|
Post by jhornbr225 on Dec 26, 2017 20:32:36 GMT -5
Well, I took the plunge tonight and flashed the BIOS in my LattePanda to the auto-power on. I followed the instructions at link. I got the BIOS from the LattePanda Github at link. You have to read the sticker on the side of the Panda to figure out your version. It took a little faith as I have the Panda disconnected from my display at the moment, as I have decided to move the LattePanda into the Double DIN case, and install it in my Tacoma pickup. I hooked up the Panda to an HDMI screen, which stayed blank during the entire process. I let it run for about 20 minutes to be sure, the instructions say it takes about 5 minutes. Once I restarted the Panda, it did some sort of Windows update which took a few minutes, then it booted normally and went right into Driveline. I then did a shutdown and disconnected the power. I plugged in the power and started the stopwatch. It was 1:32 until I got the "Good Evening" greeting, but there was no input from me to get it to boot. So I guess that's a win, even if the boot is kinda slow. No more messing around with trying to get a servo to press the power button. Inside the double DIN I have also decided against using the built in Arduino on the LattePanda, as it takes 8 seconds to boot, and applying power to the LattePanda starts the boot process anyway. I'll use an Arduino Nano inside the Double DIN to control the Power Supply, Amp mute, Digital Volume Control, etc. Arduino's typically boot almost instantly and start running your code. I'll be using an ADS Instant FM RDX-155 inside the Double DIN as well. Some reading on the web suggested adding a fat ground wire from the USB Ground side to the antenna Ground side as they appear to be isolated from one another. Another mod that was suggested was to put it in a grounded case to shield it from interference. I figured that it can't hurt, so I ordered a small Aluminium case. It's gonna be tight after I remove the USB plug and replace it with a short cable, and do the antenna mod. There are a couple raised portions in the case to hold circuit boards that I'll have to grind off, and I'll have to grind a few thousandths of an inch off each side of the plastic case of the Instant FM to get it to fit down in the case. After that it should fit OK.
|
|
|
Post by redheadedrod on Dec 27, 2017 10:55:42 GMT -5
Any idea if those units can run Android well or not?
|
|
|
Post by jhornbr225 on Dec 28, 2017 11:39:42 GMT -5
A quick search on the LattePanda site shows that some people have, but complain about the usual driver support and BIOS suckiness. I myself have no interest in Android as my $29 Amazon Fire tablet runs it, and I'm not the biggest fan. I find it best for phones and media consumption. I feel that there is enough cheap hardware out there that runs Android, I'm not going to muck up my WIN10 Lattepanda to try to run it.
It could also be that I'm just not that familiar with it. My brother bought a used Ford Ranger lately that had un Umpkin Android Head unit. I can't say that it was bad, but it's tough to use. It seems to think it's still in China, and there was no way to change that. We finally got it turned on to the Wifi at the house, but it really didn't help us any. Like I said, my dislike of it could be just from a lack of familiarity.
|
|
|
Post by jhornbr225 on Jan 8, 2018 16:32:31 GMT -5
Finally got my SOP8 to DIP8 adapters from China today. I ordered them 11/29/17. Since I'll be using 1 chip each for both the Accord PC and the Tacoma PC, I soldered up a couple chips. They are a little small, but not too bad. Now to build up the reference circuit, and program an Arduino Nano, and see if they work.
|
|
|
Post by jhornbr225 on Jan 9, 2018 20:38:01 GMT -5
Well, I spent the day working on the electronic volume control. After quite a bit of trial and error, i got it figured out. I was basing my work on an example I found that used the same chip, along with an IR remote, and a couple seven segment displays. I took out the stuff I didn't need in order to simplify the code. No matter what I did, I could not get any sound to come out. Also it seemed to be locking up the I2C bus. I then took a good look at the chip I soldered. Figures, I soldered it on backwards. After reversing the chip around, it works! It didn't even damage the chip.
Those of you with sharp eyes can see the mistake in the magnified picture of the chip. You can see that the corner with the dot on the SOP8 chip has a trace that goes to, you guessed it, Pin 5 of the DIP8 adapter.
Now I'm ready to add the rotary encoder to control the volume, and the pushbutton on the rotary encoder will be my power button for the computer!
|
|
|
Post by jhornbr225 on Jan 12, 2018 2:43:14 GMT -5
I spent the morning working on adding the rotary encoder. That part of the code went in very easy. So then I started working on all the I/O to control the Enable line on the Power Supply, The Amplifier Mute, Control of a power relay, etc. Tonight I finally got that all nailed down and tested so that it works the way I like it.
I then went to add in the keyboard commands to shut down the PC under certain conditions. I figured since there was a keyboard example in the basics section of the Arduino tutorials, that it would work. Wrong! A little bit of research shows that it only works on Arduinos that have native USB in the chip, so not any boards with a serial to USB chip on them. The Nano that I am using has such a USB to Serial chip. It looks like I need an Arduino with a 32u4 processor, not a 328P. Back to Ebay...
The Nanos I bought were like $2 each, The Pro Micros that I ordered are just under $5 each. Now back to waiting for parts from China. Good thing I'm not in a hurry. Also, I believe my code should port directly to the new board with no changes.
|
|
|
Post by jhornbr225 on Feb 6, 2018 16:10:28 GMT -5
Ok, the Arduino Pro Micro's arrived and I've got one programmed the way I want. I've got it so that a quick push (50 ms) of the rotary encoder toggles the mute of the PT2257 chip. A long push (3 seconds) will either turn on or off the power to the PC. If you try and turn off the power before a specified time after booting up, it will set a flag, wait till that time expires, then request a shutdown. Shutting the key off for more than 5 seconds acts like a long push to cause the PC to shut down. It also controls the power relay for the whole thing and the hardware mute for the 4 channel amp. It also has two diode isolated power inputs. One is Ignition, and the other is the output of the relay so that once turned on, it all stays powered up until the PC is shutdown and the Pro Micro shuts the relay off. The only thing that needs soldered in yet is the audio input and output, which will go in the corner under the 8 pin PT2257 chip, in that empty space under the 4 blue capacitors.
|
|
|
Post by jhornbr225 on Feb 8, 2018 12:18:56 GMT -5
Spent the day yesterday manufacturing. I got the circuit board mounted in the case, and drilled the holes in the side for the LM2678 switcher and the MBR735 Diode, both TO-220 packages. With them bolted into place I then soldered a couple connections on each to hold them in place so then I could remove the board and do the final soldering. I then cut up a metal soup can as a shield between the pre-filter and the switcher board. I'm going to use some power supply binding posts as penetrations through the case, but they are too long, so a quick cut with the Dremel is required. Also the hole running through the binding post is too small, so I have to carefully drill them out for a #8 screw. Next steps are to make another small board for the pre-filter, that has a couple inductors and caps on it. I also need to cut the holes in the side for the fuseholders. Manufacturing is hard! Well, time-consuming at least. You find all sorts of little problems as you attempt to assemble things. And the smaller and more compact that thing is, the more problems you will have. Every little part is custom, and you have to find all sort of little screws, nuts, lockwashers...etc. So I have two more binding posts to install, the 12V in and the 5V out. I will use the hole in the side that I drilled for the shield as a ground lug. The green binding post is for the Power Supply enable. The fuse holder sitting on the side is to give me an idea of how big the pre-filter board can be.
|
|
|
Post by SNOtwistR on Feb 8, 2018 14:45:14 GMT -5
Looking good! SNO
|
|
|
Post by COMpulse on Feb 8, 2018 15:02:27 GMT -5
Man, some of you guys on here are really good at soldering.
Do you use a magnifying lens and clips to hold the board, or what? Fine-tip soldering iron I assume. Flux?
My hands aren't steady enough for this kind of soldering. Not even close. I soldered some capacitors into a dead LCD a few years ago and brought it back to life, but even that was sketchy.
When I was in school, my electronics teacher would say to me: "Yeager, you maggot, this solder job looks like the north end of south-bound horse. I should put you in a headlock and throw away the key. Do it over."
If mine had looked like these, I would have gotten an A+
|
|
|
Post by jhornbr225 on Feb 8, 2018 16:04:43 GMT -5
My eyes were never really that good, but as I age, I find that it is tough to see the small stuff. Last year on Amazon I picked up a Magnifying headset. Best $8.88 that I ever spent. It's got an aimable LED light that's nice and bright. it's got a total of three lenses. This product installs 3 Lenses multiplying power of lens:1.5x 3x 8.5x 10x. I'm not sure which lens is which. There is the primary lens, then a flip down one on the inside. Those both allow stereo vision. Then there is the single Loop for the right eye. That one gets pretty close, but you have to adjust the LED to the one side. The LED tilts up and down, and there is a little control to go left and right. This one is currently unavailable but there is another on that looks just like it, without the "Beileshi" text for $11.74. www.amazon.com/dp/B003UCODIA/ref=psdc_2236129011_t1_B018DLMHFI At the same time I bought a magnifying camera. It's kinda junky though, and would not recommend it. The focal length is very short, and the software is very buggy and it locks up. It does have a very high level of magnification though when you can get it to work. The pictures I took of my broken Lattepanda latch were done with that camera. Also the pic of the PT2257 chip a few posts back. For the headset, obviously the higher the magnification, the shorter the focal length, so at the highest magnification, if you were trying to solder, the iron would be very close to your face. Generally I just use the standard lens, or I flip down the inside one too. After I'm done soldering, and want to really inspect and look for bridges, I'll flip down the single eyepiece. I did put the smallest tip in my iron. A Weller Model WLC100, adjustable heat. They are about $40. As far as flux, I don't use any additional, other than what's in the solder. I do keep some sandpaper nearby, in case something I want to solder is a little oxidized. Or if a wire I want to solder is oxidized, I either cut down to fresh shiny wire, or scrape it with a razor blade. I'm currently using some Kester "44" Rosin Core .80mm. A pound of solder goes a long way. I also have a 5 pound spool that I quit using a couple years ago. I had gotten it for free back in 1990 because it was junk. Every time I wanted to use some, I had to scrape off the outer layer that was oxidized. And even then, it didn't solder very well. Finally one day I decided to try a fresh spool, and what a difference. That stuff from 1990 was really junk. In the picture, the tip is cold and a little oxidized. A little heat, a clean sponge, and a good tinning, and she cleans right up. I also have been doing this a couple years. High School Vocational Electronics was in 1987....
|
|
|
Post by jhornbr225 on Feb 9, 2018 21:22:21 GMT -5
|
|
|
Post by jhornbr225 on Feb 13, 2018 20:05:49 GMT -5
I finished up the power supply yesterday. Really, there are no ground wires inside the case. Every ground is bonded to the case with star washers. The only ground wires that exist go from a lug with a star washer to each of the caps tied directly to the 12VDC input lug, and the 5VDC output lug. The shield for the feedback wire is also soldered to ground on the switcher board. Both circuit boards have a star washer under each mounting screw. The switcher board is double sided with large ground planes, and the filter board is a single sided Vero strip board, using 12 strips. The center 4 strips are cut into sections for each of the filter's inductor and caps. I kept the inductors as far apart as possible. Both outer sets of 4 strips are ground, soldered together with strips of wire running the length of the board, then bonded with star washers to the case. Today I brought it in the house and tested it. I'm a little disappointed. At 3.5A of load, the voltage sags to 4.89VDC. Unloaded it's 5.12VDC. I used the 330Ω resistor in the feedback loop. A little investigating revealed that at the 3.5A draw, the output at the switcher board is actually goes up to 5.14VDC. I'm losing the 0.25 Volt across the 5V Filter and the fuse, about 12mV each. It might just be my OCD, but I feel that I have 3 options: 1. Leave it alone, and live with it. I probably won't draw 3.5A, and 4.89VDC at that level is close enough. 2. Increase the value of the feedback resistor to 560Ω. This should increase the output enough to compensate for the loss. The problem with that is, at lower current draw, the voltage goes up, and I probably don't want to exceed 5.25VDC. 3. Change the location of taking the feedback to right at the exit lug of the power supply. The switcher circuit would them compensate for the losses of the filter and the fuse. But, this may introduce switching noise into the output. I am currently using a shielded cable with the shield grounded at the resistor end. If I was going to move it to the output lug, I imagine I should ground that shield close to the output lug, not at the switcher IC. Although, I do recall that signals in shielded cables should have the shield grounded at the receiving end, and the switcher IC is the reciever. I have not hooked a scope up to it yet to look at the output, I should probably do that. Mickz, if you're out there, (Or anybody else for that matter), what is your opinion?
|
|
|
Post by jhornbr225 on Feb 13, 2018 20:18:36 GMT -5
Another idea that I have that maybe some of you PC guru's can comment on.
Seeing that this Beelink and the LattPanda are booting quite slow, would I gain anything by taking an image of the current built in EMMC drive, and then flashing that image to an SSD attached to a USB3 to SATA converter, then setting my boot order to boot from the USB3? I have not checked on the Panda yet, but I know I can get into the BIOS of the Beelink and change the boot order.
I don't want to have to purchase another copy of Windows, as the Panda and the Beelink both came with a supposedly legal copy of WIN10.
If so, what (hopefully free) software could I use to take the image? Are there any gotcha's to the process? I want to see if the boot times are better with a 256GB Samsung SSD that I have that I could attach to the USB3 port. But I would not want to lose the ability to go back to the built in drive, if doing this does not lower the boot times.
I'm used to the old days where I had installation media for WINXP. But when you buy a computer now, at least from China, you don't get any installation media.
|
|
|
Post by redheadedrod on Feb 14, 2018 8:59:16 GMT -5
If it is in EMMC I can't imagine it would be any faster on an SSD as the interface should be significantly faster for EMMC ram. You can always boot from Linux and image the drive assuming you can boot from a flash drive. I also have a combo hard drive/ dvd player I have used to image computers. I run my imaging software from the dvd player and copy the image to the hd
|
|