50s Japanese am/fm tube dual speaker radio
 

Picked this up tonight for $20. Was filthy, took apart and cleaned. Am works well, fm was weak. Replaced 5uf capacitor across output of fm detector and fm came alive, by although stations are all strong it seems they are all slightly distorted, like it's always slightly out of tune. I like the looks of it, and AM sounds great, it will be my new morning listening radio once it works properly.

More pics - iPhone let's me post one per message for some reason

More

Inside

Better photo - daylight would help

The front end seems to be mounted on its own chassis. And the schematic diagram doesn't show the audio output tube type, probably a 35C5.

Yes, it's a little tiny sub chassis. The voltage amp is a 12AV6 and the output is exactly a 35C5. The schematic is glued to the bottom of the radio. The dial light comes on strong when you first turn on the radio, and then extinguishes. I am quite sure it is a neon light, but it's the physical size of a normal 47 pilot light - something may be causing low B+, which may be responsible for the distortion on FM. The 17EW8 is a weird tube, I don't have any spares on hand. The little speakers actually sound pretty good on AM, I am looking forward to figuring out why FM isn't working properly, hopefully it is something simple like bad filter capacitors causing low B+.

The lamp you have is probably NE-51 or more likely NE-51H (high brightness, high current version), will only glow on one electrode because it's got DC on it. It takes 90 Volts or so to get it to strike but will run at 67 Volts after that. The dropping resistor controls the current and brightness. If B+ drops below about 70V the lamp will go out. You should have about 110V on the filter capacitors; if not, one may be lacking in capacitance. They are supposed to be 60 microfarads and I am guessing they aren't close to that. When you get good capacitors there, the lamp will stay lit and the FM should work better. In fact, everything should work better.

Still, a NEAT little set.. Maybe it was from the very early stereo FM era, was advertised as a "Twin-Speaker" set to try to fool unwary consumers into thinking it was stereo...

Hmm that makes a lot of sense - I will order some new capacitors and go from there

Update: filter capacitors seem to be fine, no characteristic hum, and b+ on all three sections is over 110v. On AM the radio plays like new, on fm the volume is much quieter, and when you turn it up the sound is distorted, although it does pick up a lot of stations. The AGC voltage is only 0.6v with the strongest fm station. There are no paper or oil capacitors in the radio circuit only audio, and it works fine on am so I assume they are good.

All tubes original and if transformers still have undisturbed drops of white paint on alignment screws... Got to be something wrong in the fm section which isn't obvious to my amateur radio repair skills

All tubes test good on eico 667 including for h/k leakage

The problem is either in the V4 circuit or the FM detector. Check those germanium diodes, the OA-70s. Replace with something like 1N270 or other germanium if they aren't good. Also the detector transformer might have a problem, like misalignment or even a bad internal capacitor.

Your advice is much appreciated, thank you for following the thread - will do exactly that tonight. Is it common for Ge diodes to fail? I will check them with a digital multi meter, diode test function, is that an adequate test for small signal diodes? I could also use the EICO 667 to check them, I suppose.

I acquired a B&K 415 sweep marker generator some time ago, but have never used it. Am thinking towards setting it up and checking the IF bandpass and ratio detector S curve (youtube has some awesome videos by current and former VK members), but having never done that type of work before and not knowing the condition of the B&K I want to approach it with caution - definitely don't want to make things worse!

Just so you know who you are talking to, I would say my skill level is moderate, I am an electrical engineer, but focus on power systems and not electronics professionally. From a hobby perspective I have done a lot of work with building and repairing tube and solid state audio and guitar amplifiers, and have been poking around with old televisions for a while. I've also aligned and restored antique AM radios, but IF and RF work at higher frequencies (such as an FM radio) is kind of new ground for me, so I really do appreciate the guidance from experienced people. I do plan to replace all the electrolytic capacitors in this radio, but would like to make it work first, and then do preventive maintenance, so I don't end up introducing a new problem which makes the existing one even harder to find.

As for the neon indicator, it must just be bad, it lights up and then goes off, but the voltage does not fall below 100 at the socket base in any case. The lamp is the size of a 47, but with a screw base rather than a bayonet.

nice radio

Thank you! I think your photography skills would do it a little more justice, especially the reverse painted plastic front panel. My iPhone photos are quite crappy - although I am impressed how well the schematic diagram photographed.

I never saw a screw base NE-51 so you may have to do some surgery to replace it. You can temporarily parallel a good neon lamp to see if that's the problem. Old neon lamps do go bad from gas leaks.

I'm an engineer also but get excited about everything from DC to maybe 30 MHz. I also repair musical and stereo gear and have an impressive lab. I just acquired an HP 3456A DVM but it doesn't work so I'm negotiating with the seller. I also used to own a stereo repair shop. And I have a ham ticket.

You can use the sweep gen for education but frankly I never found one useful for FM alignment. Better to use an FM generator. You can use the DVM to check the forward drop of the diodes; it should run in the vicinity of a quarter of a volt (as opposed to silicon, about two thirds of a volt). Reverse leakage can be a problem though and your meter won't measure that unless it's drastically leaky.

Hm I need to get my hands on an fm generator. The diodes have a 0.37V drop according to my meter and nothing in the other direction but as you said not entirely useful. The neon lamp has a metric looking threaded base, not the usual type. I have an amateur radio license too but never actually used it. I would not
Call my "shop" impressive, more like a room in the basement with a work bench and a bunch of obsolete test equipment, but it gets me by.

All my test equipment is obsolete. But it all works, and that's what I care about. I've got good stuff, HP and GR and Tektronix well represented.

I use my ham license all the time. I have worked over 300 countries and have two antenna towers with yagis. I work all the HF bands except 60m; I wasn't having fun on there so I took down the dipole. I even have a gf in Indonesia and probably will never meet her. She is a fast CW operator. We also use Skype and while I play piano, she sings.

You probably already have an FM generator. Any cheap AM signal generator is an FM generator because the usual method is to modulate the oscillator. Even fairly expensive generators did that, like my old Marconi and the Measurements Model 80 (which I no longer use). Of course you have no way to know exactly the deviation but you can come fairly close by comparing the audio level to a broadcast station. So you adjust the modulation to get about the same audio, or a bit more, and look at the audio on the 'scope. You need to use external modulation because the internal one is not very close to being sinusoidal. Adjust the discriminator in the receiver to get the nicest looking wave and you are done.

Your diode measurements indicate good diodes. I assume you can measure them independently, not in-circuit.

To replace the neon you will either have to change the socket or take the base off a new lamp and glue it into the old base taken from the bad one. A better solution might be just to parallel the two bulbs, with the new bulb just taped into place. You can't tell from outside.

I was lucky recently and acquired an HP digital synthesized signal generator that covers from 10 kHz to 2.06 GHz with AM and FM. It has an OCXO and readout to one Hz so it's as accurate as I will ever need. For audio I have several units, all HP. I also have a distortion analyzer, counter, and some other goodies. It's fun just to interconnect all this stuff and see what happens. One of my favorites is the GR 1658 Digibridge which measures passive components very accurately.

Since I am getting on in years, I am concerned that my heirs won't know what to do with all this stuff. On the other hand, it's their problem. I am having fun and as long as I am healthy I will continue to play.

I told a lady friend that I bought myself a present of a new voltmeter. I asked how many people would like a voltmeter as a present. Some people just don't understand us nerds.

Man that's a pretty cool background - all my stuff works too otherwise why have it . I hooked up the b&k 415 for fun tonight and aligned the if pass band with the avc disconnected, it looks much better now. Also aligned the detector s curve, it looks good too. I notice that the maximum avc voltage does not correspond with the maximum audio strength, I think the if oscillator frequency didn't perfectly match 10.7 MHz, but the market is in the centre as it should be, so I tweaked the oscillator a bit so peak audio corresponds with peak avc voltage,

The radio is more sensitive now but still some distortion, sounds like IM distortion possibly with power supply ripple. This radio has such small speaker there could be a fair bit of ripple on the power supply before it's audible as hum - Will check that next.

I think at this point I need to do some reading about fm receivers and how to align them, my theory is a little too weak for my comfort level.

As for relatives and friends appreciating this stuff, a friend of mine asked me why I wouldn't just buy a $20 radio instead of investing the effort... I said because that would be boring. Also, one other point - you only live once, and international travel is surprisingly easy these days... Go for it what's to lose?

Nice find, Max. Never heard of the brand before.

Some distortion may be due to inadequate bandwidth, causing the signal to deviate beyond the linear portion of the discriminator curve. Not much you can do about that without compromising sensitivity.

Another source may be mismatch of the diodes, causing asymmetry of the S curve. You can judge that with an FM generator.

Check me out on QRZ.COM - callsign K6DDX.

Who thought a 60 year old radio would open such a can of worms... S curves, IF pass bands, etc.

On further reflection one thing the receiver is doing seems a little odd to me.

When the signal of the sweep frequency generator is applied to the grid of V4, the output of the discriminator transformer looks decent - when it is applied to the grid of V3, it is the same, but with more gain... when it is applied to the grid of the mixer half of V1, the signal is not much stronger than when it is applied to V3, as though there`s not much gain in that stage. I actually opened up the FM-A IF transformer, and inside the capacitors are ceramic disks, so no GE `silver mica disease` to deal with here. The IFT also tests fine for continuity, and the 17EW8 tests good. Is the mixer supposed to amplify the signal significantly, or is it normal operation what I have noticed - I wish I had a voltage chart to accompany this set, to make sure the 17EW8 is biased properly.

I found the following book, and am working my way through it, it is a good help in understanding how the receiver is supposed to function, but the description of converters is dated and doesn`t reflect how my radio is set up. https://ia700402.us.archive.org/5/it...nReception.pdf

By the symbol on the back and the schematic, it's made by Ida Densen. I had a magazine years ago that was called JEI, Japanese Electronic Industry. It was from 1964. It got lost somewhere, but it was great for identifying the various Japanese private labeled radios.
The radio shown was also sold under the Woolworth's Audition brand.

Don't expect much gain from a mixer. It seems that everything is working as intended, although you may have to dig to fix the distortion.

It was almost shocking to me, to hook up the dusty and cold 415, and have the IF response curve pop up on the screen of my 1962 vintage Stark Instruments oscilloscope looking like it jumped off the page of a textbook - same with the S curve. Talk about an easy to use setup! It gives me some confidence about using it to check the colour circuits of my Zenith TV (the reason I got it in the first place), but that`s a totally different topic.

I will have to figure out what you meant in a previous message by using an external signal to frequency modulate my RF generator, I have an ex-military AF sine wave generator which is built like a tank, like a TV-7 tube tester, it doesn`t contain a single electrolytic capacitor and is fully regulated, and employs RCA red series tubes inside. It`s very low distortion - the RF generator is a Stark, but is some cheap Japanese thing from the 60s inside, nothing like the quality of my Stark oscilloscope.

When I loose couple the Stark`s RF output to the antenna of the radio in question, a lot of signal breaks though and I can hear it in the speaker of the radio quite strongly, so the AM rejection must be less than wonderful - again as an FM alignment newbie I am not sure if that is normal or not. The Stark`s internal modulator is way off being a sine wave, not even close.

My interpretation of the S curve is that it is useful for showing only harmonic distortion - i.e. it is the transfer curve of the ratio detector. If it was curved rather than straight, that would represent harmonic distortion. The distortion I am hearing sounds more like IM distortion, which wouldn`t necessarily show up on the S curve since the test signal is probably a single sinusoidal waveform. I really should substitute another speaker, I have wasted time previously troubleshooting strange forms of distortion in radios and TVs and had it end up to be nothing more than a bad speaker. The fact that it sounds okay on AM may be throwing me off, since most of the program content on AM is talk radio which really wouldn`t be affected by IM distortion as much as the music stations on FM.

The internal modulation oscillator on most low cost signal generators doesn't generate a low distortion sinusoid; in fact most of them are terrible. That's why I suggest modulating the generator with an external audio oscillator.

When you check distortion with such a setup, you need to use enough signal to cause limiting; otherwise the AM will interfere with the FM and you only want FM.

IM and harmonic distortion are related. If the system is linear, both will be low. If it's nonlinear, both will increase. Which parameter you measure isn't usually important. You can measure harmonic distortion with a single source but need two sources to measure IM. You won't, in general, find a system where one reading is high and the other low. Further, in a low end receiver, you can't expect low distortion, although distortion will increase with modulation level. Most low end units work fine with small deviation but when you get full 75 kHz deviation they won't be able to handle it.

Of course, if you listen to rock n roll you won't be able to tell whether the distortion is from the radio or in the source. Classical and jazz are better sources, since the idea in both cases is to present the musical characteristics rather than the tonal characteristics.

And this may just be the definition of "low end receiver", although it does have the extra tube for AFC. Hopefully I can find some time tonight to fiddle with it more... I would like to get it working half way decent and put it back together. One thing I would like to do also is verify that the 415's marker is accurate.

-Max

PROGRESS!
I decided hey let's try swapping the 17EW8s and see if anything changes (both tested good), suddenly the volume on FM massively improved, and the radio actually sounded good. I cleaned the pins and sockets really well, and put the tubes back in their original locations, and the radio works 100% better, lots of volume and much more sensitive. These have got to be the worst tube shields I have seen in my life, they're made of thin sheet metal and fit so tight they have to be practically forced in place. Anyway the radio alignment is now messed up (oscillator and RF trimmer), since I was working on aligning a broken radio before, but finally I see light at the end of the tunnel with this one...