Inside a door entry system voice unit
ฝัง
- เผยแพร่เมื่อ 11 ส.ค. 2022
- This is a very standard door entry system voice unit that is often found in the main door button panel of door entry systems in the UK. The Bell System units are the most popular in the UK, and when you look at the circuitry in them it becomes clear why. The design is simple, but clever and has good protection against short circuits and installer error.
This module is the main part that deals with amplifying audio from the built in visitors microphone and also the microphone output from each flat/apartment's phone.
There is a separate video showing the circuitry in the version 5 phone here:-
• Inside a door entry phone
I specifically avoid using intrusive mid-run video adverts for a more relaxing viewing experience.
If you enjoy these videos you can help support the channel with a dollar or two for coffee, cookies and random gadgets for disassembly at:-
/ bigclive
Patreon supporters get early access to advert-free videos as they are made, and also regular live streams.
The Patreon support keeps the channel independent of TH-cam's advertising algorithms allowing it to be a bit more dangerous and naughty.
Other contribution options are available at:-
www.bigclive.com/coffee.htm
#ElectronicsCreators - วิทยาศาสตร์และเทคโนโลยี
Info: 'T' and 'R' are traditional telephone letters dating back to the first manual switchboard, and mean 'Tip' and 'Ring' on the plugs and jacks.
On manual exchange switchboards the tip and ring carried the two conductors of a balanced audio circuit. This was a two-way circuit. The sleeve connection was used to operate relays to indicate that the circuit was in use (engaged)
Ex 10 by 50 GPO switchboard operator here...I remember that!!!
And I thought it was Transmit and Receive...
@@kapegede The old 4 pole jack plug too📵
@@alexmarshall4331 You're probably thinking of the Plug 404 which was big and heavy, it was replaced by a smaller version. Our manual exchange in Yeovil was still being expanded in the 1960s and second-hand switchboard positions from other exchanges were used. Replacing the big headset jacks with two of the smaller type was one of the necessary jobs.
Nice, no corners cut while being as basic as possible. That's how units should be designed.
Remove as much as possible but not more.
Good to see the designer had Murphy's Law in mind when they created it.
No corners cut? But There are literal cuts in the corners
Of course corners have been cut. The traditional system is that the call would go to your staff and they would handle any callers or visitors…
Trim away the fat, but keep just enough for flavor.
12:16 The RC filter is known as a Zobel network, or a Boucherot cell, and is used to dampen high frequency oscillations that may damage the speaker
This oscillation is also not good for the amplifier. It will get hot even if you here nothing in the speaker.
The rc network is need to get audio amp run stable.
The purpose of the Zobel network is to compensate for the rising impedance of the speaker, with increasing frequency, due to its voice coil inductance. While there could be potential for oscillation if the amplifier stability criteria are compromised it's good practice to maintain a level impedance curve over the intended operating frequency range.
For your amusement here is a story about door entry systems, but first a bit of background.
Many years ago I attended a class in electronic music and Musique concrète. To give you an idea of the era we used an EMS VCS 3 synthesizer to create sound effects. (look it up!)
In my city were three tower blocks of luxury apartments, each tower had 60 apartments So I went there with my tape recorder and pressed all 60 buttons on tower number one. As they answered they could hear the next apartment to respond. So you'd have 60 people all saying "Hello" and thinking that the next person to respond was outside and trying to gain entry.
Repeat for towers two and three!
It was hilarious! I edited the tape to combine all three recordings, I still have that tape but no longer own a reel to reel :(
Those line resistors are very similar to a traditional CO setup. The 48 volt supply has a resistor that drops the voltage to about three volts when you go off hook and start drawing current.
That shunt resistor in parallel with the center zener is diverting some of the current around the zener and provides a discharge path for the cap. Without it, the cap would always have a 9.6v charge key on it. It should also help make the supply a bit more stable by keeping enough load in it to keep the input cap from charging to peak voltage and limiting the voltage swing.
Thank you Clive for having published this video quite quickly after the first one on the subject.
I wasn't expecting that much protection on the circuit, but maybe people installing these phones often lack knowledge, or mistakes can happen.
(This reminds me of all the smart meter fires here in France, all because they were hiring unqualified subcontractors that didn't tighten the screws quite enough).
Now that you've covered the basic, analogue door phones that allow residents to have a bit of fun I think you should make videos about the concierge units aka guard stations that are basically switchboards for digital door entry systems.
They allow the concierge to call any flat, or a flat to call the concierge, or even the concierge to connect two flats not unlike phone operators were doing back a few decades ago.
I think it would be an interesting subject.
you have to allow for less than smart service tecs :)
In Germany the intercom made by Ritto seems to be more common than the Bell one. For Ritto exists a part called "Mithörsperre" (monitor lock). This is a relay module which prevents conversations between different apartments. If you have a magnet, you can bypass it ;-) By the way: From Ritto I know three generations: The first one with carbon microphone/speaker and electromechanical buzzer, the second one equivalent to the here shown Bell system and the third one where ringing and door opener run via digital bus. So if you want to doorbell prank the neighbors you only need to send the right signal into the bus ;-)
Interesting. Every country seem to have their big favourite maker of door entry systems. In Germany there are systems made by a company called Siedle basically everywhere. The systems were quite similar but have now evolved to a pure 3 wire digital system where you can't listen into conversations anymore, because of privacy protection etc.
And you also can't unlock the door if someone hasn't rung you first I suppose ?
i can find no exploits on a casual look. github has a project siedle-in-home-bus-avr to talk to that bus.
@@psirvent8 Yes it can behave like that. However, this safety feature can be turned on ot off during initial configuration.
I have opened those Siedle units before. The old ones were very well built. But now it's quite crappy quality, especially for the prices they ask for them.
@@ChipGuy That's interesting, all the digital intercoms that I know of don't allow you (Or I should say the installer) to disable security features.
Not being able to tap into your neighbours conversations, to talk to someone in the street without said person having rung you first or to open the door without having been rung seems like the norm nowadays instead of an extra security feature at least in France and Italy as far as I'm aware of.
Engineering with human frailty in mind. I like it. Yes we are all professionals who know what we're doing, but at some point, some muppet is going to stick one of the wires in the wrong hole, and that muppet may well be me. So let's design it so that the whole installation doesn't let out the magic smoke when that happens. Good thinking.
Looking at the troubleshooting manual for the Bell units, the T R and 0 (zero not O) are transmit, receive, and zero volt rails
( doorentrydirect.com/acrobat/bell/data/Bell%20801%20Trouble%20shooting.pdf ) - R and T (receive and transmit) are referenced to the function of the audio from the handset at the far end to the speech module (receive of handset, transmit to handset).
I guess C is Common and H is Hot. Anyway it's also Transmit and Receive by door unit.
Interesting. Noted a couple of things:
- Power supply: The RC on the base of the transistor is also acting as a capacitive multiplier there will be practically zero noise on the emitter.
- Audio: The capacitors on the inverting op amp inputs are there because there is no negative rail -otherwise theyd be tied to 0V. This is common in alot of devices that use 9V batteries.
Thanks Clive. Love your vids they have really helped me in the last few months dealing with my second cancer diagnosis in the last 2 years and the subsequent chemo / radiotherapy treatment. Hope this comment isn't a downer for your subscribers.
I wondered why our door entry voice system sounded like a drunk parrot. Clive, thank you lots for this! ~Wendi 🌻
I'm a little miffed. I was just browsing the [silicon-based electronics supplier] website, looking at parts for a project while watching your video in the background, when pages stopped loading and instead I get a "under scheduled maintenance" message. I mean, it's only 2:45 am, that's practically primetime for getting work done on a side project!
OK, perhaps it's a subtle hint that I should just go to bed.
I like the way you draw diagrams. Easy to read. And I like the enlarged photos, too.
After I had watched a few of big Clive's videos, I thought: "Let's reverse engineer the lamp dimmer" which was faulty. After an hour or so and many bad drawings I gave up!! It's a great skill to have, and one of the things which makes BC's videos so easy on the mind.
My favourite door entry voice system began with the introduction of a clip from Mr Humphreys from "Are you being Served? where he says in a very faux deep voice "Menswear 3rd floor" and then the standard blurb follows. They changed it from time to time and I vaguely recall that there was one that went "Inside legs measured" again another Humphreys snippet.
Typically people use "T" to represent Transmit and "R" to represent "Receive". Usually they will be referenced as "Tx" and "Rx". Also a lot of the time the "T" on one device will go to the "R" on the other device and vice versa.
I love these videos. Keep it up.
Big Clive: If Bob Ross and Stewie Griffin had a love child.
Love watching all your videos, sometimes as background to what I’m working on.
I started tearing apart electronics to see what was inside and how they worked when I was around 5 or 6 years old.
Possibly so, however I've yet to hear Clive say when he releases the magic smoke, "We don't make mistakes, just happy little accidents.”
Clive, the T stands for tip, and the R is for ring. These correspond to the 1/4" type plug used in early phone systems. Think of a switchboard plug that the operator would use back in the day to connect a phone call. The ring is the shaft or shank of the plug, whilst the tip is obvious at the end of the plug.
More likely T=TX audio from door (the call originator) and R=RX for receive audio from the called unit. Ring + Tip are primarily for balanced telephone circuits which carry both TX & RX on the same balanced pair.
I think it's TX/RX
Pretty sure the T and R wires are probably historical references to Transmit and Receive, named by older systems and re-used when new systems were installed using the old wires, so as to not confuse the installers.
It's Tip and Ring.
@@hattree Is that similar to Whip and Sting or am I think of somthing else ??
@@brucepickess8097 it's from the old plug connectors on manual o switchboards. The two connectors on a phone plug were tip which was the point tip and ring which is the sleeve behind it.
Thank you Clive
Very interesting. It’s great to see you cover audio gear. That’s my ‘thing’.
I put up a fence around my house with a gate on the driveway. I was thinking about some kind of intercom like this, since now nobody can knock on my door. But then I thought..."fantastic, now nobody can knock on my door!" I think I like it better that way...
Lots of key design principles! Tidy!
The resistor across the zener diode keeps the regulator circuit working (with a lower output voltage) when the input power falls below a value that will turn-on the zener. It keeps some voltage drop across the base-collector junction of the pass transistor so it still has headroom to smooth the incoming power. The capacitor to ground at the common node will help remove hum. It becomes a capacitance multiplier instead of a zener reference for low input voltages.
It will keep the unit working even with a poor power source.
It is a good design as you stated.
Or, T and R could be for Transmit and Receive if the designer was a Navy person familiar with sonar lingo
I was thinking the same.
I like it. This is a good circuit for learning
It is important to use twisted pair with the terminals they say or there is often quite a loud buzz on the audio. Can't think much more, currently off work ill possibly with corona :-:
Hope you recover soon.
It isn't a balanced feed, so ideally it would use shielded pair or audio co-axial cable, but that would potentially cost more to install.
Quiz will be sent this weekend??? LOL again another great video..you ROCK!! Big C. DVD:)
I've used the TDA2822 in many different projects over the years. It's very useful to make a simple headphone amplifier. Sound quality over headphones is pretty good.
Nice design.
Seeing the 4 mounting screw holes on the front with the slices cut out of the corners makes me imagine some engineer was like, "The frame keeps cracking here" and someone else goes, "Why don't we just crack it ourselves?" Love it.
It's a traditional way to fix a flat plate between 4 screws in the corners of a box. Probably not invented for this unit.
A very good Saturday morning to you all from Wellington Somerset
The network of the 22ohm resistor and the capacitor in parallel with the speaker is there to dampen any unwanted high frequency oscillations of the amplifier. It is called a Zobel network.
I am not sure why a Zobel network is not present at the output of the amplifier driving the T terminal/speaker bus, maybe because there is the 470ohm series resistor and it is driving a bunch of cabling instead of a speaker voice coil directly.
A building I lived in did an orientation about the building, one of the things was the clarity of the intercom system to the front door, if someone tried to gain entry with ''garbled speech'' you weren't supposed to let them in, you were supposed to ask, ''speak clearly please'' and if they refused, you refused entry. A Doorman is so much easier.
A doorman doesn't have the self-interest of being a resident.
Awesome Video big clive
Visatone speaker... Did not expect that.
Also Piher pots. Not a cheaply made unit at all. Quite nice actually, no Shenzhen crap like you will find in consumer electronics these days.
The resistor and capacitor in parallel with the driver are a Zobel network. It flattens the impedance so it won’t sound tinny.
Thank you.
11:50 The cap is there to tie the inv. input to ground, without shorting out its internal DC offset (bias). It's a differential amp, but not a dual-rail design like an op-amp, where there is no bias present at the inputs or outputs - just my two cents.
The RC Zobel filter at the output - it's to counteract the inductive impedance rise of the speaker towards higher frequencies, which improves the frequency response and prevents high frequency oscillations. With speakers we don't have to worry about inductive spikes though - their inductance is way to little for that.
The 22R+capacitor on the output of the amplifier form a Zobel network or Boucherot cell. It provides a load at high frequencies, when the speaker's impedance is inductive, thus reducing the risk of oscillation.
I was going to message you Clive asking if you're all alright as I've not seen any videos from you in a while. I thought maybe you might have taken a holiday!
Nope, just TH-cam doing it's thing!
Good thing anyway, now means I got a good couple of videos to binge!
I've been turning a can of diced tomatoes into some tomato soup with my immersion blender as I'm watching this, thanks for the good morning Clive
Well made
Few years ago the door units were swapped with digital Siedle ones in my flat building in West Germany.
The 33k resistor in the power cirquit makes sure the output voltage doesn't get to high when there is (near)zero load.
The 15k resistor is part of a voltage divider, the microphone is a variable capacitor it can't pass DC but it does need a voltage to modulate.
T/R = tip/ring or Transmit/Receive
O = Shield or Null
C = Common
H = Hot
The Ford and 70 on resisters to and from the long wire audio bus provide a medium impedance to minimize noise pick up and drive long wires without significant loss. We’re you to drive the speaker directly into the wire it would be an 8 to 16 ohm output and 100 feet of wire would potentially lose signal or gain noise.
This is very similar to standard wired intercoms with an additional wire to trigger the door solenoid. These haven’t changed significantly since the 1960s or before, except to replace transistor amplifiers with ICs. Prior to that would have used a centralized tube amp and possibly a single speaker switched to act as the microphone.
Great video, as always. many thanks, Clive. I'm not sure whether the diodes are there to protect against incorrect connection as much as providing a discharge path for static build up or induced voltages in the twisted pair. These diodes are very common in RF low-noise amplifier circuits connected to antennas. The speaker will be quite inductive at RF frequencies.
The IC probably has ESD diodes built in with the same connection. The external diodes will “beef up” the overload capability for both ESD and incorrect installation.
@@Chris_Grossman Agreed, although the IC's ESD diodes are normally designed to protect against damage during manufacture/assembly rather than a potentially stressful application.
Interesting always find your video's informative incidentally I saw your video titled 5g death beams found that Very useful as it explains why my 4g signal is weaker then 3g
We have a very similar system in my building, always wondered how it worked.
I assume the connection naming is probably going back to telecom stuff, tip and ring etc??
It appears to me this system would allow for the it to function as a 'party line' where you could chat with your neighbours or listen in on their convo with the people at the door? Not all such systems allow for that afaik. I grinned at the start of the video, i once had such a system fail with the feedback circuitry. Loud squeaky noises. :D
I have the exact model Clive tore down in these videos in our apartment. I get terrible feedback, but now I can try and get maintenance to fix it at last, now I know what the issue is.
Awesome, so just need to set the speaker and mic adjustments to reasonable levels to get rid of the feedback (I assume they are both maxed right now). Ima try and get maintenance to fix this as it annoys all of us that live in the apartment block.
T for TIP, R for RING (copper wire telephone terminology). I would have preferred CHORT instead of TORCH.
Clive! Maybe it would be interesting to have a video talking about speaker's impedance at some point in future, why are there 8-4-2 Ohm and things like 6 Ohm (It's a Sony.)
i didn't realize that intercoms could have such high impedance speakers. Thanks for the video!
It reminds me of the phones built by Western Electric for the Bell system. Back then, you got your phone from the phone company. They were hard wired to the system. Bell wanted the phone to last forever, since a service man would have to come to your house to replace the phone. Those phones would probably still work, unlike smart phones that aren't any good in a few years now.
That’s the same for former GPO (General Post Office)/British Telecommunications/British Telecom/BT telephones that were originally provided. Customers rented both the line and the telephone(s). They were built with reliability as a high priority. So used rugged components and strong cases. And yes, some of these designs will still work fifty plus years after being manufactured.
Aw, okay. I was wondering about the power supply in that entry unit. Pretty clever...and doesn't cost an arm and a leg to produce the whole unit. As an aside, it was typical of Bell designed phones to have a clamping device parallel to the earpiece of it's telephones. It was a dual diode package...two diodes in parallel and in reverse polarity. That would simply limit the voltage that could reach
the ear piece to keep the db down.
H is hot C is Cold? 0 is Zero T is transmit and R Recieve maybe?
I think so.
Makes the most sense to me.
@@psirvent8 Yeah slightly strange naming scheme but that happens.
@@rimmersbryggeri Still makes more sense than Tip and Ring to me.
@@psirvent8 Yep. It does.
Just thinking about when you were going to take the speech unit apart, as they say Clive, speak of the devil and he will appear 😂😂 interesting technology through
Clive, any conformal coating on the board? They have thought of everything else.
Clive....Big but NEVER rude 👉💎👈🤜❤️🤛
Good speaker. I like Visatons for mine experiments, they are pricy, but where amazon specials or Curies essential (same think but more expensive) need 60Wats just to be heard, 6Watt Visaton has better sound then mine monitor and handles full sound range of Heidens (not sure how to spell his name) Trumpet concerto. All of that directly from jack of the RPI. I love Visaton speakers. Hovever I find that in new installation ESP seem to be more common then bell (its not as good, but its cheaper)
Hi Clive I have had 2 of these door entries broken because of water getting in so they are not 100% water proof I have a new 1 installed
Presumably
T - Tip (or transmit, but that can be confusing as it would be receive at the other end!)
O - 0V
R - Ring (or receive, but that would be confusing like 'T' above)
C - Cold (+, bizarrely?)
H- Hot (-, bizarrely?)
Agree, should definitely have swapped C and H for it to spell TORCH!
T - Transmit, R - Receive, O - Open, C - Common, H - Hot. Z would likely be the return, and I is "Indicate" for the ringers/flasher.. (flasher for the deaf)
They do inded make a 7809, I converted an old 90s car stereo from a 9v portable CD player power output to a 5v output for regular USB power simply by removing the 7809 and swapping in a 7805, not as common a voltage these days, but it's definitely a part that's out there... :)
Cool..
The 10k and capacitor will give a stable supply, the 2k7 gives a series resistance to power the mic (electret mic's have a built in pre amp)
7809's are fairly common, and have a current rating of 1.5A, if you want more, you have to add a transistor to carry the current, though i don't think that's the case here, maybe they just like old school.
Or perhpas wanted a lower power consumption, as IIRC the original 78 series of regulators ran at anything from 10 to 50mA of current in the common pin, and were not too stable with no load on them. The simple zener regulator and way overkill transistor has lower current, but also has the big advantage of drooping with increased current, so the output is simply short circuit proof, and spikes on the output or input will not cause any problems, as the input it will simply charge the capacitor up to some high voltage, and bleed off, and the output can go up to around 16V, when the transistor emitter base junction will zener, and the transistor and zener will clamp to a safe level so the IC does not exceed max voltage too much.
Also the TIP110 is a little cheaper, and the resistor and zener are a cent or so each, so no big cost, and you need the capacitors anyway. Plus the transistor will not go into thermal shutdown on a hot day, when the unit might be approaching 130C inside because it normally is in the direct sun, in a nice plastic enclosure with no cooling.
Good quality 7809 regulators have a maximum input voltage of 35V, an operating temperature range of 0°C to 150°C or -55°C to 150°C, a bias current of 4mA to 9mA.
The downside is a drop out voltage of 2V to 2.5V.
It may well be that using a TIP110, a zener diode and some resistors are just slightly cheaper, or they wanted a circuit with a better performance with lower input voltages.
@@Mark1024MAK Good quality ones are not cheapest though, even if ST does make the tabs out of what looks like copper foil these days. But even the cheap ones are probably going to cost more, as the parts in the transistor unit are so easy to use equivalents for, whatever is the right package and cheap will work here, so left over reels off another batch will be used.
My voice is my passport, verify me.
Are T and R short for Transmit and Receive??
The 78xx/79xx linear regulators are available in positive and negative 5V, 6V, 8V, 9V, 10V, 12V, 15V, 18V and 24V and 1A in TO-220 packages.
Some voltages are also available in TO-92 package for 0.1A, TO-220 package for 1.5A and 2A and TO-3 package for 3A.
You left out the 78Mxx 500mA range…
@@Okurka. - no, but the 79xx are. Such as 7905, 7912 etc.
@@Mark1024MAK I have never used those, I knew they existed but I didn't know what voltages and I didn't have time to check data sheets.
And with using I mean having to replace a defective one in the devices that I fix as a hobby and as a way to recycle and save money or making a simple power supply, I am not an electrical technician or engineer so I can't design too complicated stuff from scratch.
I didn't use all of the other all that I mentioned either, but they sold them in an assorted box with one of each type plus an adjustable one, perhaps LM337, and without including the 3A ones in some electronics store (I think in Spain) at a price very affordable, making it an interesting purchase for hobbyists like me, especially considering that I live in a rural area and going to an electronics store in the city if I need one would cost me more in fuel and parking than I paid for all of them
@@Okurka. I said 78xx/79xx assuming that whoever needs to use one will know that the 78 are positive and the 79 are negative, but if they don't know it they will find it when buying them or looking for their characteristics.
@@ruben_balea - no problem. I’m amazed that you may even have electronics stores. There are no electronics stores in my town, or in the nearest large city. The largest chain used to be Maplin Electronics. But the whole chain closed. Now I either have to buy online and wait for the delivery, or visit a trade counter in the nearest large city run by RS Components.
As you know, the 78Lxx are 100mA devices, the 78Mxx are 500mA devices and the 78xx are normally 1A devices. I got to know about these back in the mid to late 1980s because I was buying electronics projects kits that used them.
Could the R and C in parallel with the zener in the PSU be some sort of ringing dampener?
The capacitor in parallel with the zener is both to reduce the electrical noise from the zener, and to filter out as much ripple as possible should the input drop below the zener voltage. Thus providing a nice clean stable voltage for the transistor to operate at. The transistor operating in emitter follower configuration, so the voltage at it’s emitter terminal always being 0.6V (or thereabouts) lower than it’s base voltage.
Could a full bridge rectifier be used as reverse polarity protection in a DC circuit?
Yes, but it would introduce a significant voltage drop (1.2 V - two diodes in series). Also more expensive than a single diode.
"So, the designer figured out how to work out all the bugs in the system?"
"Yes, that's why there's all those openings at the bottom."
"But, isn't that how the bugs get INTO the system?"
"Yes, but they do also work back out, too."
Wait until you either get a slug inside or ants decide it’s a good place to deposit soil in…
CHROT, sounds like chroot (a Unix/Linux program for temporarily changing the root directory) but it's just me and my silly associations :)
Nice circuit, not that much of that SMD rubbish. And a very old school voltage regulator too! I'd go with a 317, but still :)
I think the 33k resistor is for discharging the cap parallel with the Zener diode after the unit is unpowered, ensuring that the output voltage will rise softly but pretty quickly (time constant here is 0.2s) on powerup, eliminating unwanted transients or surges.
The 15k parallel with the mike can be used for setting the bias voltage as it forms a voltage divider together with 10k and 2k7. In this case, the bias voltage would be around 5V provided that the condenser mic doesn't have any internal parallel resistance.
BTW. are you missing your pinkulator?
The resistor in parallel with the zener could discharge the parallel capacitor, but a faster discharge path is via the transistor base-emitter junction and then into the load.
Clive, if you're doing half-wave rectification with a single diode, won't the RMS output be a lot lower than you calculated? I thought you did the full wave rectifier equation.... ?
I guess the zener is the other way around so that might account for the voltage being somewhat higher than a simple half-wave circuit.
The capacitor will still charge to the peak of the sinewave.
Surely the "T" and "R" pins stand for Tip and Ring?
Regarding the 15k parallel to the microphone. I would guess the microphone should be biased with about 5V. The 10k, 2k7 and 15k in series would then be a voltage divider from 9V to 0V, giving about 9 * 15/27.7 V over the microphone. Rest is capacitors blocking DC.
If the microphone is capacitive, without that resistor, the microphone would then take the entire 9V, which might be above its rating.
The microphone is not capacitive. It has a JFET, wired internally across its terminals. I think the resistor is to reduce the voltage swing of the microphone. Taming it a bit distortion wise and maybe even preventing oscillations.
The front plate would be perfect for making racing rain lights, I wonder how big those front speaker holes are? Would they fit a 5mm LED?? These rain lights for race cars are just a bunch of red LEDs together wired for 12v DC. These rain lights can be anywhere from 150-300 USD, and are not repairable because the LEDs are potted in resin for most brands. Being on the rear of the car, we break them often. It'd be profitable to make them rather than buy.
I would so add a line input on this thing. Curious if the circuit can be modded to output stereo audio since the amplifier chip is a stereo amp. I did that once to a gameboy advance light and amp device but the audio circuit was messed up. A couple of modifications and I now have a nice stereo amplifier, just need a case for it.
They use both amplifier channels to get sound in both directions. I've used an older intercom with no chips that used the speaker as a microphone when the inside user didn't press their speak button to reverse the entire circuit.
@@johndododoe1411 I have a toy bullhorn thing that uses a similar amp IC and it uses each channel one for the mic and the other for the sound effects. All of it comes out in mono so the two channels are Y-ed together somewhere in the circuit. And yes I have an audio input connected to that toy.
I'm still learning electronics, but I can explain one attribute of the schematic and design. First, electret microphone capsules have a FET built in; the resistor is used to bias the transistor, which sets whether the microphone is loud or has a high dynamic range.
I guess that'd make the resistors on either side of it a voltage divider? And then the resistor would set the bias relative to that... Most electrets are 2 or 3v drive and a 1v amplitude, or a little more. The good electret capsules used in pro microphones have a huge dynamic range, with an integrated power supply (powered off the 48v dc from the interface) and have a much wider amplitude.
Is it possible that the resistor and capacitor are instead a capacitive dropper-style power supply for the microphone? It's really quite low voltage relative to the 9v supply. They're fairly sensitive.
For the speaker, I can't speak authoritatively except to say the design is optimized to reproduce the higher frequency parts of the vocal range. It'll sound awful, but quite clear.
Capacitor droppers don't work with DC.
The 10k and 100uF form a low pass filter, which will further smooth out any residual ripple. Necessary because the mic circuit is very high gain. And yes - the resistor network provides DC supply voltage (not only bias, strictly spoken) to the capsule's internal JFET
@@westelaudio943 Is the 9v supply DC? I got a bit mixed up trying to keep track of the schematic
For the bias I was referring to the resistor directly across the electret leads
@@cheyannei5983
The door station works on 12Vac but the 9V internal supply is DC, coming off the linear regulator.
Since these capsules only have two pins, I would assume they are biased internally. The bias on a JFET is usually done by a resistor from source to ground, which drops some voltage, so it's on a higher potential than the gate, which is tied to ground with a resistor which can (should) have a high value, as current doesn't flow through the gate, it's a reversed diode. Tubes are biased in the same fashion.
I think the parallel resistor is there to keep the voltage low and more stable across the FET. Limiting the voltage swing a bit, otherwise there might be too much distortion, or it might even oscillate, I'm not sure.
@@westelaudio943 Assuming this isn't a custom part rolled specifically for this intercom, an external resistor across the pins is completely normal for the microphone.
My mom used to own this here in Leaminton Spa
Your phone is a suitable standin but a kink palculator it is not. Could you push up the screen brightness some next time? It’s visible but a little dim.
I find it interesting to see a dc blocking capacitor at the output of one of the amplifiers, but not the other one. Odd...
Transmit, Receive, 0V, Hot, Cold
Arse, H&C are wrong
Clive exploring a Chinese device: "That's odd. I don't know why they did that."
Clive here: All these values have been tuned perfectly to the application, and it's got input and output protection..."
Hey, has anyone mentioned that the output cap and resistor are a Zobel network yet?
How about the many possible meanings of T and R?
;-)
My guess is T= tip, R = ring (from telephone nomenclature), C = common, H = hot (used in thermostats in the US), and O = 0 volt.
In the U.K., in electronics, H normally means high, as in HT high tension meaning high voltage.
Tip & Ring
I'm surprised they bothered with a discrete regulator. I would've gone with the LM317, or the lower current LM317L. The TDA2822 is no longer made. I wonder what the newer versions use? There does seem be quite a lot of new old stock available at the time of writing, but one day they'll have to update the design. I wonder if they'll go for an op-amp such as the NE5532+a couple of transistors for the speaker and the other op-amp used to drive the earphone directly, as it's probably powerful enough to do so.
Guarantee you there are at least 2 dozen drop in replacement amplifiers they can use in place, as that pinout has become sort of a standard, and there are hundreds of other amplifiers that use it. As well they probably bought in a few hundred reels at EOL for making these, as the volumes are not going to be that high, so having a shelf of sealed reels of them is no big cost, and they likely make at most a few hundred units a month. Every part there though is jellybean, you can substitute them with no issues, just buying whatever in a wide range that is cheapest when you are running low. Most expensive single part on the whole unit is the speaker, and then the case, followed by the hand assembly of the final unit.
Looks like they’ve just implemented a party-line. The nomenclature is certainly POTS.
Her parcel didn’t happen to contain an intercom did it?
"Very impressive; no expense spared." Without looking at the datasheet for the TDA2822, I'm going to guess that the 100uF capacitor on the inverting input works with internal resistance to act as a high pass filter, limiting the need to draw currents for low frequencies that are outside the normal range of speech. Should also filter out mains hum.
That diode protection is found *everywhere* in audio amplifiers. I've seen it at the inputs for both the preamp and power amp sections of solid state guitar amps, typically specified as 1N4448. Interesting that they don't significantly clip the signal as one might expect.
The cap is there to tie the inv. input to ground, without shorting out its internal DC offset (bias). It's a differential amp, but not a dual-rail design like an op-amp, where there is no bias present at the inputs - just my two cents.
@@westelaudio943 OK, so it's basically an AC ground reference, like is often found with active volume or tone controls in the feedback loop of an op amp. That makes sense.
The TDA2822 must have an internal voltage divider like any differential amp. I know that with the LM386 the input that doesn't get the signal is almost always tied directly to ground; now I'm curious about the 2822's equivalent circuit diagram.
@@petersage5157
Yes, but only if the op-amp is fed by a virtual center point and not by a bipolar supply, say +/- 15V. Otherwise an op-amp input can be connected to GND without a problem. The input will be at a 0V potential at idle, and the GND only keeps it there.
I would have expected a capacitor at the output of the amplifier.
Isn't there any DC on the output?
And if there isn't any DC on the output, how can there be 'undistorted' audio?
Maybe the DC is removed in the receiver module?
(It could be that the drawn speaker symbol is just a bit misleading.)
I watched Clive's other video - no there is no series capacitor at the earpiece. The earpiece (which is at least a .25 watt speaker, as it also provides the doorbell sound) is probably not bothered by a bit of DC offset, as it will be low, just like the signal coming from the mic amp itself (remember there's a 470R resistor in series).
12:20 That's a Zobel network, aka a Boucherot cell. It's there so the amplifier "sees" more of a consistent impedance on its output, and reduces the chance of the amplifier oscillating at high frequencies where the impedance is high due to the inductive nature of the speaker. en.wikipedia.org/wiki/Boucherot_cell
Thanks Clive. If you'd like an URMET speech unit with popped capacitor, let me know where to send it. Ta JC
Nicely designed unit. The resistor and capacitor across the speaker provide resistive load at higher frequencies for stability of the amplifier [See en.wikipedia.org/wiki/Zobel_network, "Zobel networks and loudspeaker drivers"]
Stones and a speaking tube most effective
Ah that explains why I couldn't get a two way taxi cab intercom system to feed back on itself because the amp unit had some sort of pic microcontroller which would have acted as some sort of feedback destroyer?
Audio feedback is when a sound gets amplified, is output to a speaker, the sound is then picked up by a microphone, amplified and output to the same speaker. Each time round if the level of amplification exceeds the losses, the volume level increases, at least until it gets to the amplifiers limit. This is also known as howl-round. The normal simple solutions are to to either move the microphone, or move the speaker or reduce the amplification level.
I would love to see how this was done in the 50s or 60s!
In the late 60s, it would have been pretty much the same but with discrete amplifiers. Before that they would have used carbon mics, which are high output and don't require amplification, except at the loudspeaker output (transistors?). Tube intercoms did exist, but the warm-up time of tubes was a problem.
@@westelaudio943 - if a tube amplifier was used, it would have been permanently powered. That’s the way an intercom that was installed in 1970 in my workplace operated. Each unit had a talk/listen switch, and I think, used the loud speaker as the microphone.
@@Mark1024MAK
Yes the (in)famous instant-on feature.