Hi everybody.
First of all, I want to thank you for creating and developing this community.
Actually I'm a repairman in the field of air conditioners. Sometimes I need to look for an impairment in the electronic circuit. It's always an IC. But which one? Because of this question, I have to buy all of them and change them one by one to find the faulty one. This is neither reasonable nor good for the circuit health (due to frequent heating). So I had to buy an oscilloscope or logic analyzer or something like that. Since I was familiar to Arduino (low level newbie, off course), I was thinking about making a customized and flexible troubleshooting device with Arduino. It will help me to program my device in accordance to each IC and also it will help me to save some money!
My question is: is this even practical?
I read Gillham's Logic Analyzer topic. The setup was easy and it's working. But I'm not sure about its application for my problem. In fact, the whole thing is becoming a dead end for me. Probably, posting here is my last try.
So, can you give me a hint? I'm not looking for a prepared project. Just wondering about what I have to search.
I posted something like this before(not completely though), but I couldn't find it. So I'm sorry if it seems repetitious to you.
Thank you for reading this long post.
Industry left the on-site repair of circuit boards more than 25 years ago. The problem you have is the manufacture did not make the changes necessary to just plug in a new board and send the old one to a repair depot.
I don't know your hourly rate, but ours is $60 per hour, 1 hour minimum. We repair our mistakes as part of doing business, but don't do other peoples boards. Your board probably cost $20 to manufacture. So, how does that compare to your time?
Could an Arduino become the basis of a circuit board debugging system? Since you are dealing with 120-240 volt power, plus motor circuits, I doubt if it's practical, or even possible. Then there are the hours spent debugging what may be an error with your test system, but is really something in the Ac control board.
Hope this may be of some help.
Paul
How many different IC's do you expect to cover? If they all have SPI interfaces that would make things easier.
If when they all die, power to ground pins also don't conduct would make it easier. Most chip hacks I know start with putting ground then power to see what happens.
You've got to draw some lines on that blank page just to begin to figure out what angles you can use.
And then there's the cost to replace vs your time as Paul_KD7HB notes which may vary from country to country.
Paul_KD7HB:
Industry left the on-site repair of circuit boards more than 25 years ago. The problem you have is the manufacture did not make the changes necessary to just plug in a new board and send the old one to a repair depot.I don't know your hourly rate, but ours is $60 per hour, 1 hour minimum. We repair our mistakes as part of doing business, but don't do other peoples boards. Your board probably cost $20 to manufacture. So, how does that compare to your time?
Could an Arduino become the basis of a circuit board debugging system? Since you are dealing with 120-240 volt power, plus motor circuits, I doubt if it's practical, or even possible. Then there are the hours spent debugging what may be an error with your test system, but is really something in the Ac control board.
Hope this may be of some help.
Paul
GoForSmoke:
How many different IC's do you expect to cover? If they all have SPI interfaces that would make things easier.
If when they all die, power to ground pins also don't conduct would make it easier. Most chip hacks I know start with putting ground then power to see what happens.You've got to draw some lines on that blank page just to begin to figure out what angles you can use.
And then there's the cost to replace vs your time as Paul_KD7HB notes which may vary from country to country.
Thank you for replying so quick!
Since I'm from Iran, it's really cost effective. We do not use hourly system. When it's done, we can recieve a certain amount. It seems bad, right? But we are under economic constraints. These boards can not be imported easily. So, no matter how much high the rate is. it will be win-win.
In fact, I don't know the type of the IC's precisely. Some of them are timers. Maybe there is a programmable IC too. And drivers. I will check the trick that you mentioned and let you know the results.
Again, thank you for helping me.
The this is the perfect opportunity to begin to produce duplicate circuit boards and set up a service depot to repair boards.
I have a customer that does exactly that for old dental chair and light systems. They have been in business for more than 15 years, and have quite a few people working for them.
Paul
"It seems bad, right?"
To me it depends on what you are required to do and what that costs.
"In fact, I don't know the type of the IC's precisely. "
" It's always an IC. But which one? Because of this question, I have to buy all of them and change them one by one to find the faulty one. "
How about the ones you buy already? You know those don't you? You get datasheets?
Perhaps make your checker for the expensive ones first?
Can you salvage boards from dead AC units? Fixed, one could replace a board in a same model broken unit and let you at least work on the board taken out on a bench indoors.
Maybe you want a board-checker to see first if the board is the problem? There should only be so many different AC units.
If you have this problem then others might too? A group action may get through this quicker to benefit all. If there is no such group then you might start the first possibly to your credit.
sonofanarchy:
It's always an IC. But which one? Because of this question, I have to buy all of them and change them one by one to find the faulty one
Really, sounds rather crude ?
If you dont know electronics or the product well enough, that you cannot troubleshoot the board down to component level, you might not know enough to build an automated tester.
I used to repair Microcontroller boards for a living, and with enough knowledge and experience you could identify a fault down to an individual component in more than 95% of cases.
Did try to build an automated tester for the boards, but even though we (and me) had designed them getting an automated tester to operate even close to human capability proved to be too difficult.
What used to help a lot was to build a 'simulator' for the boards which was a light, switch, sensor replication of what the board was actually connected to.
"It's always an IC. But which one? Because of this question, I have to buy all of them and change them one by one to find the faulty one"
I can see that my message has been delivered improperly. My mistake. It was an instance for what I have to do without my intended device.
Paul_KD7HB:
The this is the perfect opportunity to begin to produce duplicate circuit boards and set up a service depot to repair boards.I have a customer that does exactly that for old dental chair and light systems. They have been in business for more than 15 years, and have quite a few people working for them.
Paul
I was thinking about that before. People want the exact appearance for their air conditioner. These boards are customized and often have a fuse for avoiding ic cloning. So I have to produce it from a blank ic. That does not work with other components properly. For example, the temperature display. So it will became completely another device with another appearance.
GoForSmoke:
"It seems bad, right?"
To me it depends on what you are required to do and what that costs."In fact, I don't know the type of the IC's precisely. "
" It's always an IC. But which one? Because of this question, I have to buy all of them and change them one by one to find the faulty one. "
How about the ones you buy already? You know those don't you? You get datasheets?Perhaps make your checker for the expensive ones first?
Can you salvage boards from dead AC units? Fixed, one could replace a board in a same model broken unit and let you at least work on the board taken out on a bench indoors.
Maybe you want a board-checker to see first if the board is the problem? There should only be so many different AC units.
If you have this problem then others might too? A group action may get through this quicker to benefit all. If there is no such group then you might start the first possibly to your credit.
Thank you for your advice. I'll keep it in mind.
srnet:
Really, sounds rather crude ?If you dont know electronics or the product well enough, that you cannot troubleshoot the board down to component level, you might not know enough to build an automated tester.
I used to repair Microcontroller boards for a living, and with enough knowledge and experience you could identify a fault down to an individual component in more than 95% of cases.
Did try to build an automated tester for the boards, but even though we (and me) had designed them getting an automated tester to operate even close to human capability proved to be too difficult.
What used to help a lot was to build a 'simulator' for the boards which was a light, switch, sensor replication of what the board was actually connected to.
Yeah, actually I intended to make an arduino based simulator befor the tester idea. A digital potentiometer for simulating thermistors, fake digital signals that played as expansion valve pressure sensors etc. . But it needs a lot of memory on the atmel microcontroller. Which it does not have.
And the last thing, I want to use arduino for this. In the first post, I asked if it is practical or not. I'm thankful for all of your helps.
Best wishes.
Ali.
With the AVR controllers you get flash memory as well as RAM. When you put non-changing data into flash, all that does not change even to text labels and prompts as well as data constants and tables -- and use it directly in pieces, not copy whole parts to RAM to use with PC methods, then even 2K can be enough to do a lot. Add SD and the range it can cover gets large.
The ATmega1284P has 16K RAM, 128K flash, 40 pins and 2 serial ports plus the usual SPI and I2C if more is needed. Can you get those, since you do have at least one Arduino and a PC?
GoForSmoke:
With the AVR controllers you get flash memory as well as RAM. When you put non-changing data into flash, all that does not change even to text labels and prompts as well as data constants and tables -- and use it directly in pieces, not copy whole parts to RAM to use with PC methods, then even 2K can be enough to do a lot. Add SD and the range it can cover gets large.The ATmega1284P has 16K RAM, 128K flash, 40 pins and 2 serial ports plus the usual SPI and I2C if more is needed. Can you get those, since you do have at least one Arduino and a PC?
Wow. It will be a great work. This is that hint I was talking about. Thanks a lot. I'll search about it and buy required elements. And probably I'll get back to you for more specific helps(actually, your kindness). As I mentioned in the first post, this is an awesome community.
You repair air conditioners only? Too busy to do much else?
Are you an engineer or a tech maybe working to become an engineer?
GoForSmoke:
You repair air conditioners only? Too busy to do much else?Are you an engineer or a tech maybe working to become an engineer?
Something like that. This was my father's job(refrigerators and old air conditioners). I wanna be more. Because technology is developing and I need to keep myself up to date.
Do you test boards or chips with meter or analyzer at all or just replace parts until it runs?
If you have test points that you check on just one board, you might automate that after much effort learning what can go wrong (debugging & Murphy's Law) for that board.
How much mechanical difference is there from one AC to the next?
What is controlled? Motors? Sensors? Switches?
What IO is used?
Reduced to that, how different are the working, non-controls parts from one AC to the next?
Do you know others doing the same as you? You can't be the only AC fixer!
What could happen is that old parts scarcity drives prices way up, a working and tested board with new parts for less cost becomes acceptable.
It might be easier to design a general purpose controller from scratch.
Can't be that hard.
After all you only have a limited number of inputs and outputs, and a fairly simple control algorithm....
I've built similar devices based on a nano.
Allan
GoForSmoke:
Do you test boards or chips with meter or analyzer at all or just replace parts until it runs?
If you have test points that you check on just one board, you might automate that after much effort learning what can go wrong (debugging & Murphy's Law) for that board.How much mechanical difference is there from one AC to the next?
What is controlled? Motors? Sensors? Switches?
What IO is used?Reduced to that, how different are the working, non-controls parts from one AC to the next?
Do you know others doing the same as you? You can't be the only AC fixer!
What could happen is that old parts scarcity drives prices way up, a working and tested board with new parts for less cost becomes acceptable.
Yes, I test simple parts with meter. But there is no complex analyzer. Meter gives me the number and brain will analyze it (if it be able for such complexity). There are many models and brands. But they are similar to a certain level.
They recieve data from thermistors and pressure sensors to controll compressor and motor fans.
Your idea for making suitable boards is reasonable. The problem is, they are not completely old. People import parts from China for low cost. Which makes "fixing" more interesting than "making". Actually the main profit is for those who importing parts illicitly(due to economic constraints and difficult transition).
As you mentioned, I'm not the only one. There are several fixers in our small city who make money as much as large businesses. I do not predict. I can see it.
And there is another problem that I mentioned in #7. The appearance. Arduino can be the base of a suitable controller. But the appearance is one of the main problems here. Arduino can not control the display and even it can not recieve IR signals from remote control. The board is customized and the code is hidden. If I could make a beautiful appearance for the project (display, remote control, etc.) I would make a whole air conditioner (more profit!). But I'm not a designer. So, a diagnoser device is more acceptable for me.
I think the ATmega1284p is good for this. A tester or a simulator. With an android GUI, it will be good enough to make my work much easier. Now, in your opinion, between a tester and a simulator, which one is better?
Can I make an arduino with ATmega1284p? Is there any tutorial? Because there is no arduino with this IC in the market. I'm sorry, I know I'm bothering you with my questions. But it will be a huge help for me.
allanhurst:
It might be easier to design a general purpose controller from scratch.Can't be that hard.
After all you only have a limited number of inputs and outputs, and a fairly simple control algorithm....
I've built similar devices based on a nano.
Allan
Thank you for your reply. Yes, it is possible. I was typing and I didn't notice your reply. You can read about my reason about this in the #14.
The situation is not so tight with imports at low prices, but are they the same chips or upgrades with fewer chips.
What you program on an Uno is the ATmega328P chip in the socket. It can be taken out and run without the board only neding clean power, a bypass cap, a weak pullup 5V to reset and then whatever pins you need to connect. It has timers and analog to digital input as well as PWM (not analog) output. It has hardware supported high speed serial ports.
You want to have a new chip to put on the Uno for the next thing, it must be bootloaded. When you get your new chips you need to burn a bootloader on those the Uno will use. Nick Gammon's make a breadboard Arduino covers how to complete.
AVR's are made by ATMEL. You get the full datasheet from their site, find 328p
sonofanarchy:
And there is another problem that I mentioned in #7. The appearance. Arduino can be the base of a suitable controller. But the appearance is one of the main problems here. Arduino can not control the display and even it can not recieve IR signals from remote control. The board is customized and the code is hidden. If I could make a beautiful appearance for the project (display, remote control, etc.) I would make a whole air conditioner (more profit!). But I'm not a designer. So, a diagnoser device is more acceptable for me.
I think the ATmega1284p is good for this. A tester or a simulator. With an android GUI, it will be good enough to make my work much easier. Now, in your opinion, between a tester and a simulator, which one is better?
Can I make an arduino with ATmega1284p? Is there any tutorial? Because there is no arduino with this IC in the market. I'm sorry, I know I'm bothering you with my questions. But it will be a huge help for me.
I don't think you can run Android on any AVR chips. You are thinking maybe of Arduinos with SAM chips or RPi.
If you want built-in extras, make your own work on breadboards first.
Arduino is Open with loads of docs and tutorials. Top of forum in the green bar, LEARNING. Keep checking links, bookmark the interesting ones for later. It is all there.
GoForSmoke:
The situation is not so tight with imports at low prices, but are they the same chips or upgrades with fewer chips.What you program on an Uno is the ATmega328P chip in the socket. It can be taken out and run without the board only neding clean power, a bypass cap, a weak pullup 5V to reset and then whatever pins you need to connect. It has timers and analog to digital input as well as PWM (not analog) output. It has hardware supported high speed serial ports.
You want to have a new chip to put on the Uno for the next thing, it must be bootloaded. When you get your new chips you need to burn a bootloader on those the Uno will use. Nick Gammon's make a breadboard Arduino covers how to complete.AVR's are made by ATMEL. You get the full datasheet from their site, find 328p
I don't think you can run Android on any AVR chips. You are thinking maybe of Arduinos with SAM chips or RPi.
If you want built-in extras, make your own work on breadboards first.
Gammon Forum : Electronics : Microprocessors : How to make an Arduino-compatible minimal boardArduino is Open with loads of docs and tutorials. Top of forum in the green bar, LEARNING. Keep checking links, bookmark the interesting ones for later. It is all there.
About android, I was thinking about user interface. Not android code on the chip. For example, with bluetooth. There are some apps which provide it.
I think ATmega1284p is good for me. Either as a diagnoser or as a replacement for the controller. Thank you again
sonofanarchy:
There are many models and brands. But they are similar to a certain level.
'Similar to a certian level' also means the boards mostly use different components.
With the boards using mostly different components what you are actually trying to design is a general general purpose component tester.
You can automate known test-point checks if you can make contact with them all.
If you solder tabs onto test points then clip leads could attach. Attach once, use over and over.
The test unit could have many clip leads and use them differently depending on the board to test.
Arduino can read digital or analog up to 5V.