Philco 48-461 radio restoration
Philco 48-461 radio restoration
After some experimenting and theorizing I am not happy with the aux input solution. Previously, I had just connected the left channel and set mono in Windows Accessibility. But most devices are going to be putting out stereo, so we need a mixer of some sort to convert stereo to mono. I tried the passive solution with two resistors, but when used in conjunction with the audio isolation transformer, there are no good resistance values to use. The input winding on the transformer is only 100 ohms (we use half of the transformer winding to get step-up). So in order not to lose too much volume, the two resistors should be around 10 ohms (signal is now 90%). But that creates so much crosstalk and back-feed between left and right that noticeable distortion occurs.
Therefore, I am pursuing an active solution that will give proper isolation between left and right and provide some pre-amplification. I'll power the circuit from the radio tapping voltage off one of the tube filaments, and feeding it into a bridge rectifier, a filter cap, and a 7805 linear regulator. Alternatively, I can try tapping 8Vdc off the resistor between the first two filter caps and run that directly into a 7805. As long as the negative of that does not have to touch B- anywhere, that should work. Great in theory, but let's see how it goes.
Therefore, I am pursuing an active solution that will give proper isolation between left and right and provide some pre-amplification. I'll power the circuit from the radio tapping voltage off one of the tube filaments, and feeding it into a bridge rectifier, a filter cap, and a 7805 linear regulator. Alternatively, I can try tapping 8Vdc off the resistor between the first two filter caps and run that directly into a 7805. As long as the negative of that does not have to touch B- anywhere, that should work. Great in theory, but let's see how it goes.
Philco 48-461 radio restoration
An active solution with pre-amp would be sweet. Use low-noise op-amps. Excited to see your results!
Philco 48-461 radio restoration
Guys, I found that 1K ohm resistors work just fine passively (turns out it is not the DC resistance that is important, but the impedance at audio frequencies). The resistors join and feed into the isolation transformer. The other side of the transformer connects to the radio B- and radio audio input (via the switch). That means the jack is now totally isolated from the AC line. We avoid the complexity of active circuitry. Also replaced the bulb with the LED. Voltages appear fine for the rectifier filament (3.9 Vac with bulb, 6.3Vac with LED, 7.5Vdc spec maximum with good bulb/LED installed, 15Vac spec maximum with burned-out bulb/LED). Gonna leave the speaker phono jack for now as the alternative feels chintzy and we want to have a robust solution. Hope DG doesn't spank me!
Philco 48-461 radio restoration
That's all fine with me.
Philco 48-461 radio restoration
Fine with me too. Ship it!
Philco 48-461 radio restoration
Here's the proof. You can see the bluetooth cable connected to the jack, the aux/radio switch, and the isolation transformer with passive resistor summing of the left and right channels. Yes, that is my favorite JBWeld epoxy holding the transformer and switch in their respective places. That black wire connecting to the bottom side of the transformer is B- (AC line neutral). The transformer totally isolates it from the aux jack and connected device. Note that the jack itself is an isolating jack (plastic, baby) that prevents chassis from connecting to the jack common.
Now Rocky wants me to work on software for a while.
![Image](https://rationalqm.us/misc/48-461/48-461%20aux.jpg)
Now Rocky wants me to work on software for a while.
![Image](https://rationalqm.us/misc/48-461/48-461%20aux.jpg)
Philco 48-461 radio restoration
Guys, my current aux input solution sucks. I mean sure, you can hear the audio, but it is way down in volume and the 1K resistors are not big enough to avoid distortion. I deluded myself thinking it was OK. So I'm going back to the drawing board. Rocky, please mark the thread unresolved.
We need an active mixer solution. No debate about that! But since people are just going to plug in directly from iPods, MP4 players, bluetooth, etc., the circuitry has to be entirely inside the radio with the audio referenced to B- (line neutral). So I have decided to isolate the entire radio with a small 35 VA isolation transformer. It's small enough to fit within the enclosure. Then we can do whatever we want with the aux input circuitry without passing line neutral back on the audio cable, because B- would then no longer be continuous with line neutral. No audio isolation transformer is required, because the whole radio is isolated. I didn't consider this earlier because I was looking at 100 VA transformers that are too big to fit. In theory, we need 110Vac at 0.2 A, which results in 22VA. So a 35 VA transformer should be fine. The transformer I ordered is about 2" x 2" x 3".
I plan to use an LM386 set up in summing mode for audio L and R. The input resistors to the summing network will be 100K ohms, eliminating any distortion due to crosstalk. Power is derived from the filament string using a diode, capacitor, and 7812 linear regulator. I have cobbled this up on the bench and it sounds awesome.
We need an active mixer solution. No debate about that! But since people are just going to plug in directly from iPods, MP4 players, bluetooth, etc., the circuitry has to be entirely inside the radio with the audio referenced to B- (line neutral). So I have decided to isolate the entire radio with a small 35 VA isolation transformer. It's small enough to fit within the enclosure. Then we can do whatever we want with the aux input circuitry without passing line neutral back on the audio cable, because B- would then no longer be continuous with line neutral. No audio isolation transformer is required, because the whole radio is isolated. I didn't consider this earlier because I was looking at 100 VA transformers that are too big to fit. In theory, we need 110Vac at 0.2 A, which results in 22VA. So a 35 VA transformer should be fine. The transformer I ordered is about 2" x 2" x 3".
I plan to use an LM386 set up in summing mode for audio L and R. The input resistors to the summing network will be 100K ohms, eliminating any distortion due to crosstalk. Power is derived from the filament string using a diode, capacitor, and 7812 linear regulator. I have cobbled this up on the bench and it sounds awesome.
Philco 48-461 radio restoration
I got the audio amp built on a piece of vector board and working well. The two Triad isolation transformers arrived but they did not satisfy. One had an open primary = busted. The other one had a secondary that was 10Vdc too high, so it output 136Vdc which is too high. When I reversed it, the output was only 95Vdc which is too low. So that one is unusable too. The busted one is being replaced, maybe it will actually meet it's own spec. Never mind, though, I ordered a couple 187D120 transformers, which should be good and are smaller. The Triad was going to be a tight fit but the 187D120 will be a bit easier to find room for.
https://www.mouser.com/ProductDetail/Ha ... J2rQ%3D%3D
https://www.mouser.com/ProductDetail/Ha ... J2rQ%3D%3D
Philco 48-461 radio restoration
Another option is to just integrate a bluetooth receiver into the radio, connected to the LM386 amp input and powered internally. Then B-/neutral does not appear outside the interior of the radio. A small disadvantage is no direct connect of cables from MP3 players, etc. No big deal. I only use bluetooth anyway and not having an external one with extra NOISY wall wart is a big plus. AA5 radios hate nearby wall warts. This idea is really appealing to me and I have ordered a couple Drok receivers.
https://www.amazon.com/Bluetooth-DROK-R ... op?ie=UTF8
https://www.amazon.com/Bluetooth-DROK-R ... op?ie=UTF8
Philco 48-461 radio restoration
Levi told me he knows a lot about audio from his long musical career. He said the LM386 sucks and anyway I shouldn't be using a power amplifier as a preamplifier. He's right. The noise and distortion are terrible. He said cobble up a simple op-amp stage. We had a little preamp sitting around and I hooked that up and it is night and day better. So I ditched the LM386 idea and ordered a couple small preamp modules:
https://www.amazon.com/Ximimark-Magnifi ... =8-29&th=1
The preamp does two things: 1) gives a high impedance input so we can have high value resistors to combine the left and right, and 2) gives gain so that the bluetooth and radio levels can be close to each other (don't want to get blown out of the room when switching back to radio, LOL).
So we'll have a bluetooth module and a preamp module, both powered from the heater string via a 1N4004 diode, a nice fat cap, and a 7812 regulator. We'll still have the bluetooth/radio switch of course, but we ditch the aux in jack and the audio isolation transformer. And because everything is inside the radio, the main isolation transformer is not needed.
Levi
https://www.amazon.com/Ximimark-Magnifi ... =8-29&th=1
The preamp does two things: 1) gives a high impedance input so we can have high value resistors to combine the left and right, and 2) gives gain so that the bluetooth and radio levels can be close to each other (don't want to get blown out of the room when switching back to radio, LOL).
So we'll have a bluetooth module and a preamp module, both powered from the heater string via a 1N4004 diode, a nice fat cap, and a 7812 regulator. We'll still have the bluetooth/radio switch of course, but we ditch the aux in jack and the audio isolation transformer. And because everything is inside the radio, the main isolation transformer is not needed.
![grateful :salute:](./images/smilies/icon_salute.gif)
Philco 48-461 radio restoration
Honestly, everything in the last post is pie-in-the-sky. Everything I tried ran into complex issues that could not be easily solved. So I just left the simple solution, i.e., aux jack, switch, passive combining network, and input audio isolation transformer. So this baby is done!