I am looking for a fairly compact way (no room for relays) of isolating 240v mains to a logic output. I can't find a suitable google search expression. It is a one-off job - monitoring what my heating system is doing. There are 3
240 volt AC switched supplies. Way back, I used a neon lamp, black rubber sleeving and an OC71 with the paint scraped off - any more modern devices? (I am aware of creepage clearances etc, thanks)-- Geo
Google "optoisolator"
JF
Geo wibbled on Thursday 28 January 2010 18:33
Something involving an opto isolator and a suitably rated resistor? If you did this, I would recommend incorporating a local quick blow very low rated fuse on the mains side. Your typical 6A breaker or 6-13A heating fuse will blow your board to bits before the fuse clears the fault.
Make sure the resistor is rated for peak voltage then perhaps add 50-100% margin for spikes. Don't forget a reverse diode in parallel with the led element.
It's always worried me having mains and ELV on the same board (unless you are set up to verify it can't track), so why not extend your original idea and build an optoisolator with a large air gap (which is basically what you did before) - LED at one end of a piece of black plastic tube, photo-thingy at the other... Same caveats re fusing and resistor.
The other option could be a ready made current sense coil which are specifically designed to insert a bit of 240V wire through the middle - assuming that measuring load rather than voltage was any use to you?
OTOH what's wrong with your original idea - neons aren;t dead yet...
-- Tim Watts Managers, politicians and environmentalists: Nature's carbon buffer.
... and a rectifier at the other side. Else you'll see roughly 10msec on and 10msec off because of the 50Hz AC (OC71 somehow sounds like Europe). There are optos with two LEDs in there, takes care of the reverse diode and covers both half-waves with very little gap.
But they tend to turn black or flicker after a number of years of operation.
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
AC opto like this.
On a sunny day (Thu, 28 Jan 2010 18:33:03 GMT) it happened Geo wrote in :
You probably mean logic *input*? Optocoupler in a brige rectifier with high value series resistor does it just fine. You can add a capacitor to make pure DC at the micro side, but a micro will detect
100 Hz or 120 Hz no problem.
Bear in mind that mains spikes can be 10x peak voltage, although (hopefully) short duration.
-- Dirk http://www.transcendence.me.uk/ - Transcendence UK http://www.theconsensus.org/ - A UK political party http://www.blogtalkradio.com/onetribe - Occult Talk Show
Dirk Bruere at NeoPax wibbled on Thursday 28 January 2010 20:38
It's a good point... Do you think it's worth putting a varistor in the mains input? Small and cheap...
-- Tim Watts Managers, politicians and environmentalists: Nature's carbon buffer.
.........................
That was an early alloy-junction germanium low-power low-frequency transistor. It's companion "power" transistor was the OC72. From the 1950s, I think. The case was glass painted black, and scraping off the paint turned it into a dandy little phototransistor. And yes, it was European. The numbering system was an adaptation of the American system for tubes with O for zero heater voltage.
I still have OC-series glass transistors. And yes, when I was a kid I did what many others did, scraping off the paint :-)
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
"Geo" schreef in bericht news: snipped-for-privacy@4ax.com...
Why seeking problems for a one-off job? Neon bulbs are still widely available. Even OC71 may be if you want to. But take a small signal transistor in a metal can - like a BC108 - file off the top of the can and you have a replacement for the scratched OC71.
You can of course look for optocouplers like the old 4N25 or similar. Problem is to convert the 220Vac to a 20mAdc as required by the input of such a device. Will be much more expensive and space consuming then a simple resistor-neon bulb combination.
petrus bitbyter
I have used a neon bulb and a darlington phototransistor, which is sufficiently sensitive even with a large distance to the neon bulb.
fine.
Better use a low current optocoupler, since otherwise, you would have to dissipate 240 V x 20 mA = 4.8 W in the series resistor.
A high voltage capacitor and a resistor in series on the AC side would keep the dissipation on a reasonable level.
On a sunny day (Fri, 29 Jan 2010 07:42:59 +0200) it happened Paul Keinanen wrote in :
Yes, but if secondary connected to for example a PIC with internal pull-up of say 100uA, and say a 10 to 1 ratio for LED current to photo transistor current, then you need primary 1mA, at 230V that makes about 230 mW.
I have one here in use with a few hundred kOhm 1/4 watt resistors in series, it pulls up a pin on the serial port against a control line on that port :-) The opto is a CNY17-3, it detects mains coming from somewhere...
Na, then you need a resistor AND a capacitor.
Optocoupler input versus output is pretty linear, and extends all the way down. Have a look at the CNY17-3 datasheet, they stop drawing at .5 mA, but it goes way lower.
Also the *peak* AC waveform counts to pull down ... for detection.
I don't think its necessary given the average response time of your typical opto is measured in uS.
At a minimum there should be a fuse and an X2 cap across the input. The RC would help to minimize any fast transients on the line. I would also use two resistors in series to feed the opto's LED for reliability.
There are some expensive opto's with integrated logic gates like Schmitt trigger etc. At 50/60 Hz lots of Darlington output Opto's are available dirt cheap to minimize power dissipation in the Resistor.
Some early transistor radios started developing hum problems when placed near artificial lights! This was traced to the black paint covering the glass body getting scratched and letting in the light.
The solution was to fill the glass body with opaque goo, which put paid to using them as photo transistors! By the time I first found out about it, they were all goo filled!
Mullard/Philips took advantage of this by producing a rather highly priced special version - the OCP71 - which was really an OC71 without the goo or black paint, but I think you also got a lens moulded into the body for your money!
No, it was the Philips valve (tube) numbering system which was also used for early semiconductors.
The O was a letter, not a number, but did stand for zero-volt heater. C meant triode, which the transistor most closely resembled. (A, for a single diode, was also used for germanium diodes, such as the OA70 & OA71.)
As semiconductor devices got more numerous and complex (the old system, for example, had no way of differentiating between germanium and silicon devices) a new system was devised:
although that article incorrectly tries to merge both systems which is impossible - the quoted AD162 germanium power transistor for audio frequency use would have a 4V heater in that case!
In the original valve numbering system, the serial number (usually two digits but occasionally 3) on the end also identified the base type,
30-39 was International Octal, for example, and 80-89 was B9A. See the reference above for the full list but the letters in most common use decode as follows:1st letter: Heater rating A 4V D 1.4V E 6.3V F 12.6V G 5V K 2V O Cold cathode (by 1955 this also included semiconductors as these had no heater) P 300mA T 7V U 100mA
Remaining letters: Type(s) of device(s) A diode (excluding rectifiers) AA double diode with separate cathodes (excluding rectifiers) B double diode with common cathode (excluding rectifiers) C Triode (excluding power and gasfilled triodes) D Power triode E Tetrode (excluding power valves) F Pentode (excluding power valves) H Hexode or Heptode (of the Hexode type) K Octode or Heptode (of the Octode type) L Output tetrode, beam tetrode or pentode M Tuning indicator N Gas filled triode or thyratron Y Halfwave rectifier Z Fullwave rectifier
Terry
No, it is a number, zero. The 0Z4, 0A3, etc. are all indexed under zero -- a search with 'o' will find nothing.
Tim
-- Deep Friar: a very philosophical monk. Website: http://webpages.charter.net/dawill/tmoranwms
"Tim Williams" schreef in bericht news:hjuu4t$l18$ snipped-for-privacy@news.eternal-september.org...
Maybe there were some American semiconductors numbers starting with zero, the European transistors of the time definitively used the 'O'. OC3, OC4, OC13, OC14, OC44, OC71 and others. In the tube- and transistor reference books they were alphabeticaly placed at the 'O' not at the '0'. Googling for "OC13" I get a lot of hits about that transistor, "0C13" also give a lot of hits but nothing about transistors. (Though I have to admit I did not check all the hits.)
petrus bitbyter
I did just that back in the 80s for some idle experiments and at least one practical application. The projectors at a local cinema owned by a friend used vacuum photocells for the sound track. When it became difficult to get spares for the tube amplifier, I replaced it with a transistorised PA amp, and the photocells with ones I made by opening the top of 2N4033 transistors and covering them with transparent cellotape. I mounted the transistors inside a phonojack and rigged up a holder for the jack socket. They used them this way until they closed down in the late 90s.
On a sunny day (Fri, 29 Jan 2010 15:02:55 +0000) it happened Terry Casey wrote in :
I remember using a AZ-1 ar rectifier :-) I had the OC13 and OC14 too, the low cost (probably out of spec) OC71 and OC72? There was the OC44 and OC45 as RF (well a few MHz) transistors, and the OC16 'power' transistor to drive for example a speaker.
On a sunny day (Fri, 29 Jan 2010 09:16:41 -0600) it happened "Tim Williams" wrote in :
No, it is 'O' as in Ontario. Try googling OA79. It is a germanium diode, frequenctly used in xtal radios in the past...
You Canadianians have no idear about those old Philips parts :-)
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.