Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
No probs updating an existing U3S chip's firmware, but can't write to a new one...
Chris Wilson 02 January 2018
Using Extreme Burner AVR I have never had a problem updating a QRP Labs written chip, but yesterday I tried to flash a brand new chip and it "failed to find device" or similar. I use Windows 7 64 bit and write to the chip mounted in a working U3S and the U3S is powered up. New chips are 328P marked and of branded origin. Am I missing something here please? Thanks. -- 2E0ILY Best regards, Chris mailto:chris@... -- Best regards, Chris Wilson (2E0ILY) |
As far as I know (the info is in the manual), you can only write the QCX hex file to a 328P chip that was prepared for the QCX by the author of the software (Hans, QRPLab). You cannot write the hex file to a brand new chip. Neverheless you should still be able to connect your ISP to the 328P of the QCX. I tried tried just for curiousity with an old U3S 328P: You will be able to write the hex file, it will verify, but when you start up the QCX with the hex file that you transferred to the brand new chip, you will get a message?? on the QWCX display that you must have to use a proper QCX control chip to upgrade the software. Your problems seem to have little to do with the QCX itself, but with the ISP software and the connection to the QCX board. However I am no specialist of microcontroller handling and can be wrong, but I know there are fancy things like fuse bits and stuff of that kind that has to be set right first before you can try to laod new firmware. I am almost sure you know that yourself.
73 Klaus DL2QB / WN2Z |
Hello Klaus,
toggle quoted message
Show quoted text
Tuesday, January 2, 2018 I *THINK* what's required is to first write the fuse bits. I have found a link on Alan's (G4ZFQ's) web site for writing fuse bits with this device, but it's for a U3, not a U3S, and I am unsure if it is valid for the U3S processor. I may just try it, I have a few new chips. Thanks for your reply Klaus and a happy and healthy New Year! As far as I know (the info is in the manual), you can only write 73 Klaus -- 2E0ILY Best regards, Chris mailto:chris@... --
Best regards, Chris Wilson (2E0ILY) |
Folks,
I planned for failure and surprisingly was able to program a new 328P last night?for my U3S using an Ardunio, avrdudess, and the instructions for updating the firmware in the QCX November 2017 QRP-Labs newsletter.??? I figured I would brick the original MCU so bought a blank one from Amazon. General notes: 1.? Read all the messages in this forum about the subject.? Lots of great contributions-- I could not spell AVR yesterday morning. 2.? The avrdude ESS GUI is outstanding!? But it will trip all of your virus scanners.? This is covered in the GitHub discussion from four years ago.? This is the location where I found it:? 3.? I set the fuses as described in?the fuse calculator?from this site:? For the 328P I used : L?0x62, H 0xD9, E 0xFF.? Several others have discussed this in the message forum - all good information. 4.? Lastly - try to read the chip before writing to it - make sure your communication is working first.? Obviously with a blank chip there was nothing to read but I could verify the hardware and software components were communicating correctly. Now to TX CW on 630M! Regards, Dave W4VU |
While I have not had to flash a new firmware to one of my uP chips yet, what I have seen explained on this group is that most of the uP chips in the various QRP Labs kits can be firmware-upgraded by the user, including the QCX if you upgraade the uP chip supplied in the kit. All firmware upgrade hex files except the QCX files can be flashed to a new bare ATMega 328(P) if you can master the technique. As mention often, the block to doing the QCX is that the fuse bits must be set correctly in the new chip, and this information is not publicly available. So, for the QCX, a new chip ordered from QRP Labs Shop is the way to go.
-- Wes AE6ZM Hereford, AZ ? |
¿ªÔÆÌåÓýOn 3/01/2018 2:43 PM, Paul Parks wrote:
For the 328P I used : L?0x62, H 0xD9, E 0xFF.? Several others have discussed this in the message forum - all good information. Are you sure?? Those look more like the 'blank' 328's default fuse settings.? For a start, the low fuse 0x62 uses the internal 8MHz clock ... the U3S will appear to run, but the calibration and possibly the UTC clock will be out.? Try? L = 0xF7??? H = 0xD9 ?? E = 0xFC 73, Bob? ZL1RS |
There are two documents in the Files section about fuse settings. One for the Extreme programmer with Atmel Studio ( /g/QRPLabs/files/Atmel%20Studio%20Programmer%20seetings%20Atmega%20328-PU )and another for AVR Dude ( /g/QRPLabs/files/AVRDude%20Programmer%20settings%20for%20Atmega%20328-PU ).
The are different. -- Wes AE6ZM Hereford, AZ ? |
On Tue, Jan 2, 2018 at 05:43 pm, Paul Parks wrote:
I planned for failure and surprisingly was able to program a new 328P last night?for my U3S using an Ardunio, avrdudess"Planning for failure" is the first sign of maturity.? Planning for failure beyond the direst limits of one's current imagination is the first sign of the Wisdom of Age.? Seems to me, anyhow. The mystery programmer I had foisted off on me on eBay instead of the USBasp I ordered has not responded to further experimentation and remains useless as of 02JAN2018.? I do not look to any happier outcome than a refund. The Arduino/AVRdudess/DuPont-jumpers approach looks really sensible to me (which doesn't in itself guarantee it will actually work, of course) and I have all the stuff already. I expect that the introduction of the new FT8-capable U3S firmware will generate about as much traffic as that of the QCX, so we'll have plenty to discuss then. [sigh!] |
I've looked at 5 different recommendations from people that have posted their successful settings for the U3.
They do vary slightly, maybe non-applicable, non-critical or depending on programming software? See below 73 Alan G4ZFQ 2 PDFs IN FILES SECTION 1) Settings For Atmega328-PU and Atmel Studio Programmer EXTENDED = 0xFC (valid) HIGH = 0xD9 (valid) LOW = 0xF7 (valid) These settings are also those for eXtreme Programmer DO NOT use them with AVR Dude those settings should be well covered elsewhere.. 2) Programming using AVRDude and the Atmega 328P-PU and 328 PU Note:Recent evidence seems to point to the low power 328P-PU being a suitable substitute for the 328 PU. More case histories would be needed to give a positive thumbs up, due to conflicting reports. Hex settings for the Fuses using AVRDude Programmer remain slightly different from those used with Atmel Studio and eXtreme Prograrammer and others. For AVRDude use. LOW = 0xF7 HIGH = 0xD9 EXTENDED = 0x04 Naturally the end result is the same as far as the 328 is concerned. What is different is the way that AVRDude holds the data. If the Atmel Studio values are used, Verification Errors will be flagged. GROUP/WEB 3) April 2014 For fuse settings for the Atmel 328 using AVR Dude See this thread "The fuse settings for 3.03b are as follows: Extended fuse:0x04 Low fuse:0xF7 High fuse:0xDF" 4) Try L = 0xF7 H = 0xD9 E = 0xFC Bob ZL1RS 5) And Andy's page here A screenshot of Extreme Burner setup for a Atmel 328. And using Windows 8. ALSO:-Files for updating Extreme Burner for the Atmel 328. "LOW = 0X F7, HIGH = 0X D9, EXTENDED = 0X FC, LOCK FUSE = 0X FF, CALIBRATION = 0X FFFFFFB9 - I do not think you require the lock fuse or the calibration." SUMMARY ALL:- Low F7 High D9 D9 D9 D9 DF Extended FC 04 04 FC FC LOCK NA AVERDUDE different from rest? (See 2) |
I have a real AVR Programmer, actually two of them. Over the years I have had various Chinese knockoffs, eventually they ended in the trash due the software not being compatible with OS upgrades so after a while I went and paid a little extra for the real thing and have had no problems since.
toggle quoted message
Show quoted text
Just in case I took a Arduino One and programmed it to be a programmer and tested it and it worked just fine with Windows 10 and since I had quite a few Arduinos I left it ready to go as a backup. With both types of programmers I use AVR Studio software and all is well, mostly I program ATTINY85 devices for use with SoftRocks Radios but occasionally someone wants a ATMEGA328P programmed to be used as an Arduino. On 1/3/2018 1:35 AM, TrueBlue wrote:
On Tue, Jan 2, 2018 at 05:43 pm, Paul Parks wrote: --
Cecil - k5nwa |
Bob, Excellent catch: -------- For the 328P I used : L?0x62, H 0xD9, E 0xFF.? Several others have discussed this in the message forum - all good information. Are you sure?? Those look more like the 'blank' 328's default fuse settings.? For a start, the low fuse 0x62 uses the internal 8MHz clock ... the U3S will appear to run, but the calibration and possibly the UTC clock will be out.?? Try? L = 0xF7??? H = 0xD9 ?? E = 0xFC? 73, Bob? ZL1RS -------- Those are what I used and I can verify the boot up display appears correct. However, it may not work.? I will have time to test in a few days. I just did not want to break a working U3S, so tried a blank chip.?? Regards, Dave W4VU |
Hello Cecil / all
toggle quoted message
Show quoted text
Wednesday, January 3, 2018 No joy, I cannot get the programmers, either Extreme Burner or another Chinese programme called progisp1.72 to see the new virgin Atmega 328 PU processors no matter what I try. Both see, read and write the similarly marked QRP Labs programmed ones, and will re-write updated firmware to them. At a loss as to why new chips are not even seen, let alone be able to be read or written... Black art! Thanks, any other ideas welcome, I can sense the frustration building and can see why others have spent hours on these things :) I have a real AVR Programmer, actually two of them. Over the years I Just in case I took a Arduino One and programmed it to be a programmer With both types of programmers I use AVR Studio software and all is -- 2E0ILY Best regards, Chris mailto:chris@... --
Best regards, Chris Wilson (2E0ILY) |
Where did you get the new chips? Are you sure they're not counterfeit parts?
toggle quoted message
Show quoted text
On Wed, Jan 3, 2018 at 9:59 AM, Chris Wilson <chris@...> wrote:
Hello Cecil / all |
Hello Shirley,
toggle quoted message
Show quoted text
From RS Components genuine Atmel ones, so I guess they are 100% legit? Wednesday, January 3, 2018 Where did you get the new chips? Are you sure they're not counterfeit parts? -- 2E0ILY Best regards, Chris mailto:chris@... --
Best regards, Chris Wilson (2E0ILY) |
If you can see and program the provided chip but can't program a new one then the next question is where did you get those devices? If you bought them from eBay or one of the other Chinese vendors then most likely you got chips that are not what they claim or are programmed with the programming pin disabled so they can not be re-programmed.
toggle quoted message
Show quoted text
The ATTINY 85 chips I program for use with SoftRocks requires that the reset pin which is used in programming to be disabled so if I program it once, it cannot be programmed again. It takes a special "high voltage" programmer to restore the device back to be able to re-program them. I would try getting a brand new chip from a reliable source. On 1/3/2018 8:59 AM, Chris Wilson wrote:
Hello Cecil / all --
Cecil - k5nwa |
¿ªÔÆÌåÓýThis might be useful Dave ...I tried those 'default' Fuse settings (L?0x62, H 0xD9, E 0xFF).? Indeed the U3S boot up display appears, but the UTC time display runs very slowly, about 8 seconds for the seconds digit on the display to increment.? The GPS info does not display.? A Low Fuse setting of 0xF7 is what came in my original U3S (as supplied in mid-2015).? According to an , a Low Fuse setting of 0x62 will invoke the internal 8 MHz RC clock oscillator and divide the clock internally by 8.? A slow time display is therefore not surprising.? 0xF7 is the correct Low Fuse setting as per Alan's summary.? As a matter of interest, I checked the Low Fuse settings in the ATmega328(p) ICs supplied in QRP-Labs VFO and QCX ... both of them also use a 20MHz external clock crystal for the ATmega328(p), and both of them also had 0xF7 set as the Low Fuse. A High Fuse setting of 0xD9 is what came in my original U3S.? The VFO and QCX both use 0xD1 for the High Fuse which, according to the on-line Fuse calculator, will "Preserve EEPROM memory through the Chip Erase cycle".? This would be useful not to loose the myriad of settings when upgrading those two units.? Alan's summary shows one used a High Fuse setting of 0xDF.? According to the Fuse calculator that will change the Boot Flash section size and the Boot start address.? I tried this High Fuse setting in an already programmed U3S, it appeared to have no effect but YMMV.? The notes at the bottom of the on-line Fuse calculator page are worth reading, especially the 6th one down which relates to the High Fuse settings and the trouble that a wrong one will cause. An Extended Fuse setting of 0xFC is what came in my original U3S.? The Extended Fuse controls the 'Brown-out detection' setting used by the U3S, and 0xFC sets that level to 4.3V.? A setting of 0xFF will disable that function.? This will not prevent the U3S from working, but could result in erroneous operation with low supply voltage or after a brief dip in the supply voltage.? I suspect that the forum posts related to 'my U3S resets when it starts to transmit' is the Brown-out detection in operation ... their weak power supply's output voltage dips below 4.3V with the extra current required as the U3S starts to TX and the brown-out detection stops the ATmega328(p) ... the U3S stops transmitting ... the supply current drawn reduces ... the supply voltage comes back above 4.3V ... and the U3S starts up again.? The VFO and QCX also use a 5V supply on the ATmega328(p) so it comes as no surprise that the Extended Fuse setting in those units is also 0xFC.? But it should be noted that a verification error can appear with Extended Fuse of 0xFC while using some programming software ... I use eXtreme Burner and don't see that effect, so ??? ? Apparently an Extended Fuse of 0x04 works for the effected programming software, and the on-line Fuse calculator indicates that this value will also invoke the required 4.3V Brown-out detection level. 73, Bob? ZL1RS P.S.? I'm not an AVR expert, just commenting on what I see when experimenting here on the work bench. On 4/01/2018 1:56 AM, Paul Parks wrote:
|
No joy, I cannot get the programmers, either Extreme Burner orChris, Is there the option to ignore the chip identity? Not with a 328 but Averdudess would not verify a chip's ID but programmed OK when I ticked the "Ignore" button. (I normally use Khazama with a USBASP, not sure if that can ignore an ID check.) 73 Alan G4ZFQ |
Hello Cecil,
toggle quoted message
Show quoted text
The new chips came direct from RS Components and are Atmel branded, as genuine as I can possibly ascertain. I am trying to programme then *INSIDE* a working U3S, powered up, using the six pin header. Ican update factory QRP Labs chips, but still unable to even get two hardware devices to even see a new chip. I have downloaded AVRDude and AVRdudess and will have another go. is there any benefit from having the chip isolated from the U3S on a breadboard? Thanks everyone, it's now become a challenge to do this :) Wednesday, January 3, 2018 If you can see and program the provided chip but can't program a new one The ATTINY 85 chips I program for use with SoftRocks requires that the I would try getting a brand new chip from a reliable source. -- 2E0ILY Best regards, Chris mailto:chris@... --
Best regards, Chris Wilson (2E0ILY) |
to navigate to use esc to dismiss