I am learning to design the sch files.
But really a new bie, hope experienced guys can help to guide how to finish it.
Basically this a IR receive board : receive the IR code and decode it and transfer to master MCU.
I divide it into 2 parts: the hard PCB and soft FPC, compose them with free : 1 PCB + N FPCs.
I can manage simple digital parts, but for analog one, I think they are too miraculous to understand.
When you usage a Motor. Then make sure it electrically isolated from your Microcontroller.
Here is the recommended schematic.
eRony - How many commercial devices have you seen that use microcontrollers and motors? Do they use opto isolators? Probably not. Optos aren't quite the magic cure you think they are.
Xiao - you need resistors to limit the base current on your transistors. 1K might be a good start. Draw the transistor the right way up - it makes the drawing easier to understand.
This vibrating motor is same like one in mobile,only 80ma at 3v, so optional isolation maybe need not.
Kartman, thank you for advice.
I will add resist for transistors.
What’s your meaning “draw it right way”？I just selected it from default library .
And importantly, in that orientation.
Kartman:
The Data pin in P1 & P2 is the wire to transfer data between master MCU & all other slave MCUs, but the speed is not high(9600bit/s).
The length of the wire (include the power source lines) is about 150cm, do I need to do something for it?
The sch changed as following png:
Consider the voltage drop across the 0V wire due to current.
Also consider EMI and ESD. You need protection and filtering.
Your transistors look much nicer. Do the same to P1 and 2
Your transistors should also have a high value resistor - say 100k - from the base to emitter. That will stop them accidentally turning on while the various processors are in reset and the port outputs might be inputs or floating.
Neil
The length of the wire (include the power source lines) is about 150cm, do I need to do something for it?
That is a good question.
A lot depends upon the environment the devices will be operating in.
Are there large motors, or high voltage or high current wires in the area?
What is the wire being used?
The wire has capacitance, and the longer the wire the more capacitance the uC I/O pin has to drive with its signal transitions.
You mentioned multiple slaves, so that increases the load on the driver.
You might get away with driving the wires with just the micro.
If you do this make sure you use a "reasonable" gauge wire, and likely you will want it to be shielded.
A better approach might be to use RS-232 for the data transmission.
This would require a chip at each end, perhaps the T.I. TRSF3222E, or similar.
The micro's UART connects to the TxD and RxD pins on the chip.
The chip converts the 0 / 3.3 V data pulses to an approx. +/- 10 volt signal.
The chip can drive much longer lines than the micro's I/O pin can handle, and it has better tolerance for electrical noise.
The chip at the "far end" converts the RS-232 data back to logic levels to feed into that end's micro's UART.
JC
thank you，JC
I want to communicate with usart at the beginning, but given up at last after trying for many days and hundreds codes. Hard to debug and independent of hardware is hard too
Kartman:
Use tvs diode to protect the input port ,that is suitable?
For each port ,1 for power source and 1 for data line.
And the connection are:
Power source to GND
Data to GND
Is that right?
Best if you draw schematic so we can be sure we understand.
On data, use a RC filter and a tvs diode. TVS diode has capacitance - choose carefully.
Kartman,I study the RC filter for a while, get something , confirm with you now, refer to the png below.
As pin of MCU, the current is very few if it was used as input / output level, not driving parts.
I think it is only few MA , 10 ma e.g.
The voltage drop on R5 is (10 * R5/1000), it will be acceptable if the resist value less than 1K.
And the frequency is less than 5000HZ, f = 1/(2*pi*C5*R5), we get the value pair : c5=0.1uF r5=300 ohm.
Is that right?
If your data rate is 9600 baud, you don’t want the filter corner freq at 5000Hz. This will affect your data. At a guess, I would start at 100R and 100pF. Put the tvs in parallel with C5. What part number for tvs diode?
For my case the freq is not stable, as showed in the picture,the highest value is 5000 bit per second.
Why the TVs need to be putted in parallel with c5? Does not it should protect the c5 too?
Fix P1 & P2..it is ridiculous to have them pointing up in the air, for no good reason. The should be right next to each other (not 5 inches apart)since they share the same wires. Their gnd MUST connect to GND symbol
Why is P7 twisted around...just mirror P6.....raise P6 & P7 upwards & get rid of their right angle lines--make them straight across...much neater...rotate the tsop36 parts 90 deg...will be even neater.
This block in the lower left seems to have no connections at all in common with the rest--do they even share gnd?
Move P4 & P5 next to each other...minimize the total length of wire drawn. (maximize efficiency ).
A filter cap needs shown at AVR pin1...maybe it is your C2, but it should be show AT PIN 1
How does you motor connect to the board (nothing is shown)...maybe just solder wires?
Bring out ALL avr pins to vias in case you need to use those pins later. Dioes this chip need a reset pullup resitor (usually recommended)
thanks so much,I will modify it soon.
Kartman:
How about this version:
Still no pull-downs on the transistor bases?
Neil
Not a good idea to have multiple LEDs in parallel, and only 1 series resistor.
the blue ones don't seem to have any series resistor?
EDIT
Does connecting the outputs of multiple TSOP36240 in parallel work?
https://www.vishay.com/docs/82568/tsop362.pdf
Not a good idea to have multiple LEDs in parallel, and only 1 series resistor.
---> I want to decrease the parts on FPC. These LED used only for show status, keep them working is enough.
the blue ones don't seem to have any series resistor?
---> I checked the datasheet, the voltage is 2.8~3.4V. The blue LED will work in "breathing" way : dark -> high light -> dark with PWM control, that means the time with high voltage is very shot. The max voltage is 3.3v even in worst condition.
Does connecting the outputs of multiple TSOP36240 in parallel work?
---> They must work :), I will lose the base otherwise.
How?
I don't understand.
You're using a 5V TVS to protect a 3V3 pin? That's not going to work too well. Besides, put it in parallel with C5. What protects the TVS? If you wire something wrong , then TVS diode may burn.
With your FPC, I think you'll be having problems with the led current switching and the TSOP devices. This requires careful design.
---> They must work :), I will lose the base otherwise.
All your base belong to us!
This looks just as awful as the earlier version, maybe even worse...aren't you willing to spend some time improving it? What is stopping the progress?'
Where are the resistors for your LEDs? T1,T2, T3 are all in parallel...for what purpose? No bypass cap at the micro? You may just wan tot use this as a learning experience & start over.
You should explain a little more on led current and tsop, I don’t know where to go 。
I think you need to go back to basics!
If you don't have a series resistor, an LED will take "infinite" current!
https://electronicsclub.info/leds.htm
If you connect the TSOP36240 outputs in parallel, they will interfere with each other!
Neither do we! I have no idea of what you want to achieve - all I see is a schematic.
all I see is a schematic
And a rather strange one, at that!
I think this may be the HW equivalent of jumping into the code editor too soon?
There needs to be a good design before you get down to the implementation-detail of schematics ...
IR receivers, leds and a vibrating motor networked. The mind boggles...... double your pleasure, double your fun....
the function is very simple,receive the IR packet ,decode and judge, flash the red led and active the vibrating motor if it’s right. And the blue led will keeping light in breathing mode to indicate the set is running.
This is used in laser war game,as the sensor part: to receive IR packet (simulate the bullet). Now,most realized method is placing 4 tsop on 1 pcb . I will move forward 1 step: isolating the reception function as a sensor node,it can be configured, it will output the decode data to master through 1 wire.
As the picture showed, you need add more pcb if you want to detect more "shot" you received, that will be expensive.
My way to decrease the cost is using TSOP strip , not block.
And make it easy to install by FPC.
I believe it is ok to work with 4 TSOP on 1 PCB,
I'm not sure it is OK to increase it to 20 on 1 set.
That is why I'm here :)
Each TSOP IR detector is supposed to have its own RC on the power pin...did you look at the datasheet?
You have not yet fixed your schematic either...you are falling way behind
The tsop devices are really sensitive to power supply noise. Having leds share a power rail is probably not a good idea. As candyman suggests, you need the RC filter at each tsop. As for having many tsop in parallel, i’m thinking it might not work as you want.
It bugs me that the ATtiny814 symbol has 6 missing pins...
I believe it is ok to work with 4 TSOP on 1 PCB
it's not about how many you have on 1 PCB - it's about how you connect them together!
My way to decrease the cost is using TSOP strip , not block
But that will only detect "shots" along a narrow line.
Maybe that's what you want?
in my case,voltage is nearly 3.3v,not below 2.2.
I changed my idea, don’t use fpc strip, I will post it later.
Changed, isolate each TSOP with sub pcb.
Help to check again , pls.
Thank you.
What pulls the DATA line low?? --NOTHING!
Ensure the 3.3V TVS (part # ???) is not tripped by your 3.3V power supply---you need some margin.
don't call the same signal three different names...3.3V Vcc & Vout ???!!!!!!!
C6 1.5uF is prob way too large (filter on a pulsed sensor output to AVR)
you run out of wire here
And have WAY TOO MUCH here
FIX ALL OF THESE DISASTERS!
tsop will pull it low,it can pull low only in fact.
Under Normal situation ,the voltage is a little small than 3.3v, just like 5.0v or so,do we need reserve for them too?
Value for c5 computed by f=1/(2*pi*r*c),c goes down and r goes up,what is the problem if c is 1.5uF?
tsop will pull it low,it can pull low only in fact.
You better look again!
what is the problem if c is 1.5uF?
1.5uF is a pretty big cap , if you want to get a digital pulse from it from the internal 30K driver...pay attention to your needs
Please start the cleanup....you'll be glad later.
But what is the right value range? 1uF? 0.1uF?
Actually I don't know how to determine the freq for RC filter.
For the TSOP32640, work on 40K IR signal, someone told me f = 40K. But I'm really confused, TSOP it self is the chao source, it will output tripple to the system, not input from other part.
It is easy to understand use the signal freq as the freq ,like pb2,pa4.
Reading the datasheet...
The TSOP receives modulated IR directly, at 40kHz - that is, a 40kHz signal (probably an on/off square wave) - and decodes the result. If the signal is there, the output is active; if it's not, it isn't.
The datasheet has two things you should be aware of, though; that decoupling capacitor/resistor combination should be 100 ohms/4.7uF; secondly the large orange words 'not for new designs'...
Neil
If the signal is there
I think we'd typically mean data bits (such a a TV adjust code). In that case, you don't want to blend them into one blurry slag lump ,using a big cap. Using 30K & 1.5uF is almost 50milliseocnds!
The TSOP receives IR signal in PCM(pulse code modulation.)
The IR signal modulated to 40K waves, and the output will be far larger wave as picture upper.
Refering to the picture below, the internal resist is used to pull up the pin, the RC filter provided by R1 & C1, the internal 31K resist will infect the RC filter?
If the signal is there
I think we'd typically mean data bits (such a a TV adjust code). In that case, you don't want to blend them into one blurry slag lump ,using a big cap. Using 30K & 1.5uF is almost 50milliseocnds!
Indeed; three levels of signal going on there - the IR itself, its modulation at 40kHz, and the further modulation of that 40kHz signal with the actual data.
The cap under discussion has nothing to do with the data rates; it's a power line decoupler/smoother and the data sheet says it should be 4u7... mind you, the data sheet also says do not use for new designs.
Neil
The cap under discussion has nothing to do with the data rates;
As mentioned, I doubt you'd typically want a 45ms time const on the data line! (see TSOP chip--30k*1.5uF) , though it might depend if you intend to send anything....just use 0.001 uF & be happy
Production of vishay is really stable,there is no any other parts in my prototype for debug program.
I used the tsop4840 for rogramme, it is stable. It should be far more stable using the latest version.
So maybe it is not necessary to add power filter for TSOP.
The pcb will be very compact as production。
the power source is filtered by caps and protected by TVS diode,the route to tsop is very short.
In the data sheet ,it said tsop is insentive to power voltage ripples and noise.
All those support my thought: don’t add other parts for TSOP.
