each piece individually, I was rather surprised to re-discover just how heavy it is (just about 45 lbs)! If you put this on a shelf for display, it needs to be a sturdy one!

Like most of my projects, this one took a few years to complete. I am rather pleased with the results!

Back to Eric's DIY Theater Projects