Huge and Powerful: The 1937 GE F-135 Repair Complete
I’ve finally had the time to finish one of my long-running projects repairing GE’s most powerful radio offering from 1937, the GE F-135. I picked it up from Craigslist back in November but other responsibilities kept me from getting much done on it until the 1st of this year. Finally, after a few months waiting, it’s finished and playing!
The radio came to me complete and in decent shape for the age. It’s missing the glass for the dial, but is otherwise completely intact and the finish isn’t in bad shape despite a few scratches here and there.
This radio is the current king of my collection. The best radio GE sold in 1937, it has a lot of innovative features – early APEX hi-fi reception, dual 6L6 output tubes putting out 20W of audio to a large and rich-sounding 12″ speaker; a total of 13 tubes including a tuned RF stage, dual IF stages, separate oscillator and “station seeking” automatic frequency correction.
Under the chassis it’s in decent shape too. It looks like it has been worked on before a few times – there are some ’40s, a ’50s and a couple of ’90s capacitors installed. There’s a sticker from a Seattle-area Jukebox Repair store on the back which probably explains the more-recent repairs; I looked up the address and they don’t seem to be around anymore.
First thing first after doing the complete set of intake checks on the radio, I gave the cabinet a thorough scrubbing and hit it with Howard Restore-a-Finish and it came out beautifully. The intake checks were uneventful so I didn’t take any photos, but the power transformer, all the IF transformers, oscillator and antenna and RF coils and the speaker transformer and coils were just fine. After applying the Howard’s, it really cleaned up the fading and covered over a couple of small scratches and really brought out the natural shine on the wood.
Then I tested all the tubes and found that most were in good shape (especially the 6L6G tubes installed) there were a few 6J5 and 6K7 tubes that needed replaced. I had these in stock and so it was easy to solve that problem.
I worked on this one under my kitchen’s vent hood as it’s currently too cold to vent soldering fumes outside. The chassis stand is the right width though so that’s perfect.
Every paper and filter capacitor needed replaced, as well as nearly every single one of the resistors which had drifted as much as +100% in value in some cases. Fortunately the coils are all intact or this could’ve been a much messier repair.
I also found a couple of places where the previous repairman who worked on the set may have been dyslexic, as there were a couple places were numbered resistors were reversed – i.e. R23 being in R32′s place and so forth. I imagine that couldn’t have made it work any better, anyway. I tested and replaced going along from the bottom up as needed. These are in-progress shots, so you might see a couple places where leads aren’t trimmed or components aren’t yet soldered. All of those issues were taken care of but might not have made it into the photo series.
Small caps dealt with, it was time to replace the multi-section caps. There is a 4-section can, insulated from the chassis, mounted up top with a set of 2 filter caps and 2 cathode-bypass caps. These all ran to a hole in the chassis where they went above. I snipped the long cross-chassis leads and moved the components close to their intended locations. In this photo, I’ve replaced the 10uF cathode bypass capacitor with its replacement. I like using bipolar caps for the large-value cathode bypasses but that’s just my preference and what I keep in stock (I keep those values around for repairing crossovers in old speakers) but you could use a standard polarized capacitor there.
Here’s the totally-complete underside shot.
There’s still a matter to deal with above the chassis, though. This is an AFC radio which uses a special and complicated transformer heading into the diode which recovers the audio. And it has a small resistor which is reading double it’s value and needs replaced as well, or it won’t align right.
Then I sealed the can back up:
I hooked the speaker and pushbutton assembly up on the bench and gave it a test run – it fired up immediately and started pulling in a few stations even on the Shortwave bands. The dial was off alignment a bit though, so it was time for that.
For the alignment, I pulled up the signal generator and started with an IF alignment before going back to the RF stages. This radio has a special IF arrangement with a procedure, so I aligned the 3rd IF primary, second IF secondary and primary, first IF secondary and primary, then went back and aligned the 3rd IF secondary that feeds into the diode. Aligning that discriminator was a maddening 10 minutes spent trying to nudge the adjustment ever so slightly. My goal was to get 0V between two segments, but it approached that point at an incredibly steep slope. I managed to get it there, though.
The original 0V spec was made with a primitive meter; I’ll take 0.01 on a more sensitive modern instrument. That’s perfect IF alignment. It was definitely worth it though. Now onto the RF, which involved tweaking something like 16 trimmers in a precise order with an RF signal generator at various frequencies.
Finally, it was all set! Time to reassemble.
At this point, the radio plays beautifully and pulls in stations from all over, and I’ve added a line input to let me hook up an audio source. The hassle of the AFC calibration was definitely worth it, it’s nearly like magic to watch it work. With the switch off, the radio tunes sharply and a station comes in over just a few degrees of rotation. With the switch activated, it’s like an entirely different radio – the same station will come in across about a quarter-turn of the knob, 2 divisions in either direction from the center frequency and it will block quieter stations from interfering.
The radio sounds great with a pretty good frequency response and more volume than I know what to do with, too. The relay for the motor is burnt out, though. I missed that on the initial checks so when I went to test the pushbutton function…I got a whole lot of nothing. I’ll make another post here when I do get the motor resolved but for now I’m going to hang this one up and start playing it. This was a very fun and enjoyable project and I have a beautiful radio with a commanding presence to enjoy for many years to come.
More Strange Coiled Wires [Mailbag]
I had an interesting call the other day with a gentleman about a radio he’s working on, among other topics:
That’s a Philco 46-420. They’re nice little bakelite radios with 6 tubes designed to receive the AM broadcast band. He’d come across some unlabeled wiring while repairing and had dealt with it but we were talking about what it’s purpose was.
I generally work on pre-WW2 radios so haven’t run into this particular arrangement personally, but I’ve read a few different articles by other collectors on this topic and recognized it immediately. The coil, wound 8 turns around the capacitor and connected at one end to the chassis, is a type of wave trap designed to cancel out the inductance of the old capacitor. This helps to prevent interference – both received, picked up through the cap as if it were an antenna, and radiated interference from the signal passing through the cap. Philco used these capacitor wave traps in most of their radios from 1946 and on. There’s an article at the Philco Repair Bench describing one style; this is a slightly variation with the same effect.
Modern caps are constructed out of metalized polymer films that have very little inductance, but these older capacitors were just concentrically coiled metal foil sheets with a lot of natural inductance.
Bypass cap from my client's Grunow 589
Filter cap from my GE LF-116
1950s "Bumblebee" Cap Exploded - MyLesPaul Forums
0.1uF 400V TubeTime.us
Modern caps don’t have that physical property, so it’s safe to replace the wrapped capacitor with any modern replacement and either shove the new cap through the coil, or remove the coil entirely.
Thanks to Bob from Old Tyme Radio for these photos of his project, and for distracting me from being snowed in for a bit!
I’m always taking mail from readers with interesting anecdotes, photos and questions so feel free to send them over either as comments or through the e-mail address I’ve posted in my Repair Services page.
Snow in Seattle! Part 2
Following up on yesterday’s post, this photo of a sign posted downtown pretty well describes how Seattleites react at the first sign of flakes falling from the sky.
This photo courtesy of my friend Jamie Swedler.
What’s that piece of wire, some kind of gimmick? [Mailbag]
I was talking with a fellow antique radio hobbyist on the phone the other day about a repair he was working on. A TrueTone battery radio, he’d taken care of all the important steps – checking out the tubes, replacing capacitors, replacing the broken wires, that sort of thing. The radio worked before the service but wouldn’t play after and so we talked through the steps to see if there was anything he’d missed and get a second set of eyes on it.
This set had a lot of rubber covered wire which has degraded over time and turned into bare wire, so some of that needed to be replaced. This started to become a problem around the second half of the 1930s up through WW2 and sometimes after…you can replace the wire or unhook one end and cover it with heat shrink tubing. Everything sounded good until I heard about replacing some wires “on top of the variable capacitor.”
RF issues in these old things are insidious and tough to locate and it takes a lot less than you’d think to throw it off. Wire diameter, wire length, physical positioning and shielding are all involved to some degree. Wiring changes in the front end are the first thing to take a look at. He’d mentioned a broken wire on top that he’d replaced.
That rang a bell. My next question was, “Were they twisted around each other?” They had been. Problem identified.
This twisted-wire “fake” capacitor is called a gimmick and was a way to save a few cents on the manufacturing cost. It doesn’t take much to make a capacitor. All you need is a two conductive charged plates separated by something non-conductive; two wires twisted around each other don’t provide much but but can make a few pF. Just enough to couple a small bit of a high-frequency signal like in the RF or IF frequencies. It’s not always obvious that a bit of twisted broken wire is actually a circuit feature, though – especially if it’s in bad shape from age, so replacing it with a new piece of wire is a pretty obvious thing to do.
Unfortunately in this case, though, the circuit as connected is shorting the oscillator and antenna sections of the variable capacitor together and the radio won’t receive anything. It was a really quick fix, though. Just twist a little tighter and snip!
Doesn’t look like much but it gets the job done – the radio fired right up and received stations after the quick change.
He sent over some construction photos to show how you can make your own gimmick replacement:
Thanks Steve for the photos! I’m glad the TrueTone is back in operation.
I’m always happy to throw out some advice about antique radios and radio repair, so if anyone reading has any questions feel free to drop me a line through my About Me page, or make a comment reply and I’ll get back to you. Feel free to share photos and stories as well, I’ll post the most interesting ones on here so everyone can benefit.
I love my chassis stand.
It’s the winter here in Seattle and it gets cold outside occasionally. Cold enough that I can’t vent my soldering fumes out an open window like I normally do most of the year, but I don’t want them in my apartment. Fortunately for me, my chassis stand is the same width as my stovetop and sits neatly under my externally-venting range hood so I can move it over there for ventilation.
The stand can hold the chassis up at any angle and supports a lot of weight. Every time I use it, I’m impressed with the build quality and flexibility. It’s not as nice working in the kitchen as it is on my normal workbench, but it gets the job done.
Unboxing my Sharp PC-4501 classic Personal Computer
It’s 1987 and I’ve been needing a new laptop. After a careful process of comparison shopping I settled on my Sharp PC-4501 from Montgomery Ward. Let’s open it up and take a look!
For only $699.99, it’s a state of the art machine. A 7.16MHz processor, 640KB of RAM, and mine comes with an optional second 3.5″ floppy disk drive! Now I can work on a program and save its data at the same time. It was designed to be especially easy to use.
Good thing they provide unpacking instructions inside the box.
It comes with an instruction manual and the operating system disk.
I’ll make sure to return the warranty card.
Power up…time to go through the BIOS setup!
One thing I don’t like so much is the screen is front-lit with ambient lighting, and is glossy. Makes it a challenge to see what’s being displayed sometimes.
We’ll try out a productivity application.
Looks productive!
I can’t wait to begin editing documents and spreadsheets from my own home. It’ll be so much easier than my old typewriter!
Hey, I Won Something. [Basic Micro ATOM Nano Development Board]
I read too much of the Internet every day, but it’s finally paid off. I happened to catch a post on Hack-a-Day about an awesome giveaway right as it was posted for a microcontroller development board. An interesting one, too – the Basic Micro ATOM Nano.
The dev boards come complete with an LCD header, a small solderless breadboard, USB connectivity, a pair of servo connectors, and more! The Nano 28 is based on the PIC16F886 microcontroller, and features an 8MHz clock, 24 I/O pins, 14K of flash memory, 368 bytes of memory, and 256 bytes of EEPROM storage.
The total value of the package is just around $50, so this is an incredible deal! Basic Micro will even ship your kit to you for free via USPS.
Not too bad. There were 55 available and I snagged one, and they’re now out of stock and won’t be replenished so I’m assuming they were clearing out these discontinued models to the community. I’ve been wanting a microcontroller board for a while, too – but the investment in an Arduino was more than I wanted to go for at once.
This changes that, and with USB connectivity and a nice looking IDE it should be pretty easy to write for. Microcontroller programming is a new thing for me and I’ll have to learn what’s possible with it. With a lot of IO pins but very little memory, this looks like it might be good for some kind of display or visualization or to send and receive data from sensors but not much more – that’s a tiny amount of memory by any standard, the equivalent of just a few sentences of text. We’ll see what’s possible. I like the integrated solderless breadboard area – this will make it easy to try out a few ideas immediately.
This might make a fun servo clock platform. I’ve had it on hand for a little while now but only just dug it out of the box I stored it in to start messing with, hopefully I’ll be able to start developing with it sooner than later.
New Arrival: 1934 Simplex Model P Dual Band Radio
I was cruising eBay randomly the other day and made an incredibly lucky find…a 1934 Simplex Model P Dual Band tombstone radio had just been listed. I have a pretty low maximum that I’m willing to pay for any radio, as I only buy fixer-uppers and not completed radios, so don’t buy them all that often but this one jumped out at me.
It’s dirty, as can be clearly seen in the photos. It could use some attention on the trim, but the finish is in pretty decent shape; there are some veneer issues on the bottom as well that will be a new skill to learn about fixing for me. And the knobs are wrong. Electrically, it’s your standard 1930s “Parallel AA5″ design: 6A7 6D6 75 42 80. But, it has a lot of potential once it’s cleaned up and fixed up and will display very nicely.
This one makes the cut on the display shelf, so I’ll be fixing it up sooner than later.
Grunow 589 Radio Repair Redux
A client came to me with another Grunow 589 needing repair. I’ve seen this model radio before in my shop and it was pretty exciting to work on another identical one. This particular one came in excellent shape – it looks like it was stored well for most of its life. The chassis arrived with no dust, no rust, a shiny coating, and not even any cadmium flaking off under the chassis.
The radio looks like it was well loved during its time. Built in 1937, it has evidence of being repaired several times all the way up through the 1950s. The UL Reexamination Service sticker is somewhat of a testament to this as well, and the fact that it’s in nearly perfect shape.
I performed the customary intake checks – all coils, transformers, tubes and controls. This one was in pretty good electrical shape. Only two of the tubes needed to be replaced after the years, testing very weak. The volume control was in bad shape, though – same as on the other 589. Unfortunately though for this particular radio, it couldn’t be saved and needed to be replaced. The output transformer connecting the radio to the speaker was also open and needed to be replaced before the radio would run again.
I sourced a volume control identical to the original (250K with a tap for tone compensation) and a functional replacement output transformer. With these parts on hand, it was a simple matter of performing the replacements. All capacitors were replaced with brand new models, and any resistors that had drifted in value by more than 10% were also replaced. On this radio, the first audio output tube (type 76) used a bias battery to establish a control voltage. As circuitry evolved, engineers discovered this was no longer necessary; I updated the circuit design to eliminate the battery by changing the 500K grid resistor to a 10 Meg grid resistor and jumping across the battery.
I also installed interference-suppression capacitors on the incoming power line, and a 0.75A inline fuse under the chassis to protect against a tube shorting and taking out the power transformer. With these modernizations, the radio will play beautiful and should perform without maintenance for many years.
A few parts were replaced during this operation:
The radio’s owner sent me a photo of it installed back in its cabinet after he received it from service. It looks – and sounds – great!
I provide affordable antique radio repair services in the Seattle Metro area and beyond if you’re willing to ship the guts. Check out my portfolio and information page for more details!
Snow in Seattle!
I’m taking a short break from writing about technical projects to share some of the recent interesting weather we’ve had lately. It snows about once a year in Seattle, and every year we’re taken out of commission for a day or two or three in this case. Most times it’s only an inch or less of snow but this time we ended up with closer to 6″ over the span of a couple days followed up by freezing rain and the accompanying downed trees, blocked roads and power loss.
Metro buses started running on chains on Monday in anticipation of a storm that didn’t really arrive until Wednesday, since we saw a tiny dusting over the weekend. This is always interesting and results in a lot of broken chains littering the roads after the snow melts. I’ve been on a bus while a chain came off and it sounds like the bus is about to rip in half, I’d hate to think of what would happen to the next car back’s bumper.
Metro bus chains (Seattle PI)
It’s always interesting when Seattle drivers try to get out in the snow. For the most part, nobody has any idea how to drive in it. People don’t use chains, or don’t know how to put chains on properly. Or just go too fast, too steep, and crash and slide into everything like this video taken from someone’s apartment on Capitol Hill. I prefer to stay inside or take the bus.
Some interesting things do happen, though. It’s always entertaining to watch buses be towed by an enormous tow truck like in this video from the Seattle Transit Blog (where they offer a convincing defense of the city’s sometimes confusing snow response) and this one is no exception with the flying electrical sparks.
I went out to investigate and took some low-quality photos of the block around my neighborhood.
In the first set of photos there was about 1″ on the ground; it accumulated throughout the day and the next to be about 6″ total. And that’s probably the one snow storm we’ll have this year, if past are any indication.
All in all it wasn’t too bad, but we did have a bit of alarmist reporting on Tuesday. Pretty much every media outlet including the highway information signs warned of a winter storm on Tuesday, but it arrived a day late. Nearly 90% of my co-worker went home early; across the region that’s a lot of lost productivity and lost wages for people. On the other hand, being stuck out in treacherous conditions in an area chronically unprepared for snow has a definite cost in terms of accidents, delays, and travel disruptions. I wonder what the total economic impact of a bad forecast is, and whether a conservative forecast predicting snow when none arrives has a smaller or greater cost than an optimistic forecast that leaves people stranded.