Simple query on voltage delay (Capacitors)

Thread Starter

Cyclicz

Joined Apr 17, 2017
62
I have a simplistic setup where I'm using LEDs and a relay (Due to higher voltages). The issue I'm wanting to solve here is that the LED and relay circuit is "too responsive" and goes from vcc to 0v in an instant. My goal is to have the led fade down from vcc to 0v when the relay opens. Now my first thought here is to add a single electrolytic "smoothing capacitor" in parallel with the load to slow the change in voltage, but my circuit fits within a moderately tight space and a 330uf 130v capacitor is moderate in size. Perhaps there's a way I can slow the capacitors charge and discharge rate with something like a resistor (the issue then arises of heat and bad efficiency) or an inductor rather than looking for higher capacitor "size" values. I tried adding lower voltage capacitors with the same capacity and it just ends up dimming the leds. Can't really seem to find the specific data I need to solve this problem.
 

DickCappels

Joined Aug 21, 2008
10,165
Here is a schematic clipping showing how you can do it with a small capacitor.

upload_2017-5-12_13-2-49.png
In its circuit, R10 goes to +3C, and the LED on/off signal is applied to C7/R5. Miller capacitance slows down the switching speed.

If you post your circuit you can probably get an answer tailored to your application.
 

Thread Starter

Cyclicz

Joined Apr 17, 2017
62
Here is a schematic clipping showing how you can do it with a small capacitor.

View attachment 126528
In its circuit, R10 goes to +3C, and the LED on/off signal is applied to C7/R5. Miller capacitance slows down the switching speed.

If you post your circuit you can probably get an answer tailored to your application.
Interesting concept. My application circuit is just a basic circuit using a LED driver and a relay. So if the input signal is deliveried through the transistor, where are the other connections? I'm pretty sure I got the connections incorrect but also, is the R10 only present for current limiting? If I'm using a constant current LED driver I might be able to remove it.
 

Attachments

crutschow

Joined Mar 14, 2008
34,390
Dick's circuit will require a much larger capacitor to get a noticeable slow down.
The values shown give a simulated rise-time of about 300μs.

How high is the "high voltage"?

What's the LED current value?

Post the complete circuit.

Do you care if there's a slow turn-on as well as slow turn-off of the LED?
 

Thread Starter

Cyclicz

Joined Apr 17, 2017
62
Dick's circuit will require a much larger capacitor to get a noticeable slow down.
The values shown give a simulated rise-time of about 300μs.

How high is the "high voltage"?

What's the LED current value?

Post the complete circuit.

Do you care if there's a slow turn-on as well as slow turn-off of the LED?
The capacitor that gives me a good delay is 200v at 330uf. That's all I really need in perspective to delay time.
I don't think having a slow turn on really matters to me. I'm working with 126vdc with a 3A LED. (A 63v cap can be used but it dims the led significantly of course). I'll be sure to post a diagramed semi-full circuit when I have access to a circuit program, I'm not sure I have enough time to reverse engineer the LED driver, but it's just a standard constant current driver.
 
Last edited:

Thread Starter

Cyclicz

Joined Apr 17, 2017
62
I'm thinking something maybe involving transistors, as they can handle high voltage in a compact form factor... maybe a capacitor resistor circuit? Is my Objective just not possible for my size limits?
 

crutschow

Joined Mar 14, 2008
34,390
Your simple circuit shows the constant-current going through the relay but nothing controlling the relay, so I still don't understand where the capacitor goes or how your circuit works.
We're kind of blowing smoke here, so really need a schematic.
 

Thread Starter

Cyclicz

Joined Apr 17, 2017
62
Your simple circuit shows the constant-current going through the relay but nothing controlling the relay, so I still don't understand where the capacitor goes or how your circuit works.
We're kind of blowing smoke here, so really need a schematic.
Apologies for the little information I was able to give at the time.
Here's a schematic for my circuit. If you have any further questions let me know.

*Refrence #16 for correct circuit*
 

Attachments

Last edited:

crutschow

Joined Mar 14, 2008
34,390
You stated it's a 3A LED but the circuit shows a 1k resistor in series with the LED which would give about 0.12A.
I'm still confused. o_O
 

Thread Starter

Cyclicz

Joined Apr 17, 2017
62
You stated it's a 3A LED but the circuit shows a 1k resistor in series with the LED which would give about 0.12A.
I'm still confused. o_O
Doh! I must have misplaced that resistor. There is no R2 since it's constant current driven to begin with.
(The led array consists of 11 leds in series but I'm not sure if that information helps you very much)

In order to get any more information I'd most likely have to de-solder from the board so I think that's all I got here.
 

EM Fields

Joined Jun 8, 2016
583
Apologies for the little information I was able to give at the time.
Here's a schematic for my circuit. If you have any further questions let me know.

View attachment 126609
1. You've shown a relay with one end of a 9 volt coil connected to a 5 volt DC supply and the other end connected to 0 volts through an N channel MOSFET with the drain and source wired backwards.

The relay is guaranteed to work with 6.75 volts across the coil, so with a 5 volt source and the drop across the [correctly wired] MOSFET subtracting from that, your circuit is iffy, at best.

2. You've not connected a catch diode across the relay coil, so even if the relay won't actuate you could still blow up the MOSFET.

3. See: https://www.onsemi.com/pub/Collateral/2N7000-D.PDF
 

Thread Starter

Cyclicz

Joined Apr 17, 2017
62
1. You've shown a relay with one end of a 9 volt coil connected to a 5 volt DC supply and the other end connected to 0 volts through an N channel MOSFET with the drain and source wired backwards.

The relay is guaranteed to work with 6.75 volts across the coil, so with a 5 volt source and the drop across the [correctly wired] MOSFET subtracting from that, your circuit is iffy, at best.

2. You've not connected a catch diode across the relay coil, so even if the relay won't actuate you could still blow up the MOSFET.

3. See: https://www.onsemi.com/pub/Collateral/2N7000-D.PDF
Even more minor.
I would say it's upside-down. :D
Very minor item: The 2N7000 is drawn backwards.
I have no idea why everyone here continues to deal with my large amount of mistakes. Anyhow the perseverence is appreciated, here's the fixed circuit. I didn't even hook up my voltage regulator on the previous circuit!

Screenshot_6.png
*Please note that V2 is a constant current source*
I have yet to measure the voltage of each individual LED in the array.
 

EM Fields

Joined Jun 8, 2016
583
I have no idea why everyone here continues to deal with my large amount of mistakes. Anyhow the perseverence is appreciated, here's the fixed circuit. I didn't even hook up my voltage regulator on the previous circuit!

View attachment 126671
*Please note that V2 is a constant current source*
I have yet to measure the voltage of each individual LED in the array.
Try this:
LED soft OFF.png
1. If you select a relay with a 5 volt coil, then you can connect it directly to the 5 volt rail and (unless you need it
for something else) you can get rid of the 3.3 volt regulator
2. If U1PB0 is a totem-pole output, you won't need R3.
3. D1 is the catch diode you'll need to clamp the relay coil's turnoff transient to the 5 volt supply.
4. If V2 is a constant current supply, then the 11 LEDs should be connected in series, the ends of the string
should be connected across V2, and V2's output should be adjusted for the desired current through the 11
LED series string when the relay is made.
 

Thread Starter

Cyclicz

Joined Apr 17, 2017
62
Try this:
View attachment 126679
1. If you select a relay with a 5 volt coil, then you can connect it directly to the 5 volt rail and (unless you need it
for something else) you can get rid of the 3.3 volt regulator
2. If U1PB0 is a totem-pole output, you won't need R3.
3. D1 is the catch diode you'll need to clamp the relay coil's turnoff transient to the 5 volt supply.
4. If V2 is a constant current supply, then the 11 LEDs should be connected in series, the ends of the string
should be connected across V2, and V2's output should be adjusted for the desired current through the 11
LED series string when the relay is made.
1. Although I would have loved a 5v relay I only have immediate access to 12v or 3v right now and I wanted to keep the supply of my circuit below 5.
2. Good to know, with my attiny I like to be safe when it comes to a logic level mosfet and esd. I'll remove R3.
3. I'm not sure what that diode is there for to be honest but I'll trust your experience here.
4. As much as I'd like it to be in parallel, I'm not going to make a custom pcb for this project, yet series wiring seems to work fine for now.

The small tweaks are appreciated although the main question on how to smooth the output from the relay using something other than a huge capacitor still stands.

Cheers :)
 

EM Fields

Joined Jun 8, 2016
583
The small tweaks are appreciated although the main question on how to smooth the output from the relay using something other than a huge capacitor still stands.

Cheers :)
It's possible that a more compact solution could be found which eliminates the relay and drastically reduces the size of the capacitor, but first you must tell us the voltage and current rating of the LED array, a link to its datasheet, or the part numbers of the individual LEDs it comprises and, assuming the LED driver is adjustable, its minimum and maximum output current available, its compliance, or better yet, a link to its datasheet.

It'd also be nice to know what kind of time you'd like for the array to fade to zero from full brightness.
 
Last edited:

Thread Starter

Cyclicz

Joined Apr 17, 2017
62
It's possible that a more compact solution could be found which eliminates the relay and drastically reduces the size of the capacitor, but first you must tell us the voltage and current rating of the LED array, a link to its datasheet, or the part numbers of the individual LEDs it comprises and, assuming the LED driver is adjustable, its minimum and maximum output current available, its compliance, or better yet, a link to its datasheet.

It'd also be nice to know what kind of time you'd like for the array to fade to zero from full brightness.
I wish I had this knowledge on hand but I don't. I've opened things up to take a peek.

Each of the 11 led "modules" actually have 2 leds in the each so a total of 22 leds all in series.

There are actually 4 wires between the driver and the led which are a little strange. It seems there's a LED+, LED-, G (I'm assuming for gate), and S (I'm assuming for source.)

On the led panel there is a MOSFET package labeled 1N60GOD. I'm assuming this to be a power mosfet. 1 side of the led array (-) is attached to the drain pin of the mosfet, source is attached to the source wire from the driver, the gate pin is attached to the gate wire from the driver, and the other end of the array is attached to the (+) from the driver.

Moving back to the driver board I see a smoothing capacitor looking about 73uf at 200v In parallel with the LED positive and negative wire. I see a tiny chip (which I think is a controller) labeled 112PL. I couldn't find a datasheet on this component.

The other peice of importance on the board is in a 3pin mosfet-like package, labeled FTL04N60D, I was able to find a product page in some Asian language for this product with some statistics here.

I'm not sure if this small problem is solvable without ripping things apart further. There may just be no solution currently available. Let me know if the information I provided helped at all (which I doubt it)
 
Top