Interfacing 5V to unknown 10-30V signal

"Fred Bloggs" schreef in bericht news: snipped-for-privacy@nospam.com...

Okay - I got a bit confused with the diode drawn horizontal. Doesn't look bad at all, really.

--
Thanks, Frank.
(remove 'q' and 'invalid' when replying by email)

"Frank Bemelman" schreef in bericht news:41c45c0e$0$98814$ snipped-for-privacy@news.xsall.nl...

Guess you have to see the transistor was accidentally drawn an npn. It should be a pnp.

petrus bitbyter

--
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.802 / Virus Database: 545 - Release Date: 27-11-2004

Fair enough. You can safely assume it's NPN o/c like in my illustration. 10 - 30 Vdc is standard range for most proximity and photo sensors, highly unlikely to have direct TTL compatibility. Which reminds me, what are they sensing?

The legacy devices would have me worried, could your client possibly upgrade whatever it is to newer sensors? Omron, Keyence, Telemecanique, SICK, Turck and others have a wide range. Try to keep to one or two types for ease of replacement and spares. In the plant the parts I've redone all field sensors are current sources (PNP o/c) and PLC inputs are sinks. Keep away from the two-wire types, in the off-state they still draw a couple of mA so they don't work with high impedance inputs and bridging each and every one of them with 2k2 resistors is a major PITA.

- YD.

--
Remove HAT if replying by mail.

Hi,

There are a few sensors; some of them are capacitative proxes, some are reed switches (okay, those are easy enough to interface :), some are Hall effect sensors, and there may be a couple of simple microswitches in there also. This is an automated food mixing and packaging application - not strictly industrial, more what I would describe as commercial - but it uses industrial components because the company that made the mix-o-pack-o-mat thingy has a comfort level with those specific sensors. Basically, they prototyped it with a PLC, and my job is to replace the PLC with a custom black box they can sell at a healthy profit.

Sorry I can't be more specific about the application - I'm sure there would be good comments here if I got into details - but I'm under the usual vow of silence.

"petrus bitbyter" schreef in bericht news:bK3xd.116463$ snipped-for-privacy@amsnews05.chello.com...

Okay, that makes more sense:

But both resistors are probably rather low values, to make the input not serve as a sensitive antenna in a hostile industrial environment.

Speff warned for injecting current and raising the +5. Which is a serious problem indeed, when inputs are raised to +24V or something. Then suggesting a series diode, presumably the original one, with the cathode connected to the resistor, anode to the outside world:

+---+-----+------------- | | | .-. | | | | - | | | ^ | '-' | | | | |< +---+---| | | .-. | | | | | '-' | -A v -K | | --> nasty stuff->o

Now the diode parallel to B-E doesn't do much anymore, not that I see anyway.

But Speff seemed to agree on this B-E diode suggested by Bill, so I figured I must be missing something about this concept.

I am just fishing a bit for entertainment ;)

--
Thanks, Frank.
(remove 'q' and 'invalid' when replying by email)

If you saw it, your comfort level would be increased, I think. There are a fairly large number of these things installed around the world, and the nature of the material they process is such that there are really no safety issues. As a ballpark example, consider those gumball machines that drop the ball into a tube, and it runs around triggering different sound and light effects before dispensing to the consumer. That isn't the actual application, but it is a very good approximation of the "danger level".

Note: I would have preferred to spec all the electronic parts in this myself, but it wasn't possible, and the Customer Is Always Right.

But that diode reverse leakage is not passing through the transistor BE junction, it is shunted through the resistor from base to Vcc. If the BE shunt resistor is 10K then a wildly exaggerated diode leakage of 50uA at Tmax and Vreverse,max would produce a Vbe of 50x10xuxK=500mV , which is absolutely nothing.

How exciting...hopefully this slipshod thing is FOR EXPORT ONLY.

I'll second that, I hope you have access to the original program listings. Sounds like one of those things a PHB dreams up to impress the upper levels and makes a point of speccing everything by himself. Sort of like the mess I'm gradually sorting out, though in this case it was a newly baked engineer trying to impress by how cheaply he could get things done.

- YD.

--
Remove HAT if replying by mail.

Google Groups doesn't automatically quote anything in its new incarnation, and Mozilla doesn't provide a "paste as quote" option when pasting into a web form. Hence my laziness, for which I shan't apologize.

in

in

degrade

BE

BE

at

is

50uA is a high leakage current for a small signal diode, so if you assume a small signal silicon diode and a lowish value resistor, you are absolutely right. The circuit diagram didn't specify anything more than "diode" and "resistors".

------ Bill Sloman, Nijmegen

^^^

Who is 'you'? It's conventional and considerate to make it immediately clear to whom you are replying. Virtually none of your replies do so. Please use either the 'quote' facilities of your software, or just copy/paste via the clipboard.

--
Terry Pinnell
Hobbyist, West Sussex, UK

I know- if you scope out some of the ss-diode specs , you will see they quote that ridiculous number even for the 1N4148 at Vr=20V and Ta=150oC

- Fairchild even rates the

1N459A at 5uA at that ridiculous extreme too -but with Vr=175V.

A followup to this: I kind of gave up on trying to guess around this issue. I decided it was just too open-ended for me ever to be able to say "Finished!", and so I told the customer that I would design the circuit on the assumption that his sensors can drive the line high with enough current to light the LED in the opto. The design I eventually put together is at

if you're interested.

The diode across the input will actually be an FDLL4148, not the power diode shown in the schematic snippet. And the resistor on the input is to be chosen according to the expected input signal voltage.