Hopefully it’s the right place for this. Please (re)move if necessary.
A friend’s grandfather had a 40-81 portable radio he asked me to get running. Did the typical replace caps and brittle wires and got it working. Sounds amazingly good, really. Part of that was building a battery box to hold the A and B supplies.
Anyway, the battery box needed a cover so decided to make some in the style of the old batteries. A little massaging of the designs was involved (not 100% accurate but fairly close). These are 3D printed to fit the compartmented box I also printed that holds the batteries. For those that care, I used PETG for the box and its bits for heat tolerance near the tubes (which don’t seem to get too hot at all) but PLA for the cover plates for easy printing of the designs.
I knew nothing about tubes when I started and had some missteps. I get the radio back this week to mod the filament supply a little. Will post more about what worked and didn’t, etc.
Anyway, for whatever it’s worth. They snazz up a battery project. My friend only wanted a personalized cover plate. These were extra for fun. The high voltage Philco isn’t historical. It was just in the style my friend wanted but personalized and looked ok.
(This post was last modified: 07-26-2026, 04:52 PM by MZip.)
Thanks! I love the retro stuff like that. And with the back in place you can’t see it. But it looks cool when you open it up.
The battery box isn’t to scale from a real battery so another deviation. I was trying to allow as much air circulation as possible but the 40-81 just doesn’t seem to get too warm at least so far that I’ve noticed.
(This post was last modified: 07-26-2026, 05:03 PM by MZip.)
Great job!!! I like the Philco and RCA ones, but the Burgess is my favorite. I remember the Eveready "9 Lives" Logo with the black cat jumping through the "9"
If the radio is not used much, don't even bother trying to make an "A" battery eliminator. Even a good 3 terminal regulator gets a little squirrelly at 1.5V.
These tubes run very cool as both A and B currents are very low. Even the Audio Output tube should be cool to the touch. Total B Current is about 14mA (0.014A) at 90V for a total power dissipation of 1.3W. Total "A" current is 20mA at 1.5V for a total of 0.03W. Audio output power of this set is about 0.1W.
So total power dissipation for this set as less than 1.4W, as opposed to the 35W power dissipation and 1W audio output from an AA5 Radio, where touching the 50C5 OR 35W4 will give you a good burn.
Since the "A" (filament) current for this set is only about 20 mA and the average Alkaline D Cell has a capacity of between 8,000 and 12,000 MA Hours, you could get as much as 400-600 hours of operation on one D Cell. This is much cheaper than building a power supply that, if it over-volts, could destroy $60 - $100 worth of tubes. Considering the cost of 60 AA batteries that would be needed to build a 90V B battery that would last 100-200 hours, building a 90V power supply does makes sense.
"Do Justly, love Mercy and walk humbly with your God"- Micah 6:8
"Let us begin to do good"- St. Francis
Hey MrFixr, thanks for the kind words and also for all the great info!
I’m with you on the Burgess and RCA covers. Definitely my favorites too.
When I started on my friend’s radio a few things were really obvious and easy but the battery box thing has been more of an adventure and at the start of it I didn’t appreciate or know all the nuances. I found information online but come from an op amp and digital electronics background.
I found the tube data sheets and they call for 1.4V and I found another article talking about care and feeding and how low voltage can be bad through poisoning. So naturally I started with a DC-DC converter to properly feed the tube filaments. Directly-heated cathodes don’t really like that switching noise it turns out so did more digging and that’s why sensitivity went way south.
I was trying to keep filament voltage pegged at 1.4 and the DC-DC converter I used had a neat feature of dropping out instead of letting the voltage sag like a battery or LM317. Since it’s my friend’s radio, he isn’t especially handy with electrical stuff and I was looking to have it work or not with no in-between.
I tried some neat little pi filters used to trap out drone motor controller noise in FPV drones and they make a big difference in knocking down the harmonics on the converter input and output but the cutoff is too close to the fundamental so it’s still there - just smaller and rounded off.
I get the radio back Tuesday and want to try the DC-DC converter still to see if knocking down the harmonics is “good enough”. I put a cap on the output as is and it really helped so hopefully the pi filters are better enough. But the fallback is maybe the LM317 or just straight 1.5V battery and a resistor.
I know what the article said about care and feeding but in practice, how tolerant are the 1.4V direct-heated tubes to filament voltage? I know the originals were battery powered but tubes were plentiful. With them getting harder to find, I want to be as easy on them as is necessary but I don’t have much feel for what range they are ok with.
All that data is very handy. Thanks! Still sorting out the details how the filament supply will ultimately go. I don’t want to appreciably change sensitivity but also, my friend may use an AM transmitter and Bluetooth receiver to make it where they can play music from phones and tablets so sensitivity may not be so important to him.
On the cover designs, I used battery images from the web to draw them up. I saw where one person had printed the Philco wave version and used it to wrap a box and that looked cool. I 3D print a lot so making the box and covers was really easy.
I’ll know more soon if I can get that DC-DC converter quiet enough. I know just the cap made a big difference in the noise but not so keen on what frequencies are actually limiting sensitivity and if the pi filters will help enough. Will post what I find. I do like the fast dropout of the converter vs a long slow ride down the battery curve. If it was mine, no problem. But it needs to stand up to use and demos without monitoring filament voltage.
The Ever Ready B-107 is a British market battery, it's likely that the original was similar to the blue and yellow Philco battery. Never had many problem with voltage regulators, but there are more then a few circuits and plans for battery eliminators online, given the current draw of 1940s era battery tubes a voltage regulator isn't going to be run at anywhere near it's full current rating on the A+ side. As a safety precaution you could add an appropriate sized Zener diode across the filament string. 1940s era battery tubes are largely inexpensive, but it's no reason to be careless with them.
Regards
Arran
These "1.4V" tubes were designed to run on a 1.5V battery; no resistor is needed. The "A" batteries were not as big as a "#6" Dry Cell (the big round thing used in crank telephones, Model T ignitions, 1920's radios like the Radiola III, etc., but were likely 2 "F" cells in parallel. (A 6V lantern battery used 4 F cells in series.) I don't think that the voltage dropped to 1.4V under load.
Like Arran says, if you design a power supply for the filaments, do put a 1.8 - 2V Zener across the output, anode to negative, cathode to positive to prevent tube damage in case of severe overvoltage. This could be done in the radio itself. Do note that unless protected by a 0.5A fuse between the output of the LM317, the load and Zener to protect overcurrent and permanent shorting of the Zener. However, the tubes are protected.
There has to be a minimum of 3.5V difference between input and output for stability of the LM317. This means that you should use a 6V supply because a 5V supply is at that 3.5V limit.
You can either breadboard the supply on "perfboard" or buy a complete regulator assembly from Amazon, eBay or (gulp) AliExpress or (Gulp Gulp Gulp) Temu.
I have not had a problem using switch mode power supplies for filaments on the 2V tubes that were used from about 1930-1938 in battery sets. I used a 3V Wall Aart power supply and the correct amount of resistance to drop the voltage to 2V depending on the total current of the filaments. However, when trying to use one for B supply, I ran into very much switching noise. It may be about shielding of the supply or whether or the output is rectified by full wave or half wave rectification (The unit that I used had half-wave rectification.)
"Do Justly, love Mercy and walk humbly with your God"- Micah 6:8
"Let us begin to do good"- St. Francis
I think the directly heated tubes are more trouble for noise on the filament supply since even the voltage gradient from one end of the filament to the other affects operation. Were the 2V tubes indirectly heated? There the cathode is grounded and even probably shields the other tube elements from noise?
Anyway, good info on the tube filament voltage and thanks Arran for the zener idea. Right now I am powering the converter from 6V (4 cells) and was going to also for the LM317 if I use that. If I go with just a single battery it also removes the zener need I think.
The modules I’m using are from Amazon but still are fairly cheap builds. But they seem to work ok besides the normal quirks they each bring. Same for the low-pass pi filters I may or may not use. Very handy for cobbling things together if you don’t want a wall wart or to draw up your own board. ?
The reason this is kind of odd and out of order is it’s not my radio and I didn’t want to power it up until I got all the caps replaced and such. Being a tube noob I just followed the lead in the battery articles I could find but then didn’t realize how sensitive direct-heated cathodes were to noise especially in the detection and amplification stages.
And I agree. At the low currents involved (about 200mA total) I wouldn’t expect much sag on a new battery.
It’s an adventure and I learned a lot so far. My friend was really hoping I could get his grandfather’s radio working again and that’s at least done. Not perfect yet but not bad. The dc converter really does cause problems in this application. Makes sense though.
Hello Mzip,
Yes, I have used LM317 since I was a kid and built tons of circuits using it along with my other favorite IC 555 timer.
I have some battery sets that I also plan to build some power supplies for, and I have bought my prebuilt supplies from AliExpress in the past, but they are cheap to build .
There was a really good plan I linked to before that involved using a relatively inexpensive mains operated transformer that you can get anywhere, it had two secondaries which is useful. I have a fairly large collection of 1930s and 40s tube battery radios compared to most people, being a teenager, back in Alberta, they were a way for me to get into pre war radios with very little money, and most of the time I could get them to play without having to do much to them. Unless the 40-81 was stored in an unheated environment it may have still worked without doing anything to it, aside from the rotten battery cable.
I also have a 40-81 somewhere, but the back is missing, I suspect that the A-B pack that was in it may have leaked as there is a patch of corrosion on the rear apron of the chassis. It's likely that the back was damaged as well and throw away, these sets are sort of weird for a portable in that the batteries were stored in the rear rather then in their own compartment under the chassis. These would be a good entry level set for a newcomer to tube radios to work on, I think that there are only 4 paper caps in the thing, and one electrolytic.
Regards
Arran
That has at least one of the plans, it's also posted on a website where a guy made one from a double size electrical "device" box like you would use with an oven outlet. The transformer is available through both Mouser, and DigiKey, and likely a few others too. In this case the B+ of the 40-81 is 56 volts, which is sort of an odd value, 22-1/2, 45, 67-1/2, and 90 volts are more typical.
Regards
Arran
Hello Arran,
I also have a bunch of battery-operated radios including an RCA console one of my Philco Battery radios is a rare Set that I got from our late leader Ron.
Sincerely Richard
![[-]](https://philcoradio.com/phorum/images/bootbb/collapse.png "[-]")
