36V Strobe Circuit

By strobe do you mean :-

1. A circuit with a very short flash, which can be synchronized to mains electricity or a shaft to make fast-turning machines look stationary?
2. A "tump a th' beat" audio thing?
3. Why 36V? Is it an AC or DC supply?
4. Is the lamp incandescent filament type? If they are labelled 1 amp, initial current may be 20 amps when flashing.
5. How long a flash duration at 5/second?

If you want a 100 microsecond flash from LED lamps, a 2N6027 "unijunction" will handle 100V at 1 amp peak 100 microsecond and be simpler than 555 + FET. Anything more complex and you do not get much simpler than 555+FET for DC.

A scrap automotive "turn" electronic/relay flasher might have accessible on/off time resistors to modify to 5/sec. You would have to fit on a resistor /zener supply for the timer and a coil ballast for the relay.

For AC supply, there are circuits with unijunctions and SCRs or Triacs or even Diacs.

"Off the shelf" in CR4 is an elastic thing - are you in California or the Sahara?

The 2N6027 looks interesting. 114W...

You think my best bet is 555/FET switching? I was just hoping someone knew of a timing chip that could handle and switch 30+V.

I guess back to the drawing board....

Thanks!!

A 555 low voltage oscillator circuit will work and strobe rate can be set to whatever desired frequency. The oscillator output can then drive a very small relay that in turn will switch on/off the power of whatever level of voltage you desired to operate on... doesn't matter be it AC or DC level.

Well, this one is simpler than a 555, but still needs resistor/zener supply and a relay or FET for 36V. Only £0.81 each

http://www.farnell.com/datasheets/29617.pdf

I don't know what the flash rate of a 36v forklift warning light is, but it should be off-the-shelf.

Some years ago, I designed a very simple DC flasher which used a CMOS 4060 as a gate driver for an N-channel FET. The design was good for 12V to 48V and kept the logic voltage down with a 8.2V zener diode. Your responsibility would be to pick the appropriate lamp and socket for your lighting needs.

Please note: At 5 flashes per second, you are going to have a problem with lamp filament fatigue. Drop the rate to 90 fpm (flashes per minute) and success will be yours. Use a duty cycle of 1:8 (the 4060 handles this easily). A good FET for this purpose is the IRF530 and no heat-sink is necessary since power dissipation is very low, even at 3 amperes.

The product is probably still available from Trilite/MARS and the purchase will save you lots of grief. They advertise 12VDC and 24VDC versions but the electronics is good to over 50VDC. The only difference between models is the selection of lamp.

Contact me personnally for more details, if needed.

A power version of a 555. Aha, yes, you can get those from IR.

A classic 555 requires an R and a C to set the frequency, and unless it's a CMOS type you have to fool around a bit the get a 50% duty cycle. If you want to make a power device from that you certainly need to add a MOSFET or two, but you often find yourself adding all kinds of other stuff, like a low-voltage regulator for the 555 chip, and maybe a high-voltage level-shifter for one of the MOSFETs. These things are provided in IR's family of self-oscillating drivers. Yes, drivers, that's right, you do have to add the usual external MOSFET(s).

To pick two parts from their lineup, first I'll suggest the IRS2153 (datasheet link). This part works from 15 to 600 volts, and can drive two MOSFETs in what's called a half-bridge configuration. That's just a way of saying it actively drives the output both low and high, so it looks like a high-voltage CMOS output. Or you can use just a single MOSFET. The 2153 has a 1.1us deadtime between low and high gate drives, so it's best used below say 200kHz. It has 180mA gate-drive capability, to deal with your MOSFET's high gate capacitance. For the "555" portion of the IC's power, it includes a 15-volt zener diode, so you just add an external resistor to power V_CC from your high input voltage.

The '2153 costs $2 at DigiKey.

If you look at the '2153 block diagram, below,

you'll recognize the 555-style oscillator circuit on the left. Then they've added dead-time delays, a level-shifter and two gate-drivers. Also note the zener diode on the V_CC pin. The '2153 costs $2 at DigiKey.

Second, I like their IR2085, or actually, AUIR2085, as they like to identify some of their power parts as "automotive grade." This is meant for use from 8 to 100 volts, and up to 500kHz (it can be tweaked to 1MHz, try C_T=22pF). For the highest frequency ranges it has a fast sub-100ns deadtime between the upper and lower MOSFET gate-drives. The '2085 provides a serious ±1A gate drive capability, so you can get really speedy rise and fall times from your MOSFETs, even down to under 20ns. They claim 25ns high and low pulse-width matching. But you are limited to 100 volts. The '2085 does not include an internal zener on the V_CC supply pin.

Both of these 555-style self-oscillating power drivers run at a 50% duty cycle. But you can add a diode and resistor to force other duty cycles of your choice. You can add a pot to get a variable duty cycle.

I have used the IR2085 to make a high-current DC step-up / step-down transformer, where the DC transformer ratio is determined by the duty cycle. And yes, the step-up/step-down works in both directions, just like an AC transformer. The above-mentioned pot set the DC transformer ratio. If you use it to reduce the output voltage, you get an increased output current, just as with an AC transformer. And vice versa if you drive the current back the other way. I called it a bus converter. But that's a story for another day.

"I have used the IR2085 to make a high-current DC step-up / step-down transformer, where the DC transformer ratio is determined by the duty cycle. And yes, the step-up/step-down works in both directions, just like an AC transformer. The above-mentioned pot set the DC transformer ratio. If you use it to reduce the output voltage, you get an increased output current, just as with an AC transformer. And vice versa if you drive the current back the other way. I called it a bus converter. But that's a story for another day."

Already well covered on CR4, Re: Why Not DC? Yes, the DC Bus Converter. (link).

So you think this will work by allowing me to power it with 36v in and pulse 36v out at 12-20hz?