4066b Quad Bilateral Switch

[Pete9]

So to save £££ on points detection switches, I have opted to go back to my youth (not that long ago as I'm only 37) and play with some micro electronics to control bits and bobs depending on the position of the points. A single PL13 costs £4.50, whereas a 4066b Quad switch, costs 16p and creates several interlocking options!

 

I'm using Peco Electro points with a PL10 motor and one PL13 switch under the PL10 Motor.

 

Intention is to wire the PL13 to two 4066b ICs and  therefore creating 4 output controls per Normal / Reverse detection.

 

Normal:

(IC 1) Output 1 - +ve Frog Polarity

(IC 1) Output 2 - Signal Control

(IC 1) Output 3 - Track Isolation

(IC 1) Output 4 - Signallers Panel Indication

 

Reverse:

(IC 2) Output 1 - -ve Frog Polarity

(IC 2) Output 2 - Signal Control

(IC 2) Output 3 - Track Isolation

(IC 2) Output 4 - Signallers Panel Indication

 

 

The track isolation is for safety when points are to be travelled in heel to toe direction. There will be a section of track that is x in length (roughly an engine length), that will have it's power removed if the points are set in an opposing direction.

 

 

Has anybody any thoughts on the above? Oh, and some pointers as to what to watch out for such as wire size, resistor values etc (it's been a while) 

http://www.rapidonline.com/Electronic-Components/4066b-Quad-Bilateral-Switches-dil-83-0392

 

I've also come across 10 x 4069ub hex inverters in my loft. I obviously ordered these for a reason, but can I think why at the moment!?  

http://www.rapidonline.com/electronic-components/4069ub-hex-inverters-83-0398

 

 

Pete

[Miss Prism]

I doubt the 4066 is rated anywhere near a PL10 inrush current.

[Dagworth]

the 4066 is not being used to control the point motor.

 

Andi

[Suzie]

The PL13s are very unreliable at the best of times, and the 4066 has a high on impedance that needs to be taken in to account - it is not suitable for switching any current at all!

 

16p is cheap, but too cheap in this situation.

 

Think about using a PL15 instead of the PL13 if you want physical detection of the point position to give the panel indication and frog switching, and add a latching relay in parallel with the PL10 to provide the other circuits - or just add a 4-pole relay to one of the PL15 contacts to give you five circuits in total which will probably be sufficient.

 

I you do not want manual operation just use latching relays and save the money you would have spent on the PL13 or PL15 for an even cheaper solution.

[Pete9]

Correct Andi, the 4066 is only for detection. 

 

The PL13 does seem a little cheap and nasty for switching. Why pay £5.50 for a dual micro switch though? We are only paying that much for convenience of locating the switch on the underside of the motor. Surely there must be something more efficient out there that can be used? Also, the PL15 only has two outputs, I'm talking 4 outputs, so even using the PL15, I would need to get creative.

 

The 4066 wouldn't control high currents, even the frog polarity would be controlled via something with more umph. This something would however be controlled by a 4066 or similar.

 

Look at the outputs, are they really high current?

 

Output 1 - +ve Frog Polarity - a higher current capacity relay / transistor

Output 2 - Signal Control - an LED with a 560ohm  resistor. Or another interlocking setup.

Output 3 - Track Isolation - a higher current capacity relay / transistor.

Output 4 - Signallers Panel Indication - an LED with a 560 ohm resistor.

[Robin2]

£5.50 wold buy you a servo to move the points. You can control several servos with an Arduino Uno and have "proper" control logic as well.

 

The Arduino Uno costs about £25 but it is very versatile.

 

...R

[Baby Deltic]

Rather than using a 4066, surely using the point switch to control several transistors would be a lot more reliable than the 4066. I sourced some twin coil relays and connected one across each switch like a flip-flop thereby allowing one set of contacts for frog switching and one set for route indication. The 4066 cannot handle much current and is much better suited as a controlled zero volt switch than most things.

[Baby Deltic]

£5.50 wold buy you a servo to move the points. You can control several servos with an Arduino Uno and have "proper" control logic as well.

 

The Arduino Uno costs about £25 but it is very versatile.

 

...R

The I/O of the Uno is fairly limited. You may need some additional logic I.C.'s to break-out the I/O into more lines, such as a 4067 or 4514 if you need loads of I/O for multiple points and indicators.

[Robin2]

The I/O of the Uno is fairly limited. You may need some additional logic I.C.'s to break-out the I/O into more lines, such as a 4067 or 4514 if you need loads of I/O for multiple points and indicators.

or get an Arduino Mega which has more I/O pins. If you have that many points and signals the extra cost of a Mega would be justified.

 

...R

[Baby Deltic]

or get an Arduino Mega which has more I/O pins. If you have that many points and signals the extra cost of a Mega would be justified.

 

...R

Indeed. The Mega also has a lot more functionality because the AVR is much larger.

[Gordon H]

I am struggling to see what a 4066B brings to this particular party, other than low price. It is a low current analogue switch which merely provides 4 identical on/off channels - and any voltages on those switch channels must be referenced to the Vdd and Vss supplies of the 4066B itself.

Instead, a 4 pole DPDT relay would give much more flexibility - and galvanically isolated contacts - and changeover capability.

[richard-g8jvm]

There are both N and P ch MOSFETS available in a 4 pin DIL package , you could use 4 of these side by side.

These are nice devices and with switch a few Amps as the drain- source resistance is very low.

 

Also have a look at the Texas Instruments ULN series of multiple transistor & Mosfet arrays.

I wouldn't recommend using the 4016/66 unless reading the application notes from the various manufacturers, they are

not designed for DC level switching, but for small signal AC , audio that type of thing.

That doesn't mean you can't use them for switching DC, but if you try to pull any current through them the voltage drop will be high.

CMOS is a high impedance logic family., even switching a LED you have to be careful, you need to look at the data sheet for the source and sink

current

The data sheet may say the 4066 will handle 10mA max, but read the footnote, as long as the voltage across the switching FET doesn't exceed

0.4V, and that is the Absolute Maximum rating.

Also the channel resistance is typically 50 ohms at 12V VDD,

If you use them or any CMOS device for source or sinking current from a load, be careful if you load is even slightly inductive, as unless the output stage is designed for switching a load it won't have any reverse bias protection, so the back emf off an inductive load will punch holes

in the drain source channel.

Like all 14/16 pin DIL ICs remember the actual device inside the package will be on a substrate only a few millimeters on its longest dimension, especially these days as the same die will be used on be the smd package and the through hole , DIL, package.

Always tempting to use devices you have a tube or two of, but if like mine they have been acquired on the surplus market they tend to sit on the shelves for years before you find a use for them.

And a foot note

Modern devices have a finite shelf life once the sealed packaging is broken, you will after a while, depending of where they have been stored,

get a moisture ingression in to the package , this results in either dendrite growth or when heat is applied when soldering the water turns to steam

and the excess pressure will either crack the package or damage the substrate, so its unlikely to function as designed.

Something also the bear in mind when de-soldering devices from old boards , that may have been kept in damp conditions.

 

HTH

Richard

[Crosland]

Modern devices have a finite shelf life once the sealed packaging is broken, you will after a while, depending of where they have been stored,

get a moisture ingression in to the package , this results in either dendrite growth or when heat is applied when soldering the water turns to steam

and the excess pressure will either crack the package or damage the substrate, so its unlikely to function as designed.

Something also the bear in mind when de-soldering devices from old boards , that may have been kept in damp conditions.

 

Only the very thinnest surface mount packages are affected like this (i.e. not DIL or SO packages) and it's more of a problem with automated, rather than hand, soldering. The device manufacturer will give instructions on how to bake them, to drive off the moisture without causing damage, before procesing.

 

Andrew

[richard-g8jvm]

Here we go again , Crossland has to disagree !

 

Before we go any further State your professional qualification and experience in the semiconductor or electronics manufacturing industry.

 

Richard

[Crosland]

Here we go again , Crossland has to disagree !

 

Before we go any further State your professional qualification and experience in the semiconductor or electronics manufacturing industry.

 

Richard

I'm not into w***y waving competitions.

 

If you think what I said is wrong, then stick to arguing the facts.

 

Andrew

[Dagworth]

Here we go again , Crossland has to disagree !

 

Before we go any further State your professional qualification and experience in the semiconductor or electronics manufacturing industry.

 

Richard

Crossland is the man behind SPROG, http://www.sprog-dcc.co.uk/ Your own qualifications please?

 

Andi

[richard-g8jvm]

OK Andy

30+ years as an electronic engineer, last was a RF design engineer for EMC test equipment manufacturer/test house, RF test equipment engineer, Systems support engineer Ericsson in Sweden,Technical manager/engineer R&S Munich, development engineer Nokia ,Production engineer NEC, plus various smaller companies in RF & EMC design, production & development engineer Motorola , now retired due to ill health .

Spent a long time at a major Telelcoms investigating lifetime failure modes and redesigning to minimise failure modes in cellular base stations, as well as tracing back to the event in production which was causing early failure.

Enough ?

[Dagworth]

OK Andy

30+ years as an electronic engineer, last was a RF design engineer for EMC test equipment manufacturer/test house, RF test equipment engineer, Systems support engineer Ericsson in Sweden,Technical manager/engineer R&S Munich, development engineer Nokia ,Production engineer NEC, plus various smaller companies in RF & EMC design, production & development engineer Motorola , now retired due to ill health .

Spent a long time at a major Telelcoms investigating lifetime failure modes and redesigning to minimise failure modes in cellular base stations, as well as tracing back to the event in production which was causing early failure.

Enough ?

All that but still can't spell my name ;-)

 

Andi

[richard-g8jvm]

Humble apologies for not spotting the "y" should have been an "i"

I've had a gut full of this forum, the majority are helpful and friendly. its spoiled by a minority that just want to troll

Richard

[Crosland]

All that but still can't spell my name ;-)

 

Andi

People in glass houses... Neither of you got mine right :-)

 

Andrew

[Dagworth]

People in glass houses... Neither of you got mine right :-)

 

Andrew

I was using the spelling in the quote I was replying to ;-)

 

A

[Dagworth]

Humble apologies for not spotting the "y" should have been an "i"

I've had a gut full of this forum, the majority are helpful and friendly. its spoiled by a minority that just want to troll

Richard

I think my ratings would prove that I'm not a troll, yours however...

 

Andi

[richard-g8jvm]

I couldn't care a monkeys about ratings, your answers show you are