Accepting 5V over the RJ-11 jack seems asking for trouble. Someone is going to plug it into a regular phone socket at some point, so it needs to be capable of surviving at least 60VAC. Even if a secondary pair is used, there are so many different wiring standards in use, it'll end up plugged into something live sooner or later.
Luckily telephone lines are very current limited. You'd be hard pressed to draw more than 100mA even with a dead short at ring voltages. One zener diode might be all you need.
Also not the most unusual thing I've seen over a RJ11, which includes keyboards, RS-232, other random digital IO and full USB with data. But these are normally on items that are not phones.
It would... if you're willing to just let the thing draw up to 0.1W all the time, and fail catastrophically if anything goes wrong with the zener or the supply is more capable than you anticipated (i.e. you plugged into a PoE jack by accident). This is a very bad idea.
It's been a very long time since I had a cable modem that didn't have a pair of 4P4C (aka RJ11 or POTS) jacks on the back. These exist to make available the VOIP service that I haven't wanted for decades.
Others using this same enormous, American ISP probably have similarly-equipped gear.
But I must admit that it has also been a very long time since I connected anything there. I don't know if they supply 48v talk battery or not when the modem has a data-only configuration.
Calling people and playing with phones. I have an internal Asterisk server for me and a few friends. The nicest thing is that when the phone rings, it's a person who I know, not a scam. (Unknown external numbers get direct to voicemail)
Also great for my friends' kids. They can call each other or their grandparents or whoever, without having to touch a cellphone that tries to distract them with YouTube or whatever.
My house uses mostly analog with the Adit 600 (6 lines, 3 using the 70s wiring in the wall, 3 dedicated to specific phones). Everyone else has cheap Polycom desk phones and uses SIP/RTP.
Lovely job, I like the "here's where I've got to so far" style. This gives people enough to work from and make a hand-wired version while he's waiting for boards to come back. Even if he runs out of time to do anything it's leaving a working design and firmware for folk to start from.
It's definitely too easy to put off publishing until it's "properly" done and then it never happens.
We have a similarly converted rotary phone at my local astronomy club's observatory, used to receive visitors. It's driven by an arduino if I remember correctly. The main issue is that the sound quality of its built-in speaker is quite poor, especially compared to modern cell phone speakers we're used to nowadays, with a narrow frequency window and frequent echoes.
In contrast, I've been messing around with some old analog phones and VOIP stuff recently, and I've been blown away with how good the quality is over what I was expecting. The audio bandwidth is low yes (it's G.711u, so 8 KHz), but the actual quality is great, with no hyper-compression artifacts or over-applied background noise filters. And yes this is on the original 60's carbon microphones.
The regular phone calls I normally make on my cell still all sound like they are underwater for some reason, except for those rare times when the better connection is actually working.
Interesting. I remember voice clarity being exceptional on old handsets and being really subpar on gsm and later. Only now we FaceTime voice and discord the quality seems to be coming back.
How can that be if the speakers in phones ubiquitously were so bad?
My understanding is that at some point the audio of phone calls ended up being compressed multiple times on its way to the receiver as it went through different voip setups and such.
Prior to that, uncompressed or compressed once calls weren't so bad on bad speakers.
For the 21700 cell if you want to do this without spending a lot on batteries, but also be sure that any misconfigured hardware can't drain the battery below a damage threshold, use a button top protected 21700 cell. Here's a good 'house brand' one that is popular with 21700 flashlights.
Mmmmmmmmmmhhh. Maybe better to damage
A battery than to have a drone fall out of the sky? Also most protected cells have modest current capabilities, and might not be enough for many drone applications.
By its nature, MUAVs tend to push batteries to their limits and lifecycles tend to be limited. A battery that does mission X must have a maximum weight and a minimum capacity, and a minimum current spec. If you increase the life cycles, you decrease the maximum performance (range, payload, etc).
When you get into bigger drones, more modest restrictions on mission capability give acceptable tradeoffs, but for MUAVs short cycle life is often seen as an acceptable cost.
Frequently , after a hundred (or often much less) cycles in a MUAV a battery is retired from stressful missions due to significant drops in capability. It’s still fine for other uses, but at 90% it’s no longer sufficient for frontier MUAV applications.
The best choice to keep drones from falling out of the sky is to make sure that your auto land is set to a reasonable battery level (calibration by height if you will be at altitude) and use cell cutoffs at or below the minimum safe voltage if at all. (They are usually set to a few tenths of a volt above the minimum safe, often as much as 500mV.) The final battery voltage should be managed manually and batteries tested if over-discharged.
At least, that’s what prevents the most hardware from falling out of the sky onto people’s heads, sometimes bursting into flame on impact. Usually that is seen as the least desirable outcomes of a battery stress situation.
So I was referring specifically to this phone hardware project and to flashlights, which don't fly... But my thoughts on drones are also that it's very different if you're building consumer class things that are meant to be re-used or what's going on in ukraine right now. They're using high amp rate cylidrical 18650 and 21700 cells in 6S2P and 6S3P packs for one way trip munition drones where you don't particularly care if you damage the li ion cells for any future re-use. In which case in the last few minutes of flight you can abuse the hell out of them with high amperage draw down to 2.5V (but beware of extreme voltage sag at the bottom end) and never have to worry.
If we're talking about something that's designed to be re-used then a totally different set of design considerations, as you outline above, become applicable.
> Frequently , after a hundred (or often much less) cycles in a MUAV a battery is retired from stressful missions
Are you talking about lipo pouch cells or li-ion cylindrical? Because if we're talking about state of the art high amp draw rated li ion cylindrical cells, 100 cycles on something that isn't heavily abused for heat and doesn't go below 15% state of charge should not significantly affect its lifespan. But if the batteries are being abused or every last bit of margin on range/payload is needed that it calls for almost "factory fresh" cells, or there's just absurd amounts of DoD money to throw at buying new battery packs, sure, I guess.
If you’re getting into bigger (kg+) drones, it starts to make sense to shoot for 100cycle+ (and for 10kg+, 1000 cycle-ish) lifespans. The economics and the physics work out. But for very tiny systems like the single cell drone featured, the equation starts to favor squeezing every last drop of energy reasonably possible, which means that you can easily get caught out downwind or “stayed to get a few too many pictures” and end up hitting the safety walls. (Partly because the risk of doing so is proportional to the size)
It is interesting to note that Kobalt 24v power tool batteries don't have any internal over-discharge protection. I guess this can make them better-suited for flying telephone use than some alternatives.
I think he might have been calling into question the wisdom of using the standard telephone network to directly place voice calls. Technically it's encrypted but your privacy is anything but guaranteed.
This project uses the same AG1171 SLIC module; it's just not enough by itself and still requires a microcontroller (or at least non trivial logic) even before you get to processing the audio streams.
There's a lot of hardware out there that does 95% of what OP's does, though you'll pay a pretty penny for it (e.g. https://www.peplink.com/products/cellular-modems-adapters/po... ) though ther's been a flood of them from noname sellers on Aliexpress and eBay under the banner of "LTE Landline Converters"
Cool! But who remembers phone numbers anymore? Maybe a small screen that you can navigate by dialling 1 (down), 5 (up) and 0 (select) would be a cool add-on (after dialling 0 or 411 to reach the operator perhaps).
Fashion bag. Dock your cell-phone inside, bluetooth pair to the purse (w/ functional receiver + off-hook sensor) and treat that as a siri-enabled headset in a phone w/ fabric coiled cord.
"Ring ring! Hello? Hey Siri, what's the weather and call my ex so I can hang up on them!"
Accepting 5V over the RJ-11 jack seems asking for trouble. Someone is going to plug it into a regular phone socket at some point, so it needs to be capable of surviving at least 60VAC. Even if a secondary pair is used, there are so many different wiring standards in use, it'll end up plugged into something live sooner or later.
Luckily telephone lines are very current limited. You'd be hard pressed to draw more than 100mA even with a dead short at ring voltages. One zener diode might be all you need.
Also not the most unusual thing I've seen over a RJ11, which includes keyboards, RS-232, other random digital IO and full USB with data. But these are normally on items that are not phones.
>"One zener diode might be all you need."
It would... if you're willing to just let the thing draw up to 0.1W all the time, and fail catastrophically if anything goes wrong with the zener or the supply is more capable than you anticipated (i.e. you plugged into a PoE jack by accident). This is a very bad idea.
also 1-Wire
Dunno about you but I don't know anyone with live RJ-11 jacks in their house anymore.
I have one. It comes from my cable modem.
We still have a landline. It's part of the cable package, but it mostly just gets robocalls.
Didn't know people still had cable packages!
Eh. My wife still watches live TV. I don't. If it were just me, I'd be Internet-only.
But live TV can be streamed via IP?
It's been a very long time since I had a cable modem that didn't have a pair of 4P4C (aka RJ11 or POTS) jacks on the back. These exist to make available the VOIP service that I haven't wanted for decades.
Others using this same enormous, American ISP probably have similarly-equipped gear.
But I must admit that it has also been a very long time since I connected anything there. I don't know if they supply 48v talk battery or not when the modem has a data-only configuration.
I've got them, powered by an Adit 600 in my basement.
What do you use that for?
Calling people and playing with phones. I have an internal Asterisk server for me and a few friends. The nicest thing is that when the phone rings, it's a person who I know, not a scam. (Unknown external numbers get direct to voicemail)
Also great for my friends' kids. They can call each other or their grandparents or whoever, without having to touch a cellphone that tries to distract them with YouTube or whatever.
My house uses mostly analog with the Adit 600 (6 lines, 3 using the 70s wiring in the wall, 3 dedicated to specific phones). Everyone else has cheap Polycom desk phones and uses SIP/RTP.
I have 4 or 6, I can't recall how many are on my Grandstream. Phones are fun!
I am not normal.
Lovely job, I like the "here's where I've got to so far" style. This gives people enough to work from and make a hand-wired version while he's waiting for boards to come back. Even if he runs out of time to do anything it's leaving a working design and firmware for folk to start from.
It's definitely too easy to put off publishing until it's "properly" done and then it never happens.
We have a similarly converted rotary phone at my local astronomy club's observatory, used to receive visitors. It's driven by an arduino if I remember correctly. The main issue is that the sound quality of its built-in speaker is quite poor, especially compared to modern cell phone speakers we're used to nowadays, with a narrow frequency window and frequent echoes.
In contrast, I've been messing around with some old analog phones and VOIP stuff recently, and I've been blown away with how good the quality is over what I was expecting. The audio bandwidth is low yes (it's G.711u, so 8 KHz), but the actual quality is great, with no hyper-compression artifacts or over-applied background noise filters. And yes this is on the original 60's carbon microphones.
The regular phone calls I normally make on my cell still all sound like they are underwater for some reason, except for those rare times when the better connection is actually working.
I've seen projects that swap out the rotary phone speaker with a speaker from a pair of on-ear headphones.
The Secret Life of Machines TV show has a fun demo of how the speakers in old telephones produce sound https://youtu.be/MCgCSMq5Xpo?si=CT5aOGttBpPk9tEM&t=458
Interesting. I remember voice clarity being exceptional on old handsets and being really subpar on gsm and later. Only now we FaceTime voice and discord the quality seems to be coming back.
How can that be if the speakers in phones ubiquitously were so bad?
My understanding is that at some point the audio of phone calls ended up being compressed multiple times on its way to the receiver as it went through different voip setups and such.
Prior to that, uncompressed or compressed once calls weren't so bad on bad speakers.
Could be that it wasn’t bad but has gotten bad because of aging of the speaker, capacitors, resistors, etc.
For the 21700 cell if you want to do this without spending a lot on batteries, but also be sure that any misconfigured hardware can't drain the battery below a damage threshold, use a button top protected 21700 cell. Here's a good 'house brand' one that is popular with 21700 flashlights.
https://liionwholesale.com/collections/batteries/products/pr...
Mmmmmmmmmmhhh. Maybe better to damage A battery than to have a drone fall out of the sky? Also most protected cells have modest current capabilities, and might not be enough for many drone applications.
By its nature, MUAVs tend to push batteries to their limits and lifecycles tend to be limited. A battery that does mission X must have a maximum weight and a minimum capacity, and a minimum current spec. If you increase the life cycles, you decrease the maximum performance (range, payload, etc).
When you get into bigger drones, more modest restrictions on mission capability give acceptable tradeoffs, but for MUAVs short cycle life is often seen as an acceptable cost.
Frequently , after a hundred (or often much less) cycles in a MUAV a battery is retired from stressful missions due to significant drops in capability. It’s still fine for other uses, but at 90% it’s no longer sufficient for frontier MUAV applications.
The best choice to keep drones from falling out of the sky is to make sure that your auto land is set to a reasonable battery level (calibration by height if you will be at altitude) and use cell cutoffs at or below the minimum safe voltage if at all. (They are usually set to a few tenths of a volt above the minimum safe, often as much as 500mV.) The final battery voltage should be managed manually and batteries tested if over-discharged.
At least, that’s what prevents the most hardware from falling out of the sky onto people’s heads, sometimes bursting into flame on impact. Usually that is seen as the least desirable outcomes of a battery stress situation.
So I was referring specifically to this phone hardware project and to flashlights, which don't fly... But my thoughts on drones are also that it's very different if you're building consumer class things that are meant to be re-used or what's going on in ukraine right now. They're using high amp rate cylidrical 18650 and 21700 cells in 6S2P and 6S3P packs for one way trip munition drones where you don't particularly care if you damage the li ion cells for any future re-use. In which case in the last few minutes of flight you can abuse the hell out of them with high amperage draw down to 2.5V (but beware of extreme voltage sag at the bottom end) and never have to worry.
If we're talking about something that's designed to be re-used then a totally different set of design considerations, as you outline above, become applicable.
> Frequently , after a hundred (or often much less) cycles in a MUAV a battery is retired from stressful missions
Are you talking about lipo pouch cells or li-ion cylindrical? Because if we're talking about state of the art high amp draw rated li ion cylindrical cells, 100 cycles on something that isn't heavily abused for heat and doesn't go below 15% state of charge should not significantly affect its lifespan. But if the batteries are being abused or every last bit of margin on range/payload is needed that it calls for almost "factory fresh" cells, or there's just absurd amounts of DoD money to throw at buying new battery packs, sure, I guess.
If you’re getting into bigger (kg+) drones, it starts to make sense to shoot for 100cycle+ (and for 10kg+, 1000 cycle-ish) lifespans. The economics and the physics work out. But for very tiny systems like the single cell drone featured, the equation starts to favor squeezing every last drop of energy reasonably possible, which means that you can easily get caught out downwind or “stayed to get a few too many pictures” and end up hitting the safety walls. (Partly because the risk of doing so is proportional to the size)
> ...like the single cell drone featured...
Featured where?
Oh that’s hilarious.
I got this article conflated with another I was reading. My apologies to all.
Age is a hell of a drug. My apologies, wrong article.
Oh that’s hilarious.
I got this article conflated with another I was reading. My apologies to all.
What are the chances of this telephone falling out of the sky due to battery failure?
I was going to say pretty low but then I realized you weren’t just be ing snarky lol
I got this article conflated with another I was reading. My apologies to all.
Still, if you don’t want your phone to fall out of the sky, Put it on a drone with proper battery limit discipline.
It is interesting to note that Kobalt 24v power tool batteries don't have any internal over-discharge protection. I guess this can make them better-suited for flying telephone use than some alternatives.
;)
Who mentioned drones? This is for a phone.
Oh that’s hilarious.
I got this article conflated with another I was reading. My apologies to all.
When I finally agree to cave in and buy a phone for my soon-to-be teenager, little does she know this is the phone I have in mind…
i wouldn't want my child making unencrypted calls :p
It's no more or less encrypted than any cellular phone making a call across your local 3GPP-whatever standard network.
Doesn't this project go straight into modern standard mobile phone systems? It'll be encrypted then.
I think he might have been calling into question the wisdom of using the standard telephone network to directly place voice calls. Technically it's encrypted but your privacy is anything but guaranteed.
Slightly off topic but you can also use a SLIC module to talk to old phones:
https://youtu.be/nibV_idJjlg
https://www.ebay.com.au/itm/305557776577
This project uses the same AG1171 SLIC module; it's just not enough by itself and still requires a microcontroller (or at least non trivial logic) even before you get to processing the audio streams.
I remember someone building this (couple of years ago) https://skysedge.com/telecom/RUSP/index.html I so wish this was an actual product.
There's a lot of hardware out there that does 95% of what OP's does, though you'll pay a pretty penny for it (e.g. https://www.peplink.com/products/cellular-modems-adapters/po... ) though ther's been a flood of them from noname sellers on Aliexpress and eBay under the banner of "LTE Landline Converters"
This is very neat.
An uglier way to do it is with an off the shelf pulse to tone box and a router with VoIP support, like an N300.
Do you think it could use WiFi calling? I suspect that can't be done is it requires one of a limited set of carrier approved phones in each area.
Love seeing old tech brought back to life like this. The tactile experience of dialing is just something else.
Cool! But who remembers phone numbers anymore? Maybe a small screen that you can navigate by dialling 1 (down), 5 (up) and 0 (select) would be a cool add-on (after dialling 0 or 411 to reach the operator perhaps).
The next move is to go further back in user interfaces: To place a call, pick up the phone and dial 0 or turn the crank to ring the operator.
Have a bot do the rest, exactly like a telephone operator.
https://www.reddit.com/r/esp32/s/eSlEHcXbJ0
That would be funny you're at a club with this phone and doing phone calls
Fashion bag. Dock your cell-phone inside, bluetooth pair to the purse (w/ functional receiver + off-hook sensor) and treat that as a siri-enabled headset in a phone w/ fabric coiled cord.
"Ring ring! Hello? Hey Siri, what's the weather and call my ex so I can hang up on them!"
or in the corporate office
What would be worse this rotary dialer or cherry blue mechanical keyboard