That chip you refer to at 6:33 minutes in is a Linear Tech Triple Monolithic Step-Down Regulator with LDO. Just search for LT 3507. There is just one chip, with different temperature specifications. It has one 2.4 Amp output and two 1.5 Amp outputs (Switching). Pins 2,& 3; 30 & 31; 35, & 36 - +4Volts - +36Volts (this would be your +12Volt input) Pins 37 & 38; 28 & 29; 33 & 34, are the three outputs. Pins 10, 11, & 12 are the Power Good signals - Open Collector, active high. Suggest you download the datasheet. Omegaman
Yeah, exactly, it was very easy to find once you know the logo of Linear Technologies, and it's a very recognizable brand so I'm very surprised that any electronics related person doesn't recognize the company.
a few days ago i've tried to fix a laptop battery for my mom, a dead battery and i had no success with it, because a can't reprogram de IC from de BMS, until i saw your video about BMS and battery status. Today i saw this video and it help a lot with the frustration that i had. Thanks from Brazil!
I had a Korg DW8000 synthesiser I had on and off the bench for months, the microproessor was not running. I thought I had reached my limit after checking all power and reset lines etc. Then one day I decided to swap the crystal and after that it all came back to life. It sometimes it takes persistence to climb the tree of knowledge.
True. When you run into an MCU, that's normally the end of the line. But in exceptional cases, a reflow will fix it or a crystal replacement. In my experience, it's very few and far between that crystals fail, but it does happen. It has happened after resigning myself to failure and on the verge of giving up, that I've replaced the crystal as an afterthought, and that has solved the problem. It aint over till it's over!
*Excellent example here as far as knowing your limitation. As an Automotive Locksmith myself, I also fix different car electronics. Even though I started recently doing this and need more experience in removing/replacing MCUs, there are programmers which can read this micro controllers, extract their software and write them on to a new micro controller. That would have fixed the unit. The most important thing here, is to understand what does what, which you have given us a fantastic explanation. I just subscribed to your channel and I believe I will learn a lot from you sir. Thanks for your contribution to humanity, because this is what this is. Helping others. God bless!!!*
Yes, 100% true, basic electronic knowledge and skills, good programmer, oscilloscope, giving light in the tunnel, after checking power supply, eliming the shorts, the logic often going corrupted, caused by low battery voltage, poor grounds, electric spikes, best practices is reversed engineering and start from very basics, automotive electronics repairs is very pleasure business and worth investing time.
Sometimes filtered osciloscope enough for power supply and noises :) to the board along with rest :) catch the frames with raised and lower spikes speaking :)
From my experience, if the MC is damaged or has a corrupted firmware reading could fail or you could read corrupted firmware which obviously not gonna help. It's really hard at this point.
Even if firmware read protection in chip is not enabled (which at this case certainly was enabled), there is no reason to use already corrupted firmware. Also how you will read firmware from physically fried chip?
@@KrotowX I don't think that the chip is fried. It looks in excellent condition. I believe the problem is caused by something else that needs further investigation. Even though the chip might be faulty it doesn't mean the firmware is no good. It's worth giving it a go as this particular chip is not very expensive.
Being a hardware and firmware designer, who also did repair sometimes, I highly approve your point: the complexity and raising integration into ICs saddly makes it really hard, sometimes even impossible to fix. It's kind of playing lotto... Especially without the schematics. If the issue isn't obvious, doesn't pop up any soon, you're goign to loose. In my country - depending on the device's price of course - for the avarage you've got max 15 to 20 minutes for a full repair. If you can't succeed within that time, you lost on that one. It's a really hard, unforgiving, greedy bussiness... Either you've access to some helpfull data or you always face the same product, gaining knowledge for the next repair (like Louis Rossmann, only fixing macbooks) so you get quicker or know what usually fails... And when it comes to failure on ICs, especially those holding unaccessible firmware, just forget it, be sorry, next. Its sad, I don't like this, things should be more fixable, but it's the hard truth... PS: minus isn't necessarly ground, you can also have a plus, a minus and a ground; or even more.
@Jeffrey Wolsieffer "and even Companies like "Digi-Key" and "Motorola" simply Gave Up on Selling Directly to the Public." Do you mean Digikey stopped selling the specific chips you are talking about? Digikey does sell to the public. Always did. Mouser is another source well respected. Digikey is an interesting model of customer service. I remember back in the late 1970s, they were a hobbyist surplus seller. They would buy parts over purchases from manufactures and re-sell them. These were OEM new parts. By the 1990s they became a serious electronics parts distributor. Today they have parts manufactures lining up at their door to carry their line. IMO, they accomplished this by providing outstanding customer service. And they still do. Mouser has a similar history but is bit smaller than Digikey.
@@alfa-psi oh, right, i even know... I'm not an english native, neither did i ever live in an english speaking country, so sometimes it happens i do a misspelling or such ^^
I feel your pain ... working without schematics or any support from the manufacturer it's a real drag. To repair that board you would need a new microcontroler a JTAG adapter and the firmware image eventually with a reseted ID , an interesting project for a Sunday afternoon at your garage but not for a repair shop with fixed costs. Keep up the good work , i always take a few minutes to watch your videos (when i am not dealing with the same s****).
That blue connector is an LVDS video connector. I work on these all the time at work. When we install Aftermarket carplay units they interupt that LVDS and inject the Carplay video image into the LCD instead of the OEM infotainment image. Check out ZZ-2 carplay, Grom or NAVTV and you will see how that signal works. BMW, Audi, and othes use this style connector. GM uses what looks like a USB but they are all LVDS. I hope this helps some.
Hi, Sorin. Nice explanation about our limitations. We should never got trapped into the rabbit hole. The unknow IC is probably Linear Technology, and LT 3501 is indeed triple regulator. All the best and keep on doing a great job ❤
Hello Sorin. I believe that the power supply IC is manufactured by "LT" - Linear Technologies. The Atmel IC is an FPGA - Field Programmable Gate Array, and like you clearly explained, without additional information (which determines what each of the programmed Atmel IC pins is supposed to do), you really cannot move forward any further with any additional diagnosis, as to why this remotely located tuner board is not functioning as intended, or as in this case, at all. Full stop. Good call sir. Fred
Important Note: I originally misidentified the Atmel QFT IC used on your UUT tuner PCB Sorin. I have several projects currently going through my mind at this time and I misidentified the UUT microcontroller as an FPGA (Field Programmable Gate Array), which it clearly is not! As I am sure you know, the Atmel IC used on your UUT remote tuner unit, is in fact, a 32 bit AVR flash microcontroller, also manufactured by Atmel technologies, however the IC is a totally different animal. As we all know, on the internet, it is a Mortal Sin to misidentify a semiconductor! So to hopefully right my wrong, please accept my mea culpa/excuse my absentmindedness Sorin and also viewers of the channel. I will try harder to not make any more errors when I place a comment in the future Sorin. My bad... Fred
Hello ,@@ChupoCro! : The short answer is that I am a moron, and in my haste to console my dear friend Mr. Sorin, I had a brief memory fart that exited my mouth (through my keyboard) whilst I was still thinking through the Xilinx FPGA unit, up on the logic analyzer back at the office. My mind was still on a custom Altera based FPGA contestant/prototype, to be used in a new, custom, bidirectional, multi-protocol UART "bridge" product, being debugged and validated for full-duplex, BACnet protocol backbone translation. The finished unit(s) (products) will be implemented in the new N.Y. Mets stadium, building automation solution/product. There is an existing, legacy, resident, Xilinx (FPGA) currently installed which is resident inside of a Siemens "Main Control" panel, which needs to be able to read from and write to the existing BACnet data buss, in order to be able to communicate with our new products. This system is in its final debugging stage and I had been working on one of the UART PCBs and I managed to import and by mistake, reference that component in my head whilst I quickly wrote that half a$$ed comment to Sorin. That particular Siemens UART, which is an FPGA centric solution also supports Modbus TCP-IP, and is required to be preserved for cost-saving, and unfortunately is still stuck in my head... even this morning... So much for not bringing work home with oneself. A Big Fail again. WooHoo!!! I hope that you have an awesome weekend ChupoCro! Fred
Well done. If I was the customer, I would be ok with this final decision. Repair or replace? Replace, an opinion presented by an electronics technician. I believe you could repair the board, however parts and labor might exceed the cost of a replacement. THIS IS VALUABLE KNOWLEDGS. Good Show.
Thanks Sorin, I understand your limitation. I have more but still learning. Most off the time from you. This is one I learned a lot thank you for sharing
People should know there limits in all things.. Sometimes physically and even mentality it is good to try to push beyond your limits. But in understanding and knowledge not.. Excellent teaching Sorin..
I was reading all comments about how to keep testing that board; also I commented about a USB protection issue happens to me. But the original purpose of this video was to explain when you need to stop maybe for technical or personal limitations or maybe because is impossible to find a component to fix it. Greetings
I also would have checked the X-tal oscillator to see if there is a valid clock signal on the µC. But indeed once you arrive at the software level it becomes a dead end pretty much. I used to repair TV 's back in the the 90's and it was a different story with most analog electronics.
I second that. I've repaired several items that even had the clock running at the right speed but the scope showed a fuzzy sign wave. Changed the x-tal, sign wave clean, and microprocessor is talking again.
I believe what frustrates most of us is having a constant stream of unrepairable items, most of the time it is just very bad luck, I guess the best way to avoid that is to "filter" items on reception and as you said at the beginning, just reject some jobs.
New to all this PCB diagnostics since as service technicians we hardly ever have to replace or resolder individual components on a PCB. We just swap them out en-bloc...Too bad. What an utter waste...Today I was real bummed out having to replace a pretty expensive main board (worth at least Eur 3K) which had an on-board power supply. It had given up the ghost due to m/p a capacitor-related failure, but I could see no broken or damaged traces. Nor could I detect any suspicious odors suggestive of anything having been burnt. I wish I had your mad skillset and your experience to take a proper look at that main board and find out what the issue was. Thanks for a great video!Subscribed:)
Great video Sorin, i have come across a few of these lately but i always keep going and end up frustrated, questioning myself. Its good to see others have issues like this too and that all boards cant be fixed!
Very Very difficult to diagnose circuitry if you don't have schematics of the chips in question. I subscribed right away as I watched how you troubleshoot the circuit board, even though you don't have schematics of all the components. Great Video. Thank You.
Hi Sorin. You are right; sometimes we need to stop and that's all. The board you was working makes me remember a keyboard based on Arduino chip. It was having a erratic problem with USB communication; all voltages was fine but sometimes was recognized and sometimes not. In my case the problem was the USB protection chip SR3.3.TCT. Maybe in your case one or both of the 6 pins USB in series with USB lines are killing USB comms. Thanks for your videos.
I watch a few electronics engineer's get out their microscope and I know right away that my limit. I have worked on a few circuits using SMD's etc, but got away with using a head worn magnifier - but it becomes quite taxing for me. It's good fun watching you work on them though. We used to deliver a tv for the customer to use while we repaired their tv. That was in the 70s when we went up to 3 channels.
I felt frustrated many times when i trip up to fix pb some times i go beyond my limitation that result messing up some devices, thank you for this lesson
this hthe best Episode so far! a lot of people when watching videos repair ,,, once doing same they think they r already technicians and this lead to a lot of computer repair shops...but no real repair just creating problems!
While I think you're giving good advice on knowing your limits, I also believe it depends on the person. I remember working on a power supply board for a TV and constantly failing to fix it. After 3 months of random attempts, I finally found the issue and fixed it. It was a free TV and I was too stubborn to give up. Most would give up - but the difference is knowing your limit and pushing through it. Time and patience is needed to learn - and while some things cannot be fixed, some can. I will tell you the sense of accomplishment was so awesome. Try, fail, try again, fail again, and repeat until you're happy. Sometimes it works out, other times it doesn't - accept that fact and I think most people will be OK. As always Sorin, thank you for all the knowledge you share!
This is true, when I see something giving me serious difficulty I suspend the repair, I mean a very difficult one, and I have come to find out that coming back to it 2 weeks later, you must have grown mentally while fixing other things. And when I get back it that isuue boom it's fixed and I'm happy.
I didn’t see this in the comments. Most all automotive modules have a communication statement to the buss when they power up to let everyone else on the buss know they are ready. The point is, look at the comm pin on the atmel device when power is applied. Not the main power, the ON power from the radio. If it doesn’t spit out something on the comm bus, you are done. Looks for serial data. If you have something, then check the two six pin canbus converter chips next in line. Data must be on the connector port when power or it will never work. It’s a short burst. These are universally true for most modules.
The chip in 6:38 is a LT3507 step down converter from Linear Technology. You can get it from Aliexpress, but there are quite a few different variants around and I could not identify the exact version on your board (my limitation...).
@@angryman9333 I've had counterfeit chips delivered from ali. Aliexpress is just a marketplace where many sellers offer their products. It all depends on the seller.
Salut Sorin ! Tocmai ce ți-am descoperit canalul de TH-cam. M-am abonat instant pentru că sunt un entuziast al electronicii. Multă sănătate și câți mai mulți abonați!
The funny thing is, where you stop, I take over. I know how to flash chips, pull firmware, and understand digital communication very well. I'm just trying to learn your side which seems impossible to me but I'll never stop learning so I can help others!
Sorin that's wonderful, I will work with the information.. Sometimes you can fix it, and sometimes not...knowing what can be done is always satisfying to me! cheers
The Session_valid signal is high when the voltage on the USB_VBUS pad is higher than or equal to 1.4V. For UC this is input signal. so in this particular case the problem could be not in UC. You have 2 more components attached to 49 pin. one of them is MLCC which can be cracked and shorted it is the most common problem in electronics under stress like vibrations and falls.
"A man's got to know his limitations." Clint Eastwood, from the movie "Magnum Force". Thanks for sharing your approach. So, my takeaway is that the microprocessor cannot be replaced and even if it could, the information to program it is unavailable. Correct?
Is there a clock signal? looks like possibly two xtal/oscillator positions on the board, one has capacitors but no oscillator 16:31 top left around pin 99,100 also looks like there could be another oscillator at the right hand side around pin 47, 48 ish? I've had similar where the oscillator has failed. Worth a shot...
I spent 14 years repairing electronics (mainly before too many SMDs). People would bring me amplifiers or synthesizers or whatever and say 'it was working fine until it went wrong - I'm sure it's something simple'.
What content, theme or topic is covered here! And if we add to that the feedback from colleagues with their comments, then of course it makes us reflect and situate ourselves in reality. I saw it twice translated into my native language. The best video by far (and only here) to date in my opinion. Thank you very much for sharing. All the best.
I am getting old.. I love electronics and I it's a great hobby.. I have found lots of limitations.. One of these day I will get one of thoes microscopes... Wish you well
Great point! I sometimes stop to late and put it aside where it stays for weeks and months. Knowing it is a no fix is much better. Keep up the good work. How is your ham license going?
What you say at @08:21 is very important, many of these cards will be impossible to test without the other things that are supposed to be connected to it. Sometimes.
great work to save time on processing how the PCB works and most circuits doesn't have a schematic diagram (the factory do not want you to repair it) so I do a checkup on most likely to get defect if the system not working them the CPU chip got damaged from esd shock (modern design want to save components cost ) like filter inductors and transient diodes at removed so using continuety test using a multimeter on MCU pins to gnd and vcc if you find very low ohm like 200 ohm then it's likely damaged from esd .
A very nice vidéo as usual! Wisdom starts when you understand your limitations! That said, as the client has the same working board, there is probably a way to read the ATMEL and copy it in the non working board. Of course we need the right programmer to proceed… my two cents!
i was about to say the same. Its a good chance, that as you know your limits, you can also learn a bit from 2 boards. 1 of them we know works.. so we could use it to just check voltages and stuff around and compare.
This is a lost trade. My fathers a 3rd gen repair technician. My mom stole me from my father when I was 8 so I lost the chance to learn. These videos make me want to try anyway!
this title basically screams my skills in a nutshell...i have been told by a couple of folks here and there that my repair skills are "embarrassing" yet i haven't been doing it for long
Have a similar experience from automotive OEM audio amplifier. After owner shorted 2 dynamics somehow it switched off one of two 4-channel amplifier chips. All the components are working fine except chip not received i2c command to start. Schematic uses some switching chip for i2c because all two amplifiers is the same so has che same address in i2c. Switching line come straight from the main FPGA and it not switching. If I physically connecting two i2c buses after switch - all two amplifiers starting. But balance-fade will not work for sure so it make no sense. Looks like pure software bug on double-error stored somewhere in the FPGA EEPROM.
The logo on the chip is from a very well known chip manufacturer: Linear Technology. Linear Technology was acquired by Analog Devices but apparently the brand was kept.
That chip manufacturer is Linear Technology. What I usually do to identify an unknown chip is Google a list of semiconductor manufacturer logos, then match up that to what's on the chip.
I think u need a osciloscop to check the data or signal pin. Try to looking pulse signal from ic. Present or not. If not present the ic may be faulty. If present. In this case cannot comunicate may be the software corrupted. This just for curious job. But btw. In the end You are right. This is the limit.
From time to time, I get funny jobs like this coming in. The problem is, a lack of information and, of course manufacturers never release the schematics. I generally don't take them on unless I know I can fix them. You can spend a lot of time and then you can't bill it 😐. Most things these days are made to be replaced, not repaired. It's more money for the manufacturers.
thats true. i bet there is a way to test it out... Would data ports need the actual radio to be tested tho? can you test data output with oscilloscope only? I'd like to learn more about that tester.
@Mr Guru Your arrogance, ("Guru" word in the nic gives a hint) is something. So, because you can search for a clock pulse that, by the way can can be stopped by the firmware in some of the sleep modes of the chip or you use a logic analyser on a MC that you have no idea on program structure and port functions, you can get more info than Sorin has? I bet you can read the locked firmware, reverse engineer it, fix it, put it back in and fix the thing in your morning meditation. And then preach how good mast***bation is for gurus.
@@nikspanakis yes he smells ignorance. He doesnt have any respect this man. He already recorded about 700 videos to this platform and he tries his best to explain people how he fix things and inspiring them. If you this "guru" better record hundreds of videos like him instead of empty talks then we can listen to his voice.
As a last resort - I’d put a heat gun over it / or in an oven to heat board up just in case there’s a bad solder ball under Put a resistor on the output connections - just in case there’s a sense resistor / current reqd in any of the processor chips
Think anyone saying Sorin gave up doesn’t understand his running a business , whilst he may repair this with clock signal check etc , it’s much more profitable and easier to repair a power fault than these intricate issues. Go see northridge fix for examples , good content doesn’t always mean happy bosses 👍
My first career was electronics and I then switched to IT. So the question is why are we spending our time? And the answer is to earn a living.At some point the cost of the time exceeds the cost of replacing the unit with a new one. Or in the case of running a repair business, your time cannot exceed the amount you can charge for the repair.Time is the finite resource. Electronics is much harder than IT, and typically in IT someone is paying for your time (salary/contract day rate) vs in electronics repair your income is related to how many repair jobs you complete. So it must be time boxed. I had a faulty inverter and I could not fix it; I spent so much time on it and I then stopped and bought a new one (had already repaired it once). I had a Samsung phone water damage; I fixed the hardware and got the phone working in a couple of hours and then the owner could not remember the password and I spent loads of time trying to get round that and failed on that; hardware fixed but thwarted by the Andriod security.
I just posted similar. Looks like there is a position for one around pin 99 (left of the board) as two capacitors are there, but no oscillator. There seems to be one present at the right side connected to around pin 49 or so with cap either side. I'd get the scope on it to check.
Heah, its your time, I wouldn't stopped there if the microcontroller clock works I would extract the hex file from it, and load in order new microcontroller, but the price would be too high anyway.
I am not superman 🙂, aways better with, besides some industrial PCB makes are erasing the chips names and serial numbers, so you can't fix it without their knowledge.
As a last resort, you can hose down the board with flux, and reflow every solder joint, or the ones you determine to be involved. The SMT process is not always perfect. A required signal from an open joint can prevent passing power on self test and appear dead when really it’s only waiting.
My Camry clock illumination faded. Google says there are dry joints on a few resistors. I look and solder a few even they look perfect soldered. Still no fix. I persevered and flowed over whole board and it came back to live. Last step on checklist will be reflow all over if I'm desperate to fix a board . Unseen bad joints is always a possibility.
That chip you refer to at 6:33 minutes in is a Linear Tech Triple Monolithic Step-Down Regulator with LDO. Just search for LT 3507. There is just one chip, with different temperature specifications. It has one 2.4 Amp output and two 1.5 Amp outputs (Switching).
Pins 2,& 3; 30 & 31; 35, & 36 - +4Volts - +36Volts (this would be your +12Volt input)
Pins 37 & 38; 28 & 29; 33 & 34, are the three outputs.
Pins 10, 11, & 12 are the Power Good signals - Open Collector, active high.
Suggest you download the datasheet.
Omegaman
Awsome!
correct LT 3507
Yeah, exactly, it was very easy to find once you know the logo of Linear Technologies, and it's a very recognizable brand so I'm very surprised that any electronics related person doesn't recognize the company.
@@InTimeTraveller Clearly he's not good enough.
@@InTimeTraveller didnt even know about the brand, just searched for LT 3507 and the next line of the chip
a few days ago i've tried to fix a laptop battery for my mom, a dead battery and i had no success with it, because a can't reprogram de IC from de BMS, until i saw your video about BMS and battery status. Today i saw this video and it help a lot with the frustration that i had. Thanks from Brazil!
I had a Korg DW8000 synthesiser I had on and off the bench for months, the microproessor was not running. I thought I had reached my limit after checking all power and reset lines etc. Then one day I decided to swap the crystal and after that it all came back to life. It sometimes it takes persistence to climb the tree of knowledge.
True. When you run into an MCU, that's normally the end of the line. But in exceptional cases, a reflow will fix it or a crystal replacement. In my experience, it's very few and far between that crystals fail, but it does happen. It has happened after resigning myself to failure and on the verge of giving up, that I've replaced the crystal as an afterthought, and that has solved the problem. It aint over till it's over!
@@Rendon276 Haha that was exactly the same as my experience. Got lucky and robbed one out of a dead STM32 dev board lol
Let me give you another of your limitation : you have no idea how much you're inspiring me ! Thank you !
*Excellent example here as far as knowing your limitation. As an Automotive Locksmith myself, I also fix different car electronics. Even though I started recently doing this and need more experience in removing/replacing MCUs, there are programmers which can read this micro controllers, extract their software and write them on to a new micro controller. That would have fixed the unit. The most important thing here, is to understand what does what, which you have given us a fantastic explanation. I just subscribed to your channel and I believe I will learn a lot from you sir. Thanks for your contribution to humanity, because this is what this is. Helping others. God bless!!!*
Yes, 100% true, basic electronic knowledge and skills, good programmer, oscilloscope, giving light in the tunnel, after checking power supply, eliming the shorts, the logic often going corrupted, caused by low battery voltage, poor grounds, electric spikes, best practices is reversed engineering and start from very basics, automotive electronics repairs is very pleasure business and worth investing time.
Sometimes filtered osciloscope enough for power supply and noises :) to the board along with rest :) catch the frames with raised and lower spikes speaking :)
From my experience, if the MC is damaged or has a corrupted firmware reading could fail or you could read corrupted firmware which obviously not gonna help. It's really hard at this point.
Even if firmware read protection in chip is not enabled (which at this case certainly was enabled), there is no reason to use already corrupted firmware. Also how you will read firmware from physically fried chip?
@@KrotowX I don't think that the chip is fried. It looks in excellent condition. I believe the problem is caused by something else that needs further investigation. Even though the chip might be faulty it doesn't mean the firmware is no good. It's worth giving it a go as this particular chip is not very expensive.
Being a hardware and firmware designer, who also did repair sometimes, I highly approve your point:
the complexity and raising integration into ICs saddly makes it really hard, sometimes even impossible to fix. It's kind of playing lotto... Especially without the schematics. If the issue isn't obvious, doesn't pop up any soon, you're goign to loose.
In my country - depending on the device's price of course - for the avarage you've got max 15 to 20 minutes for a full repair. If you can't succeed within that time, you lost on that one. It's a really hard, unforgiving, greedy bussiness...
Either you've access to some helpfull data or you always face the same product, gaining knowledge for the next repair (like Louis Rossmann, only fixing macbooks) so you get quicker or know what usually fails...
And when it comes to failure on ICs, especially those holding unaccessible firmware, just forget it, be sorry, next. Its sad, I don't like this, things should be more fixable, but it's the hard truth...
PS: minus isn't necessarly ground, you can also have a plus, a minus and a ground; or even more.
Great explanation :) no words after, great people 😊
@Jeffrey Wolsieffer "and even Companies like "Digi-Key" and "Motorola" simply Gave Up on Selling Directly to the Public."
Do you mean Digikey stopped selling the specific chips you are talking about? Digikey does sell to the public. Always did. Mouser is another source well respected. Digikey is an interesting model of customer service. I remember back in the late 1970s, they were a hobbyist surplus seller. They would buy parts over purchases from manufactures and re-sell them. These were OEM new parts. By the 1990s they became a serious electronics parts distributor. Today they have parts manufactures lining up at their door to carry their line. IMO, they accomplished this by providing outstanding customer service. And they still do. Mouser has a similar history but is bit smaller than Digikey.
@@alfa-psi oh, right, i even know... I'm not an english native, neither did i ever live in an english speaking country, so sometimes it happens i do a misspelling or such ^^
The step-down IC is a LT3507 from Linear Technology.
Hi flo, we came to same conclusion (nowadays Linear Technology merged with Analog Devices).
What an absolute legend, Sorin is hands-down the teacher, the mentor, the boss, and the friend that everyone would love to have ❤❤❤
No...
Absolutely agree mate
I second that.
I feel your pain ... working without schematics or any support from the manufacturer it's a real drag.
To repair that board you would need a new microcontroler a JTAG adapter and the firmware image eventually with a reseted ID , an interesting project for a Sunday afternoon at your garage but not for a repair shop with fixed costs.
Keep up the good work , i always take a few minutes to watch your videos (when i am not dealing with the same s****).
Sometimes even when you have a schematic it is not easy to find faults.
That blue connector is an LVDS video connector. I work on these all the time at work. When we install Aftermarket carplay units they interupt that LVDS and inject the Carplay video image into the LCD instead of the OEM infotainment image. Check out ZZ-2 carplay, Grom or NAVTV and you will see how that signal works. BMW, Audi, and othes use this style connector. GM uses what looks like a USB but they are all LVDS. I hope this helps some.
Nice history but isnot electronics
Ķ
he still cool teacher and watch him all every day
Hi, Sorin. Nice explanation about our limitations. We should never got trapped into the rabbit hole. The unknow IC is probably Linear Technology, and LT 3501 is indeed triple regulator. All the best and keep on doing a great job ❤
Hello Sorin. I believe that the power supply IC is manufactured by "LT" - Linear Technologies. The Atmel IC is an FPGA - Field Programmable Gate Array, and like you clearly explained, without additional information (which determines what each of the programmed Atmel IC pins is supposed to do), you really cannot move forward any further with any additional diagnosis, as to why this remotely located tuner board is not functioning as intended, or as in this case, at all. Full stop. Good call sir. Fred
Important Note:
I originally misidentified the Atmel QFT IC used on your UUT tuner PCB Sorin. I have several projects currently going through my mind at this time and I misidentified the UUT microcontroller as an FPGA (Field Programmable Gate Array), which it clearly is not! As I am sure you know, the Atmel IC used on your UUT remote tuner unit, is in fact, a 32 bit AVR flash microcontroller, also manufactured by Atmel technologies, however the IC is a totally different animal. As we all know, on the internet, it is a Mortal Sin to misidentify a semiconductor! So to hopefully right my wrong, please accept my mea culpa/excuse my absentmindedness Sorin and also viewers of the channel. I will try harder to not make any more errors when I place a comment in the future Sorin. My bad... Fred
@@electronicengineer lol. No worries friend.
Hello ,@@ChupoCro! : The short answer is that I am a moron, and in my haste to console my dear friend Mr. Sorin, I had a brief memory fart that exited my mouth (through my keyboard) whilst I was still thinking through the Xilinx FPGA unit, up on the logic analyzer back at the office. My mind was still on a custom Altera based FPGA contestant/prototype, to be used in a new, custom, bidirectional, multi-protocol UART "bridge" product, being debugged and validated for full-duplex, BACnet protocol backbone translation. The finished unit(s) (products) will be implemented in the new N.Y. Mets stadium, building automation solution/product. There is an existing, legacy, resident, Xilinx (FPGA) currently installed which is resident inside of a Siemens "Main Control" panel, which needs to be able to read from and write to the existing BACnet data buss, in order to be able to communicate with our new products. This system is in its final debugging stage and I had been working on one of the UART PCBs and I managed to import and by mistake, reference that component in my head whilst I quickly wrote that half a$$ed comment to Sorin. That particular Siemens UART, which is an FPGA centric solution also supports Modbus TCP-IP, and is required to be preserved for cost-saving, and unfortunately is still stuck in my head... even this morning... So much for not bringing work home with oneself. A Big Fail again. WooHoo!!! I hope that you have an awesome weekend ChupoCro! Fred
Well done. If I was the customer, I would be ok with this final decision. Repair or replace? Replace, an opinion presented by an electronics technician. I believe you could repair the board, however parts and labor might exceed the cost of a replacement. THIS IS VALUABLE KNOWLEDGS. Good Show.
Thanks Sorin,
I understand your limitation. I have more but still learning.
Most off the time from you.
This is one I learned a lot thank you for sharing
People should know there limits in all things.. Sometimes physically and even mentality it is good to try to push beyond your limits. But in understanding and knowledge not.. Excellent teaching Sorin..
@Mr Guru WOW I am glad you are not my guru.. you missed every point of this video.. But thanks for the comment..
@Mr Guru I am so happy for you..
You are the father I never had! Thank you for everything and keep it up, you are the best! ❤️
I have learned so much from you and could never thank you enough!
I was reading all comments about how to keep testing that board; also I commented about a USB protection issue happens to me.
But the original purpose of this video was to explain when you need to stop maybe for technical or personal limitations or maybe because is impossible to find a component to fix it.
Greetings
Thanks for your encouraging words...from a novice in electronics!
I also would have checked the X-tal oscillator to see if there is a valid clock signal on the µC. But indeed once you arrive at the software level it becomes a dead end pretty much. I used to repair TV 's back in the the 90's and it was a different story with most analog electronics.
I second that. I've repaired several items that even had the clock running at the right speed but the scope showed a fuzzy sign wave. Changed the x-tal, sign wave clean, and microprocessor is talking again.
I had bad crystals too... And of course Bad crystal/ceramic filters in RF circuits
I believe what frustrates most of us is having a constant stream of unrepairable items, most of the time it is just very bad luck, I guess the best way to avoid that is to "filter" items on reception and as you said at the beginning, just reject some jobs.
New to all this PCB diagnostics since as service technicians we hardly ever have to replace or resolder individual components on a PCB. We just swap them out en-bloc...Too bad. What an utter waste...Today I was real bummed out having to replace a pretty expensive main board (worth at least Eur 3K) which had an on-board power supply. It had given up the ghost due to m/p a capacitor-related failure, but I could see no broken or damaged traces. Nor could I detect any suspicious odors suggestive of anything having been burnt. I wish I had your mad skillset and your experience to take a proper look at that main board and find out what the issue was. Thanks for a great video!Subscribed:)
Sorin is the best easy to understand teacher you can have
Great video Sorin, i have come across a few of these lately but i always keep going and end up frustrated, questioning myself. Its good to see others have issues like this too and that all boards cant be fixed!
Perfect example Said about limitations...Building Beautiful lives through knowledgeable personnels like you Sir, God Bless you Sir 🙏 😊
Very Very difficult to diagnose circuitry if you don't have schematics of the chips in question. I subscribed right away as I watched how you troubleshoot the circuit board, even though you don't have schematics of all the components. Great Video. Thank You.
thank you for clarifying and encouraging me to specialize on one thing to avoid disappointments.
Hi Sorin. You are right; sometimes we need to stop and that's all.
The board you was working makes me remember a keyboard based on Arduino chip.
It was having a erratic problem with USB communication; all voltages was fine but sometimes was recognized and sometimes not.
In my case the problem was the USB protection chip SR3.3.TCT. Maybe in your case one or both of the 6 pins USB in series with USB lines are killing USB comms. Thanks for your videos.
BIG BOSS SORIN, YOU CAN'T JUST GIVE UP ON THIS BOARD COME ON, CHECK OUT FOR SIGNALS, RESET LINES. CLOCKS AND ALL. you taught me everything I know.
I watch a few electronics engineer's get out their microscope and I know right away that my limit. I have worked on a few circuits using SMD's etc, but got away with using a head worn magnifier - but it becomes quite taxing for me. It's good fun watching you work on them though. We used to deliver a tv for the customer to use while we repaired their tv. That was in the 70s when we went up to 3 channels.
I am disappointing when cant fix sth and this video give me the power to continue electronics!!Thank you very much nice explanation video!
I felt frustrated many times when i trip up to fix pb some times i go beyond my limitation that result messing up some devices, thank you for this lesson
this hthe best Episode so far! a lot of people when watching videos repair ,,, once doing same they think they r already technicians and this lead to a lot of computer repair shops...but no real repair just creating problems!
While I think you're giving good advice on knowing your limits, I also believe it depends on the person. I remember working on a power supply board for a TV and constantly failing to fix it. After 3 months of random attempts, I finally found the issue and fixed it. It was a free TV and I was too stubborn to give up. Most would give up - but the difference is knowing your limit and pushing through it. Time and patience is needed to learn - and while some things cannot be fixed, some can. I will tell you the sense of accomplishment was so awesome. Try, fail, try again, fail again, and repeat until you're happy. Sometimes it works out, other times it doesn't - accept that fact and I think most people will be OK. As always Sorin, thank you for all the knowledge you share!
This is true, when I see something giving me serious difficulty I suspend the repair, I mean a very difficult one, and I have come to find out that coming back to it 2 weeks later, you must have grown mentally while fixing other things. And when I get back it that isuue boom it's fixed and I'm happy.
I didn’t see this in the comments. Most all automotive modules have a communication statement to the buss when they power up to let everyone else on the buss know they are ready. The point is, look at the comm pin on the atmel device when power is applied. Not the main power, the ON power from the radio. If it doesn’t spit out something on the comm bus, you are done. Looks for serial data. If you have something, then check the two six pin canbus converter chips next in line. Data must be on the connector port when power or it will never work. It’s a short burst.
These are universally true for most modules.
The chip in 6:38 is a LT3507 step down converter from Linear Technology. You can get it from Aliexpress, but there are quite a few different variants around and I could not identify the exact version on your board (my limitation...).
The only p/no variations for the LTC3507 chip are temperature related.
Getting a chip from aliexpress is like playing lottery with counterfeit chips. Good luck.
@@killer_x_treme wrong ali isamazing
@@angryman9333 I've had counterfeit chips delivered from ali. Aliexpress is just a marketplace where many sellers offer their products. It all depends on the seller.
Very good Sorin. I had a radio with a similar processor. It was running hot after a lightning strike - nothing to be done.
No ya didn't
Learning to let go of a repair and when to do it was probably the hardest part of it. Good advice!
I would check the data lines with a scope if some signal is present :)
One basic thing I would check is clock to MPU. Scope USB pins on MPU, if I see activity, then follow to USB protection circuity (if any present).
Salut Sorin ! Tocmai ce ți-am descoperit canalul de TH-cam. M-am abonat instant pentru că sunt un entuziast al electronicii. Multă sănătate și câți mai mulți abonați!
The funny thing is, where you stop, I take over. I know how to flash chips, pull firmware, and understand digital communication very well. I'm just trying to learn your side which seems impossible to me but I'll never stop learning so I can help others!
I am watching your amazing videos and i learn alot, for the specific circuit maybe a hit with a hummer can fix it.
Simply the best intro I heard. Know your limitations
thanks for this explanation now i know my limitation , if i cant fix just RTO and your problem will be solv
Sorin that's wonderful, I will work with the information.. Sometimes you can fix it, and sometimes not...knowing what can be done is always satisfying to me! cheers
The Session_valid signal is high when the voltage on the USB_VBUS pad is higher than or
equal to 1.4V. For UC this is input signal. so in this particular case the problem could be not in UC. You have 2 more components attached to 49 pin. one of them is MLCC which can be cracked and shorted it is the most common problem in electronics under stress like vibrations and falls.
"A man's got to know his limitations." Clint Eastwood, from the movie "Magnum Force".
Thanks for sharing your approach. So, my takeaway is that the microprocessor cannot be replaced and even if it could, the information to program it is unavailable. Correct?
This time learn something? We always learn something. You are the best!
Is there a clock signal? looks like possibly two xtal/oscillator positions on the board, one has capacitors but no oscillator 16:31 top left around pin 99,100 also looks like there could be another oscillator at the right hand side around pin 47, 48 ish? I've had similar where the oscillator has failed. Worth a shot...
So true video mr. Sorin. We all so understand your message. Thanx
thank you sorin , it's a good lesson you learned to us today
thank you sorin. Keep up the good work. your lessons and explanations and time to make the videos is appreciated.Thank you again,.
”You Can't Fix It, Because You Are Not Good Enough.”
That should be on the wall of every person, not just repairman.
I spent 14 years repairing electronics (mainly before too many SMDs). People would bring me amplifiers or synthesizers or whatever and say 'it was working fine until it went wrong - I'm sure it's something simple'.
This tutorial is amazing and you are really good at teaching !! great job sir !
What content, theme or topic is covered here! And if we add to that the feedback from colleagues with their comments, then of course it makes us reflect and situate ourselves in reality. I saw it twice translated into my native language. The best video by far (and only here) to date in my opinion. Thank you very much for sharing. All the best.
I am getting old.. I love electronics and I it's a great hobby.. I have found lots of limitations.. One of these day I will get one of thoes microscopes... Wish you well
LT3507 is a triple, current mode, DC/DC converter with internal power switches and a low dropout regulator
Great point! I sometimes stop to late and put it aside where it stays for weeks and months. Knowing it is a no fix is much better. Keep up the good work. How is your ham license going?
Check the filter next to the orange diode by power connector, left to left side right to right side, should be no crossover left to right vice versa
Does Atmel chip need a clock to work? Is it derived from a crystal? Is crystal oscillating?
Is chip enabled?
What you say at @08:21 is very important, many of these cards will be impossible to test without the other things that are supposed to be connected to it. Sometimes.
Super tare video, gandire. Daca medicii ar invata sa diagnosticheze, sa gandeasca, de la electronisti, ar fi un pas imens in medicina.
great work to save time on processing how the PCB works
and most circuits doesn't have a schematic diagram (the factory do not want you to repair it)
so I do a checkup on most likely to get defect if the system not working them the CPU chip got damaged from esd shock
(modern design want to save components cost )
like filter inductors and transient diodes at removed
so using continuety test using a multimeter on MCU pins to gnd and vcc if you find very low ohm like 200 ohm then it's likely damaged from esd .
A very nice vidéo as usual!
Wisdom starts when you understand your limitations!
That said, as the client has the same working board, there is probably a way to read the ATMEL and copy it in the non working board. Of course we need the right programmer to proceed… my two cents!
i was about to say the same. Its a good chance, that as you know your limits, you can also learn a bit from 2 boards. 1 of them we know works.. so we could use it to just check voltages and stuff around and compare.
Knowledge is everything. If you are academically sound then you can repair almost everything. You just need some experience...
Thanks bro finally someone who isn't posting malware or fake stuff, you deserve my subscribe!
I love you sorin am doing a lot because of you.
This is a lost trade. My fathers a 3rd gen repair technician. My mom stole me from my father when I was 8 so I lost the chance to learn. These videos make me want to try anyway!
this title basically screams my skills in a nutshell...i have been told by a couple of folks here and there that my repair skills are "embarrassing" yet i haven't been doing it for long
Yes I can ! ...Cause my name is David ! :) ...;) pour les français.
Thanks for the video, Sorin
Have a similar experience from automotive OEM audio amplifier. After owner shorted 2 dynamics somehow it switched off one of two 4-channel amplifier chips. All the components are working fine except chip not received i2c command to start.
Schematic uses some switching chip for i2c because all two amplifiers is the same so has che same address in i2c. Switching line come straight from the main FPGA and it not switching.
If I physically connecting two i2c buses after switch - all two amplifiers starting. But balance-fade will not work for sure so it make no sense.
Looks like pure software bug on double-error stored somewhere in the FPGA EEPROM.
Thanks for the helpful advise Sorin
@5:15 those are the most gorgeous tweezers I have ever seen
The logo on the chip is from a very well known chip manufacturer: Linear Technology. Linear Technology was acquired by Analog Devices but apparently the brand was kept.
That chip manufacturer is Linear Technology. What I usually do to identify an unknown chip is Google a list of semiconductor manufacturer logos, then match up that to what's on the chip.
I think u need a osciloscop to check the data or signal pin. Try to looking pulse signal from ic. Present or not. If not present the ic may be faulty. If present. In this case cannot comunicate may be the software corrupted. This just for curious job. But btw. In the end You are right. This is the limit.
From time to time, I get funny jobs like this coming in. The problem is, a lack of information and, of course manufacturers never release the schematics. I generally don't take them on unless I know I can fix them. You can spend a lot of time and then you can't bill it 😐. Most things these days are made to be replaced, not repaired. It's more money for the manufacturers.
Yes Sorin
The limitation is stopped on Atmel microcontroller test.
I could be tested by special tool to check whether it responds or not.
thats true. i bet there is a way to test it out... Would data ports need the actual radio to be tested tho? can you test data output with oscilloscope only? I'd like to learn more about that tester.
@Mr Guru Your arrogance, ("Guru" word in the nic gives a hint) is something. So, because you can search for a clock pulse that, by the way can can be stopped by the firmware in some of the sleep modes of the chip or you use a logic analyser on a MC that you have no idea on program structure and port functions, you can get more info than Sorin has? I bet you can read the locked firmware, reverse engineer it, fix it, put it back in and fix the thing in your morning meditation. And then preach how good mast***bation is for gurus.
@@nikspanakis yes he smells ignorance. He doesnt have any respect this man. He already recorded about 700 videos to this platform and he tries his best to explain people how he fix things and inspiring them. If you this "guru" better record hundreds of videos like him instead of empty talks then we can listen to his voice.
This dude is my TH-cam teacher!!!!!
The chip is LT3507, Yes it is a DC-DC
Thank you, ur right, strange, i searched on google but nothing pop up
@@electronicsrepairschool and i think that on the blue port(4pin) the LT3507 trought that transistor and resistor is outputing current to the radio!
As a last resort - I’d put a heat gun over it / or in an oven to heat board up just in case there’s a bad solder ball under
Put a resistor on the output connections - just in case there’s a sense resistor / current reqd in any of the processor chips
Man, I've first thought that this is a channel of Ray Liotta. You're like twins !
Think anyone saying Sorin gave up doesn’t understand his running a business , whilst he may repair this with clock signal check etc , it’s much more profitable and easier to repair a power fault than these intricate issues. Go see northridge fix for examples , good content doesn’t always mean happy bosses 👍
"A man's got to know his limitations" Dirty Harry (Clint Eastwood) as inspector Harry Callahan in Magnum Force.
The chip is a linear technology lt3507. You can find the datasheet online.
My first career was electronics and I then switched to IT. So the question is why are we spending our time? And the answer is to earn a living.At some point the cost of the time exceeds the cost of replacing the unit with a new one. Or in the case of running a repair business, your time cannot exceed the amount you can charge for the repair.Time is the finite resource. Electronics is much harder than IT, and typically in IT someone is paying for your time (salary/contract day rate) vs in electronics repair your income is related to how many repair jobs you complete. So it must be time boxed. I had a faulty inverter and I could not fix it; I spent so much time on it and I then stopped and bought a new one (had already repaired it once). I had a Samsung phone water damage; I fixed the hardware and got the phone working in a couple of hours and then the owner could not remember the password and I spent loads of time trying to get round that and failed on that; hardware fixed but thwarted by the Andriod security.
Isi gaseste locul langa celelalte placi defecte pentru piese intr-o cutie pe care o avem toti electronistii :))
There is one thing more to check if the microcontroller doesn't have a internal resonator, we can check for the external clock or Cristal.
I just posted similar. Looks like there is a position for one around pin 99 (left of the board) as two capacitors are there, but no oscillator. There seems to be one present at the right side connected to around pin 49 or so with cap either side. I'd get the scope on it to check.
Heaa, if you work with time budget, you had to stop. It's not easy to repair something without schematics.
Heah, its your time, I wouldn't stopped there if the microcontroller clock works I would extract the hex file from it, and load in order new microcontroller, but the price would be too high anyway.
If you never try you never know!
I am not superman 🙂, aways better with, besides some industrial PCB makes are erasing the chips names and serial numbers, so you can't fix it without their knowledge.
thank u sorin, i didnt get anything on this video, accept about the limitation. thanks for reminding
The chip manufacturer was Linear Technologies. They have since been bought up by Analog Devices.
Great show! I’m still outlining my limitations.
As a last resort, you can hose down the board with flux, and reflow every solder joint, or the ones you determine to be involved. The SMT process is not always perfect.
A required signal from an open joint can prevent passing power on self test and appear dead when really it’s only waiting.
My Camry clock illumination faded. Google says there are dry joints on a few resistors. I look and solder a few even they look perfect soldered. Still no fix. I persevered and flowed over whole board and it came back to live. Last step on checklist will be reflow all over if I'm desperate to fix a board . Unseen bad joints is always a possibility.
Another great vid dear Sorin! God bless you & your family!
It worked. Thanks a lot
The chip at the beginning is an LT3507 from Linear Technology now a subsidiary of Analog Devices.
Correct me if i'm wrong but on the board you have a jtag interface that can help you identify the nature of the problem
Know your limitations and improve by getting experience and over time you'll be more successful in your repairs.
Those are CAN hi and low lines used to communicate. The microcontroller has a built in Can capabilities. You can sniff the can communication
no. unfortunately there was USB.