Go Down

Topic: TTL / CMOS high voltage low current sinking (Read 1 time) previous topic - next topic

jug5525

Hi!

Can anyone tell me if it is possible to control a high voltage (170V DC), low current (3mA) load by sinking it's current with a 5V powered TTL or CMOS chip?  (0= load is on, 1 = load is off)
 
- Would current flow across load when there is 0 (0V) on the IC's output without burning the chip?
- Would there still be current flowing across load when a logic one (5V) is applied to the output since there is still 165V difference?

The "circuit" can be seen on the this link.


 

allanhurst

No - but a suitable high voltage transistor or MOSFET would do the job

regards

Allan

Wawa

An external transistor is always needed when load supply voltage is higher than the logic voltage.
Use a high voltage transistor with suitable base resistor.
Back in the days we use a BD232 for that, with a 33k resistor for 3mA collector current.
Leo..

Grumpy_Mike

Quote
- Would current flow across load when there is 0 (0V) on the IC's output without burning the chip?
Yes the current would flow across through the  load when there is 0 (0V) on the IC's output without burning the chip. No harm would be done.

But the as soon as you set the output to 5V the chip would be instantly destroyed.


Jiggy-Ninja

Yes the current would flow across through the  load when there is 0 (0V) on the IC's output without burning the chip. No harm would be done.

But the as soon as you set the output to 5V the chip would be instantly destroyed.


Would it really? It's a 3 mA load, I'd expect the ESD diodes would shunt that away from the output buffer.

Grumpy_Mike

Would it really?
Yes.

Quote
It's a 3 mA load, I'd expect the ESD diodes would shunt that away from the output buffer.
While there is no official current rating for those diodes, in an application note it says to keep the current below 1mA. The point is that they are ESD diodes they are not designed as over voltage protection diodes. That is to say they are designed more for discharge pulses than a continuous conduction.

jug5525

Thank you!
I was looking to build a nixie tubes clock using logic ICs like the one on the following link. So would the circuit on the link not work and I would have to use a bunch of transistors or is there something I'm missing?

http://www.allspectrum.com/store/images/6-digit-nixie-controller-schematic-pic.jpg

Grumpy_Mike

#7
Sep 14, 2016, 09:50 am Last Edit: Sep 14, 2016, 09:50 am by Grumpy_Mike
Quote
So would the circuit on the link not work
Yes it will work, if you can get hold of 74141 chips, because they are designed to take such a high voltage and are protected internally. The input to these chips can be connected directly to the Arduino.

jug5525

Oh, okay
I guess I will use those ICs then.
Thank you for taking time to explain :).

Jiggy-Ninja

Yes.
While there is no official current rating for those diodes, in an application note it says to keep the current below 1mA. The point is that they are ESD diodes they are not designed as over voltage protection diodes. That is to say they are designed more for discharge pulses than a continuous conduction.
I'm well aware of what they are and are not designed for. An app note recommendation (AVR182: Zero Cross Detector?) isn't the same as an actual datasheet spec, so there's no telling if a sustained 3 mA is enough to cause instant destruction. I've never tried to use the ESD diodes as a continuous clamp like that, but I would lean towards 3 mA causing a more prolonged failure of the diode rather than immediate frying.

If you say something that definitive about something you might be wrong about, I just have a vision of someone ignoring what you said and doing it anyway, and if it's not instantly destroyed thinking that you don't know what you're talking about. While Jug is being sensible and listening to the advice, others do not, as I'm sure you remember from that recent LED thread.

Grumpy_Mike

Quote
I just have a vision of someone ignoring what you said and doing it anyway, and if it's not instantly destroyed thinking that you don't know what you're talking about.
:)
So two mistakes then. But point taken thanks.

Some beginners don't believe you can fry a chip and there be no smoke. I even met someone last week who said the moon landings were faked and all the stuff they brought back consisted of 50% Helium 3. But he failed to even understand that if they were faked then they couldn't bring any stuff back. No accounting for folk.

Basically a few years ago I talked to a friend who had a lot of experience in fabrication of ICs. He said that the geometry was different on the ESD diodes and he would never consider using it as over voltage protection, no matter what the current.

charliesixpack

#11
Sep 15, 2016, 03:27 am Last Edit: Sep 15, 2016, 03:31 am by charliesixpack
You will not just cause the ESD diode to conduct if you exceed the positive rail voltage.  Any p diffusion tied to the output will forward bias (p diffusions are always diffused into n type material).  This includes pnp collectors, p diffusion resistors and PFET drains.

By the way, an ex coworker just released his 10th book on ESD.  ESD diodes are not "normal" structures you would use in the design of circuits even though they were built with normal diffusions.  I designed everything in my off chip drivers except the ESD device.

Jiggy-Ninja

By the way, an ex coworker just released his 10th book on ESD.  ESD diodes are not "normal" structures you would use in the design of circuits even though they were built with normal diffusions.  I designed everything in my off chip drivers except the ESD device.
Link? I'm always up for new learning.

charliesixpack


Go Up