Constant current continuity tester?

[throbscottle]

Is it a good idea to use a constant current source (I have a j-fet in mind) as the basis of a continuity tester?
Thanks in advance

 

[crutschow]

It would work, but why go to the complexity of a constant-current source when a simple resistor and buzzer works fine to detect continuity?

 

[rjenkinsgb]

I do not see any advantage with that?

For a true low-ohms continuity test, I'd think a simple current source - resistor from a known voltage - and an adjustable voltage comparator across the test connections, so only if the current causes a voltage drop below a certain voltage threshold (= below a set resistance) is the circuit considered continuous.

The voltage threshold detector is the "basis" for an electronics tool, I'd think?
The current only needs to be reasonably consistent over a variation of a fraction of a volt, so not at all critical.

Add a Schottky diode across the test leads and you should have a continuity tester than will not damage semiconductors.

 

[AnalogKid]

Nothing wrong with it, but simple continuity needs only a resistor.

ak

 

[throbscottle]

Thanks for that everyone. I just wondered if it was better than just using a resistor. Don't know about being extra complex, it's just a j-fet with with the gate connected to source! I like that schottky diode idea though
Ok next question - what is considered to be a suitable/safe current for continuity testing? Mainly for beeping out home-brewed PCB's and rats-nest perf-board, as well as fault-finding.
(edit) actually 1 advantage - it doesn't care what voltage the supply battery is!

 

[crutschow]

what is considered to be a suitable/safe current for continuity testing?

I'd keep it below a mA.

 

[Nigel Goodwin]

If you're going to use constant current, you may as well make a full low-ohm meter, and use a comparator to sound a buzzer if it's below a certain value.

 

[throbscottle]

Nah, I'll just use a resistor like everyone else does. Maybe have a high and low setting. A low-ohm meter doesn't justify the time it would take!

 

[unclejed613]

when i worked at a prototyping house, they needed to build continuity testers to verify wiring of the test fixtures they built. since there were often ICs and other semiconductors in the circuits of the fixtures, they wanted a tester that could be preset to something like 1 or 2 ohms as the threshold, had a max current of 0.5mA, and a probe voltage of less than 0.3V.

 

[unclejed613]

i kinda had to reverse engineer from memory 30+ years ago the circuit used used a TL072 op amp, with LEDs and a sonalert "mini" sounder (shown by I1)... it was something like this...

UUT is the wire being tested... the resistors, obviously for this kind of balance on a resistance bridge are 1% tolerance, and we had a 10 turn pot for the center part of the resistor string of R4, R7, and R6 to set the resistance the device triggered at... (just replace R7 with a 47 ohm 10-turn trimmer with the wiper connected to the noninverting input of the op amp)... this could be set to trigger at 2 ohms or less... and the sonalert would tell you if there was a loose connection because it would break up if you wiggled the wire and it was loose... the voltage across the probes is very low, and the current less than 50uA so it won't damage semiconductors

 

[Tony Stewart]

A simple CC generator with an Op Amp is all that is needed to convert resistive current into voltage using a Holland current source. Scaled by decade in resolution starting from 1mA https://tinyurl.com/ydwo2gor

 

[unclejed613]

using a Holland current source.

actually it's "Howland"... here's a really good app note about Howland current sources:
AN-1515 A Comprehensive Study of the Howland Current Pump (Rev. A)

 

[Tony Stewart]

actually it's "Howland"... here's a really good app note about Howland current sources:
AN-1515 A Comprehensive Study of the Howland Current Pump (Rev. A)

Thanks to iOS it respelt it for me

 

[Tony Stewart]

It depends if you want a trivial continuity tester or a real short circuit microohm short circuit finder and PCB short blower using a CMOS Howland Converter with emitter followers on a single battery supply. Or both....

 

[mab2]

Why re-invent the wheel? the diode test on my very cheap digital multimeter is a 1mA cc, and reads the voltage across the test leads: so a diode shows 0.7v forward, reverse connected is overrange, shorted leads reads zero and a 1k resistor shows 1v, etc.

I assume the resistance ranges use a range of constant currents and measure voltage too.

 

[unclejed613]

Why re-invent the wheel?

a continuity tester with a sensitive response and an audible sounder can tell you when you have an intermittent connection when most DMMs have a sample time of a second, and you really can't easily tell if a connection is intermittent...

 

[throbscottle]

Why re-invent the wheel? the diode test on my very cheap digital multimeter is a 1mA cc, and reads the voltage across the test leads: so a diode shows 0.7v forward, reverse connected is overrange, shorted leads reads zero and a 1k resistor shows 1v, etc.

I assume the resistance ranges use a range of constant currents and measure voltage too.

Whew! Lot of activity here since I last looked!
It's because I don't want to have to use my meter. On the low ohms range it beeps up to about 50 ohms, but I might have to change whatever setting I was using to get it on that range, and disconnect whatever leads I was using and plug in it's test probes - it's just inconvenient.
Anyway, re-inventing the wheel is fun - and sometimes you get a better wheel.

 

[throbscottle]

You lot have posted some ideas whilst I was wheel-reinventing! Here's what I came up with after ignoring everything (else) I found on the 'net...
I wanted a tester with some hysteresis to avoid "scratching" when it makes contact. The design here has about 100mV, which is probably too much but ya gotta start somewhere. The 470 ohm preset sets sensitivity from about 100 ohms to under 5 ohms.
It's going to be a 2 probes with a wire between type of affair, 1 probe holds the circuit. Not sure how I'm going to power it yet. Maybe some button cells.
Not a constant current source in sight....
Can't build anything until after house move, so just exists in my computer for now...