1. Home
  2. Electronics Design
  3. high voltage blinker

high voltage blinker

This might work. The idea is to build a static Schmitt trigger that switches on its own supply voltage. Q1C is low below VK=4 volts, high above 12 volts, has hysteresis between, and doesn't depend on any capacitors or charge storage or anything tricky like that. The parts values are futzed, so it probably needs tuning, but that's the general idea.

I verified that a white Cree LED makes a healthy flash if you hard-discharge a

1uF cap at 10 volts into it. So I guess R4 isn't necessary, but it makes the sim a little nicer.

One nice feature is that all the current from R7 must flow through the LED. So the blinker circuit could fail various ways and the LED will still glow at dangerous voltages.

Version 4 SHEET 1 1232 680 WIRE -288 -416 -352 -416 WIRE -224 -416 -288 -416 WIRE -48 -416 -144 -416 WIRE 32 -416 -48 -416 WIRE 144 -416 32 -416 WIRE 304 -416 208 -416 WIRE 528 -416 304 -416 WIRE 656 -416 528 -416 WIRE 784 -416 656 -416 WIRE 1120 -416 784 -416 WIRE 1120 -352 1120 -416 WIRE 304 -336 304 -416 WIRE 528 -336 528 -416 WIRE 784 -336 784 -416 WIRE 32 -304 32 -416 WIRE -352 -272 -352 -416 WIRE 1120 -224 1120 -272 WIRE -352 -144 -352 -192 WIRE 32 -144 32 -240 WIRE 304 -144 304 -256 WIRE 368 -144 304 -144 WIRE 784 -144 784 -256 WIRE 784 -144 448 -144 WIRE 912 -144 784 -144 WIRE 1072 -144 912 -144 WIRE 784 -80 784 -144 WIRE 1120 -64 1120 -128 WIRE 528 -32 528 -256 WIRE 640 -32 528 -32 WIRE 720 -32 640 -32 WIRE 784 64 784 16 WIRE 528 80 528 -32 WIRE 304 128 304 -144 WIRE 384 128 304 128 WIRE 464 128 384 128 WIRE 304 192 304 128 WIRE 528 224 528 176 WIRE 304 336 304 272 FLAG -352 -144 0 FLAG 784 64 0 FLAG 304 336 0 FLAG 528 224 0 FLAG 1120 -64 0 FLAG 32 -144 0 FLAG 640 -32 Q1B FLAG 912 -144 Q1C FLAG 384 128 Q2B FLAG -288 -416 V1 FLAG -48 -416 VC1 FLAG 656 -416 VK SYMBOL voltage -352 -288 R0 WINDOW 0 57 38 Left 2 WINDOW 3 52 72 Left 2 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V1 SYMATTR Value SINE(100 100 10m) SYMBOL res -128 -432 R90 WINDOW 0 76 58 VBottom 2 WINDOW 3 85 59 VTop 2 SYMATTR InstName R7 SYMATTR Value 3Meg SYMBOL npn 720 -80 R0 WINDOW 0 122 28 Left 2 WINDOW 3 97 68 Left 2 SYMATTR InstName Q1 SYMATTR Value BC847C SYMBOL res 768 -352 R0 WINDOW 0 75 52 Left 2 WINDOW 3 62 88 Left 2 SYMATTR InstName R1 SYMATTR Value 2Meg SYMBOL res 288 176 R0 WINDOW 0 -78 21 Left 2 WINDOW 3 -93 59 Left 2 SYMATTR InstName R2 SYMATTR Value 400K SYMBOL res 288 -352 R0 WINDOW 0 63 65 Left 2 WINDOW 3 44 95 Left 2 SYMATTR InstName R3 SYMATTR Value 10Meg SYMBOL npn 464 80 R0 WINDOW 0 105 29 Left 2 WINDOW 3 83 67 Left 2 SYMATTR InstName Q2 SYMATTR Value BC847C SYMBOL res 512 -352 R0 WINDOW 0 73 61 Left 2 WINDOW 3 48 92 Left 2 SYMATTR InstName R5 SYMATTR Value 20Meg SYMBOL res 464 -160 R90 WINDOW 0 76 54 VBottom 2 WINDOW 3 85 51 VTop 2 SYMATTR InstName R6 SYMATTR Value 4Meg SYMBOL nmos 1072 -224 R0 WINDOW 0 -71 107 Left 2 WINDOW 3 -95 139 Left 2 SYMATTR InstName M1 SYMATTR Value BSS123 SYMBOL cap 16 -304 R0 WINDOW 0 -58 27 Left 2 WINDOW 3 -58 58 Left 2 SYMATTR InstName C1

SYMBOL LED 144 -400 R270 WINDOW 0 -41 32 VTop 2 WINDOW 3 -47 31 VBottom 2 SYMATTR InstName D1 SYMATTR Value QTLP690C SYMATTR Description Diode SYMATTR Type diode SYMBOL res 1104 -368 R0 WINDOW 0 -67 54 Left 2 WINDOW 3 -73 91 Left 2 SYMATTR InstName R4 SYMATTR Value 100 TEXT -192 176 Left 2 !.tran 200 uic TEXT -240 56 Left 2 ;SCHMITT BLINKER TEXT -216 104 Left 2 ;JL July 17 2014

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation
Reply to
John Larkin
Loading thread data ...

hes on

s, has

e or

eds

The BSS123 gate threshold is too low to guarantee Q1 is conducting enough a t low voltage on C1 to prevent it from hanging the Schmitt.

rge a

the sim

D. So

t
Reply to
bloggs.fredbloggs.fred

Nice idea - might work! R4 drop just enough to keep M1 enhanced long enough to be visible.

piglet

Reply to
piglet

I'd say it is essential otherwise when M1 enhances it will starve itself of gate drive.

piglet

Reply to
piglet

Is a 60mA, 170us-wide pulse enough to see?

Reply to
John S

John Larkin wrote:

I changed the input to 25 meg and 100nf to give a reasonable blink rate at 48V.

The minimum voltage was quite high, around 220V. I increased R2 to try to reduce the minimum oscillation voltage. Got it down to 110V, but it puts out one pulse and hangs. The pulse width above 1 mA is around 82 us. About the same as mine, but mine went down to 31V.

See if you can fiddle with it.

Version 4 SHEET 1 1232 680 WIRE -80 -416 -96 -416 WIRE -48 -416 -80 -416 WIRE 64 -416 32 -416 WIRE 96 -416 64 -416 WIRE 128 -416 96 -416 WIRE 256 -416 192 -416 WIRE 464 -416 256 -416 WIRE 496 -416 464 -416 WIRE 656 -416 496 -416 WIRE 832 -416 656 -416 WIRE -96 -400 -96 -416 WIRE 96 -400 96 -416 WIRE 256 -400 256 -416 WIRE 464 -400 464 -416 WIRE 656 -400 656 -416 WIRE 832 -400 832 -416 WIRE 96 -320 96 -336 WIRE -96 -304 -96 -320 WIRE 832 -304 832 -320 WIRE 256 -224 256 -320 WIRE 288 -224 256 -224 WIRE 656 -224 656 -320 WIRE 656 -224 368 -224 WIRE 688 -224 656 -224 WIRE 784 -224 688 -224 WIRE 656 -192 656 -224 WIRE 832 -192 832 -208 WIRE 464 -144 464 -320 WIRE 560 -144 464 -144 WIRE 592 -144 560 -144 WIRE 464 -112 464 -144 WIRE 656 -80 656 -96 WIRE 256 -64 256 -224 WIRE 368 -64 256 -64 WIRE 400 -64 368 -64 WIRE 256 -48 256 -64 WIRE 464 0 464 -16 WIRE 256 48 256 32 FLAG -96 -304 0 FLAG 656 -80 0 FLAG 256 48 0 FLAG 464 0 0 FLAG 832 -192 0 FLAG 96 -320 0 FLAG 560 -144 Q1B FLAG 688 -224 Q1C FLAG 368 -64 Q2B FLAG -80 -416 V1 FLAG 64 -416 VC1 FLAG 496 -416 VK SYMBOL res 48 -432 R90 WINDOW 0 -7 53 VBottom 2 WINDOW 3 39 53 VTop 2 SYMATTR InstName R7 SYMATTR Value 25e6 SYMBOL npn 592 -192 R0 WINDOW 0 72 11 Left 2 WINDOW 3 72 76 Left 2 SYMATTR InstName Q1 SYMATTR Value BC847C SYMBOL res 640 -416 R0 WINDOW 0 43 25 Left 2 WINDOW 3 33 92 Left 2 SYMATTR InstName R1 SYMATTR Value 2Meg SYMBOL res 240 -64 R0 WINDOW 0 42 35 Left 2 WINDOW 3 38 65 Left 2 SYMATTR InstName R2 SYMATTR Value 820e3 SYMBOL res 240 -416 R0 WINDOW 0 44 29 Left 2 WINDOW 3 38 92 Left 2 SYMATTR InstName R3 SYMATTR Value 10Meg SYMBOL npn 400 -112 R0 WINDOW 0 64 17 Left 2 WINDOW 3 60 64 Left 2 SYMATTR InstName Q2 SYMATTR Value BC847C SYMBOL res 448 -416 R0 WINDOW 0 45 25 Left 2 WINDOW 3 42 93 Left 2 SYMATTR InstName R5 SYMATTR Value 20Meg SYMBOL res 384 -240 R90 WINDOW 0 2 60 VBottom 2 WINDOW 3 33 59 VTop 2 SYMATTR InstName R6 SYMATTR Value 4Meg SYMBOL nmos 784 -304 R0 WINDOW 0 64 -2 Left 2 WINDOW 3 59 68 Left 2 SYMATTR InstName M1 SYMATTR Value BSS123 SYMBOL cap 80 -400 R0 WINDOW 0 25 11 Left 2 WINDOW 3 23 48 Left 2 SYMATTR InstName C1 SYMATTR Value 100nf SYMBOL LED 128 -400 R270 WINDOW 0 27 76 VTop 2 WINDOW 3 -4 58 VBottom 2 SYMATTR InstName D1 SYMATTR Value QTLP690C SYMATTR Description Diode SYMATTR Type diode SYMBOL res 816 -416 R0 WINDOW 0 46 35 Left 2 WINDOW 3 44 71 Left 2 SYMATTR InstName R4 SYMATTR Value 100 SYMBOL voltage -96 -416 M0 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 WINDOW 3 -49 96 Left 2 SYMATTR Value 110V SYMATTR InstName V1 TEXT 240 -496 Left 2 !.tran 2 uic TEXT 240 -520 Left 2 ;'JL's Schmitt Blinker 110V Input TEXT -96 -224 Left 2 ;PULSE(0 50e3 0 2 1u 1u)

Here is the PLT file

[Transient Analysis] { Npanes: 4 { traces: 2 {524294,0,"V(q2b)"} {34603012,1,"I(D1)"} X: (' ',1,0,0.2,2) Y[0]: ('m',0,0,0.05,0.5) Y[1]: ('m',0,0,0.002,0.02) Volts: ('m',0,0,0,0,0.05,0.5) Amps: ('m',0,0,0,0,0.002,0.02) Log: 0 0 0 GridStyle: 1 }, { traces: 1 {268959749,0,"V(vk)"} X: (' ',1,0,0.2,2) Y[0]: (' ',0,0,1,13) Y[1]: ('_',0,1e+308,0,-1e+308) Volts: (' ',0,0,0,0,1,13) Log: 0 0 0 GridStyle: 1 }, { traces: 1 {268959746,0,"V(vc1)"} X: (' ',1,0,0.2,2) Y[0]: (' ',0,0,1,14) Y[1]: ('_',0,1e+308,0,-1e+308) Volts: (' ',0,0,0,0,1,14) Log: 0 0 0 GridStyle: 1 }, { traces: 1 {268959747,0,"V(v1)"} X: (' ',1,0,0.2,2) Y[0]: (' ',0,0,20,220) Y[1]: ('_',0,1e+308,0,-1e+308) Volts: (' ',0,0,0,0,20,220) Log: 0 0 0 GridStyle: 1 } }
Reply to
Steve Wilson

Replace the sine generator with a DC set for 40V and it will run as John L has designed it.

Reply to
John S

Sure. I just used the parts in the LT Spice standard library. I'd use an FDV301, lower gate threshold, which we have in stock here. As I said, the parts values here were just guesses, and proper analysis would need to be done before this is finished.

It's the principle that matters now: a capacitor-free, latchup-free, no PNPN, Schmitt trigger. Inspired by my great respect for the NE-2, which has a peak point current of picoamps and a valley current sufficient to melt the electrodes and explode the glass.

--

John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

Have you figured out how to stop it from hanging below 110V?

Might as well use a NE-2. It would be much brighter discharging a cap from

120V than trying to drive a LED at 82us pulse width.

Of course, you might try having the control circuit monitor the HV decay before shutting itself off.

Then you can blink the LEDs any way you wish.

Reply to
Steve Wilson

Starving itself is OK. By the time the VK voltage gets down to the lower Schmitt threshold, about 4 volts, we've dumped the desired energy into the LED. As I've noted, I'd use a lower threshold mosfet in real life, and check all the math.

Point is, it's the first good circuit I've seen for this little problem.

--

John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

Should be. A 10 mA, 1 ms pulse looks good into the Cree white LED. And if I charge up a 1uF mylar cap to 10 volts, and slap it across the LED, it makes a nice fat flash.

R4 could be higher, to spread out the LED pulse, too.

--

John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

On Thu, 17 Jul 2014 15:20:55 -0500, John S wrote:

It looks a little more sane with ramps...

Again, the values were guessed and need proper engineering.

Version 4 SHEET 1 1232 680 WIRE -288 -416 -352 -416 WIRE -224 -416 -288 -416 WIRE -48 -416 -144 -416 WIRE 32 -416 -48 -416 WIRE 144 -416 32 -416 WIRE 304 -416 208 -416 WIRE 528 -416 304 -416 WIRE 656 -416 528 -416 WIRE 784 -416 656 -416 WIRE 1120 -416 784 -416 WIRE 1120 -352 1120 -416 WIRE 304 -336 304 -416 WIRE 528 -336 528 -416 WIRE 784 -336 784 -416 WIRE 32 -304 32 -416 WIRE -352 -272 -352 -416 WIRE 1120 -224 1120 -272 WIRE -352 -144 -352 -192 WIRE 32 -144 32 -240 WIRE 304 -144 304 -256 WIRE 368 -144 304 -144 WIRE 784 -144 784 -256 WIRE 784 -144 448 -144 WIRE 912 -144 784 -144 WIRE 1072 -144 912 -144 WIRE 784 -80 784 -144 WIRE 1120 -64 1120 -128 WIRE 528 -32 528 -256 WIRE 640 -32 528 -32 WIRE 720 -32 640 -32 WIRE 784 64 784 16 WIRE 528 80 528 -32 WIRE 304 128 304 -144 WIRE 384 128 304 128 WIRE 464 128 384 128 WIRE 304 192 304 128 WIRE 528 224 528 176 WIRE 304 336 304 272 FLAG -352 -144 0 FLAG 784 64 0 FLAG 304 336 0 FLAG 528 224 0 FLAG 1120 -64 0 FLAG 32 -144 0 FLAG 640 -32 Q1B FLAG 912 -144 Q1C FLAG 384 128 Q2B FLAG -288 -416 V1 FLAG -48 -416 VC1 FLAG 656 -416 VK SYMBOL voltage -352 -288 R0 WINDOW 0 71 75 Left 2 WINDOW 3 31 119 Left 2 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V1 SYMATTR Value PULSE(0 100 0 50 50 10) SYMBOL res -128 -432 R90 WINDOW 0 76 58 VBottom 2 WINDOW 3 85 59 VTop 2 SYMATTR InstName R7 SYMATTR Value 3Meg SYMBOL npn 720 -80 R0 WINDOW 0 122 28 Left 2 WINDOW 3 97 68 Left 2 SYMATTR InstName Q1 SYMATTR Value BC847C SYMBOL res 768 -352 R0 WINDOW 0 75 52 Left 2 WINDOW 3 62 88 Left 2 SYMATTR InstName R1 SYMATTR Value 2Meg SYMBOL res 288 176 R0 WINDOW 0 -78 21 Left 2 WINDOW 3 -93 59 Left 2 SYMATTR InstName R2 SYMATTR Value 400K SYMBOL res 288 -352 R0 WINDOW 0 63 65 Left 2 WINDOW 3 44 95 Left 2 SYMATTR InstName R3 SYMATTR Value 10Meg SYMBOL npn 464 80 R0 WINDOW 0 105 29 Left 2 WINDOW 3 83 67 Left 2 SYMATTR InstName Q2 SYMATTR Value BC847C SYMBOL res 512 -352 R0 WINDOW 0 73 61 Left 2 WINDOW 3 48 92 Left 2 SYMATTR InstName R5 SYMATTR Value 20Meg SYMBOL res 464 -160 R90 WINDOW 0 76 54 VBottom 2 WINDOW 3 85 51 VTop 2 SYMATTR InstName R6 SYMATTR Value 4Meg SYMBOL nmos 1072 -224 R0 WINDOW 0 -71 107 Left 2 WINDOW 3 -95 139 Left 2 SYMATTR InstName M1 SYMATTR Value BSS123 SYMBOL cap 16 -304 R0 WINDOW 0 -58 27 Left 2 WINDOW 3 -58 58 Left 2 SYMATTR InstName C1

SYMBOL LED 144 -400 R270 WINDOW 0 -41 32 VTop 2 WINDOW 3 -47 31 VBottom 2 SYMATTR InstName D1 SYMATTR Value QTLP690C SYMATTR Description Diode SYMATTR Type diode SYMBOL res 1104 -368 R0 WINDOW 0 -75 66 Left 2 WINDOW 3 -78 99 Left 2 SYMATTR InstName R4 SYMATTR Value 100 TEXT -192 176 Left 2 !.tran 100 uic TEXT -240 56 Left 2 ;SCHMITT BLINKER TEXT -216 104 Left 2 ;JL July 17 2014

--

John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

But surely without R4 that time will reduce to the schmitt bjts storage and miller delays? I say keep R4.

piglet

Reply to
piglet

Your requirement was 1 watt dissipation in the bleeder. At 5KV, that is

25 megohms.

You are using a 3 megohm resistor. The dissipation at 5KV is

5e3^2/3e6 = 8.3 watts.

I posted your circuit with a 25 megohm resistor. It hung at 220V. It fails to meet your requirement to work at 48V.

You said you would prefer a circuit that did not hang if you could avoid it. Yours hangs.

Your discharge cap is 1 uF. That would give a pulse at approximately every 10 seconds. No tech will wait that long to see if the high voltage has decayed. A more reasonable flash rate is 1 Hz. At that rate, the LED shows a pulse of about 82 us above 1 mA. As you have stated, it will probably be too dim to see.

Please apply the same criteria to your circuit that you have applied to everyone else.

Reply to
Steve Wilson

[LTSpice ciruit]

150usec flashes would be hard to see. I'd make R4 larger. 1k results in

1.4msec flashes.

I've posted another one in the original thread.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

My sim seems to work at 40 volts.

--

John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

It doesn't hang; it just quits working at some minimum voltage. This circuit could be juggled to work at 48V and 25M, but it can't generate energy so there's a fundamental limit on average LED output.

It looks like I can use a lower HV supply, 2000 or maybe 2400 volts, so I can stand a 3M resistor string. The whole system design is still evolving.

The criterion that I have applied is that the circuit be reliable and have no hangup states. And be reasonably simple.

--

John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

I tried connecting a 10-volt-charged, 1 uF mylar cap right across the Cree LED, and got a nice bright flash, so the value of R4 doesn't seem to matter much. But it could certainly be tweaked for max visual brightness. Efficiency probably drops at high peak currents.

--

John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

I will, so that I can control peak LED current. But with no R4, it's still got to blink.

One concept of this circuit is that it does NOT depend on charge stored in caps or semiconductor junctions or SCR peak/valley ratios or things scurrying around in circles.

This circuit is still an idea to be developed, not a perfect posted thing. I don't mind evolving things in public.

--

John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

Yeah, it's a matter of trial and error, see what gives the most visible flash.

It's a lot of parts though. With the NCP302 you'd save two resistors and probably one SOT23 part because it's got the threshold detector and inverting function built in. But your circuit should work as well.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.