-
Someone gave me an old speaker from the 40s: Hushatone Pillow Speaker model BE-301. Unfortunately, the TS connector is gone and it came with no manual and I haven't been able to locate a manual online.
I tested the ohms of it and it seems to register 23 M ohms, which doesn't seem right. So I don't know if it's messed up inside or I'm reading it wrong.
I found one Ebay posting where it said it was 30k ohms and 5 volt speaker, but not sure if that's accurate.
Besides that, what would be a way to connect this thing to the output of a modern audio interface with a 1/4 TRS jack, or would I need to rig something up to get this speaker power, etc?
Images are attached.
-
If it has a moving coil transducer, then it is open circuit (hopefully connection rather than coil). On the other hand, if it's really thin then it might be some sort of piezoelectric transducer (hopefully not an electrostatic speaker of some sort).
Try measuring its capacitance instead.
If it is piezoelectric, you might get away with just a parallel resistor to prevent it from looking like a purely capacitive load. -
If it has a moving coil transducer, then it is open circuit (hopefully connection rather than coil). On the other hand, if it's really thin then it might be some sort of piezoelectric transducer (hopefully not an electrostatic speaker of some sort).
Try measuring its capacitance instead.
If it is piezoelectric, you might get away with just a parallel resistor to prevent it from looking like a purely capacitive load.
Thanks.
I checked capacitance. It took a few seconds and registered about 10nF. I could hear clicking sound out of the speaker when the probes touched it, so it seems to work.
I found an excerpt from an article about it. It says it has a "crystal element". Apparently these were used in hospitals to put under patient's pillows so they could listen to the radio and not disturb others. My guess is it may have been plugged into a headphone jack, though not sure if old radios used one TS plug for both headphones and a speaker.
Image attached of an ad I found in an old magazine. Seems to only use .01 watt of power.
-
30K @1KHz, 0.01W, and P=V2/R gives V of 17.3V RMS or 49V pk-pk It seems likely that it wont be sufficiently excited by a modern headphone jack. OTOH you probably don't want to push a vintage speaker that hard, so 8.7V RMS, 24.5V Pk-Pk is probably plenty. Its still going to be a PITA to do that without a matching transformer. Some testing with a signal generator capable of at least 10V pk-pk is called for before you commit to anything.
IMHO the original installation instructions probably called for a dummy load resistor on the output transformer and a coupling capacitor chosen to drop the signal voltage to what the BA-301 could handle, from the output valve anode, or possibly one from each anode if the set had a push-pull output stage. -
The 0.01W is presumably a maximum rating. If you could hear clicking during a DMM capacitance test then it's probably reasonably sensitive.
You need very little volume for a pillow speaker (that's kind of the point). I'd just try it and see if it is loud enough. From the screenshot, it just came with an ordinary phone (headphone) plug. Headphone outputs of the time didn't output much voltage if they were mainly designed for magnetic headphones.
EDIT. If you are wanting to experiment with connecting it to a normal stereo headphone output, you could try a series 33R resistor from each channel to one lead and return the other lead to the barrel connection. That will give enough resistance to provide a stereo to mono conversion and help isolate the capacitive load. If it works, fine. If not, then you'll have to decide whether it's worth going to more expensive solutions for greater volume. -
The 0.01W is presumably a maximum rating. If you could hear clicking during a DMM capacitance test then it's probably reasonably sensitive.
You need very little volume for a pillow speaker (that's kind of the point). I'd just try it and see if it is loud enough. From the screenshot, it just came with an ordinary phone (headphone) plug. Headphone outputs of the time didn't output much voltage if they were mainly designed for magnetic headphones.
EDIT. If you are wanting to experiment with connecting it to a normal stereo headphone output, you could try a series 33R resistor from each channel to one lead and return the other lead to the barrel connection. That will give enough resistance to provide a stereo to mono conversion and help isolate the capacitive load. If it works, fine. If not, then you'll have to decide whether it's worth going to more expensive solutions for greater volume.
Thanks man. I will give it a go.
-
The 0.01W is presumably a maximum rating. If you could hear clicking during a DMM capacitance test then it's probably reasonably sensitive.
You need very little volume for a pillow speaker (that's kind of the point). I'd just try it and see if it is loud enough. From the screenshot, it just came with an ordinary phone (headphone) plug. Headphone outputs of the time didn't output much voltage if they were mainly designed for magnetic headphones.
EDIT. If you are wanting to experiment with connecting it to a normal stereo headphone output, you could try a series 33R resistor from each channel to one lead and return the other lead to the barrel connection. That will give enough resistance to provide a stereo to mono conversion and help isolate the capacitive load. If it works, fine. If not, then you'll have to decide whether it's worth going to more expensive solutions for greater volume.
Well I have good news and... other news. I made a cable with resistors as you suggested and the speaker works
Tip and ring conductors I kept together on the speaker side and the shield by itself. Unfortunately, I grabbed the wrong pack of resistors off the bench (I really need to do this stuff when I'm more alert and not at 5AM). I'd rather not peel layers of heat shrink, so it's staying this way unless it might break the speaker. Used 6k8 instead of the 33R you suggested. Not sure wtf that might do, but it works. 
They're rated for 1/2 watt. With more resistance, I have to turn the gain higher which is probably safer for the speaker so it slowly gets louder. Or not, idk lol. At least it gets louder gradually.
To try and work out the polarity since I can't see the speaker cone, I didn't solder the new piece of cable together with the existing one, and touched the resistor leads to the side plugged in and listened. One way sounded more distorted with less bass so I figured that was out of phase and wired them the other way.
At any rate, it works. As one would expect it's heavy in the mids and high mids. You can make out the harmonics of bass and drums in certain music. Interesting sound. Gonna mic the speaker tomorrow, analyze it and mess with some plugins to shape the audio a bit. I like new toys.
Anyway, thanks for the help. I give myself a C-
P.S. soldering small resistor leads to audio cable conductors and thick shield is a pain in the *** -
Its HF -3dB cutoff is almost certainly determined by your choice of resistors and the capacitance of the speaker. The resistors are effectively in parallel, giving a cutoff frequency of slightly under 4.7KHz.
-
I was just wondering last night whatever happened to that pillow speaker! Good to hear that you're at least getting a sensible volume, even if the frequency response needs some work.
I was a bit worried that it would be a bit insensitive - Coming from way back, it may use Rochelle Salt for the piezo element rather than more modern ceramics.These areIt is sensitive to moisture and heat, but they probably considered humidity when they designed it as a pillow speaker. It's good to hear that it hasn't degraded anyway.
https://en.wikipedia.org/wiki/Potassium_sodium_tartrate -
Its HF -3dB cutoff is almost certainly determined by your choice of resistors and the capacitance of the speaker. The resistors are effectively in parallel, giving a cutoff frequency of slightly under 4.7KHz.
Thanks. I have to learn audio formulas so I can calculate these things. If you know any good books or online resource I'd appreciate it.
Do you mean it's acting like a low pass filter cutting out the high end or a bandpass filter?I was just wondering last night whatever happened to that pillow speaker! Good to hear that you're at least getting a sensible volume, even if the frequency response needs some work.
I was a bit worried that it would be a bit insensitive - Coming from way back, it may use Rochelle Salt for the piezo element rather than more modern ceramics.These areIt is sensitive to moisture and heat, but they probably considered humidity when they designed it as a pillow speaker. It's good to hear that it hasn't degraded anyway.
https://en.wikipedia.org/wiki/Potassium_sodium_tartrate
Yeah, I'm happy enough that I didn't fry it. My guess is if it was mass-produced for hospitals back then, and based on the era, they weren't too concerned with people getting nice juicy bass response out of a pillow speaker. Unless the resistor is acting as a bandpass filter and killing the low end and high end simultaneously.
Interesting about the crystals. Will read up more on them.
One thing I failed to mention is that the original cord detaches from the speaker, though it is a very tight fit. Might not be obvious from the bad pictures because the top of the cord is curved to follow he speaker shape. I need to pull it out and clean the contacts. When I tested it I could tell the contacts were dirty or corroded as it messed with the signal or made it cut out intermittently. -
The series resistors from L and R are effectively in parallel for in-phase (mono) content and for single channel content form a potential divider with Thevenin equivalent resistance of half the individual resistor value.
That resistance then forms a single stage RC low-pass filter with the transducer capacitance. See: https://en.wikipedia.org/wiki/RC_time_constant#Cutoff_frequency
Sorry, no book recommendations - I picked this stuff up out of long out of print electronics magazines the better part of a lifetime ago. -
Yeah, I'm happy enough that I didn't fry it. My guess is if it was mass-produced for hospitals back then, and based on the era, they weren't too concerned with people getting nice juicy bass response out of a pillow speaker. Unless the resistor is acting as a bandpass filter and killing the low end and high end simultaneously.
No, I don't think you should get your hopes up regarding bass response. A piezoelectric transducer is going to have a fairly high LF cut-off frequency. Being capacitive, it consumes less and less power as the frequency drops. These things would have been aimed at speech rather than modern bass heavy music - think easy listening and hospital headphones of the time.
At best, you could try filtering out some of the HF, but then you might end up significantly reducing the overall volume level and getting a more of a pronounced midrange (land line telephone) feel. Just try tweaking the series resistor for best sound. -
No, I don't think you should get your hopes up regarding bass response. A piezoelectric transducer is going to have a fairly high LF cut-off frequency. Being capacitive, it consumes less and less power as the frequency drops. These things would have been aimed at speech rather than modern bass heavy music - think easy listening and hospital headphones of the time.
At best, you could try filtering out some of the HF, but then you might end up significantly reducing the overall volume level and getting a more of a pronounced midrange (land line telephone) feel. Just try tweaking the series resistor for best sound.
Oh, I wasn't having my hopes up for bass. Part of the reason I like modifying old devices like this is to record degraded signals that are mostly mids and highs. Was just curious if the resistance would mess with the low end. As Ian said, it seems to behave as a low pass, which is fine (I don't need all the high end anyway). If anything, it's filtering out some of the HF noise.
I'm pretty good at filtering audio like this in post so it's useful in music. I usually start with a spectral noise filter, but not overkill. Some noise is fine and too much filtering produces weird underwater-sounding artifacts. The harsh peaks can be tamed with surgical EQ, dynamic EQ or multi-band compressors. And hum and excessive noise can be reduced with a shelf or L/H pass. Then maybe a slight EQ bump in the mids with a low Q so it's a hump more than a spike. After that I'll add some compression to boost the signal so it's not lost in the mix. Is a rather simple procedure if you know the tools.
From what I'm hearing from this speaker, I'll probably mic it for background vocals with FX or play synth melodies through it and then delay them in conjunction with the normal synth recordings.
Thanks for all your help in getting it to work guys. I appreciate it. If I have extra time, I'll post a recording so you can get an idea of the sound.
-
Just a quick update: I took some 220 grit wet/dry sandpaper to the speaker contacts (they were coated in something white ??), polished them up with 0000 steel wool, used contact cleaner on the cable jacks, wiped everything down with IPA and took a sponge and some Brasso to the bakelite shell. No more crackling or cutting out and even though the case has nicks, it shines