A very silly circuit

I looked at that exact problem recently and didn't do a complete design, but it looks pretty easy to me.

There seems to be a standard socket for the relays that fit in the boxes in a car. But the relays are slightly different sizes, especially the height. I was looking at inserting a wireless connection in a relay package to allow the motor to be killed without cutting into the wiring.

--

Rick

On Saturday, October 3, 2015 at 2:09:48 PM UTC-4, snipped-for-privacy@yahoo.com wrot e:

That's not an SSR.

The real low-dough fix is about a 3" piece of 12 gauge wire with two

1/4" push-on terminals crimped on to it, to plug in where the contacts were. Don't forget to unplug it when you park the car, though. :)

If it's something that can't stand to run all the time, put a two- terminal turn signal flasher where the contacts were. The duty cycle is approximately 50% but will vary with load, temperature, moon phase, etc.

If it's the "standard" automotive relay that comes in about a 1" cube with four or five 1/4" push-on tabs on the bottom, you may be able to get a "generic" one for $10 or so at the car parts store, or possibly at a car stereo and alarm store. Sometimes the car parts people call this a "Bosch relay", and I think Bosch originally came up with the package, but they are now made by many other companies. Like everything else, name-brand ones work well and cheap Chinese ones work for a while.

The four-terminal ones are SPST, and *most* of the five-terminal ones are SPDT. *Some* of the five-terminal ones are SPST, with two terminals for one side of the contacts. These get used for things like fog lights, where you want to power two things from the relay contacts, and don't like splices for some reason.

The terminals are almost always numbered with the Bosch numbering system. 85 and 86 are the coil. 30 is the "battery" or "feed" side of the contacts. 87 is the NO contact and 87a (if present) is the NC.

You can get them with or without a mounting tab. Usually they are rated 30 A but some are rated for more. If they are over 30 A, the coil terminals tend to bigger than 1/4" push-ons.

When they first started, the coil was just a coil. After computers in cars happened, the relays grew diodes or resistors across their coils, to help get rid of the spike at switch-off. 12 V is the most common coil voltage but you can get 24 V.

Digi-Key has Potter and Brumfield VF4 series, Omron G8JN series, and Panasonic CB series relays like this, for about $3 to $9 quantity one.

There are some other "standards". There is one that is about 1" x 1" x 0.5", with more closely-spaced 1/4" push-on terminals on the bottom. Some Japanese cars have relays that have two rows of up to three push- ons each in a "connector housing" that is part of the relay body. Digi-Key also has some of those, but not as much selection as the 1" cube ones.

Standard disclaimers apply: I don't get money or other consideration from any companies mentioned.

Matt Roberds

It took me two summers to figure out why my car would boil sometimes.

Dead fan motor: disection revealed EMI supression choke separated from brush carrier board

I fitted a generic fan (20% of the price of genuine), and now the AC works better too.

--
  \_(?)_

Or IR2153 with its HO and LO ouput stages driving a tiny transformer directly, in the full bridge configuration. Just 3 parts and allows multiple isolated output voltages if needed. Unfortunately, I didn't invent it. ;-)

Best regards, Piotr

AUIR2085S is higher current and higher speed +/-1A, 500KHz

-Lasse

Thanks Matt.

It's approximately a Bosch format with a mounting tab, but half the standard thickness.

Electronic Goldmine and BG Micro have the standard Bosch format for about $1.50. My fallback is to get one of those and kludge a wiring harness to mate the replacement to fit the original connector.

Cheers, James Arthur

If the terminals are 1/4", maybe something like:

If the terminals are tiny, maybe something like

I haven't seen any of these "half size" ones with a mounting tab... it seems like they came out after all the relays got moved to a central relay box. I'm sure an automaker can get them that way, but I don't know if humans can get them in quantity 1.

Digi-Key sells the sockets to go with the Bosch relays, but they come as a housing and loose crimp terminals. Sometimes it's easier to get something like this

with a foot of wire already on it. The local car alarm shop will definitely have these kinds of sockets in stock.

Standard disclaimers apply: I don't get money or other consideration from any companies mentioned.

Matt Roberds

Maybe the trick of generating a negative bias current to pull an opamp output down though ground?

--

John Devereux

Something like that, where you need a half a volt of negative voltage and don't want a charge pump or anything like that.

We use some cool Ixys self-protecting SSRs as current limiters, where we don't need either isolation or switching.

The PV optocouplers are under 1% efficient. Lots of other uses would appear for one that was, say, 10%. More than that would be difficult.

You don't even need an opto-coupler for that. As Pease wrote: just use a

2N2222 (or similar) with base grounded, zener the emitter junction with plenty mA and see a small negative voltage at the collecter. The B-E zener emits light that excites the B-C junction as a photodiode.

piglet

Since it's not April, the 1st, so.. please explain. :-)

Best regards, Piotr

I thought my description did explain it. Go on, try it:

Have fun.

piglet

Weird, I got -0.37V with a 3904, but 10 k ohm load killed it,

100k ohm and it fell to ~ -0.037.

George H.

Yeah, power conversion efficiency is terrible but that can be OK if you only need a micro-amp. I believe someone has used this as a way of getting a few tens of millivolts for op-amp offset nulling!

Watch out how you use the transistor after your experiment, the sustained Veb breakdown will have reduced beta and increased noise. I tried this again just now and initial beta of 553 fell to 525 after jsut one minute of experiment.

piglet

That's a new one on me. I imagine that is even less efficient!

--

John Devereux

Applied for bias cancellation...

Cheers, James Arthur

Probably noisy. And maybe drifty.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Dunno. Zeners are still lower noise than band-gaps, right?

Woodward's one of the best, but it has been eighteen years since he wrote that. And he was souping up what is basically a high-test LM324, so it just might not matter that much.

Cheers, James Arthur