Test board with multiple MCU?

[trastikata]

Hello,

I have some peripherals which I'd like to test with different PICs. I've designed one board with soldering places for different MCUs and all connections from the peripheral devices (except ICSP ports) are running parallel to each MCU.

The idea is to solder one PIC at the time and test the peripheral devices, then remove it and test the othe PIC type.

But I was thinking - if solder all the PICs but place a DIP switch at MCLR for each PIC and hold all PICs in reset state except the one I'd like to test - do you think this could work?   

[top204]

As I found out many years ago, even if a device is not powered up but is in parallel with other devices, a load is placed on the pin. Even if the parallel devices are both powered and one has its pins set to input, there is still a load on the pins.

This may stop devices seeing logic levels correctly and may effect quite a few peripherals, such as the USART and I2C and SPI, as I found out to my alarm with a project many years ago, and to one more recently where the USART was transmitting to more than 2 devices in parallel, and the loads on the devices, even when set as input, was too much for the USART TX.

[trastikata]

Thank you Les, then I'll have to solder and de-solder.

[See_Mos]

I am sure that this has been thought of before but in this situation I would have a motherboard for the peripherals and put the microcontroller chips on individual daughter boards that are then pin compatible with the motherboard.

[trastikata]

I am sure that this has been thought of before but in this situation I would have a motherboard for the peripherals and put the microcontroller chips on individual daughter boards that are then pin compatible with the motherboard.

Thank you See_Mos, this is only a test board for code development thus designing and ordering several boards for each device and peripheral would be too much .

I can live with soldering and de-soldering the MCU, it's easy with the heat gun and low-temp solder.

[charliecoutas]

I'll second the low temp solder. It was Trastikata who introduced me to bismuth/lead solder; I've used it many times since. Thanks!

[diebobo]

I'll second the low temp solder. It was Trastikata who introduced me to bismuth/lead solder; I've used it many times since. Thanks!

Does one has an ebay/Ali link perhaps ? Last time I searched everything was advertised as low melt, but only to find normal range melting temperature every time..

[RGV250]

Hi,
Am I missing something, as it is not for production why not sockets?

Bob

[trastikata]

@RGV250

The PICs have QFP package.

@diebobo

I just bought 50gr Bismuth from Ali. Then melted some in a pot with normal 60-40 solder in proportion 30%/70% Bismuth/Solder. 

This alloy is not eutectic, has lower mechanical strength and eats up the soldering iron tip quite fast but it melts around 130 degrees C and allows for easy prototyping and differential (selective) IC de-soldering.

[diebobo]

@diebobo

I just bought 50gr Bismuth from Ali. Then melted some in a pot with normal 60-40 solder in proportion 30%/70% Bismuth/Solder. 

This alloy is not eutectic, has lower mechanical strength and eats up the soldering iron tip quite fast but it melts around 130 degrees C and allows for easy prototyping and differential (selective) IC de-soldering.
[/quote]

Ah, that's an idea.. Got those Bismuth bricks laying around for some while to experiment making nice crystals. Making low melt solder instead .. Thanks for idea..

About your own problem, why not make an adapterboard for the MCU and get a socket on the mainboard.. Just switch adapterboards to get another cpu.. Different challenges arise with pin count and prob. with pinnames etc.. Me personaly just prefers to make another demoboard if the cpu layout is different then i have already, pcb's are cheap. Just design, order and do something else in the meantime...

[Dolci]

i have made a development board with a zip socket.

[RGV250]

Hi,
They do make QFP test sockets but I have just seen the price so not a good idea.
You could make an adaptor though, something like these.
https://cpc.farnell.com/proto-advantage/pa0096/adaptor-qfp-64-to-dip-64-0-5mm/dp/PC01828?srsltid=AfmBOorzOETTI09W3PrHeYpaD69lhuTrL5bWtcNQ1Uyg1EaO03keE-83F4Y

Bob

[John Lawton]

...I was thinking - if solder all the PICs but place a DIP switch at MCLR for each PIC and hold all PICs in reset state except the one I'd like to test - do you think this could work?   

My understand is that, yes, ports are in input mode when the device is in reset, but of course powered, otherwise the protection diodes would sink as much current as they could into the chip body.

I don't know whether this holds with all the peripherals stuffed into the devices these days, but it might work. I'd be interested to know the results if you try it.

John

[trastikata]

Making low melt solder instead .. Thanks for idea..

I am not sure where I found this ternary diagram, it was more than 10 years ago, showing the melting point of Bi-Pb-Sn alloys, but it is an interesting stuff.

Shows that you can get it bellow the boiling point of water. Which is fun because I actually tried it to disassemble PCBs in boiling water 

[trastikata]

Hi all,

some update about the original post ... just received the PCB last week and populated it recently.

The board is meant for TFT display library development and tests. It has pin headers for 3 different TFT displays - one with parallel and two with serial interface. The board has one test button and test LED.

There two MCU's - PIC18F46J50 and DSPIC33EP128MC504 where all displays are connected to in parallel, thus each MCU is connected to each display and each MCU has its own ICSP port. The both MCUs share the same crystal. Schematic is attached.

Naturally there can be only one display and one MCU active at the time. The other MCU is being held in RESET by pulling down the MCLR pin.

I was a bit skeptical and curious if this will work but after few days testing, everything works, which makes the library development for 16b and 8b MCU at the same time much easier.

[basparky]

Nice work! can we see the board in action?

[trastikata]

Nice work! can we the board in action?

Thank you.

For the second part I am not sure if I understand the question?

[basparky]

Nice work! can we the board in action?

Thank you.

For the second part I am not sure if I understand the question?

Perhaps some pictures or movies with a working display?

[trastikata]

Perhaps some pictures or movies with a working display?

I see, here's something

[charliecoutas]

Nice work Trastikata.
Charlie