CMOS buffer outputs - UK Vintage Radio Repair and Restoration Discussion Forum

Oldcodger · 25th Sep 2015, 11:52 pm

From one data sheet for this sort of device (more accurately CD4049), I noticed that more than one output can be coupled to increase the current drive.
I'd suspect that in the write up I saw (but like as usual, can't find now I need it), that the outputs would need to be diode coupled to prevent problems. Anyone confirm?. I could just add a transistor, but with the drive capacity shown as circa 26mA, I'd like to power a reed relay of 500ohm from 12v = circa 24mA and this is in car, so to be safe I'd like to use the other two unused stages. (circuit is ONLY powered, via a tried and tested 12- 5-12v converter, when the ign is off).

julie_m · 26th Sep 2015, 12:08 am

Provided you tied the inputs together so the outputs could never assume opposite states, it ought to be safe just to parallel the outputs. All the devices are in intimate thermal contact and connections are short, so you can expect them to switch simultaneously and even if there is a brief fraction of a microsecond while they are in opposite states, the decoupling capacitors you fit to each IC (you did fit decoupling caps, didn't you?) will maintain the supply for the duration of the short.

You could be paranoid and use diodes to share the load current between outputs while preventing the possibility of shorting the supply -- but then you'd have to choose whether you wanted sourcing or sinking, you couldn't do both.

Leon Crampin · 26th Sep 2015, 10:26 am

It is within the maker's recommendation in most applications sheets to parallel drivers within devices, as long as the inputs are tied as detailed above, and also as long as all of the stages connected together are on one chip. Connecting together separate devices is not recommended.

I have done this many times commercially (4049/50 driving reed relays directly) without any problems.

Leon.

Oldcodger · 26th Sep 2015, 7:26 pm

thanks folks for the replies.
I was certain I'd seen this in a data sheet, but as i couldn't find it was beginning to think I'd imagined it ,putting it down to a senior moment.
Leon - that's my idea. i.e. driving a 500ohm reed from a nominal 12v supply . To be on safe side, I'd sooner use the unused inverters to drop the current through each one.

mods- can't see any reason to keep this open , given two prompt and accurate answers.

Radio Wrangler · 26th Sep 2015, 8:29 pm

Driving an inductive load from CMOS... be sure to put in a catching diode for the transient on turn-off!

CMOS has an implicit thyristor across the power pins, and driving an input or output beyond the rails can trigger it, resulting in general destruction.

David

Oldcodger · 27th Sep 2015, 1:23 am

I ALWAYS factor that in to circuits. But thanks ,anyway. The L*di/dt can result in serious voltage ,which without a diode WILL result in a junction breakdown.

25th Sep 2015, 11:52 pm	#1
Oldcodger Nonode Join Date: Oct 2014 Location: West Midlands, UK. Posts: 2,181	CMOS buffer outputs From one data sheet for this sort of device (more accurately CD4049), I noticed that more than one output can be coupled to increase the current drive. I'd suspect that in the write up I saw (but like as usual, can't find now I need it), that the outputs would need to be diode coupled to prevent problems. Anyone confirm?. I could just add a transistor, but with the drive capacity shown as circa 26mA, I'd like to power a reed relay of 500ohm from 12v = circa 24mA and this is in car, so to be safe I'd like to use the other two unused stages. (circuit is ONLY powered, via a tried and tested 12- 5-12v converter, when the ign is off).

26th Sep 2015, 12:08 am	#2
julie_m Dekatron Join Date: May 2008 Location: Derby, UK. Posts: 7,735	Re: CMOS buffer outputs Provided you tied the inputs together so the outputs could never assume opposite states, it ought to be safe just to parallel the outputs. All the devices are in intimate thermal contact and connections are short, so you can expect them to switch simultaneously and even if there is a brief fraction of a microsecond while they are in opposite states, the decoupling capacitors you fit to each IC (you did fit decoupling caps, didn't you?) will maintain the supply for the duration of the short. You could be paranoid and use diodes to share the load current between outputs while preventing the possibility of shorting the supply -- but then you'd have to choose whether you wanted sourcing or sinking, you couldn't do both. __________________ If I have seen further than others, it is because I was standing on a pile of failed experiments.

26th Sep 2015, 10:26 am	#3
Leon Crampin Octode Join Date: Nov 2005 Location: Surrey, UK. Posts: 1,874	Re: CMOS buffer outputs It is within the maker's recommendation in most applications sheets to parallel drivers within devices, as long as the inputs are tied as detailed above, and also as long as all of the stages connected together are on one chip. Connecting together separate devices is not recommended. I have done this many times commercially (4049/50 driving reed relays directly) without any problems. Leon.

26th Sep 2015, 7:26 pm	#4
Oldcodger Nonode Join Date: Oct 2014 Location: West Midlands, UK. Posts: 2,181	Re: CMOS buffer outputs thanks folks for the replies. I was certain I'd seen this in a data sheet, but as i couldn't find it was beginning to think I'd imagined it ,putting it down to a senior moment. Leon - that's my idea. i.e. driving a 500ohm reed from a nominal 12v supply . To be on safe side, I'd sooner use the unused inverters to drop the current through each one. mods- can't see any reason to keep this open , given two prompt and accurate answers. Last edited by Oldcodger; 26th Sep 2015 at 7:30 pm. Reason: close request.

26th Sep 2015, 8:29 pm	#5
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,901	Re: CMOS buffer outputs Driving an inductive load from CMOS... be sure to put in a catching diode for the transient on turn-off! CMOS has an implicit thyristor across the power pins, and driving an input or output beyond the rails can trigger it, resulting in general destruction. David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

27th Sep 2015, 1:23 am	#6
Oldcodger Nonode Join Date: Oct 2014 Location: West Midlands, UK. Posts: 2,181	Re: CMOS buffer outputs I ALWAYS factor that in to circuits. But thanks ,anyway. The L*di/dt can result in serious voltage ,which without a diode WILL result in a junction breakdown.