<<< Universal Electronic Switch >>>

Author Topic: <<< Universal Electronic Switch >>> (Read 25062 times)

Topic Options

Matt Watts

<<< Universal Electronic Switch >>> « on: 2014-08-20, 17:43:01 »

Newbie

Posts: 809

This project is something I determined was needed to allow us to focus on other principles without the distraction of building custom pulse drive circuits. My goal was to develop an electronic switch that has the following characteristics:

Digital logic control input
High speed operation
Fully isolated signal and power
High voltage and current capability
DC and AC switching similar to a true mechanical switch

What you will see in the attached images is the result of this effort, minus the heatsink.

<<< Universal Electronic Switch >>>

UniSwitch.jpg (176.5 kB, 1552x1020 - viewed 1596 times.)

<<< Universal Electronic Switch >>>

UniSwitchPCB1.6A.jpg (213.95 kB, 1274x901 - viewed 1078 times.)

<<< Universal Electronic Switch >>>

UniSwitch1.6A-Ready.jpg (134.8 kB, 647x459 - viewed 1095 times.)

« Last Edit: 2014-08-20, 18:18:19 by Matt Watts »

Matt Watts

Universal Electronic Switch -- Future Revision « Reply #1 on: 2014-08-20, 18:01:31 »

Newbie

Posts: 809

A couple things I have bumped into during the development cycle of these UniSwitches that I will likely add into a future revision:

A slightly larger board with corner holes drilled and nylon standoffs screwed in place
R3 that limits the drive current to the opto-coupler should be a 2k adjustable pot
Test point headers need to be added for easier scope probe attachment
Some routing work to get the switched signals as far away as possible from the control signals
Redo hole sizes so that component removal is slightly easier

Matt Watts

Universal Electronic Switch -- Components « Reply #2 on: 2014-08-20, 18:27:52 »

Newbie

Posts: 809

Attached are the component spec sheets I worked from for the design.

Most everything was acquired from Digi-Key and the boards manufactured at Bay Area Circuits. Schematic capture and layout was done using DipTrace.

Component source and pricing:

D1/2. http://www.digikey.com/product-search/en?KeyWords=STTH6004W 2 @ $5.83
D3. http://www.digikey.com/product-search/en?KeyWords=CP41B-WES-CK0P0154-ND 1 @ $0.34
Q1/2. http://www.digikey.com/product-search/en?KeyWords=FGH60N60SMD-ND 2 @ $7.93
U1. http://www.digikey.com/product-search/en?KeyWords=516-1651-5-ND 1 @ $3.92
U2. http://www.digikey.com/product-search/en?KeyWords=CLA356-ND 1 @ $5.12
U3. http://www.digikey.com/product-search/en?KeyWords=DG06S1212A 1 @ $20.16
J1/2/3. http://www.digikey.com/product-search/en?KeyWords=ED1605-ND 3 @ $0.75
0.1uF Caps http://www.digikey.com/product-search/en?KeyWords=399-4610-ND 3 @ $1.28
socket http://www.digikey.com/product-search/en?KeyWords=ED56083-ND 1 @ $1.58
1000uF Cap had on hand. 1
Resistors < $0.50 each 4

PCB ~ $25.00 each

I come up with just under $100 a pop at $92.73. Obviously this would be cheaper to buy components in bulk and to do a little more shopping around. Still though, if you can find a $100 hockey puck type switch that can perform this well, by all means, let me know.

DS_DG06S & D_03152012.pdf (212.38 kB - downloaded 521 times.)

STTH6004W.pdf (82.26 kB - downloaded 657 times.)

HCPL-2212.pdf (326.47 kB - downloaded 388 times.)

FGH60N60SMD.pdf (570.43 kB - downloaded 1134 times.)

IXD_614.pdf (2743.37 kB - downloaded 572 times.)

CP41B-WES,WGS.pdf (845.21 kB - downloaded 482 times.)

« Last Edit: 2014-08-22, 04:53:19 by Matt Watts »

Matt Watts	Universal Electronic Switch -- Project Files « Reply #3 on: 2014-08-20, 18:40:55 »
Newbie Posts: 809	Attached is a zip file containing the DipTrace project files needed to replicate this project. UniSwitch.zip (37.9 kB - downloaded 359 times.)

poynt99	Re: <<< Universal Electronic Switch >>> « Reply #4 on: 2014-08-20, 19:46:15 »
Group: Administrator Hero Member Posts: 3538 It's not as complicated as it may seem...	Just curious, is the switch completely isolated? Just wondering as I see a cct gnd on the igbts. --------------------------- "Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa

Matt Watts

Re: <<< Universal Electronic Switch >>> « Reply #5 on: 2014-08-20, 20:30:54 »

Newbie

Posts: 809

Quote from: poynt99 on 2014-08-20, 19:46:15

Just curious, is the switch completely isolated? Just wondering as I see a cct gnd on the igbts.

It is isolated on the signal side with the optical coupler and on the power side with the DC to DC converter. So the GND you see is only internal and no way connected to anything external. GND may not have been the best choice of bus type. I suppose I could use V_EE or V_SS depending upon the final drive devices.

I did some tests already switching AC mains and found that it works as intended. The heatsink will need to be electrically isolated with something suitable though.

poynt99	Re: <<< Universal Electronic Switch >>> « Reply #6 on: 2014-08-20, 22:33:54 »
Group: Administrator Hero Member Posts: 3538 It's not as complicated as it may seem...	In terms of voltage, how low of a sine wave will it pass undistorted? --------------------------- "Some scientists claim that hydrogen, because it is so plentiful, is the basic building block of the universe. I dispute that. I say there is more stupidity than hydrogen, and that is the basic building block of the universe." Frank Zappa

Matt Watts

Re: <<< Universal Electronic Switch >>> « Reply #7 on: 2014-08-20, 23:05:13 »

Newbie

Posts: 809

Quote from: poynt99 on 2014-08-20, 22:33:54

In terms of voltage, how low of a sine wave will it pass undistorted?

Good question. I'll have to run some tests and see. The diodes drop around 0.8 volts, so one can expect at least that much distortion for sure. I certainly wouldn't attempt to use these to switch audio signals, but with some tweaks an alternate version could probably be built to do much better.

I looked at a lot of those so-called hockey-pucks and none of them provided what these switches offer; the ones that came close were prohibitively expensive.

Matt Watts	Universal Electronic Switch -- Intro Video « Reply #8 on: 2014-08-21, 03:37:23 »
Newbie Posts: 809	Video intro: http://vimeo.com/103968642

Centraflow

Re: <<< Universal Electronic Switch >>> « Reply #9 on: 2014-08-21, 09:40:20 »

Group: Experimentalist
Hero Member

Posts: 3914

Very nice Matt

What is the shortest pulse width it can do? or did I miss that somewhere?

Really is nice and compact

regards

Mike

---------------------------

"All truth passes through three stages. First, it is ridiculed, second it is violently opposed, and third, it is accepted as self-evident."
Arthur Schopenhauer, Philosopher, 1788-1860

As a general rule, the most successful person in life is the person that has the best information.

Peterae	Re: <<< Universal Electronic Switch >>> « Reply #10 on: 2014-08-21, 10:00:32 »
Group: Administrator Hero Member Posts: 4635	Indeed Nice work Matt, i would also be interested in the rise and fall times when switching a resistive load.

Matt Watts

Re: More Testing Needed « Reply #11 on: 2014-08-22, 05:06:17 »

Newbie

Posts: 809

Thanks guys.

I'll need to setup my pulse generator and scope with the external sync and try to find the limits this thing has. I would suspect voltage and amperage will change the rise/fall times, so do you guys have any specific value of interest you would like me to focus on?

I think I'll start with rectified and filtered DC at about 160 volts using the same lamp as the load. As I recall the initial amperage against an extinguished tungsten filament is pretty high. I'll work at it until I'm certain it can do 200ns. May take some time so please bear with me.

Peterae	Re: <<< Universal Electronic Switch >>> « Reply #12 on: 2014-08-22, 08:55:12 »
Group: Administrator Hero Member Posts: 4635	A scope shot of a pulse of between 30-50V into a 10 Ohm resistor would show me what i need to know for my application. I'm always on the look out for a fast driver and fet/igbt, the specs on the data sheet looks good Thanks Peter

Matt Watts	Re: <<< Universal Electronic Switch >>> « Reply #13 on: 2014-08-23, 00:21:36 »
Newbie Posts: 809	Quote from: Peterae on 2014-08-22, 08:55:12 A scope shot of a pulse of between 30-50V into a 10 Ohm resistor would show me what i need to know for my application. Setup using 30+ volts with 8 ohm load. http://vimeo.com/104149605 Cheers!

TinMan	Re: <<< Universal Electronic Switch >>> « Reply #14 on: 2014-08-23, 03:27:06 »
Group: Elite Experimentalist Hero Member Posts: 5677	Nice build Matt. Would this work as an AC induction motor speed controller?. --------------------------- Never let your schooling get in the way of your education.

Matt Watts

UniSwitch -- Inductive Motor Speed Control « Reply #15 on: 2014-08-23, 04:47:00 »

Newbie

Posts: 809

Quote from: TinMan on 2014-08-23, 03:27:06

Nice build Matt.
Would this work as an AC induction motor speed controller?.

Thank you TinMan.

I see no reason why it would not. If you look at my first video where I was strobing the AC mains into a light bulb, clearly the power is being chopped into little discrete pieces but the main AC 60 cycle signal is still there so the motor "should" still sync on that.

I was asked by another gentleman if the inductive kickback would destroy the UniSwitch. To be quite honest, I don't know. But I did propose putting two of these switches on each side of the inductor and keep their input signal synchronized, that way when they both turn off, the inductor is completely disconnected on both legs from the power source--just like a double pole, single throw mechanical switch. In this scenario, it would seem to me electrostatic capacitance would be the only means of destroying the semi conductors. Get that far and the next thing you could try is adding a third switch that shorts the inductor--turns on when the other two switches turn off. Seems to me this would hold the magnetic field in the inductor during the off-time transition. Time it right and the inductor would never get the opportunity to develop that high voltage inductive kickback. What would actually happen...? Someone needs to try it and found out--might lead to a reduced Lenz Effect motor controller.

Peterae	Re: <<< Universal Electronic Switch >>> « Reply #16 on: 2014-08-23, 10:38:01 »
Group: Administrator Hero Member Posts: 4635	Hi Matt Thanks for doing the test, good results That's a handy switch you have designed & built Cheers Peter

TinMan

Re: UniSwitch -- Inductive Motor Speed Control « Reply #17 on: 2014-08-23, 14:46:26 »

Group: Elite Experimentalist
Hero Member

Posts: 5677

Quote from: Matt Watts on 2014-08-23, 04:47:00

Thank you TinMan.

I see no reason why it would not. If you look at my first video where I was strobing the AC mains into a light bulb, clearly the power is being chopped into little discrete pieces but the main AC 60 cycle signal is still there so the motor "should" still sync on that.

I was asked by another gentleman if the inductive kickback would destroy the UniSwitch. To be quite honest, I don't know. But I did propose putting two of these switches on each side of the inductor and keep their input signal synchronized, that way when they both turn off, the inductor is completely disconnected on both legs from the power source--just like a double pole, single throw mechanical switch. In this scenario, it would seem to me electrostatic capacitance would be the only means of destroying the semi conductors. Get that far and the next thing you could try is adding a third switch that shorts the inductor--turns on when the other two switches turn off. Seems to me this would hold the magnetic field in the inductor during the off-time transition. Time it right and the inductor would never get the opportunity to develop that high voltage inductive kickback. What would actually happen...? Someone needs to try it and found out--might lead to a reduced Lenz Effect motor controller.

Well that would be handy if it did work as a speed controller,but thinking on it a bit more,i believe that the hurtz(frequency) of the AC wave form actually has to be reduced in order to slow the speed of an AC induction motor?.I think by keeping the 60 Htz and chopping the wave,the motor would stay synced until we got rotor slip. The result would probably be lots of noise and vibration from the motor-along with a rapid loss of torque.

You could just place a 150 volt neon across the inductor to burn off most of the inductive kickback. This should kill it enough so as it dosnt do any damage else where.

---------------------------

Never let your schooling get in the way of your education.

muDped

Re: <<< Universal Electronic Switch >>> « Reply #18 on: 2014-08-23, 21:51:44 »

Group: Tinkerer
Hero Member

Posts: 3325

Excellent project Matt Watts!

High speed isolated switches such as the one you've
fabricated are very useful devices.

Switching with variable pulse width at a high frequency
such as shown in your video is an excellent way to
emulate electronically the voltage control capability of
the Variac.

Your switch could easily, as you hinted at in your video,
be adapted to make an effective desulfator/charger for
lead acid batteries. Narrow pulses at a frequency of
One to Three KiloHerz does the job well. With very
narrow pulses the average power delivered to the battery
would be low and quite safe.

Yes indeed, lots of creative possibilities...

---------------------------

For there is nothing hidden that will not be disclosed, and nothing concealed that will not be known or brought out into the open.

forest	Re: <<< Universal Electronic Switch >>> « Reply #19 on: 2014-11-21, 11:53:47 »
Hero Member Posts: 512	Excellent I also concerned about flyback spike (kickback) possibility to damage circuit. What can be done to avoid it yet still allow to experimenting with BEMF ? Now I think would be good to have a simple square wave generator. Just 50% duty cycle is eanough for me , but what I think is more important is to have a way to release just one single pulse with pressing a button

Matt Watts

Re: <<< Universal Electronic Switch >>> « Reply #20 on: 2014-11-21, 12:15:25 »

Newbie

Posts: 809

Quote from: forest on 2014-11-21, 11:53:47

Excellent

I also concerned about flyback spike (kickback) possibility to damage circuit. What can be done to avoid it yet still allow to experimenting with BEMF ? Now I think would be good to have a simple square wave generator. Just 50% duty cycle is eanough for me , but what I think is more important is to have a way to release just one single pulse with pressing a button

Thank you for the compliment. I've used these (built two of them) quite a bit lately and they seem pretty robust. Unlike a typical MOSFET switch, these are dual and in series so they switch the current in both directions; that could be why I haven't blown them up yet. With a flyback spike, the energy has to go somewhere and with these switches, they completely disconnect from the circuit when they turn off, so I don't think they get smacked like a single transistor would.

I've tried them on some pretty big inductors with success so far. Can't quite understand why Mr. John B. doesn't use something similar with his SG. From the way Mr. Lindemann explains the SG, it's real easy with their design to pop semiconductors.

Using IGBTs, they do warm up a bit if you punch 30 amp spikes through them, but you also have the flexibility to push 300 volt at a 1/2 amp instead if that better matches your output impedance. They're about as Universal as I know how to make them. Not super fast, not super high current; not super high voltage, just something in the middle that can be used for lots of testing.