Hi Mark
The short on the "B+" side of J4..? The B+ goes to pin 2 of the FETs, all the associated 10nF bypass caps and the two 470uF caps.
Possibly a etching fault on the PCB.? Or one of the bypass caps is shorted.? Maybe one of the 470uF caps wrong way round although it may not show a short unless damaged by application of reversion polarity initially. TP4056 put in wrong way round?

I have seen a couple of the FET's on a build overheat. No known cause but replaced and OK.? Your case melts might be? a bad sign. I recommend a hot air gun to replace any smd part by the way. By the nature of tight space, the tracks in that area are pretty narrow and easy to damage.
Afraid I haven't checked the data sheet for the TP part too closely but would make sense it current limits as it would be part of the battery charging.

glenn

开云体育

Dave

The ATU-10 is a 'QRP' version of the ATU-100. Power is limited to about 10W and it uses latching relays for low consumption.? It fits into a small extruded case, 100 x 71mm available from usual places.

The Schematic link is on N7DDC's Github page.

Glenn

vk3pe

Just built my second AT-100 but display is upside down in the case.? Flipping it over does not work, bottom line is not visible.? Is there a way to flip it over?
Thank you? 73? Emory? WM3M

Glenn,

Thanks for the suggestions. Double checked the FET's under a USB microscope, and all are correctly oriented.

There is a direct short from the B+ side of J4 to ground.

Am I reading the schematic correctly that B+ only goes to pin 2 of the fet's? If so, I assume that this means one of the FET's has an internal short from Power to Ground? If so, any suggestions on how to locate the bad one? There are a couple with case melts that probably happened when I put them in, obviously could start with them....

BTW: Apparently, the TP4056 has some sort of current overload protection. Fuse1 (PTC) is *not* going high (still see usb power to Vcc pin of TP4056. But the Bat output (pin 5) that should be 4.2, goes to zero when I bridge j4. (No batteries in my current configuration).

73
Mark
AJ6CU

was the atu discussed on this list built to the atu100 schematic or is there a difference.
also what is the atu10?

dave k7da

Glenn

thanks. The operation of that fuse explains what I saw. Will check the orientation of those chips. That and the adjacent caps were the area where I was most concerned about the quality of my work.?

73
Mark
AJ6CU

开云体育

In the process of incrementally assemblying my ATU-10. Got to the first step where it booted with the two pics programmed and installed. Moved on to putting in the relays.? I have *not* installed any of the toroids, swr, etc.? Figured I would see if I smoked anything when I powered up the relays.

So I installed j4 to power the relays...? Screen is blank. When I remove the jumper (power still on), display comes back and it works.

I assume this means that the power to the relays is shorted to ground through (fortunately) a resistance sufficient to prevent burning up the 3.3 regulator?

Checked the regulator with the jumper on and it is still at 3.3...

Expected or do I have a problem?

73
Mark
AJ6CU?

开云体育

No problem Mark. That catches many, including myself?

Unless you have a copy of the original compiler (Mikro-C at about US$300) then no way to re-compile it.

BUT i think a couple of people here maybe trying to move it to Arduino IDE platform?

Glenn

Normally the OLED should be on at power up, but after 30secs? it powers down to save power.? A long push should bring it up again.

Duh.... PB-1 must be installed to get to display.? All set (and embarrassed).

73
Mark

Now I am really looking like the noobie I am...? Apparently 0x3C i2C address is 0x78 (on the bus)....

Back to ground zero here.

Mark

Glen,

Recompiling looks hard. Any pointers on how to change the I2C address in the hex? Or alternatively, (probably better) a link to a OLED with a 0x78 address in the .91 size? They all seem to be 0x3c...

73
Mark

Glen,

Thanks for the suggestions. The progress bar did move across the dialog, couldn't see it hit the end as it was too fast.

But I think I found the issue...? Scanned my OLED's and they are at 0x3C. Took a peak at the oled_control.h and the address used there is 0x78.

Looks like I got to read up on that thread about recompilation...

73
Mark
AJ6CU

Mark,
2.? I have found that if you use one USB cable from pc to the USB connector on the ATU-10 to provide power, then there is no need to do your other points (Advanced etc)

ie You need two USB connections to the PC

• 1 to the USB socket on ATU which will provide power.
• 2 to your Pickit.

(If you don't have a battery installed make sure the output of the TP4056 is about 4.1V or so.? Occasionally I have seen this chip not provide any output voltage.?? If i try again, it's OK.)

Mark,

Did you see the progress bar while programming the PIC16F18877, move right across the box.??? It can happen that it only moves a small percentage of the bar graph indicating its not programmed. I have seen this.

I'm unsure why this happens sometimes.?? Try a few more times.
EDIT as mentioned in my build notes, this can also happen if the 2 data lines from the USB connector do not connect to the PIC. Or they are shorted together. Check with ohm meter against the Sch.

Its also possible of course your OLED is faulty.

?

glenn

vk3pe

?

Hi,

Using the ATU-10 v1.2 board and got to the point where both MCU's are apparently programmed.

1. Programmed the 1454 with the MPLAB IDE
2. Plugged in the board, it came up as a disk on my PC, and dragged and dropped the hex file for the 18877 to that disk. I got the copy progress bar, and afterwards I checked the size vs the original and they matched. Unplugged the board, and plugged it in and it mounted as a disk, and only the htm file was there (I assume this is normal...?)

Unplugged the board again, plugged in the OLED (yes i confirmed polarity) and nothing... No backlight, no nothing.?

Swapped OLED's and cables, still nothing.

Confirmed continuity between the 18877 and the OLED headers, and the sda, sdl, and Vcc checked. Plugged the board in without the oled connected, and there was ~4.3V on vcc..

Could this be an I2C address issue? Or perhaps the 18877 is dead?? Suggestions on where to go from here?

Thanks

73
Mark
AJ6CU

Looks like I solved it.? For future reference:

1. I installed MPLAB X IDS 5.45
2. I connected the usb to power the chip. Note that pin2 of the ICSP is NC. Since that is where VDD comes in, you need to provide power and you cannot use the programmer.
3. Enable Advanced settings
4. In the power section, select power from PICkit3 (yes, I know, this is a NC). and click the Powertarget from PICKit 3 box
5. I set the voltage to 3.35V
6. Under programming options, click use low voltage programming (for high voltage programming, need different set of Pins!)
7. I selected Apply Vdd before Vpp

Not sure if any or all of these are needed, but after using this configuration, I could erase and program the chip. And it verified on read!

On to install the other MCU and see if this really programmed things....

73
Mark
AJ6CU

Tried MPE IDE 5.4 and 6, they can report a bad id for the the 1454, but are able to erase it. When I try to program, they come up with an error that expected byte didn't match.

The IDE's both say unknown chip IDE. Have set the voltage to 3.3.? Also tried the V3 of standalone Pic3 programmer. It recognizes the PICkit3 but never sees the chip.

I am sure this has been discussed, before, any pointers? (Searched but did not find the a solution).? The instructions on carnut shows a pickit4...

73
Mark
AJ6CU

Hi, Marek.?
The ATU does not care what SWR at tuning process time. The minimum power should be 1 Watt and more for tuning .