Forum Replies Created
Is that 16MHz crystal for both the CH340 and the ATMega644?
sorry for the late reply….busy weekend.
I’m going to be inquiring at my local maker-space this evening about the availability of an oscilloscope. In lieu of that, I have my eye on a Tektronix TBS1052B-EDU 50 MHz, 2 Digital Channel Oscilloscope at Amazon or a PC-based unit. So, what we’re looking for with the scope is not necessarily specific data that is being exchanged on any particular data line, but merely the quality of the signal being exchanged. Specifically, I want to see nice edges from all address and data lines (i.e. – google+images “square wave”) and no signals that look like they are ramping up to near Vcc (i.e. – google+images “capacitor charging square wave”). If there is something up with a data line or multiple lines, trace it/them until you find the issue. It’s quite possible to find solder bridges that tie data lines together or a data line to some other part of the circuit. Solder bridges love to hide under connectors and other parts. If all of the µC pins look OK, I would measure all of the MOVs (V0-V9) for resistance and also remove the LCD and see if the valves actuate using the web interface. My next step would be to start isolating the sections I mentioned in my last post. Depending on your skill with a soldering iron, you could easily lift pins to the µC on IC5, IC4, U1, IC3 & resistor RQ (in that order). Post back your findings and if you have any more questions.
Howdy thisisit. I have been exceptionally busy over the last month, so I haven’t been able to troubleshoot my PCBs to my satisfaction. They have been sitting idle for a few weeks with no resolution.
I don’t know your technical expertise, but I can tell you what my approach will be when I resume troubleshooting these PCBs in a few weeks. My ethernet section (the ENC28J60, the RJ45 jack and all the discrete components) appears fully functional, as does the power supply section. I’m not concerned that the triacs don’t actuate at this point. I am concerned about the data lines on the µC that are driving the LCD, the 74HC595N and the relay (the sections that are non-functional). It would be very interesting to wire the µC (and associated circuitry) and the LCD separately on a breadboard and see if the LCD functions. It may be that there was something on the PCB that was dragging down a data line or two. That could also be simulated by carefully bending pins up on the µC and reinserting it into it’s socket. My USB section (or rather it’s communication with the µC) is also suspect. The CH340G ICs I bought from China came in a non-ESD bag from the vendor, so its possible that some have been damaged, including the one I installed.
I wanted to thank you for your inquiry, as it has renewed my interest in repairing my PCBs and rehashing my previous efforts has also given me a few more things to try, including isolating the µC from the:
• Relay circuit
• SD card socket
• Valve driver circuit
Good luck and please post your findings. It may just help someone else.
I finally got the opportunity to try to troubleshoot this issue. I removed the µC & LCD sockets. While I had the sockets out, I cleaned up the PCB and inspected the area. I did find a slightly damaged trace, which I repaired. Everything looked pretty good, so I soldered in new sockets and replaced the µC & LCD. I still have the same issue with improper function of the LCD. With the small amount of using the web interface to actuate the device, some functions seem to be working, but others do not. I am not able to operate the triacs, actuate the relay, change the LCD contrast or brightness. I was able to update my location and give it a static IP address. In addition, it appears to sync with NTP and it also got a notification that there was a firmware update (I have 2.1.2). When I tried to update with the osFWUpdater, a device was NOT detected and the update failed. I wish I had a o’scope or logic analyzer so I could do a bit more troubleshooting, as it appears that the µC is partially functioning.
I think what I’ll do at this point is to declare defeat and put them away for now…rather disappointing.
I buzzed-out the connections from the LCD connector to the µC, and everything seemed OK. I also buzzed-out connections from all of the µC & LCD connector pins to adjacent pins and to any of the traces that run adjacent to the pins. I did not find any shorts. I tried another LCD from the lot of five I bought, and two other LCDs I had from other projects. That’s five LCDs that I tried in the board and none of them worked. What I find most unusual is both boards exhibit the exact same characteristics. I’d be willing to bet that if I built up another PCB, it would do the same thing. It almost makes me think that I either installed a part incorrectly on both PCBs, or that I got a bad batch of one of the parts from my source.
I think that I’m going to use the web interface to exercise as much of the circuit as I can, then make an educated guess as to what is installed on the board that has not been tested (i.e. – SD card socket, relay circuit, µC buttons) and start removing them.
On a whim, I plugged in an ethernet cable after reading a post in a different thread. It appears that I have indeed gotten the µC programmed correctly, as I found it’s IP address listed on a network scan and I was able to access its web interface. I still need to replace the DS1307, 12MHz crystal and, of course, get the LCD working properly. I have additional LCDs that I can try, but at least I can concentrate my efforts in a particular area. If you have any additional suggestions, they would be greatly appreciated. If I happen to get it repaired, I will post my findings.
Also, I received the three enclosures I ordered late last week. I will say this about your company: Your shipments and Very, Very prompt! Thanks!
I am experiencing the same issue as stated above, except that I am having this problem on two PCBs. Neither of the PCBs will power-up and run correctly. In fact, they both act exactly the same. Vin and Vcc are both within spec with both USB and a 24VAC Adapter. It’s the third or forth time these PCBs have been inspected by me. Unless there is a whisker or ball of solder under a part, these PCBs are good to go. I’m guessing that the µCs in the DIY kit were already programmed before being sold. Since there was a period in recent history where the OpenSprinkler DIY kits were not available, my only option at that point was to buy bare PCBs and source the parts myself. I assembled one OS Pi and two OS 2.2u boards. The OS Pi works great and is currently installed and running. I’m having lots of problems with the ATmega644 based PCBs. Since I sourced all the parts (including the ATmega644), the µCs had no firmware or bootloader installed. After two days of trying different schemes for programming, I successfully programmed the µCs by using an old Duemilanove as an ISP with the 644s in a breadboard with a 16MHz crystal.
I have been through the forum posts looking for anything that might help me. I’d really appreciate your assistance before I pitch these in the circular file and build up a couple more OS Pi’s.