555 Lesson Learned

Phil Pilgrim (PhiPi)

I'm partners in a sailboat (a T-bird 26). It's been having mold/moisture issues, and I've been tasked with creating a fan to circulate the air in the cabin. I have a couple old 12V computer fans, but they draw about 200 mA, which would ultimately drain the boat's battery ahead of the solar charger's ability to keep it charged. So I designed a 555 circuit to let the fan run for five minutes and rest for 20 minutes. It worked great on a solderless breadboard. But when I soldered up the circuit on a perfboard, it refused to start the fan.

It turns out that the problem was leakage current, due to residual flux between pins. I'm using a 3300 uF cap, along with more than 500K total charge resistance. Without adequately cleaning the board, the charge on the cap would never exceed 5.41V, and it takes 8V to turn the fan on. After meticulously cleaning the board with isopropyl alcohol and a toothbrush, plus an X-Acto knife between pads, the circuit works fine now.

I probably should've just used a BS2, but I had an NE555 in the scrap drawer and wanted to do things on the cheap. Oh, well. Lesson learned.

-Phil

Tracy Allen

Or dead bug wiring at least for the hi-Z node. Here's an example posted by our friend @Heater.

Phil Pilgrim (PhiPi)

That reminds me of a Bob Pease ratsnest, but at a much smaller scale:



Somehow, by carefully avoiding shorts, he got his prototype circuits to work!

-Phil

jmg

Phil Pilgrim (PhiPi) wrote: »

.... It's been having mold/moisture issues, .... Without adequately cleaning the board, the charge on the cap would never exceed 5.41V, and it takes 8V to turn the fan on. After meticulously cleaning the board with isopropyl alcohol and a toothbrush, plus an X-Acto knife between pads, the circuit works fine now....

Hmm, 'works fine now' in an environment known as having mold/moisture issues, sounds the the triumph of optimism over experience....

There are divider '555' like parts - HEF4541, HEF4521, HEF4060 & even 74HC5555 that run faster Osc and divide to the long times..
- and close on the heels of those parts, are the low cost MCUs that now approach Logic-MSI prices.

Here is an 8-pin one, 4k Flash, ADC, UART, 256R, 1.8 V – 5.5 V (there is a SOT23-6 SN8F57011 mentioned, but I can't find prices on that )
https://lcsc.com/product-detail/SONIX_SN8F5701_C141725.html

Phil Pilgrim (PhiPi)

jmg wrote:

Hmm, 'works fine now' in an environment known as having mold/moisture issues, sounds the the triumph of optimism over experience....

And that's why I'm going to conformal coat the entire circuit before I encase it in shrink tubing. Been there, done that.

You do know what conformal coating is, right?

-Phil

laurent974

It's always interesting to find 555 applications.

By the way i just discovered the Sonix SN8F5701. it seems interesting, but i can't find a DIP version of it to test it on a breadboard.

jmg

laurent974 wrote: »

By the way i just discovered the Sonix SN8F5701. it seems interesting, but i can't find a DIP version of it to test it on a breadboard.

Dip these days is quite rare.
For an easier to obtain part, not in S08, the N76E003 comes in TSSOP20, and you can get break out boards on eBay, and eval boards here https://direct.nuvoton.com/en/
SO8 parts fade a little, as TSSOP20 parts are getting very close in price
Or, if you really want DIP ... (seems no ADC ? )
https://lcsc.com/product-detail/STC_STC15F101W-35I-DIP8_C183063.html
https://lcsc.com/product-detail/STC_STC15W104-35I-DIP8_C78255.html
https://lcsc.com/product-detail/STC_STC15W202S-35I-DIP16_C183076.html

kwinn

According to the data sheet it is available in a DIP8 pckage.

ceptimus

Phil Pilgrim (PhiPi) wrote: »

jmg wrote:

Hmm, 'works fine now' in an environment known as having mold/moisture issues, sounds the the triumph of optimism over experience....

And that's why I'm going to conformal coat the entire circuit before I encase it in shrink tubing. Been there, done that.

You do know what conformal coating is, right?

-Phil

Aren't you still likely to get problems when the electrolytic cap ages and becomes more leaky? I wouldn't trust a 555 circuit with such a long timing period to age well.

Genetix

Phil,

Why did you need to X-acto your PCB?
I prefer RA flux over the No-Clean stuff and 99% IPA with an acid brush always did the trick.

You might want to conformal coat a junk PCB that has some caps on it to be sure you won't get the same problem again.
Anything with RF shouldn't be conformal coated just so you know.

erco

Ha, I just used a 555 this week to toggle a relay. They are rated at ~200 mA output and the coil current was 75 mA, so I drove it directly. Only ever worked with the relay on the high side (relay wired to +5, active low). But still I got very erratic results. Worked one day, not the next. Drove me nuts. Finally added a switching transistor and that solved the problem.

And yeah (hangs head), it was one of my 4-cent Ebay China 555 timers!

Heater.

Just be sure not to use the conformal coating of death, which after some years sucks up water thus shorting your circuit and rotting the components on the board:

Genetix

Heater,

I would think that anyone paying big bucks for something wouldn't want to see black gunk used all over the place.
The only exception would be the underside of a car.

Heater.

"Black gunk" is a common thing when electronics needs to be proof against harsh environmental situations.

For much the same reason as the black gunk on the underside of cars.

Phil Pilgrim (PhiPi)

ceptimus wrote:

Aren't you still likely to get problems when the electrolytic cap ages and becomes more leaky? I wouldn't trust a 555 circuit with such a long timing period to age well.

It doesn't need to. It's a short-term solution.

Genetix wrote:

Why did you need to X-acto your PCB?

There were a couple places where the alcohol-soaked toothbrush didn't reach. Same applies to teeth. That's why, even with regular brushing and flossing, dental hygenists still have jobs scraping tartar.

erco wrote:

Ha, I just used a 555 this week to toggle a relay. They are rated at ~200 mA output and the coil current was 75 mA, so I drove it directly. Only ever worked with the relay on the high side (relay wired to +5, active low). But still I got very erratic results. Worked one day, not the next. Drove me nuts. Finally added a switching transistor and that solved the problem.

Yup, same maddening problem here; same solution. I'm driving the fan with a 2N4403 PNP. The 555 output is "rated" at 200 mA, but seems not to have the necessary oomph to drive the fan directly.

heater wrote:

Just be sure not to use the conformal coating of death, which after some years sucks up water thus shorting your circuit and rotting the components on the board:

I used Konform. It's a hydrolytically stable, elastomeric silicone spray.

-Phil

Phil Pilgrim (PhiPi)

I took some measurements. The fan runs at an interval of three minutes on and 18 minutes off. When the fan is running, the unit draws 190 mA; when not running, 11 mA. That calculates to a mean battery draw of 36 mA. These numbers are close enough to the figures I was hoping to hit.

-Phil

Genetix

Close enough for military work.

erco

Tracy Allen wrote: »

Or dead bug wiring at least for the hi-Z node. Here's an example posted by our friend @Heater.

Dead bug wiring is awesome, as is point-to point wiring on vintage radio chassis. Most LED cubes are just neatly soldered LEDs hanging in space. It would be fun to build a robot (walking or rolling) using no structure other than necessary components and conducting wire.

One more thing added to my bucket list...

erco

erco wrote: »

Ha, I just used a 555 this week to toggle a relay.

Here's a video of the wacky prop I helped on. I added the amplified MP3 sound and the 555 timer to toggle a relay that cycles a linear actuator to move that big silver hose back & forth.