Intruder alarm panel project (retrofit)

[877]

Hi,

I want to add some features to my existing intruder alarm panel. Its an Accenta 8, and I will be using a Wemos D1 mini and some relays (please see below for exact details):

Alarm panel schematic:

Voltage readings in unset, set and alarmed states:

My project requirements and information:

12V relay
5V relay

The main problem I have is the "D6 sounder on/off" part.

• The terminal marked B is 5V when sounder is on, and 0.5V when the sounder is off.
• The sounder has a permanent + supply, when terminal B is pulled down to 0V it's activated.

So I need help figuring out:

If I trigger the 12V relay from terminal B, will that work? (as I'm writing this I think not)

i.e.
Relay is powered from 14V and 0V terminals

sounder is off - trigger voltage is 5V
sounder is on - trigger voltage is 0.5V

I have read the relay specs, I'm just a bit confuddled

or

Should I just use a 5V relay for the sounder part, and trigger the 5V relay from terminal B

BUT the relay needs a permanent 5V supply from somewhere.

I will be powering the Wemos for a good quality 5V PSU, would that be acceptable to give the relay 5V supply (in effect the supply would be from a seperate PSU than the trigger 5v).

Any advice is much appreciated on the overall project...

[PaulRB]

I don't think you need any of those relays.

What voltage do you measure at the keyswitch terminals? (Measure both against the 0V terminals).

Voltages higher than 3.3V can be reduced down with a simple voltage divider. This calculator will work out suitable values, example below.

Power the Wemos, via its 3V3 pin, from the 13V terminal via a small dc-dc convertor.

[877]

PaulRB:
I don't think you need any of those relays.

What voltage do you measure at the keyswitch terminals? (Measure both against the 0V terminals).

Voltages higher 3.3V can be reduced down with a simple voltage divider.

Power the Wemos, via its 3V3 pin, from the 13V terminal via a small dc-cd convertor

Hi Paul thanks for the reply,

At the keyswitch terminals it's 4.5V when measured against 0V (it's got a link fitted as present so same on both terminals). All of the zones are 4.5V as well (even unused ones).

I did think about voltage dividers earlier on, you mean just use actual resistors? I'm ok with the maths, I guess just use high value resistors as it's only a signal voltage?

One thing, the Wemos can draw around 800mA when using wifi, the alarm panel fuses are quite low values at only 1A for the whole panel. So I was planning to use a seperate 230V > 5VDC 3A power supply.

If I did this, would I need to common up the 0V outputs between the alarm panel and my 5V psu?

That is why I went down the relay route, to create some segragation.

But, I would like to hear your thoughts.

Thanks

[PaulRB]

Remove the link and re-measure.

The Wemos will draw about 80mA, not 800mA. There are spikes of 100-200mA during WiFi transmission, but these are very short indeed and will be absorbed by the capacitors in the power supply, the DC convertor and the Wemos itself, so no need to worry. The Wemos' current consumption can be reduced further still, if needed, by using a power saving mode. Assuming the Wemos current is 80mA, and the convertor is 80% efficient, the current draw will be around 25mA at 13V. So you don't need to use another PSU, but if you did, yes, you would need to connect the grounds.

A small transistor may be needed to interface with the keyswitch terminals.

The good thing about the calculator is that it finds suitable pairs of resistor values from the standard values. You can set the "Resistor Scale" to 10,000 to minimise the current, but I wouldn't go higher than that because the Wemos' input pins could be triggered by noise interference.

[877]

PaulRB:
Remove the link and re-measure.

The Wemos will draw about 80mA, not 800mA. There are spikes of 100-200mA during WiFi transmission, but these are very short indeed and will be absorbed by the capacitors in the power supply, the DC convertor and the Wemos itself, so no need to worry. The Wemos' current consumption can be reduced further still, if needed, by using a power saving mode. Assuming the Wemos current is 80mA, and the convertor is 80% efficient, the current draw will be around 25mA at 13V. So you don't need to use another PSU, but if you did, yes, you would need to connect the grounds.

A small transistor may be needed to interface with the keyswitch terminals.

The good thing about the calculator is that it finds suitable pairs of resistor values from the standard values. You can set the "Resistor Scale" to 10,000 to minimise the current, but I wouldn't go higher than that because the Wemos' input pins could be triggered by noise interference.

Hi again Paul,

OK I have done more voltage checks (all to 0V):

With the keyswitch link in, it's 4.5V on both terminals
With the keyswitch link out, it's 5.0V on top terminal, and 0V on the bottom terminal.

It behave like this:

Remove the link - the alarm beeps and sets as if code was entered.
Insert the link - the alarm instantly unsets as if code was entered.

So I am thinking I will still need one relay for this function, I would say the Wemos shield relay, which I already have several. Or do you still think a transistor? (of which I have none).

I definitely read the Wemos pulls 800mA, but checking my own it is indeed only 80mA, and that is with wifi enabled and a 1.44" tft screen.

What do you think about this step down converter? Amazon link
Being Amazon I can get it nice and quick...

Regarding the resistor values, thanks for the link. I'm wondering what value of current is a good value to aim for? e.g. 1mA per input? I want to minimise current draw on the alarm panel and minimise wasted energy across the resistors obviously.

I have the EE Tooliit Pro app which seems ok:

Many thanks for the help!

[PaulRB]

If you short the top keyswitch terminal to 0V, does it behave like the link is there? I was hoping that the bottom keyswitch terminal is just another 0V. But that 0.5V difference makes me wonder. (Apologies if doing that sets the alarm off! The 0.5V might be an anti-tamper thing)

Using a relay for that would in some ways be the easiest thing, because the relay contacts would act just like the link or an actual keyswitch. But... you seem to be concerned about current consumption. Relays use lots of current, when the contacts are closed. Many many times more than a voltage divider. You plan to activate/inactivate the alarm this way, so the relay would be activated for long periods.

As for the voltage divider's current, the 22K+6K8 I suggested before would draw less than 0.5mA. 220K+68K would be a tenth of that. Neither sounds like a problem.

I don't like those DC converters much. 10 times the size of the ones I suggested, far more output than is needed here (which may mean they are at the lower end of their efficiency range) and you have to adjust them yourself to 3.3V. Those do at least seem to have multi-turn pots, which will be easier to adjust and less easy to knock and change the output. But I have similar ones and the terminals are, annoyingly, not on a 0.1" grid, making them difficult to use while prototyping on breaboard and soldering onto stripboard.

PS. Please stop quoting my entire post in your response. It just makes the thread uneccessary longer. Either just quote the parts you want to ask about, or just hit Reply instead of Quote.

[877]

Thanks Paul!

No quote this time not sure why I was doing that tbh.

I think I would prefer a relay, I’ll have it so it’s only active when the alarm is unset (approx 10% of the time).

I’ll order those converters you recommended too just to be safe.

I’ll update my project plan tomorrow and update it here.

Again thank you!

[877]

Hi Paul,

Here is my updated project plan with voltage dividers, and relay for the key switch. I decided on a separate psu for the wemos while I wait for the buck converters.

Couple of things I wanted to check:

Are zener diodes needed for the inputs to protect against over voltage?

Is the ground bond wire corrrect on my drawing?

Thank you

[PaulRB]

You missed the 5V & GND connections to the relay module. But as I said, I would look to replace that with a transistor, or if that cannot be made to work, maybe an opto-isolator. Other than that, it looks ok.

I don't see any need for zener diodes. They are often recommended for use in auto applications where the alternator and ignition systems can cause voltage spikes, but that's not an issue here.

The ground wire looks ok to me. You are correct to connect the ground wire from your temporary psu to ground of the panel.

I suggest you prototype this on breaboard before soldering anything. That way you can, for example, test the output voltage from your voltage dividers before connecting them to the Wemos pins.

[877]

Hi again,

I didn't show the relay power as it's stacked on top of the Wemos, I only showed the D7 as a dashed line as it's easier to understand for me.

Like this:

Regarding the keyswitch, I am reticent to short the terminal down to 0V in case it damages the circuit board somehow. I will hopefully find some more info on why one is 0.5V when the link is out.

Your alternatives are interesting and I would like to learn more:

Transistor - I understand the basics i.e. base, emitter and collector, I just don't have the knowledge to choose the correct one, or how the circuit should look.
Opto coupler - same as above, I understand it in principle but building a working one is another matter.

I don't want to take advantage of your help, so if you are busy I will research them myself in due course. I will definitely breadboard the voltage dividers before connecting any pins!

Really appreciate the help, thanks

[PaulRB]

This is the bit I'm not sure about:

With the keyswitch link in, it's 4.5V on both terminals
With the keyswitch link out, it's 5.0V on top terminal, and 0V on the bottom terminal.

The only thing I can imagine is that, behind the scenes, its something like this:

Remove the link - the alarm beeps and sets as if code was entered.
Insert the link - the alarm instantly unsets as if code was entered.

So if the wire between the keyswitch and the panel is broken, the alarm is triggered.

But why two resistors? I don't know. Seems to me it would work just as well with one. But I'm not a security engineer ..

So the idea of either a transistor or a opto-isolator is to act like the keyswitch or relay (but using much less current to operate than a relay).

The potential problem with a transistor is what voltage at its base will be needed to trigger it to switch on. If the circuit is wired like my guess above, then no problem, the 3.3V from a Wemos pin will be enough to trigger the transistor to switch on and act like the link is in place.

If that can't be made to work for some reason, then an opto-isolator should do the job. It contains a transistor too, but instead of being triggered by a voltage on its base pin, it is triggered by infra-red light from an led inside the chip. So the question of what voltage is required goes away. Only enough voltage to light the ir led is needed.

For a transistor, you could try an ordinary npn bjt type like BC337.

For an opto-isolator, a basic model like 4N35.

[877]

I honestly don't know how it works, I got curious and took the pcb out.

I am attaching photos of:

-Front view of pcb
-Rear view of pcb (mirrored horizontally to match the front)

I tried drawing out the circuit but quickly got lost. It look like the 40pin IC is a multiplexer after some googling.

CP82C50A-5

Stückzahl / quantity: 1

-- NOT RoHs  --


case: DIP-40

manufacturer: Harris

manufacture Ident Code: CP82C50A-5

product number: 04671

Datasheet / Datenblatt: CP82C50A-5.pdf

High Performance Programmable Universal Asynchronous Receiver/Transmitter

Ebay link

Datasheet

If you are interested, please let me know what you think. If you have had enough then I totally understand and appreciate the help up to now.

Thanks

[877]

Also here is a photograph of the rear, partly marked with components.

Green = link
Red = resistor
Yellow = capacitor

And my attempt at following the circuit (probably wrong).

[PaulRB]

[hammy]

Have you read the manual ? I have a Honeywell ( not same as yours ) but the outputs , as it states , on mine were open collector

[877]

Hi hammy,

Yes I have read the manual, it only says the zones are ‘positive loop’.

Each zone needs to be shorted out and motion detectors are normally closed.

The key switch terminals are what we were looking at, it looks quite complicated.

How it works in practice is beyond me.

[PaulRB]

My schematic was the wrong way up. I meant to put 1r at the top connecting to 5V and 9r at the bottom connecting to ground.

Your pictures show that it is 1r = 1 X 10K = 10K (brown, black, orange) connecting to 5V and 10r = 10 X 10K = 100K (brown, black, yellow) to ground. So I was pretty close!

I think a bc337 with its collector and emitter connected to the top and bottom keyswitch connections, and it's base connected to the Wemos pin with perhaps 100K should work. With all those high value resistors, the currents involved are very very small. For example the current flowing through that link is around 5uA. A similar current will flow through the transistor's base and emitter. With such thing currents, I can't see any damage being done to the panel by trying it.

[877]

Yes you were very close!

I think I will give the transistor a try, but I only have S8050 NPN in my toolbox. Would that suffice?
Datasheet

Updated keypad schematic, albeit very simple:

[PaulRB]

Sounds like that might work. Give it a go. If 100K does not work, try 10K or 1K.

[877]

Will do, might take a little while to build but I will 100% report back.

Thanks Paul