Help, please, selecting some Electronic Components

[John HaineParticipant @johnhaine32865]

The diode is totally unnecessary since it is always reverse biased! I suspect that the "designer" adapted a circuit originally intended to drive an inductive load which would need a diode to catch the inductive kick on switch off. Not understanding why it was there they left it in just in case. From other artefacts of the design understanding was thin on the ground.

The MOSFET seems to be a fairly generic power device and other types would be suitable.  I don't know where your pulses are coming from but the threshold voltage of the device specified can be as high as 4v which is a bit marginal for a 5v CMOS drive.  

Edited By John Haine on 17/03/2023 11:52:12

[Michael GilliganParticipant @michaelgilligan61133]

Goodness Gracious !!

I thought RS was a little pricey, but:

**LINK** https://www.ebay.co.uk/itm/171948255344

MichaelG.

[Martin KyteParticipant @martinkyte99762]

Posted by Joseph Noci 1 on 17/03/2023 10:21:50:

Posted by Martin Kyte on 17/03/2023 09:37:15:

So C3 not connected to the input then.

The limit on repetition rate is going to be dependent on the time constant of the 100 ohm resistor R1 and the capacitor C2. Which with the values listed is 0.33s giving you about 3 Hz.

You could split C2 into two or more components which would shorten the time constant of each.
I think Joseph Noci said as much above.
Regards Martin .

Caps in parallel add, so final RC remains the same.

Yes I know. Muddled thinking for a bit. Engage brain before opening mouth and all that. It really wants a switching device across the 100R.
Not the most brilliant cct I’ve seen there are loads of High Power led drivers that would be more suitable.

regards Martin

[Robert Atkinson 2Participant @robertatkinson2]

Posted by Clive Steer on 17/03/2023 11:00:22:

White LEDs are usually constructed using a blue/UV led covered with a white fluorescent material and the shade of white depends on the materials recipe. I suspect that the material may have a degree of persistence that would moderate 10Mhz operating.

I used a similar circuit to make an LED strobe light to visually check the operation and timing of watch escapements but these operate at relatively low frequency.

CS

I'm late to this thread,

Clives description of a white LED is correct and thoughts on persistence reasonable. Indeed I made a similar assumption when looking for flash lamp for a day job instrument design.

I was surprised to find the LED was very fast. It actually had faster turn of and less afterglow than a xenon tube. Even a short arc tube with quench circuit was slower. The limiting factor on measurement was the photodiode I was using.
So don't dismiss the white LED.

Robert.

Edited By Robert Atkinson 2 on 17/03/2023 12:36:37

[Michael GilliganParticipant @michaelgilligan61133]

It’s obviously of no great consequence, but …

I have to admit that in several posts I have mentioned Fig. 4(a)

… this should, of course read Fig. 3(a)

MichaelG.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Robert Atkinson 2 on 17/03/2023 12:35:49:

[…]

So don't dismiss the white LED.

.

Don’t worry, Robert … That is most unlikely in my case !!

MichaelG

[GeorgineerParticipant @georgineer]

Posted by Michael Gilligan on 17/03/2023 11:23:49

.… I mourn the passing of those helpful little ‘hump-back bridges’ that we had on old drawings !

MichaelG.

That and the four-way wire junctions made me suspicious from first reading of the circuit. That style of drawing was superseded by straight crossings and 'blobs' at junctions in… what?.. the 1960s?

George

[Maurice TaylorParticipant @mauricetaylor82093]

Hi, Why not get the components and make the circuit and try it.,then you will know if it’s suitable.

Maurice

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Maurice Taylor on 17/03/2023 16:09:38:

Hi, Why not get the components and make the circuit and try it.,then you will know if it’s suitable.

Maurice

.

Because some of the components are not adequately specified

I asked for advice here in the hope of avoiding problems caused by sub-optimal choices.

… Type and brand-name of capacitors with the lowest available ESR, and that sort of thing.

Hoping to benefit from the experience of others.

MichaelG.

[Michael GilliganParticipant @michaelgilligan61133]

[ UPDATE ]

I have made contact with Susan Parker [who designed the Xenon Flash part of the system] and although this little circuit is not her design she has kindly commented regarding the Diode:

I suspect the D1 is there as the LED is on a cable and it snubs cable reflections. There LEDs can get quite expensive, so it’s a just in case belt and braces part. Whether it ever actually does anything is of course another matter.

I hope to have further discussion sometime over the weekend.

MichaelG.

[Sonic EscapeParticipant @sonicescape38234]

I tried a simple discrete mosfet driver out of curiosity. If I understand correctly the goal is to drive the LED with pulses of 1-2µs? This is my schematic. I just used components that happened to have around:

I decreased the value of the drain resistor. I think it make sense to have more current through LED. The circuit is supplied at 5V. This are the waveforms for 1µs pulse:

Yellow is in the drain of T3 and the blue trace is from the base of T1 and T2. This is with a 200ns pulse:

I didn't switch the normal ground wires of the oscilloscope probe with the low inductance springs. I guess that might cause that ringing. This is the circuit:

The rising/falling edges looks good. Increasing supply voltage doesn't improves them. But increasing the driving voltage amplitude improves the edges significantly.
     T3 turns off very fast almost completely. But then it need a lot to fully switch off. Since the LED emission vs voltage is non-linear this long tail might not be noticed.  

 

Edited By Sonic Escape on 17/03/2023 17:09:42

Edited By Sonic Escape on 17/03/2023 17:10:17

[Robert Atkinson 2Participant @robertatkinson2]

OK ,

I've desigined and built exactly this type of circuit.

1/ The 0.1uF on the input of the driver chip is incorrect. Probably a misdraw of the datasheet test circuit which has two power input bypass capacitors.

2/ The two 0.1uF capacitors on the driver chip should be connected directly across the IC pins. One between 1 and 4 the other between 5 and 8. Use ceramic capacitors.

3/ The diode is required. At these switching speeds the parasitic inductances can cause enough bacj EMF to damage the LED. Connect the diode directly across the LED pins.

4/ I would not worry too much about the inductance of the 100 ohm resistor, just don't use a wirewound. A film type is good choice.

5/ Any modern low ESR electrolytics will be OK.

You need a neat layout with short lead lengths and a good earth. I'd do it "dead bug" style on a bit of plain copper board.

Robert.

[Michael GilliganParticipant @michaelgilligan61133]

Thanks, both

Definite progress in my comprehension

MichaelG.

.

Still a little bewildered by the specified component for R1 … I wonder if they were just lucky enough to have one in the store-cupboard

MichaelG.

[Robert Atkinson 2Participant @robertatkinson2]

Posted by Michael Gilligan on 17/03/2023 11:51:38:

Goodness Gracious !!

I thought RS was a little pricey, but:

**LINK** https://www.ebay.co.uk/itm/171948255344

MichaelG.

That seller is just listing RS aand Farnell parts at huge markup. They hold no stock, just order as required. Ther are a number of similar sellers. Bizzarly they actually have lots of sales.

If you want a low inductance 100R thick film power resistor PM me and I'll send you one. It might be used depends what I find first….

Robert.

Robert.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Robert Atkinson 2 on 17/03/2023 18:38:34:

If you want a low inductance 100R thick film power resistor PM me and I'll send you one. It might be used depends what I find first….

Robert.

.

Done !

Thank You

MichaelG.

[Sonic EscapeParticipant @sonicescape38234]

The self inductance of R1 doesn't matter here. It will work exactly the same way even a wire resistor

Edited By Sonic Escape on 17/03/2023 19:26:07

Edited By Sonic Escape on 17/03/2023 19:26:18

[Robert Atkinson 2Participant @robertatkinson2]

Inductance in R1 will make a difference. It will slightly increase the voltage on the capacitor. But given the relative values I'd agree it can probably be ignored.

Robert.

[Clive SteerParticipant @clivesteer55943]

Thanks Robert for the update about the rise and fall rates for light emitted from a white LED.

CS

[Michael GilliganParticipant @michaelgilligan61133]

Another update :

Probably my final one on this thread; but those who spotted the big error in the Fig. 3(a) schematic are welcome to proclaim “I told you so !”

Over the weekend, Susan Parker kindly had a look at the schematic,

and, in answer to my questions, advised me:

1. The drawing is indeed wrong, C3 [at the input] is a decoupling cap.
2. If the diode is to do anything useful, then a Schottky part should be used.
3. The resistor should be low inductance as a matter of principle, but otherwise is immaterial as to the tolerance etc. since is is the far side of the main reservoir cap. It does need to be of sufficient power as it is recharging the 3300uF cap.
4. Note however this circuit is NOT suitable for your application, the 100R and 3300uF time constant is way too slow for your frame/strobe rate.

So … my understanding is that the circuit is good for the originally intended purpose: BUT the input is wrongly drawn and needs to better reflect the diagram in the data-sheet [which I posted earlier].

But unfortunately it is not fast enough for my proposed frame-rate.

Thanks, everyone, for your comments.

I will probably build one anyway, just for the short flash-duration, even if it doesn’t cycle at the rate I want.

MichaelG.

[Robert Atkinson 2Participant @robertatkinson2]

If you want a faster cycle time reduce the value of the storage capacitor and increase the voltage at the supply end of the 100R resistor. Keep the energy the same. Bear in mind the stored enery is half the capacitance times the SQUARE of the voltage so if tou double the voltage use a quarter of the capacitance. So for 12V you want about 560uF. In practice the preferred value of 470uF will be fine. This just sets the flash duration IF the pulse is longer than the RC time constant.

In practice you can have a high flash rate as long as you keep the pulses short. Which is what you want anyway. This is not xenon strobe, you can turn it off. The original designer did not do a good job.

Robert.

[Neil WyattModerator @neilwyatt]

I agree with Robert, except if the repetition rate is much faster then you don't want to risk pumping too much energy through the LED. As a rough guide the resistor should be chosen to limit the average current through the LED to its normal operating current or rather less.

The LED will get a rough ride, but the idea is that the heat generated in the LED by the narrow pulses will dissipate between the pulses enough to avoid overheating, but the pulse energy integrated over time should not exceed the normal operation or it will overheat.

Neil

[Clive SteerParticipant @clivesteer55943]

I think the bonding wire that connects the LED terminal to the top side of the diode die may be the feature that limits the max pulse current that can be achieved.

CS

[Author]

Posts