ไม่สามารถเล่นวิดีโอนี้
ขออภัยในความไม่สะดวก
EEVblog
ฝัง
- เผยแพร่เมื่อ 27 พ.ย. 2012
- 555 T-Shirt: www.zazzle.com....
Hacking the Mantis Elite microscope to control the LED brightness with a 555 PWM circuit.
Forum Topic: www.eevblog.com...
EEVblog Main Web Site:
www.eevblog.com
EEVblog Amazon Store:
astore.amazon.c...
Donations:
www.eevblog.com...
Projects:
www.eevblog.com...
Electronics Info Wiki:
www.eevblog.com...
I love easy projects like this! They're great for learning the basics of things like timers, dimmers, etc.
Dave should do a video series called 555 things you can make using a 555.
Oh yeah!
Bless you, what a great idea...
EEVblog Hear that Dave, Your People want you to do this series.
A million views per video at $300 / million = $166,000 in TH-cam revenue. I would only ask for a meagre 25% cut.
jesus doesnt need money =]
I like that!
Dave...you just screwed Mantis out of doing a $300 accessory light dimmer for their microscope.
What he's actually done is give them the blueprints necessary to steal the idea, manufacture and trademark it, and then take legal action against him. #CapitalismBaby
And this video will demonstrate prior art!
They could do all that but "prior art" prevails and their action would fail. Besides, whilst the specific (clever) usage is for the Mantis in this instance the basic circuit exists in probably 10 million posts on the web ... do you think Dave really cares?
@@boblewis5558 I have seen the second version even on Great Scott's channel. Only difference was that I didn't quite understand the fact that pin 7 is used as an output. 18:50 is quite a jaw dropper for people just learning.
@@lamjeri a quick look at the internal circuit block diagram of a 555 would explain that confusion. BOTH pin 3 AND pin 7 are effectively output pins.
Pin 7 being labelled "discharge" doesn't mean it cannot be used as an "output" pin because internally it is basically just an open collector of an NPN transistor, which when switched on provides a discharge path to ground for the timing capacitors.
Pin 3 conversely has an output drive circuit which has its own "pull up" internally, hence the lack of need for the external pull up when used as the "discharge" pin.
A clever use of basic understanding of the internals to "reuse" the chip in a clever component reduction. An engineering solution to the ever present "bean counters'' desire for cost saving, no matter how trivial!
It might surprise people starting to learn, but even engineers with decades of experience can be surprised by novel uses of a circuit because someone has realised the "detail" and how it can be used alternatively. A classic example is the use of 3 logic inverter gates to produce an oscillator. Not something that might easily occur to someone just starting out in electronics but very old hat to someone of experience.
Cool stuff. My preference would have been to use circuit 1 with an N-channel MOSFET as a low-side switch. Circuit 2 was a clever solution though and I never would have thought of it.
That would be my preference also. Thanks.
Be aware that NE555's output is not rail to rail, especially in logic 1 state. A resistor(let's say 1kΩ) betweeen base-emitter(BD136) would greatly improve transistor's switching time and would reduce collector cut-off current. A maximum limit current resistor(330Ω or more) at discharge pin(7) or at the wiper is a good idea too.
This is not relevant for the CMOS 555 timer versions.
Dreat educational project and demo of making a PWM circuit from 555. Many thanks.
I had subtitles on for a few of these videos, as I didn't have my headset around. The way you're voice is automatically made into subtitles had me cracking up give it a try some time, just turn captions on when watching a video back.
I wanted to do something exactly like this myself, I'm going to copy the final circuit and wire it up to my own mantis. Amazing they did not include something like this for such an expensive price of equipment!
Thanks Dave, I've seen a lot of your stuff and keep finding ways to implement things from videos you do. I think I am going to use a slightly modified version of this to make a dimmer for the under-cabinet LED strip lighting I'm installing in the kitchen.
On a barely related note, I'm going to buy one of those t-shirts for my wife merely so I can comment on how nice the threshold and output pins look when they're cold :-P
Thanks for showing the traditional version of the circuit before optimizing it. That's very helpful.
I like the idea of having each side's LEDs adjustable in brightness so you can get some control of the direction the lighting is coming from. An XY pot would be friggin sweet. Run the stick along the diagonal for equal overall brightness and off diagonal for fading from one side to the other. User friendly as heck!
Informative, clearly explained, outstanding as usual. I have 555 ckts I still use that I made about 15 years ago. I'm now encourage to dust off my scope, meters, etc and build the LED dimmer, which I need... .
Thank you for posting. I didn't notice until the second day that BOTH sides of the potentiometer were connected to power through a 1K resistor.
You should connect a 1k resistor between those two diodes opposite to trimmer. I built the circuit on breadboard and it works fine but when i made it on pcb as a prototype, my smd 555 chips worked good for several minutes and then stopped working. I blew 2 chips and then i searched some formulas, and came to know that the capacitor on pin 6 is getting charged through the resistor, diode and potentiometer and discharging through the diode only directly to ground through an inbuilt transistor on pin 6. Due to very low resistance on the discharge path, when the duty cycle is at maximum, the capacitor is discharging in a very short time and a lot of current is flowing through inbuilt transistor which will eventually blow the chip making it unusable. I dont know how long the dip version of chip lasts with this circuit but the smd chip will blow instantly. So add a 1k resistor in the discharge path!!!
Good to hear you're still working on it. One of my favorite EEVblog segments!
Thanks Dave for this fun little video. It was great seeing you working on stuff in a bit more real time and hands on.
Very good. I like coming of pin 7 . Used that pin in a simple 555 time base circuit on a very simple scope and a 0.1 out from the transistor.
.
You can always save this pull-up resistor in 555 timer astable multivibrators by using its output pin to control charging and discharging of the capacitor. You can also save the base resistor by use of a p-channel MOSFET for switching of the LEDs also from the output pin.
This T-shirt sould have DaveCAD written under the picture!!!
This is one of your best vids Dave
Love how elegant and small the final solution is. Considering doing something similar with a logarithmic pot to get a variable boost converter.
Nice, simple solution. A fancier solution might use a logarithmic pot, since the human perception of brightness is more or less logarithmic.
Would be nice if this could have dimmed one or the other side independently so you could get "side-ish" lighting. That's sometimes useful for component recognition and adds a bit more 3D depth to the image.
+PIXscotland I think it's just a matter of build 2 units and placing one each side. Good luck
I don't quite know enough about the math to explain it here but the values are based on a formula that determines the values for the fewest parts that will give the most complete spread of values. The formula is also used to determine things like what currency values a country should have for the best use of resources and transaction efficiency.
I'd like to see the voltage out of that LM317 when you use the dimmer!
I suppose it's jumping all over the place.
KISS, just change the existing lm317 to a voltage reg with the pot and be done. the led's have individual resistors up to the plug packs voltage for there max draw. Unless you wanted to do a 555 tutorial that is ;). Take care Dave, love your vids.
Oh wow, you left LEDs as the steering diodes. This is a lazy hack with a legendary level of purist utility in the design, I love it.
"We don't really care, we just want to get this thing working."
- Dave Jones
Skip the 1K resistor, and swap pins 3 and 7 on the 555. Use pin 7 as the output (it's open-collector) and pin 3 (push-pull) to drive the steering diodes to the sides of the pot.
The video explains it. The P.S. drives a constant current device, so more resistance would simply increase the voltage (across the pot) and deliver the same current to the LED's. A PWM is the perfect answer to this situation, although a non-inverted output would have allowed a higher PWM cycle (10% - 99.9% is better than 0% - 90% for LED control).
PWM doesn't change the voltage/amplitude like a voltagedivider/pot, it changes the fequency - is like switching on and of very fast. The brightnes is determined if "the switch" is longer on or off (longer on -> brighter and longer off -> less brighter)
Definitely the video I have enjoyed the most from you this month.
Loving the ChestCAD Dave!
Good Vid. Dave. I was using your first Bread Board design and changed it to your second design, as its for a differant use I have differant resistors but thanks the new board works much better.
15:52 "drop that bass resistor" ... sample gold
Looking at "The Big Test" starting at 24:58 I can't see why you would want to dim the LED's at all!!
Nice video Dave!! I applied your circuit as a comparision between Analog PWM vs. Microcontroller PWM.
Pretty interesting behaviour using constant voltage source. I can reach 26% as minimum and 96.76% as maximum.
I do not why, but, pretty intersting thoug.
Thank you.
Nice simple circuit but when ever you drive a PNP this way it may not shut off because the base voltage is always a little lower than the supply. A 20K from base to emitter will ensure it shuts off reliably and will reduce power dissipation in the transistor.
And i tryed to explain that i missunderstood that. You are right. It should be well within the kHz range.
Very nice use of 555 and nice hack. Should should contact them and donate the schematic so they than add it to next version or maybe charge or get free one.
The thing that amazes me the most is that the manufacturer of the microscope did not include the dimmer function. Would not have affected the manufacturing price by to many fractions of Dollars I imagine. It could even have been used as a selling point.
Love the solution. :)
Yes, I was thinking that, but the wires are paralleled inside the optic head, so you'd have to have a separate pot and circuit inside each LED half. It would be pretty ugly.
"should" - no "could" - yes
if he wants to share his knowledge, it his choice, and would be a great gift from him. It's not his obligation tho.
Nice one, Dave. Would love to see more episodes like this one where you actually MAKE stuff!
That's the hacker's spirit I wanted to ignite with my proposal of the PWM device in between the Mantis and the power supply.
Dave, When I do this circuit I connect the pin 4 to pin 5 this eliminates the need for the small cap and makes for an easier PC board layout.
What might be good also it to be able to switch one side of leds off so you can get shadows, can be handy some times.
Hi, Dave, great video as usual. You can omit the 220R too, replace the BD136 with a IRF520N. Did this works great.
Thanks Dave.
I will certainly get down to modding mine.
And yes, different types of LEDs could well be useful. In fact vision engineering can supply different led banks. Not sure how much they cost though.
Yep, I guess 0.1uF is nothing to be worried about, still i filter most of the PWMs before i feed them in to LEDs or DC motors simply because additional high frequency in a DC motor is 1- more noise and 2- bad for non-dipped coils. I know it's overdoing it, but i think for me it is worth it. I also have slight light induced headaches when the light is too bright or there is some kind of flicker.
They need a controlled current and controlled voltage. Voltage is easy. Since current is the number of electrons that flow in a given amount of time, you can see how PWM controls current, and does it more efficiently than anything else since it has an off state where no electricity is being used. A resistor just converts the current to heat.
Add i was giving 2 examples a) frequency of light b) frequenzy of flashing where somebody can be abnormalyl good at detecting light. I see those light flashes in neon tubes too. I just say that there are limits. And that tim was right - pwm sould be in the kiloherz range so you cannot see it.,
Look up "Renard series".
Basically it comes down to the fact that you don't really care (I choose arbitrary numbers here) if you got 900 or 950 Ohms, but you do care if it's 100 or 150! While the difference is the same, the ratio is quite different.
So you choose values (starting with 1) where every number is the previous number times a constant factor, preferrably in a way that after a certain number of repetitions you get to 10 (so the whole series can repeat with just a zero added).
I agree with you %100. Also I wish to Dave starts to teach about Microcontroller programming such as open source Arduino...
Lovely. Could a bypass switch be used to get that additional 100 lux.
lol. I didn't even noticed they are supposed to have different colors (same problem as David)... Anyway, I don't think it matters that much :D Good job with 555. It's one of few circuits that survived huge period of time without any major changes, still great..
Dave,
I think it would be great if you made more videos that were like tutorials for electronics beginners and/or intermediate. For example, teaching us how an adc works, or maybe beginner embedded electronic design, or maybe even teaching about microcontrollers, what a dc/dc converter is. Stuff like that.You have so much knowledge that you should share with us.
Yeah, it was basically real time. It's a video I could have shortened a lot in editing, but I figured there was more value in the real-time stuff.
i was referring to frequencies of "flashes" due to the PWM signal, not colour frequency of light.
for example, there are times where i have been in office buildings and been able to "see" the flashes of the flourescent lights overhead. obviously its fast as hell but sometimes it's perceptible to me.
I perfectly understand your design.
robot797 asked if he can use LM317 to controll an amount of light produced by LEDs.
TheLawnWander said LEDs need PWM to do that.
I have met a few people who mistakenly believe LEDs must be powered only in pulsed manner - otherwise LEDs will work in wrong mode. Bullshit. Actually, an amount of light produced by LED depends from current through it, so you should control current flowing through LED by any method you can do it, but not only by PWM.
Your device has a pretty old National Semiconductor Logo :-)
Well, anyway they are part of TI now...
Nice Blog Dave!
Ground bounce from inductance of the test lead. No ground clip on CH2 (so uses CH1)
The LED PCB power connectors should be the same on both sides. Just bypass the existing wiring from the jack to the PCB plugs. 2 555s and an XY pot would sure be cool and easy to use. There is the problem of where to run the wires to the outside of the case without modifying the case.
Not only that, you have the balls to make them make it standard issue :D
A Power Potentiometer is an expensive part and will dissipate a lot of waste power as heat. Dave is in Australia so he is warm enough. By switching on and off to control brightness the design will be much more efficient.
You can replace the PNP transistor with a N channel MOSFET and the 0.1uF capacitor with 47nF capacitor .
That's effectively what a dropper resistor does, converts the constant voltage into a known constant current.
I fully agree. UI accessibility can be really hard if you have to target people with serious disabilities, but color blindness is so easy to make allowances for, and so common, there's really no excuse not to.
I would've said replace the regulator resistors on the regulator with a potentiometer, but that works too.
Now with much more experience, I can confidently say the 555 Timer was the right move. The potentiometer wouldn't be very accurate.
The PWM version is more efficient. Yes, you still need the series resistor to limit the maximum current. But at let's say 50% duty cycle the resistor also only drops voltage 50% of the time. No current for the other 50%, no losses then (ignoring power needed by the 555)!
Err, it's a 0.1uF bypass cap... And the regulated 9V input (minus the LM317 drop) limits any possible maximum voltage. Nothing to worry about.
i have a 12vdc pc fan i need to turn on and I have a 24vdc source. I know I could just use resistors to run it ,but its just wasteful to do it that way. So from the start I have wanted to use pwm mainly for efficiency,it would use at least half the ma with PWM. But watching your Vid Dave I think I might add a pot and drive it between 75% to 125% , so i can see if at the low end it would cool with less noise or if its not cooling enough can drive the fan higher till I get a bigger fan. Thanks Dave for all your Vids. I have not only learn how to do alot of things, but also a few I now know how to do different or better .
125% duty cycle? x.x Well that would be something, wouldn't it.
No worries, I'm not red-green colorblind and the light green is pretty similar to the yellow for me, too ^^
That's something that has been bothering me with quite a few scopes I've used in the past.
Actually, my first thought was to use a potentiometer with a switch, and using it to bypass the dimmer.
However, finding linear taper pots with a switch is tough....most of the time, they are audio taper.
If you are willing to order from a parts supplier, then I'm fairly sure you can buy them.
A potentiometer (wired as a variable resistor) in series with a constant current varies precisely NOTHING except the series resistance. It's a CONSTANT current - that's the whole idea - it doesn't matter if you have five or 50 LEDS in series (provide the drive voltage is high enough) it will be the same i.e. CONSTANT current and an extra series resistance will do nothing.
Great learning video Dave really enjoyed it ..................BUT.............a gross little hack thing covered in shrink sleeve dangling from a Mantis. Are you mad!!! You're an electronic genius and a mechanical Phillistine Dave. Stick the thing inside, drill the casting and stick the pot shaft out the hole with a nice knob on it. Beauty. Aesthetics Dave, Aesthetics.
you can also use this circuit for controlling motor speed...
Note that if you are anywhere that honours US patents, someone holds a current patent on using PWM to alter LED brightness.
555 is such a versatile chip
I thought I saw the frequency swing swap when the 555 circuit had the last modification. It was at first lower frequency at higher duty cycle and then it went to higher frequency at higher duty cycle.
if you want to see some car LEDs in action, watch some of top gear's tests of the newer ferrarris and lamobs. they all have obvious PWM in their LEDs when you see their highspeed cam footage in the road tests.
it's pretty interesting to see it show up in slow mo like that.
Every scope is that "sensitive". Welcome to the wonderful world of probing!
Also, since it was only 1/2 amp current, why not just put the pot on the 317 and make it vairy the current linerally?
Great little hack Dave; you are going to have to name it something like the Mantis "DUD" (Down-Under-Dimmer;)
Having to use what I have on hand, I used an IRL3803 (140 amp rating is a tad overkill) instead of BD136. Between pin 3 and the gate I used 47k and since I have a CMOS version of the 555 as well the cap from pin 5 really isn't needed. I read it somewhere on Talking Electronics that pin 5 need not be connected in CMOS versions for PWM control, or perhaps I misread something, but it works well without the cap... The voltage going to pins 4, 7, and 8 get an additional 220ohm resistor but the voltage going to the FET remains unchanged. I got VERY close to 0-100% according to my crappy sound card oscilloscope.
Great video as usual Dave, but as a red-green colorblind viewer I thought you should know I can barely tell the difference between the green and the yellow on the oscilloscope. Any other color combination would be fine.
I think Mantis did 317 on purpose to supply stable current. remember old microscopes having bulbs? being relativly slow those can be PWMed without light flux pulsing. Don't get me wrong, pulsed leds are great for torch lights, but in my opinion they are bad for microscope backlight.
Read the part where you talk about your keyboard. The sentence afterwards combined with the sentence where you say your keboard uses visible pwm frequenzys can be read like this. And yes i looked up how car leds work. You are correct on the leds.
You never said what the trails u see look like. SO i imagines if they ahve a big gap it might be that you blinked with ur eyes.
i was in a shop the other day and they had christmas lights up that were obviously PWM'd way too slow compared to the other sets around them. the frequency of the flashes was making me really uncomfortable but my dad who was with me at the time couldnt notice the difference.
i guess some people are able to pick up higher frequencies than others.
The light mater enclosure seems quite common, I have a thermocouple thermometer in one of these
No, just blinky lights, or the same would apply to incandescents driven by DC.
Have you ever considered using coloured filters to help you discriminate between colours? Pause the video and hold a red filter up to the screen. The yellow trace will appear brighter, because of the red component. The reverse will happen with a green filter. Just a thought to aid your independence in these matters. I'm not colour blind myself, but it seems like it might work.
well I never did... never even bothered to think about the loverly 555 as having 2 outputs...cheers.
convert the LM317 from current source to voltage regulator for 9 Volts or whatever you want and then u use the 555 circuit for dimming.
Was trying to fiddle with a charge-discharge relaxation oscillator using single supply, didn't work, I have to do some hand calculations for it first.
Thanks Dave, my favorite type of video.
Very good, but I have one question, because use 2 leds and non-conventional diodes, such as 1N4007 or the like, or failing that the typical 1N4148?
An HM2007 IC would be nice touch for hands free control. Brighter... dimmer... dimmer... brighter...
very easily 3D print a case using your maker bot and that case would just fit inside the square opening where the power jack socket is on the mantis head
the values are chosen so there is an even spread of values available over the entire range. Component values come in a series of 12, 24, or 48 values per decade (Sometimes higher for high tolerance values). ( Search for "E24 series" and you'll get some tables.).
The values chosen are so that there is a value close to every, say, 5%. This means that they track a log series.
Finaly, it just so happens that some values are very common. 22 and 47 are two of those very common values!.
i watched it all the way to the reacharound.
Of Course there are people who can see higher blinking frequenzoes or a bit further into the uv regine. But i would hardly doubt that you can see flashing lights in teh kHz regime.
Brilliant Dave