Re: Buck converter vs power from Pi question... #12vsupply


Jack Spitznagel
 

Hi Bryan,

 

Thanks… but this board is the “hulk” of a board you replaced back a while after the problem occurred. The GPS chip and chips around it got very hot after a cat had knocked the GPS puck off the window casing ledge. I noticed after it was too late that the puck rattles – perhaps something got knocked loose. I bought a replacement puck from you, which did not stop the issue.

 

I pulled the original Pi3B and the old board out the other day and found that the pi will boot (no hat), but will eventually become heat unstable. It would boot (hat on) while powered from the pi, but not when powered from the DRAWS hat. When it booted (with a configured SD card), the GPS is “unreachable” with the new puck attached. (Puck works on another unit.) That section of the board around the GPS chip warms up again, but not like it did for the failure.

 

I’ll be happy to send in the board and if we can figure out whether a refurb is possible, you are welcome to return it to service and resell if you think it is worth the trouble. I have two working systems now so I really don’t know what I would do with a third.

 

Jack Spitznagel – KD4IZ

Cell: (410) 207-5123

Biomedical Consulting Services

 

 

 

From: udrc@nw-digital-radio.groups.io <udrc@nw-digital-radio.groups.io> On Behalf Of Bryan Hoyer
Sent: Sunday, January 31, 2021 19:34
To: udrc@nw-digital-radio.groups.io
Subject: Re: [draws and udrc] Buck converter vs power from Pi question... #12vsupply

 

Hey Jack,

 

We have had zero failures in the Buck Converter, except for unfused reverse voltage blowing the protection diode. If you power the Pi separately we use the recommended isolation circuit. In fact, I have hot switched between the 2 supplies without problems.

 

All ICs  on DRAWS, including the GPS, operate on the 3.3V supply, which is provided by the Pi. We have had zero GPS failures, we did have an I2C to serial failure on 1 unit. The GPS receiver is on the I2C serial port.

 

Email me at sales@... and I’ll arrange for a refurb unit for you.

 

73,

Bryan K7UDR



On Jan 24, 2021, at 11:56 AM, Basil Gunn <basil@...> wrote:

 


Hey Jack,

Bryan should comment on this:

after my experience with one of the early experience boards where the
buck converter and the GPS chip got very hot and apparently one or
both failed. Happy to send the hat back for a post-mortem


I have been running multiple different versions of DRAWS boards on 12V
supplies for over a year & I have not seen your symptom. However, I have
a friend who has seen 2 of his RPi/DRAWS units experience much elevated
temperatures. I actually fixed one of his units by updating his kernel &
swapping out one of the heat sinks.

1. I power my RPi/DRAWS units with Alinco DM-330MV power supplies set to
13.8V except for the one I run in my truck.

2. All heat sinks for the RPi are not equal. My experience is that the
super cheap ones (they are all pretty cheap) have inferior thermally
conductive paste.

3. If you have an RPi 4 be sure to run the latest Raspberry Pi OS
package update as that also contains current firmware for the RPi. If
you are seeing temperatures > 80C then you need to update your RPi
firmware.

4. If you are interested in tracking your temperature, graph it out. I
have a repository that plots RPi core temp, ambient temp & CPU activity:

https://github.com/n7nix/rpi-temp-graph

Other:
I supply a script on the NWDR image that displays current RPi core temp
& throttled state. The following script was run on: 
Pi 4 Model B, Rev 1.1, 4GB mem with NO fan.

pi_throt.sh

temp=53.5'C
throttled=0x0

kd4iz de n7nix

Jack Spitznagel <
kd4iz@...> writes:


John, Bryan, and Basil,

Take a look at the close up in
https://nw-digital-radio.groups.io/g/udrc/message/5393. In that set up
am not using 12V through the DRAWS buck converter, rather powering the
whole beast from the Pi4 USB 3. I am still a bit nervous after my
experience with one of the early experience boards where the buck
converter and the GPS chip got very hot and apparently one or both
failed. Happy to send the hat back for a post-mortem (on my dime) if
you think it would be helpful to have a failure analysis. Frankly I
would love to fix it if I can identify the bad chips, but not exactly
sure where to start looking to analyze...

Jack Spitznagel – KD4IZ

Science River LLC

Biomedical Consulting Services




 

Join udrc@nw-digital-radio.groups.io to automatically receive all group messages.