EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: bumba000 on September 25, 2016, 05:13:59 pm
-
Hi All,
I am working on a project here, but I've just come across something in a schematic that I've not seen before.
It looks to me like there's an LED and a standard diode applied in opposite directions at the same location. Please help me understand this.
Thank You, John
(http://i.imgur.com/0eU2ide.png)
-
Does it have a part number?
Tim
-
It's marked as 2 LED's? Maybe just an indicator for reverse polarity? Green in the correct direction, red in the opposite?
-
Thank you all for your super quick responses. I figured that the image I posted may cause some confusion because of that double mention of LED. Attached is the schematic I'm working from.
Again, Thank you.
... There are no part numbers mentioned. Only general ideas of what can be used.
-
An internal ESD protection diode?
See page 4:
http://www.osram-os.com/Graphics/XPic6/00102535_0.pdf/ESD%20Protection%20while%20Handling%20LEDs.pdf (http://www.osram-os.com/Graphics/XPic6/00102535_0.pdf/ESD%20Protection%20while%20Handling%20LEDs.pdf)
-
Well I of course don't have a ton of experience with all of this. I'm learning as I go and so I greatly appreciate your input (all of your inputs). That being said, I don't see any reason for a standard diode in reverse direction, so I'll go with the ESD idea. Since this is an audio spectrum analyzer most of the LEDs will be on/off/on repeatedly, quickly and for long periods of time. If ESD can be an issue here ( sounds like it can be ) then the standard diode in reverse directly seems like the solution.
Thanks again,
John
-
umm, any ideas on which diode should be used for the ESD protection?
-
umm, any ideas on which diode should be used for the ESD protection?
The whole symbol (2 diodes) is a single part. The one that is not light emitting is there for ESD protection (potentially) of the one that is. Knowing exact part number would be helpful.
Schematic is drawn haphazardly, so it is possible that whoever did it, just picked a diode with required PCB landing pattern and it ended up being an LED with embedded back diode. There is no real need for them in this particular schematic.
-
I don't have any part numbers, only recommendations on transistor bc547 and op-amps lm358. In regards to the LEDs and other diodes, I'm being left in the dark so to speak...
-
Since the schematic symbol of the diode across the LED is very small compared to all the other diodes in the schematic, and that it has no part identifier, I would think it is part of the LED as in the Osram reference posted by eliocor (http://www.osram-os.com/Graphics/XPic6/00102535_0.pdf/ESD%20Protection%20while%20Handling%20LEDs.pdf (http://www.osram-os.com/Graphics/XPic6/00102535_0.pdf/ESD%20Protection%20while%20Handling%20LEDs.pdf)).
As for ESD protection, I don't see that you would need it for this circuit.
Most LEDs have a very small maximum reverse voltage (about 5V), and this diode could protect against this, but again I don't see that you would need it in this circuit.
Where is this circuit originally used? Could it be in a really electrically noisy environment?
Maybe they just had a bunch of these types of LEDs lying around and they used them.
I don't know.
-
I don't have any part numbers, only recommendations on transistor bc547 and op-amps lm358. In regards to the LEDs and other diodes, I'm being left in the dark so to speak...
Then just use whatever diodes you like.
Reverse diodes will also protect LEDs from applying excessive reverse voltage, but it is not a concern in this schematic. It is quite possible that originally selected diodes were for some special applications.
-
Hey Gary,
The application of this circuit is an audio spectrum analyzer. There will be a lot of amplified music going through the circuit.
Hey Alex,
How about the shottky diodes as in D1, D11, D12, D13, D14, D15, D16, D10 and D17. What would I use here and why would the designer have gone 11, 12, 13, 14, 16, 10 and 17?
Thank you,
John
-
How about the shottky diodes as in D1, D11, D12, D13, D14, D15, D16, D10 and D17.
They actually set voltage thresholds at which transistors will open. Those are required.
What would I use here and why would the designer have gone 11, 12, 13, 14, 16, 10 and 17?
Again, schematic is poorly drawn, so numbering is also poor, there is no specific meaning behind this.
Also, the whole thing can be implemented with way fewer parts using MCU.
-
I know about the MCU methods. I have MSGEQ7's and many tiny duinos. I just wanted to analog it. I made a couple myself , but couldn't seem to get the voltage meter part working, so I bought some AN6884 before I found the schematic I've posted here. Now I'm going to try building this circuit. The diodes are confusing me though...
So then D1 might be .3Vf, D11 .4Vf, D12 .5Vf and so on??
-
So then D1 might be .3Vf, D11 .4Vf, D12 .5Vf and so on??
No, they all must be the same to get a linear response. They need to be on a log scale to get logarithmic response (what you typically want from an equalizer).
-
Wondering if it something like the littlefuse PLED series
http://m.littelfuse.com/products/led-protectors/led-protector/pled-5-qfn/pled5q12.aspx (http://m.littelfuse.com/products/led-protectors/led-protector/pled-5-qfn/pled5q12.aspx)
I am using these with CREE WLEDs as ESD and open circuit protection.
-
I am using these with CREE WLEDs as ESD and open circuit protection.
With 5V supply and BC547 driving them, they must be regular indication LEDs.
-
For your application you surely do not need ESD protected LEDs.
Take a look at those two videos (they also talk about what happens to damaged LEDs) :
https://www.youtube.com/watch?v=NjKgPLeJ79Q (https://www.youtube.com/watch?v=NjKgPLeJ79Q)
https://www.youtube.com/watch?v=_wR_mFWeI6Y (https://www.youtube.com/watch?v=_wR_mFWeI6Y)
If you want go dig deeper regarding what damages can an ESD discharge do to LEDs and why protection is needed, take a look at some link I have found (in some few minutes):
https://en.wikipedia.org/wiki/List_of_LED_failure_modes (https://en.wikipedia.org/wiki/List_of_LED_failure_modes)
http://sinogerman.brinkster.net/brite-led/AENESD.htm (http://sinogerman.brinkster.net/brite-led/AENESD.htm)
http://assets.sylvania.com/assets/documents/ESD.a9e90e9d-c91e-4ea3-9ebe-0be6e5570cd7.pdf (http://assets.sylvania.com/assets/documents/ESD.a9e90e9d-c91e-4ea3-9ebe-0be6e5570cd7.pdf)
http://www.cree.com/~/media/Files/Cree/LED%20Components%20and%20Modules/XLamp/XLamp%20Application%20Notes/XLamp_Elec_Overstress.pdf (http://www.cree.com/~/media/Files/Cree/LED%20Components%20and%20Modules/XLamp/XLamp%20Application%20Notes/XLamp_Elec_Overstress.pdf)
http://www.forge-europa.co.uk/knowledge-zone/electrostatic-discharge/ (http://www.forge-europa.co.uk/knowledge-zone/electrostatic-discharge/)
https://www.google.it/search?q=led+esd+damage&source=lnms&tbm=isch&sa=X&ved=0ahUKEwjbyJDJqKvPAhXBbRQKHSTVABgQ_AUICCgB&biw=2880&bih=1527 (https://www.google.it/search?q=led+esd+damage&source=lnms&tbm=isch&sa=X&ved=0ahUKEwjbyJDJqKvPAhXBbRQKHSTVABgQ_AUICCgB&biw=2880&bih=1527)
http://www.littelfuse.com/products/led-protectors.aspx (http://www.littelfuse.com/products/led-protectors.aspx)
http://www.dfrsolutions.com/uploads/white-papers/LED_Failures.pdf (http://www.dfrsolutions.com/uploads/white-papers/LED_Failures.pdf)
https://books.google.it/books?id=uRoreBF61o4C&pg=PA120&lpg=PA120&dq=led+esd+damage&source=bl&ots=zxEYXh_AI-&sig=BUnn8iYHiiWb0ItS5fywW9r_DNo&hl=en&sa=X&ved=0ahUKEwjbpbbMqavPAhXG8RQKHd0OA4I4FBDoAQghMAE#v=onepage&q=led%20esd%20damage&f=false (https://books.google.it/books?id=uRoreBF61o4C&pg=PA120&lpg=PA120&dq=led+esd+damage&source=bl&ots=zxEYXh_AI-&sig=BUnn8iYHiiWb0ItS5fywW9r_DNo&hl=en&sa=X&ved=0ahUKEwjbpbbMqavPAhXG8RQKHd0OA4I4FBDoAQghMAE#v=onepage&q=led%20esd%20damage&f=false)
https://www.roodmicrotec.com/fileadmin/user_upload/pdfs_Broschuere_Praesentation/Downloads_Papers/Papers_2015/SMT05-2015_LED_Failures.pdf (https://www.roodmicrotec.com/fileadmin/user_upload/pdfs_Broschuere_Praesentation/Downloads_Papers/Papers_2015/SMT05-2015_LED_Failures.pdf)
http://www.leds.de/out/media/Elektrische__berlastung.8c687d4.pdf (http://www.leds.de/out/media/Elektrische__berlastung.8c687d4.pdf)
-
ok Alex. I see how my D1 .5Fv , D2 . 4Fv is incorrect. I guess using all the same Fv make much more sense. .5 to turn on the first, an additional .5 to turn on the second and so on.
I will leave out the ESD protection for now and see what happens. I'm still trying to wrap my head around why the LED anode would connect to VCC (5v) and the LED cathode via transistor would connect to VCC2 (10v). Sure doesn't seem like it's going to work, but then again - what do I know? not much apparently! ;D
-
would connect to VCC2 (10v).
It connects to VCC/2 = 5V. and VCC is 10 V, so voltage across the transistor-resistor-LED is 5V.
And it connects to VCC/2 because it is a virtual ground in this design.
-
LOL, I'm off to a great start ehh? Already confusing voltages. Well thanks for pointing that out.
Virtual ground huh? Ok. I'll be back. I've some research to do on this virtual ground thing...
Thanks, John
-
Virtual ground huh? Ok. I'll be back. I've some research to do on this virtual ground thing...
That's just so you could properly amplify audio signal for filtering. The input signal is decoupled though C1 and biased at VCC/2 via R3. So VCC/2 is a "zero" level of the audio signal. This way you don't need a negative supply voltage.
-
I was thinking the circuit designer had made a mistake with the voltages and non-use of ground. Thank you very much Alex.
-
This will likely be obvious to you all, but why is IC1B seemingly not connected to anything? Is it really that half of the IC isn't used?
-
Yes, that's the second half from the IC1A. Also, don't leave inputs of IC1B floating, short them both to ground.
-
Oh. I've made some breakout boards for the LM358 as I purchased a bunch of the mini so8's and have gotten good and bad results. But mostly bad when using only half the chip. When I've done this I do nothing with the other half. I guess this could be why I've had such mixed results.
Please tell me I'm also correct on the audio/signal input. The Audio neg goes to ground and the Audio + goes to IC1A +in. Right?
-
I guess this could be why I've had such mixed results.
That depends on what was the experiment. The worst that can happen is oscillations on the output, since inputs are floating.
Please tell me I'm also correct on the audio/signal input. The Audio neg goes to ground and the Audio + goes to IC1A +in. Right?
Audio in is a real ground and C1 (where it says INPUT).
-
Audio in is a real ground and C1 (where it says INPUT).
All that we've discussed and this is what has me puzzled. What do you mean by this?
-
Audio input is SGND and IN signals.
This circuit will require a lot of tweaking, it will not work out of the box. Also, breadboard is no the best way to prototype this, stray capacitance of the breadboard will make it a nightmare.
Also, start building it from the end. Build one indication chain and see how it works if you feed it with different voltages, then build filters, then buffers, etc.
-
If no part number is given, then the symbol is nearly meaningless.
It could be a rectangle with wires coming out, and the schematic would be identical. No? :)
One would suppose the author intended the symbols, that you should use ESD-protected types. But this kinda begs the question why................
Tim
-
Yeah, I'm really not sure. It was an instrucables that I found. Unfortunately they want visitors to pay $35 for membership to get the rest of the files, so that's not happening.
What else is not happening is anymore PCB etching until I get this completely figured out. You all have done a great job getting me on track with this hokey schematic and Alex said just what I was thinking - start at the end.
I notice that the LED portion is essentially what's inside an LED driver such as AN6684 and others I had purchased. So I like the idea of building a driver circuit and I will do just that. First.
P.S. Anyone PCB etching using the inkject, muriatic acid, hydrogen peroxide method... I've found the absolute best paper is Q1987A HP glossy brochure paper instead of photo paper. Rarely have a trace not transfer. Much faster and less expensive!!
-
Yeah, I'm really not sure. It was an instrucables that I found. Unfortunately they want visitors to pay $35 for membership to get the rest of the files, so that's not happening.
They want $35 to view full article on one page and download PDF. You can click "Next Page" many times and get all the information.
What else is not happening is anymore PCB etching until I get this completely figured out.
Use prototyping boards. And then OSHPark.
-
uhhh, that doesn't work for me Alex. Maybe in a few years I'll look at all them parts and pieces and make sense of it. I'm having trouble with the entire schematic! LOL. Many small pictures doesn't help me unfortunately. :-//
Thanks again. Thank you all very much.
-
It was an instrucables that I found.
Give us the link to the instructable.
-
http://www.instructables.com/id/9-Channel-pure-analog-graphic-equalizer/ (http://www.instructables.com/id/9-Channel-pure-analog-graphic-equalizer/)
-
So now we know that LED used is GM5BW96385A and it has the following note in the DS:
Similar part to GM5BW96380A, but includes self-protection against static electricity
So the mystery of the diode is solved.
-
Where in the heck did you come up with GM5BW96380A? I don't see that mentioned anywhere.
-
On indestructables click next until you get to the last page, there will be schematic and PCB files in Eagle format.
-
oh. well that explains it. I don't have a program to open either of those two files. I did grab that eagle file, but yeah, couldn't open it. So I guess it would have been pretty helpful to us all had I linked to that originally.
Ahhh, sorry and thank you. :)
-
oh. well that explains it. I don't have a program to open either of those two files. I did grab that eagle file, but yeah, couldn't open it. So I guess it would have been pretty helpful to us all had I linked to that originally.
Eagle is free, you can just download and install it.
Also, in this specific case, I don't have it installed either, but those files are human-readable XML files, I just opened .sch in a text editor and searched for one of the LED names on your schematic.
-
I did try to open the file using acrobat, firworks and notepad++ all to no avail. hrmmm. I'll try again? Heck maybe just install eagle?? LOL I thought I remembered it being a trial. It could just be that it's been a while and I don't remember clearly. Either way. Thank You.
-
Notepad++ must be able to open it. The worst case scenario - rename them into .xml.
Eagle is completely free with limited board size and number of layers. Free version can open any board, but won't let you edit outside the limited area. I'm pretty sure everything you'll find on the instructables site was created in a free version.
-
Oh Yeah, it all works. Not sure why I couldn't open that .sch with notepad++ before, but I can now. I also have re-installed eagle and went with the freeware license instead of trial. I guess I'll have to learn eagle at least a bit, but why can't I see the part models in the eagle, but when I open the same file in ++ there it is?!
NutZ! every part used is listed! :-DD
-
but why can't I see the part models in the eagle
They are not normally shown on the schematic, but if you open component properties, you will see all the info you need.
-
Oh yeah I'm starting fresh here. I first had to figure out which tool I needed to be able to select the components. But then... Oh only then are all of the juicy details revealed. All them details...
Thank You Alex.
-
Would you expect output pin 7 of the first filter to always measure 4.92 V? It does until the frequency is between 40 and 60 hertz during that time the output pin 7 then drops to around 4.7 V. Is this how the circuit should behave? Even without using the first half of the first op amp I still get 4.9...
-
Would you expect output pin 7 of the first filter to always measure 4.92 V?
It should be VCC/2, so yes, ~5 V sounds about right.
It does until the frequency is between 40 and 60 hertz during that time the output pin 7 then drops to around 4.7 V.
What is the amplitude of the input signal and what is the source?
The first half is a simple 1:1 buffer, so make sure that is works as such and leave it in place for the following experiments.
-
Okay so I wasn't quite sure what are one and are two should be and didn't want to bother asking you I used 1M and 10K so I have some crazy game going on there. As for amplitude I'm not quite sure. Here's what I've got I'm using my cell phone with a signal generator app every time I turn it on it cranks my volume up to 100 and 50% so I turned it back down to 75% then the signal generator has a game and I said it to -9 to -7 DB. Any lower the output voltage stays at 4.92 no matter what higher and I don't remember what happens...
-
You can't just feed the signal from the phone directly to IC3A input. Remember, the signal needs to be centered around 5V. That's what IC1 does. So if you want to use the phone as the source, you need to pass it though IC1 circuit.
-
using IC1 I believe a buffer. at first I figured might be causing a problem which is why the output voltage was so high I was not expecting that. so I eliminated it to see if there would be any difference and I don't think there was. I guess I was kind of expecting to get next to nothing out of it until the frequency was right and then there would be enough voltage to pass the first couple of diodes turning on the LEDs. I have several types of LED drivers that I planned to use but in this case they would always be on until the frequency matched and then they would go off. I think I need to spend a little time and figure out exactly what's going on here because I apparently don't understand. I have it working but I don't understand...
-
holy smokes! it works! I've built up the next step and it works. I know you said to start at the end and work back and I didn't do that, but once I got all the way to the end it did work as expected which after testing out the first half without the second half is not what I expected. :o
I added two diodes, one capacitor 1uf, two resistors 1k, one 2n2222 transistor and it works perfectly! Perfectly.
Thank You Sir.
-
Alrighty then moving on, it works with the second transistor LED set and it works with a third if I remove one of the resisters from the third set. If I just use an6884 LED driver it works great. So I haven't tried the next frequency yet but what I have noticed is that below 30 Hertz the LEDs Flickr intensely until down to about 5 Hertz then they go off.
-
So I haven't tried the next frequency yet but what I have noticed is that below 30 Hertz the LEDs Flickr intensely until down to about 5 Hertz then they go off.
As I said, the circuit will require tuning and it would be really hard to debug this without proper equipment. You really need at least a scope and a signal generator.
-
Lol. Well I have an oscilloscope but it's one that I made and it plugs into the computer audio card. It works pretty well but I'm sure not as well as the real deal. I've noticed a little interference from the computer internal components whatever sometimes, but yeah I do not have an oscilloscope yet.
-
Ok. So I'm still working on this. For the first three filters I've used electrolytic capacitors. I'm beginning to think I should be using Tantalum. The part used in the schematic is C-EUC0805K. The sizes used are 1uf, 100nF and 10nF.
The circuit is working with electrolytic, but I don't have and can't find 10nF electrolytic. Researching part C-EUC0805K I'm finding that it's a surface mount tantalum.
Is this correct? Should all caps in the project be tantalum?
Thank You,
John
p.s. Ceramic capacitors don't work at all.
-
The circuit is working with electrolytic, but I don't have and can't find 10nF electrolytic.
That's just way to small to make electrolytic.
I see no reason why ceramic capacitors won't work in this case. That would be something I investigate first.
-
Using ceramic capacitors in just one of the three filters makes the other two filters not work. All of the capacitors in the Schematic have a line added to one side making me think that means the negative side which is what I've done using electrolytic where size matches stock.
-
What is the source of your ceramic capacitors? Are you sure they are actually capacitors?
There is no way caps in one filter will affect any other filters, that's why there is a buffer before each filter.
-
Sold as, purchased as, look like and taste like capacitors. 8) Heck they even measure up with my handy LCR meter.
(http://i.imgur.com/iHo7SYU.jpg)
Here's what I've got so far. I'm not a fan of the jumpers, but I'm still fine tuning my layout skills and I also don't have my double sided boards yet. I know you can't diagnose a circuit from a picture like this, but...
(http://i.imgur.com/rz8iInX.jpg)
Here's the bottom
(http://i.imgur.com/f9o1X4I.jpg)
-
I've just purchased these SMD Tantalum Capacitors in 0.010uF, 0.1uF and 1uf. All for this project. Sure hope they work!!
http://www.ebay.com/itm/Kemet-0-01uF-01-50V-X7R1206-SMD-Tantalum-Capacitor-C1206C103K5RAC-10nF-40pcs-/142128076359?hash=item21177e5647 (http://www.ebay.com/itm/Kemet-0-01uF-01-50V-X7R1206-SMD-Tantalum-Capacitor-C1206C103K5RAC-10nF-40pcs-/142128076359?hash=item21177e5647)
-
I'll have ceramic capacitors here tomorrow or the next day that are large enough to use in this project. I have the 10nF and 100nF covered, I just don't have any 1uF ceramics right now... I do have polyester, but they're 100v and none large enough.
Hey, while I'm on this. I've read time and again that capacitor voltage doesn't matter. Just that it needs to be greater than the voltage in the circuit. If I'm running a 5v something I can use a 6.3v or 10v cap, but I can also use 1000v cap. As long as the capacitance value is right for the job, the voltage won't have any negative effects? At all? There is no advantage in this case in using 6.3, 10, or a 1000 other than cost??
Thanks, John
-
As long as the capacitance value is right for the job, the voltage won't have any negative effects?
That is correct in many cases.
At all?
Well, capacitors have many parameters apart from the capacitance value and rated voltage. ESR would be one of those parameters.
But for this project, none of that should matter and any capacitors should work.
-
I'll have ceramic capacitors here tomorrow or the next day that are large enough to use in this project. I have the 10nF and 100nF covered, I just don't have any 1uF ceramics right now... I do have polyester, but they're 100v and none large enough.
Hey, while I'm on this. I've read time and again that capacitor voltage doesn't matter. Just that it needs to be greater than the voltage in the circuit. If I'm running a 5v something I can use a 6.3v or 10v cap, but I can also use 1000v cap. As long as the capacitance value is right for the job, the voltage won't have any negative effects? At all? There is no advantage in this case in using 6.3, 10, or a 1000 other than cost??
Thanks, John
Sort of! It depends on the type of capacitor...
Short Answer: As long as the maximum voltage you place across the capacitor is lower than the capacitor's rating, it won't blow up or anything. The voltage rating is literally the working voltage of the capacitor. (So, like you said, you can put 5V across a 6.3, 10 or 1000V capacitor and you're fine; however you shouldn't put 1000V across a 6.3V cap!)
Long Answer: Depending on the dielectric of the capacitor, what you're using it for and the temperature you'll be using it at, you may need to derate the value.
Aluminum Electrolytic: These are generally used for bulk storage and smoothing ripple. As a rule, the higher the voltage rating of this type of cap, the larger the ripple current it can tolerate. This shouldn't be something you have to worry about unless you're making a huge linear power supply or something.
The voltage rating of AE caps is also fairly conservative; you can generally go over the rated voltage without trouble (assuming you're using a brand name cap and not a questionable off-brand). If you do exceed the voltage rating, you'll lose some capacitance (e.g., if you run a 10uF 6.3V cap at 9V, you might only get 8uF of capacitance).
Ceramic: These are generally used for decoupling. They tend to be very temperature sensitive, and quickly lose capacitance as temperature changes when run near their rated voltage. Say you've got a board inside an enclosure that gets up to ~85c. You've got a bunch of 1uF and 0.1uF 6.3V Z5V ceramic caps on the board to decouple the 5V rail. Well, because you're running them near the rated voltage and at a temperature above ambient, the capacitance could be down by as much as 80% (meaning that 1uF cap is really a 0.2uF cap)!
The way to avoid this is to use better dielectrics like X7R (which will only lose 15% of its rated capacitance over a temperature range of -55 to 125c). As a rule you should also derate the voltage, I like to make sure my ceramic caps are rated for at least double the voltage they'll see (so, a 10V cap if I'm decoupling a 5V rail); this also helps to lessen the loss of capacitance over temperature.
MLCC style caps should be avoided anywhere they'd be exposed to frequencies in the audio range. Due to the nature of their construction, they can actually resonate and "sing"! The inverse is also true, so additionally they should be avoided in equipment that will be exposed to regular vibration or shock, as this can cause the capacitor to actually generate a voltage on its own! (This is called microphonics and I believe Dave did a video on this, where he taps an SMD cap and shows the voltage on a scope.)
Tantalum: These have similar uses to AE caps. You absolutely *must* obey the voltage rating of tantalum caps! They have a nasty habit of literally exploding (violently, too) when their voltage rating is exceeded. Tantalum is also a conflict mineral, which is sort of interesting from a philosophical point of view.
Okay. So that's the long version of the basics. :)
-
Thank you very much!
-
Okay folks. I have the multilayer ceramic caps, I've re-re-built using them and it works. It works very well. Very Well. I did have another problem too. Each filter (seems to me anyways) is a voltage divider then two caps in parallel one having a resistor after it. I had placed the divide R2 after the two parallel caps. This is where the re-re comes in...
Hey I'm super green, what else can I say but thank you guys.