SOP8 to DIP8 Adapter Socket for programming?

[kalel]

I was wondering on the cheapest way to use easily programmable MCUs in small experimental projects.


E.g. ATTINY13A SSU comes for ~2.75 at 10pcs with shipping from China. DIP8 packages are a lot more expensive (~$10).


There's something like this for easy programming ($2, probably available cheaper):





But it needs a DIP8 package. Would something like this fill the void?
It's called "SOIC8 SOP8 to DIP8 EZ Programmer Adapter Socket Converter Module" and available for <$1






I'm just curious whether or not that would allow me to program directly those SMT packages and then probably use something like this to actually use it in the project with protoboards and breadboards:





It's still a lot more work than using DIP8 directly, but if it's much cheaper and I didn't get something wrong (it could work using those adapters), might be the way to go (and practice some soldering along the way).


Edit: If soldered using the last image adapter, I'm not sure the converter from middle image is even needed.

[CJay]

Why not just solder the SO-8 package to the DIP adapter and then program it?

[Rbastler]

There are these Sot8 clamps:
https://www.aliexpress.com/item/Free-Shipping-update-version-SOIC8-SOP8-Test-Clip-For-EEPROM-93CXX-25CXX-24CXX-in-circuit-programming/32358875720.html
I bought one to change the eeprom content from my VC921.

edit: typo

[kalel]

Quote from: CJay on Today at 05:47:20 AM

Why not just solder the SO-8 package to the DIP adapter and then program it?

I think that is a good point I realized myself after writing the initial post, there might not be any advantage.


The clamp might be more useful.

[KL27x]

These sockets are almost worthless for programming chips.

If you are interested in 8 pin SOIC MCU, specifically, you can buy a clip to program these chips. It looks like a little like a clothes pin. They make them for DIP and SOIC. For the life of me I can't remember what they are called or how to search for them on Mouser/Digi, but I know I bought mine on Mouser.

They clip from the top of the chip and they work on loose chips as well as chips that are already soldered onto a pcb (provided that there's no pin conflicts). And for SOIC, they work on soldered chips provided that there aren't huge solder blobs/spikes on the pins.

If you have ICSP pins available with no conflicts, you can also put a programmer header onto your circuit board. For 8 pin chip, you can even make small mcu board with 8 pin header for connecting to the circuit and another 6 pin header for  programming ICSP where possible (or unplug the mcu board for programming if you have a conflict). I have a little MCU board like this for 2 of the 8 pin PICs I have traditionally used, frequently. This is because for 8 pin chips, I will often not have ability to ICSP, and adding ICSP header and putting jumpers on every ICSP conflict in the circuit is somewhat cumbersome, if it's not going to be in the final board. It's easier to just socket the entire chip board to the dev circuit, which is only 2 or 3 more pins.

[KL27x]

Here's a pic of some stuff, left to right:

Fly probe ICSP adapter
DIP chip clip
Soic chip clip
MCU dev board (internal clock, only)
MCU dev board clipping into ICSP cable and into 8 pin logic probe (sometimes I can dev and debug a simple firmware with just the logic probe; I can also plug in 8 pin logic analyzer into the bottom of the probe).

You can also just hold a loose SOIC chip down over a breakout board wired to your programmer, in a pinch.

Most of these things are relics that I haven't used in 10 years, except for the pogo pin adapter on the right. In many cases it is easier to just use a larger MCU, where necessary, to have ICSP pins available to a pad header for interface with such an adapter.

[stj]

use these.

http://www.ebay.co.uk/itm/272494288750

[macboy]

These sockets are almost worthless for programming chips.

Because ....

I disagree. It's easy to pop the chip in, program it, and pop it out. The one I have (came with a TL866 kit) makes good contact, and I've use it for I2C EEPROMs, SPI FLASH, and micros. The big disadvantage, which makes it less useful but not useless, is that it can only be used on a bare chip, before soldering to the board. For development where you need to change the data frequently, it is not the tool for the job.

The DIP-to-SOP adapter is a good choice if you might use the chip on a solderless breadboard, or prototype PCB with 0.1"  holes. The combination will cost less than the DIP packaged chip and is equivalent.  I always use machined-pin (round pin) headers on these, not the 0.025" square post type. The round pins work very nicely in breadboards and IC sockets. The square posts are too big for either and can damage whatever you plug them into... unless you are just plugging the proper mating connectors into them (so-called "Dupont" jumper wires) or soldering them.

Adding ICSP connections to the board is always an option. Probably a good one. This is where you can use those 0.025" square post headers.

I have no opinion on the quality of the cheap chineseum SOIC clips... how well they grab or how durable the contacts are, etc.  I do have some nice Tek branded ones that work a treat.

[kalel]

The DIP-to-SOP adapter is a good choice if you might use the chip on a solderless breadboard, or prototype PCB with 0.1"  holes. The combination will cost less than the DIP packaged chip and is equivalent.

That is what I was wondering mostly regarding those adapters. I am more likely to work with protoboards or other improvised solutions than affording PCBs for every project.

[stj]

if your developing, you really need a programming header on the board.

the sockets are good for making really compact stuff where your code is finalised, and putting initial data into eeproms or serial flash.
(also handy for laptop repair and multimeter modding!!  )

[KL27x]

I disagree. It's easy to pop the chip in, program it, and pop it out. The one I have (came with a TL866 kit) makes good contact, and I've use it for I2C EEPROMs, SPI FLASH, and micros. The big disadvantage, which makes it less useful but not useless, is that it can only be used on a bare chip, before soldering to the board. For development where you need to change the data frequently, it is not the tool for the job.

Well, I erred. I tend to think in binary. For coding/dev an 8 pin micro, it is almost useless. For flashing more than a very small number of chips (in a highly labor intensive fashion), it would be useless. For flashing just a few bare chips, here and there, it would be useful... particularly for large ROMs.

[Howardlong]

I disagree. It's easy to pop the chip in, program it, and pop it out. The one I have (came with a TL866 kit) makes good contact, and I've use it for I2C EEPROMs, SPI FLASH, and micros. The big disadvantage, which makes it less useful but not useless, is that it can only be used on a bare chip, before soldering to the board. For development where you need to change the data frequently, it is not the tool for the job.

Well, I erred. I tend to think in binary. For coding/dev an 8 pin micro, it is almost useless. For flashing more than a very small number of chips (in a highly labor intensive fashion), it would be useless. For flashing just a few bare chips, here and there, it would be useful... particularly for large ROMs.

It depends. I’ve programmed a thousand in one session (PIC32 QFN28 but in a similar ZIF contraption). It took about three hours. If you need them quick, because your line is stopped, sometimes it’s the only way. A ZIF socket for a QFN28 is a couple of orders magnitude more expensive than those 99c SOIC8 units though!

Except in exceptional circumstances, I always provide a programming/debug port on my designs, and this occasion was no exception, however the assembler had no means of using it, or training or documentation to do so. It was faster to give them the devices I preprogrammed myself, and dropped off later in the day form the line to get going again.

[KL27x]

I’ve programmed a thousand in one session (PIC32 QFN28 but in a similar ZIF contraption). It took about three hours.

That's only 10.8 seconds per chip. Depending on program length, flashing could take longer than that! Anyway you cut it, that is excellent output and probably a very long and tedious 3 hours!  Did you make any special tools/trays to aide this process? Vacuum pickup tool essential, I presume? Did you have 2 set up so you could unload/load while the other was flashing? That's excellent overall time from taking bare chips in tube and getting flashed chips back into the tube, even if the flash took only 2 seconds!

sometimes it’s the only way

Maybe for you.

[Howardlong]

I’ve programmed a thousand in one session (PIC32 QFN28 but in a similar ZIF contraption). It took about three hours.

That's only 10.8 seconds per chip. Depending on program length, flashing could take longer than that! Anyway you cut it, that is excellent output and probably a very long and tedious 3 hours!  Did you make any special tools/trays to aide this process? Vacuum pickup tool essential, I presume? Did you have 2 set up so you could unload/load while the other was flashing? That's excellent overall time from taking bare chips in tube and getting flashed chips back into the tube, even if the flash took only 2 seconds!

sometimes it’s the only way

Maybe for you.

It was tedious, but then there are many jobs in one-man-band manufacturing and distribution which are like that, you just put the radio on and knuckle down. Almost always the thought of doing such repetitive and seemingly mundane tasks is worse than actually doing it. You usually fine tune your technique over time to become faster and faster: improving the process is part of the challenge and what keeps you sane.

I made a jig for the ZIF socket, and have a foot switch to start the combined programming/verify procedure. I place the parts one at a time in the ZIF with tweezers, but rather than closing the ZIF, I just apply a little pressure on the device down firmly but gently onto the ZIF contacts. This is far faster than having to open and close the ZIF door. I do a tube at a time, 65 or so at a time, and after each tube has been programmed I mark each chip with a silver marker pen, and put back in the tube.

Programming/verify is only 5s per chip, and another few seconds per chip for my manual tweezer, tube decanting and marker pen skills.

Most of the time I am able to plan and get Microchip to program them for me, but on this occasion they had no stock so I had to scrounge various compatible SKUs from a number of sources and program them myself.

Sourcing parts in volume was particularly bad around the time of the tsunami in 2012, it took about a year to settle down.

[KL27x]

Most of the time I am able to plan and get Microchip to program them for me,

I have never had to do that. I routinely flash my own up into the thousands, but i use a custom diy solution.  For 1000 chips u have to pay setup fee of 50.00, off the bat, right? Plus maybe 4 cents per..... up to 90 bucks already.

[Howardlong]

Most of the time I am able to plan and get Microchip to program them for me,

I have never had to do that. I routinely flash my own up into the thousands, but i use a custom diy solution.  For 1000 chips u have to pay setup fee of 50.00, off the bat, right? Plus maybe 4 cents per..... up to 90 bucks already.

There’s no setup fee once you are happy with your program. There is a one off programming fee for each test run. Once you have known working code it’s per chip. I find if you have a known working .hex file, there’s no need for a test run.

Whether the per unit programming cost is worth it to you depends on how you value your time. You could also ask why you get an assembler to place the parts for you when you could do that yourself (I’ve been theor too!).