Вставка
- Опубліковано 21 лип 2024
- JLCPCB PCBs for $2 & SMT assembly service: jlcpcb.com/
Patreon: / sdgelectronics
A big thank you to all of my Patreon supporters!
-------------------------------------------------------------------------------------------------
Soldering Tools and Equipment:
Solder Paste: bit.ly/2E1P5jR
Solder Flux: bit.ly/2MQn7we
Quick 857DW+ Hot Air Station: bit.ly/2OWDvf4
T12 Soldering Station: www.banggood.com/custlink/3KK...
AC Version: www.banggood.com/custlink/3GK...
T12 Iron Tips: www.banggood.com/custlink/3DG...
Daniu Tip Cleaner: bit.ly/31VHwWM
KSGER T12-BCM3 Tip: bit.ly/2xgYXCQ
Test Equipment:
Mustool MDS8207 Multimeter: bit.ly/2kKHr7A
FY6900 Generator: bit.ly/2yZI40n
Differential Probe: bit.ly/31xNTOF
Microscope Equipment:
Eakins Barlow Lenses: bit.ly/3aBY5tR
Eakins Black Microscope: bit.ly/2VEcYaW
Eakins Camera Microscope: bit.ly/2x86hnT
Camera Adaptor for Microscopes: bit.ly/2KBZQMQ
Amscope Microscope Head (US): amzn.to/2YmZWNx
Amscope Trinocular Microscope (US): amzn.to/2YdBE8k
Amscope Trinocular Microscope (UK): amzn.to/2V8w9Gh
Swift SS41-WF20 Binocular Microscope US: amzn.to/2VLWEpb
Swift SS41-WF20 Binocular Microscope UK: amzn.to/2GYGYFV
Visit my website: www.sdgelectronics.co.uk/
#led #constantcurrent #schematic #ElectronicsCreators - Наука та технологія
A full design from start to end is always great to see. You can learn from watching how other people go through the process.
You should try Octopart when you know a part number and want to search for stock and prices
Indeed, I always forget about octopart.
Excellent, perfect timing, about to make something similar, good detail.
Great work and great explanation. Always good to see a professional at work
Enjoying following this project and the design intent behind it a lot. I’m currently working on a tiny-moving head driver based on the AL8860, and I feel ya about the limited data sheets. But for the price and availability, not a horrible option - I’ve been doing burn in tests at 1A on the 8806 recently, have yet to have any failures.
I Love it, Please do more of these.
Nice! Thank you so much for the explanation
thanks for your led driver vids. we are currently designing a low-cost chlorophyll meter and realized that we have a lot of things to consider. we are currently using a buck/boost dc-dc converter from monolithic power that accepts a single-cell lithium battery as led driver for the high power red and NIR LEDs. not yet sure if we would need the lenses. hope you could also make a video of that in the future. the concept of the design is well explained by apogee instruments. cheers!
You didn’t hit the “No Thanks” button in the JLCPCB website to make the Thickness changes stick, this adds $33 to the cost. Crappy UI from JLCPCB that’s caught me off guard before!
It is nice to see you sharing your talents & experience and learning from you, as opposed to watching less than quality 3rd party oriental products' reviews (no offence lol). Glad to be a member of your viewers (and now subscriber). Top quality content, and engineering. thanks.
Can someone explain why the potentiometer output is opamp buffered? Thanks.
To drive all of the LED drivers without affecting the potmeter adjust voltage (some chips/inpus can have non-linear loading profiles, resistive and capacitive loading)
I.e.:It works a voltage/current buffer, so the voltages through the divider is 'always spot on', no matter how many drivers, etc is connected.
Or more simply.....an ideal voltage source has zero output impedance..if any current is flowing to the next stage the voltage will change...A simple pot is a long way from ideal
Thanks, for your replies. Would it really make a noticeable difference on a brightness control like this? It's sourced from a 5V reg which I would have thought would be pretty stable. I'm not criticising by the way, just trying to learn :)
@@ScottMarley @24:00 we learn that 'only 250µA is going through the voltage divider network. Since the input is 5V, the output voltage is generated/halved over 10k+2k4 = 12400 ohm. In a series connection the same current flows through all components, where in parallel the lowest resistance/impedance 'takes relatively all'
Apparently there's no data about the pin loading (which would be a parallel loading on the halving output) in the datasheet (www.diodes.com/assets/Datasheets/AL8805.pdf), and (excessive) pin loading can turn out to be a 'trap for young players' - I suggest reading more about the problem in this well written piece about temp measuring (the output impedance there is 20k, or just 'a bit over 12k4 used here - some graphs provided) It probably won't do much in terms of overall performance, you'd turn the potmeter a bit more or less to obtain the light intensity wanted - but there's no guarantee it won't load the circuit completly down, requiring a re-design to compensate.
www.nutsvolts.com/magazine/article/february2016_AnalogFrontEnds
@@ScottMarley Pot is used as a voltage divider. Dividers output level is highly dependent on load current. Take any calc like ElectroDroid and see for yourself to get the idea of influence. To workaround you need current amplifier circuit like op-amp buffer or common collector BJT amplifier (also called emitter follower).
Even if the load is not that high, adding buffer is worth in order to gain circuit stability in cases when there is uncertainty as mentioned above.
Superb. Thanks!
Any opinions on ditching the regulator and Opamp, and just running the potentiometer diviver from a constant current source?
@23:30 - Using the 78L05 with a 10µF cap can cause oscillations. The datasheet recommends 0.33µF, however a 100nF should do just fine - and you can use a similar to the one on the output.
You already have the large 470µF on the input near the jack, so another 10µF won't do much difference in my experience, especially if they're all electrolytic caps.
Talking about decoupling, I can't see you've decoupled the MCP6001 op-amp - if it was me, I'd put another 100nF cap across its supply.
A thought about using ENIG - Electroless nickel immersion gold (ENIG) consists of an electroless nickel plating covered with a thin layer of immersion gold, which protects the nickel from oxidation. You're having it manufactured and assembled in China in 'one go' - why not save a LOT of money and use HASL (which won't oxidize THAT fast, in fact - it quite lasts) ??
Interesting project, quite inspirational - and certainly looking better than the CFL ring light I currently have. Thanks, looking forward to the next episode :ø)
Ah, the ENIG was for the LED boards which I will have to assemble myself. I'll have lots of spares so this will keep them usable in case I need some LED boards for something else in the future. I didn't add another 100nF capacitor as the 78L05 is only powering that op-amp and the resistor divider. They will be physically located next to each other. Thanks for your thoughts
@@sdgelectronics Layout is an important thing, it can solve (or create) a lot of things.
I would advise that you change the 10µF cap into a 100nF - I'll tell you a little story, and it might help get my point across:
A little bird drinks water from a small ceramic cup, all was good - for some time, until the cup broke and the bird stopped singing (for no reason. It just cracked, poor cup).
A well meaning man put a ceramic bathtub in the cups place, thinking the bird can now drink and shower at the same time. The bird song started again, as the tub filled up.
The well meaning man also had 4 cows, which were also drinking out of ceramic tubs - the cows had plenty of water, and gave off a lot of white milk when milked each day.
After the man installed the bathtub for the bird, the cows started giving sour milk and the bird song got gobbled at times. The man started investigating, and found the water level in the tubs nearly always matched. That couldn't be, the bird couldn't drink that much water - he was mystified, where did the water go.
Searingly he found that the bathtubs for the cows had begun sucking water out of the birds bathtub, and in the process the sucking vortices nearly drowned the bird for a moment. The water pipes hadn't changed, gravity and water was the same - how could this be. Bernoullis was the mans name, he published this - and it became known as the Bernoullis princip. I don't know what happened to the bird or the cows, but the man became quite famous.
muRata (not muCow) later on invented a 'one-way valve for filling up bird tubs' - EMIFIL® NFM41P Series
www.digikey.com/catalog/en/partgroup/emifil-nfm41p-series/7429
My view and experience on surface finishes - it rarely matters, unless the pcb's are stored in a horrid place and/or has some extreme requirements.
Google returns: "ENIG has several advantages over more conventional (and cheaper) surface platings such as HASL (solder), including excellent surface planarity (particularly helpful for PCBs with large BGA packages), good oxidation resistance, and usability for untreated contact surfaces such as membrane switches and contact points."
Should the surface become all gray and unsolderable, they are (often easily) restored by applying solder and wicking it off again. But, I won't stop you - it's your money :ø)
Ok, you might not Bernoulieve a word of the story above. (It was actually aCat that invented the filter to avoid chugging bird ♪hit)
25:10 for anyone else wondering, the opamp has 3 more pins on the other side :D
Thanks for all the great content. Note that the Sinotech 78L05 had a recommend operating of 20V. Not sure if your using a 24V plug pack but figured you want to avoid the potential of getting the control out of regulation. Cheers.
Good spot, I'm not sure what that really means though given the other specifications in the datasheet. I'll probably stick with an ST or Ti part which doesn't have that condition.
@@sdgelectronics It is better to add 5.1V zener like SMAZ5V1-13-F between the 24V and the input of the 78l05
Wouldnt a zener plus Potentiometer be even simpler and possibly cheaper than using the linear regulator plus opamp for vcrtl?
A zener diode is definitely another option, but you waste a bit more power as heat on the board.
@@sdgelectronics TL431 is a very good temp compensated 2.5V reference, and cheap as dirt!
What's the purpose of the MCP6001?
What about a 2.5V zener diode (plus series resistor to limit current to a couple mA) instead of the 5V linear regulator? Gets you 2.5V straight off, and no need to divide further.
Thats what I thought aswell surely the input impedance for vctrl will be negligable
I still need to buffer the output as the loading on the input pin to the driver ICs will result in a non-linear brightness control. Rail-to-rail op-amps are normally 'almost' rail-to-rail, but not quite.
Does under-driving the LEDs change their CRI/coloUr temp?
If you under drive by limiting the current it can change the color temp. PWM modulation with the right current would be best. But the changes in color temp is usually not noticeable to the eye.
Yes, and it adds more variability between other LEDs. The spec on these LEDs is pretty good though if you look at the graphs in the datasheet, so I'm not overly concerned.
have you considered tme.eu for sourcing the components ?
Yes, I often use them. I will likely combine orders for the GPSDO and this project at the same time, so it'll be whichever supplier has most parts in stock. I get free delivery at RS and Farnell, so other bits can come from there.
Hey Steve had a few thoughts about this one. Its actually really interesting project for those of us with such microscopes & we thank you for pursuing this project further. Of my questions some are probably redundant , others perhaps you cannot integrate or better not to. My thoughts: 1) were the satellite boards are the thing you are going to make the slight inwards angle with? I guess that makes sense. I hope you still intend to make such an angle. 2) If I zoom in with the dial, then the amount of light getting through the lens is reduced. This makes the whole image dimmer at the higher zooms. So it may be desirable to have the ability to adjust the led brightness dynamically while operating. Rather than have just some tiny trimmer pot that requires a screwdriver to adjust because it was only ever intended to be setup once. But was that actually your intention with the design? I think ATM these cheap ring lights we all buy are very convenient with a sturdy pot dial to adjust... its just they are under powered to begin with so we always have it on maximum setting all of the time. . In fact an even more elaborate and complicated design would be mechanically linked into the microscopes optical zoom knob itself. Thats too ambitious for this project but maybe could be a possible future mod or upgrade if was ever found to be useful enough by anybody. I am not sure. 3) Driving from 24v is conveniently the same voltage as many soldering irons. And perhaps certain common led dc supplies. But it is also maybe a little bit high since it excludes (without introducing a buck converter) other PSU sources such as usb power source (which is 20v max). Or a laptop power brick which is 18-20v. So maybe that aspect is worth considering to make sure from an interoperability standpoint. 4) if there is no protective housing then a white pcb for white microscope. Or a black pcb for a black one... that would look pretty cool. 5) Your original mounting design for the pcb is great for those with a 3d printer. But is there an good alternative for those of us who dont have one? Thinking maybe something really simple like sandwiching it between 2 hose clamps and using foam as a buffer to prevent scratching and help to keep a good grip on the pcb. ATM i have been using strips of esd rubber mat, for my current ring light. And low density foam. This also helps stop my metal microscope head from rotating, or getting all scratched up by those 3 grub screws. Anyhow good luck with this project... so much cheaper than dropping 200 bucks on those polarizing ring lights. Looking forwards to seeing how it goes. Also I was feeling maybe? its a negative point not to be able to dim below 25% of max. Since that affects the total adjustment range and also the choice of alternative leds. For example if those ones caanot be sourced so easily by others. Or if somebody also want to add polarizing filter (in addition). Then that might become an issue in those broader situations. But we shall see for your particular use, its not likely to matter.. But you know... its a limitation of that specific driver so to get around that would then require changing the led driver. Which is really my only question if they all do that, or if its specific to only some led drivers. And not for others, well maybe we cannot know without doing a more exhaustive parts search.
Thanks for your comments. I may change the potentiometer, or at least do a combined footprint if possible so that there are some options. I'm just planning out the best way to do the angled PCBs - I want it to be quite robust, but that probably means custom or 3D printed parts. I think if other LEDs are likely to be used, I would suggest anyone to change the sense resistor rather than relying on the adjust pin.
‘In Stock’ of course, but why filter on ‘Active’? Unless you’re a factory all you care about is if you can buy them now.
I'll put the files on the website afterwards so others may want to build it some time in the future.
Sure, but isn’t the point of the series to teach the engineer to fish? Sorry for the mixed metaphor 😊
Diodes Inc. datasheet usually sucks. I had a lot of issues with support from them in my job as well.
Using satellite PCB for the leds is a expensive solution, have you checked if it is possible to just angle the lenses?
To make a angled lens mount I would use a 3D printer.
I did take a look, but the collimating lenses require precise positioning of the LED flat to the lens, so it's not possible to do this without compromising optical efficiency.
Talking about overengineering.. there's a definitely simplier way to do CTRL voltage, with fewer components, therefore less assembly cost. 10uF and 22uF, haven't you seen Daves video on BOM consolidation? Just messing around - nice video!
Although it is a good practice, when you make 10 PCB without a large number of components it does not have a great impact as David also said in that video. xD
feel free to make you own designing video/series better in any way! And refrain from insulting Steve by schooling him based on EEVBlog's latest content.
Dave is an entertainer, choosing an easier way to produce content, he gives out tiniest bits of knowledge, that you won't even know how to use in real applications because his videos can not possibly contain bigger picture.
On the other hand Steve is working his ass hard to deliver projects covering the whole design process from scratch and he is doing it in a very informative and effective way. His channel is two heads above quality wise comparing to latest EEVBlog posts..
I actually missed the 22uF, it was supposed to be 10uF anyway!
Informative as always! Why the hell did someone dislike this video?
I'm guessing when you said white silkscreen you meant soldermask?
yes, I missed that :D
First? 😀