[SOLVED]How do you label ATTiny's?

Go To Last Post
17 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi all,

For my LiFePO4 cell management I am flashing a bunch of ATTiny85's with different millivolt correction factors to account for the 5% tolerance of the voltage divider. So I need to label them -50, +50, +100, etc.

All the colored felt-tip pens I have tried make dingy lines that are hard to read, very similar to the existing chip markings! Scratching works but can't be erased. Sticky label/rubber cement is messy and will probably fall off over time. Velcro attachments seem like tremendous overkill.

I don't have any nail polish but will buy an assortment of colors if that is the best solution. I'd really like to write on them; is there some fine-point erasable grease pencil designed for IC marking?

Last Edited: Mon. Sep 7, 2009 - 07:48 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I think nail polish or model paint is the best solution. You could use the same colours as used on resistors to code the offset.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

dak664 wrote:
I don't have any nail polish but will buy an assortment of colors if that is the best solution. I'd really like to write on them; is there some fine-point erasable grease pencil designed for IC marking?

For temporary markings I use labels, either printed with info, or simply different colour dots. If I need something slightly more permanent, but still removable, a white grease pencil works. It's too fat to really write, but ticks & bars work pretty effectively.

Writing code is like having sex.... make one little mistake, and you're supporting it for life.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you use a pencil to writing in IC, you could erase it after...
Sometimes I write on a protoboard or a board if I don't want to write on the IC..

Regards,
Brunomusw

Regards,

Bruno Muswieck

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Paint pens designed for marking metal would work. Different colors and tip sizees are available. Here is a selection page from Enco. Harbor Freight may have something similar.

http://www.use-enco.com/CGI/INSRAR?PMSECT=2003287

Rick

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

For semi-permanent quick marking, we've been using the Sharpie markers in the "metallic" silver (grey) color to mark on dark surfaces.

For a permanent color dot/marking, use a paint marker as suggested. In front of me are Sanford "uniPaint opaque, oil-base paint marker". We use the fine point. Any office supply outlet should have them.
http://www.officedepot.com/catal...

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Inexpensive monochrome laser printers can print very small text, digits, or symbols on standard paper. Use an image program like PhotoShop, PaintShop Pro, or GIMP to prepare a range of the labels that would be placed on the IC top. Then shrink the image as much as possible while preserving the ability to read the font and print it. Use a pair of sharp, small scissors to cut out the label and then use either hot glue or a tiny amount of superglue to affix the tiny piece of paper to the IC top surface. Or get a young child (or company intern) to do it.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Perhaps this is heretical, but why bother compensating for the resistors in the first place? Use 1% metal film resistors instead - they aren't any more expensive than cheap and crappy carbon resistors.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes, inkjet labels secured with clear postal tape are my preferred labeling method when there is sufficient margin for adhesion. I jot down the needed labels and print them when I have a sheetful.

Nail polish makes excessively large dots for me, I'd be lucky to get 6 on an 8 pin DIP. Maybe I need more practice :)

In this case I need to record calibration and cell ID on a one-off basis, +50 HAJ for example. Preprinting all the possible combinations would be tedious and wasteful of supplies. So far a #3 rapidograph pen with blue-black ink works the best, it is readable with a flashlight at an angle and erases with when dabbed with a paper towel moistened with isopropyl alcohol. I will explore other colors, and maybe spray paint a batch flat white for contrast if I can find a good way of masking the pins.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

AllPraiseBob wrote:
Perhaps this is heretical, but why bother compensating for the resistors in the first place? Use 1% metal film resistors instead - they aren't any more expensive than cheap and crappy carbon resistors.

Well, I have a lot of cheap and crappy carbon resistors to use up. I am using two of the '85 pullups as the high side of the divider and don't know what variation to expect from them. I am prototyping and don't know what value I will ultimately use. And the local Radio Shack doesn't stock metal film resistors.

So far (6 units) the variations have been within 110 millivolts out of 3.6 volts. I need <50 millivolt accuracy.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Here is how we approached similar requirements: In the unit test rig, have "good" signals--voltage, timing--available. Put the unit into calibration mode to check the timing, etc. levels for this particular unit and then store in EEPROM to use during operation.

This can often be as simple as using an accurate (or known) Vcc level in the test rig.

Depending on the characteristics, I use either single-point calibration knowing that (0,0) is in fact that or close enough--no offset--but in the general case two-point calibration and derive rise/run and intercept from that. In practice that works well for the linear signals I commonly encounter in my industrial apps.

Lee

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I am trying to stay away from EEPROM since the power supply is one cell in a battery that could brownout on the shelf or under load, and it is my impression that EEPROM reads are more sensitive to brownout (?). I might use the EEPROM to store the number of charges at the end of a charge cycle, but otherwise I'd like each startup after an 8 second sleep to be bulletproof.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Writing to the EEPROM can be a problem, but not reading.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you're worried about EEPROM, then have the calibration utility program the data into flash.

Writing code is like having sex.... make one little mistake, and you're supporting it for life.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

Writing to the EEPROM can be a problem, but not reading.

Well, it can--when the supply voltage gets lower than what the AVR is spec'ed for, the CPU may keep running but all bets are off on correct operation.

Remember that you only need to read the cal values to SRAM locations at startup. ;)

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

OK, I will patch in a representative 5.6K resistor, set the STK500 target voltage to 3.5 volts, and have the '85 automatically write a correction factor to EEPROM after leaving programming mode. I can then reenter programming mode to read it and see what the variation is. Will report back on that.

Normally in my circuit, every 8 seconds an LED flashes 1-5 times to signal battery voltage of 3.1-3.5 volts. Above that the '85 doesn't sleep but activates one or more current bypass shunts to prevent overcharge. As a diagnostic, if waked using the reset pin I have it flashing the tenths followed by hundredths of a volt. Maybe I could also use the reset pin to optionally trigger a calibration mode at a certain voltage...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

dak664 wrote:
OK, I will patch in a representative 5.6K resistor, set the STK500 target voltage to 3.5 volts, and have the '85 automatically write a correction factor to EEPROM after leaving programming mode. I can then reenter programming mode to read it and see what the variation is. Will report back on that.

That worked very well! I selected 12 5.65-5.85 KOhm resistors for the sense circuits with a 5.75K on the STK500 and stored an additive offset in EEPROM on the first run. Those corrections varied by 84 millivolts across 12 MCUs.

Initially I thought it wasn't going to work since one of the divider pins was jumpered to xt1 which drove it high. But programming went fine without that jumper since the chip was using the internal clock, so it could be disconnected.

Along the way I found the solution to the labelling problem, wrapping strips of sticky label tape around the chip, which also forms a pull tab for hot removal and insertion. The paper is strong enough for a few cycles, but wrapping a clear tape strip over that protects the writing and allows a good hard pull.

Here's a picture of the battery management board showing all the labels. The tight layout is not as goofy as it appears since ground for each '85 is Vcc for the next. The c program is attached.

Thanks for all the suggestions.

Attachment(s):