Author Topic: RGB Amplifier  (Read 616 times)

Offline bowhunter3125

  • Full Member
  • ***
  • Join Date: Jan 2016
  • Location:
  • Posts: 233
  • Kudos: 0
RGB Amplifier
« on: March 20, 2017, 09:40:24 PM »
Not happy with the Amplifiers currently on the market. I run around 1500 dumb nodes in my show and last year i went through a lot of the amplifiers you can get from ebay and amazon. Issue i have seen with them through testing is they do not like to be turned off and on a lot. Tunr them on and leave them on they are ok, but not for what we do. SO last year i started on designing a new amplifier, one that will work for what we do. This will fit inside a TA 200 enclosure. I am currently looking at using either a LM324 or an OP482 operational amplifier and 6 or 10 amp mosfets. I am unsure why there is such a huge price difference between them. I dont quit understand how to read data sheets, but another designer suggested the OP482.


I dont think there will be a very high demand for these, most folks use mainly smart pixels for their shows, i use smart for my tree and matrix, but everything else is dumb nodes. Was hoping for some input on my design from you smart designers out there. I did not include the 47 ohm or 10k ohm resistors between the OP Amplifier and the Mosfets on my design, they will included in the final PCB drawing. I also have plans to include an LED to show it is getting power. I am up in the air about including a fuse as all my power from the PSU's runs through a fused distro board.




Offline AAH

  • Sr. Member
  • ****
  • Join Date: Jul 2014
  • Location: Australia
  • Posts: 481
  • Kudos: 14
  • Blinky blinky blinky
    • I love blinky lights
Re: RGB Amplifier
« Reply #1 on: March 21, 2017, 03:11:53 AM »
  There's all sorts of things that you need to be aware of. The voltage at which they work will govern the choice of mosfet and potentially the op amps. The maximum drain source voltage (Vds) is likely to be 30V or higher for most mosets that you are looking at so that probably isn't an issue. The 2 gate source voltages are potential problems. The Vgs-threshhold is where the mosfet turns on. This is typically 2.5-3V and that's where is starts turning on hard and acting more like a switch. You will need to ensure that you're above the threshhold voltage but below the Vgs-maximum voltage. A 5-10V Vgs will typically turn on most mosfets quite hard. The op amp that is used will preferably have rail to rail output voltage swings or close to it. The LM324 will come fairly close to what you want. It's important to have 47R or similar gate resistors or you can get funkiness happening (been there/done that). You will need some sort of pullups on the input side of the op amp to provide the signal inversion. It may actually be better using somewhat lower resistance that 10k as the output capacitance of the mosfet whose signal you're amplifiying needs to be overcome. If you're planning on using 12V and 1/4W resistors then I would think about maybe using 1k pullups. The gamma curve, particularly at low dimming values, should be noticably better. If you're looking at switching 5A for instance I wouldn't use less than 20A mosfets as the extra capacity will allow them to drop less voltage and hence stay cooler. Going to much higher current mosfets like the SSF7509 ones I use on most of my dimmers is a trade of. They tend to have higher gate capacitance and hence turn on and off slower but they have much lower on resistance and therefore stay cooler and don't need any heatsinking.
  I'm too lazy to do some maths on slew rates and rise and fall times for the mosfets but assuming that the source PWM signal is switching at 200-400Hz then the mosfet needs to be going rail to rail faster than 255 times that switching rate to allow for the full 0-255 dimming range. The LM324 has a slew rate of 0.5V/us and the OP482 has a rate of 9V/us. Given that these are quite slow speeds I would be thinking about whether the voltage range that the amplifier was being designed for was within the operating range of 4000 series CMOS logic as it's a helluva lot faster but has lower output current and is 15 or 18V depending on manufacturer.
You can call me AAH, you can call me Al, or you can call me Alan.
http://www.hansonelectronics.com.au/wp-content/uploads/Catalogue.pdf Catalogue of my blinky bits
http://www.hansonelectronics.com.au/

Offline bowhunter3125

  • Full Member
  • ***
  • Join Date: Jan 2016
  • Location:
  • Posts: 233
  • Kudos: 0
Re: RGB Amplifier
« Reply #2 on: March 21, 2017, 09:21:57 PM »
I am fairly new at this. What is the purpose of the pull up between the input at the Op Amp? The mosfets i chose are the same ones we use in our dc rgb controllers.

Offline AAH

  • Sr. Member
  • ****
  • Join Date: Jul 2014
  • Location: Australia
  • Posts: 481
  • Kudos: 14
  • Blinky blinky blinky
    • I love blinky lights
Re: RGB Amplifier
« Reply #3 on: March 22, 2017, 01:09:23 AM »
The pullup is between the incoming channel and the DC+. Without the resistor the input will end up sitting at zero volts. With too high a value it takes too long for the resistor in conjunction with mosfet capacitance to get back up to 5V, 12V or whatever. It's not an instantaneous thing.

Offline bowhunter3125

  • Full Member
  • ***
  • Join Date: Jan 2016
  • Location:
  • Posts: 233
  • Kudos: 0
Re: RGB Amplifier
« Reply #4 on: March 22, 2017, 09:48:12 PM »
The pullup is between the incoming channel and the DC+. Without the resistor the input will end up sitting at zero volts. With too high a value it takes too long for the resistor in conjunction with mosfet capacitance to get back up to 5V, 12V or whatever. It's not an instantaneous thing.


So I need to put a 1k ohm resistor between the RGB wires and the -A, -B, -C pins on the Op Amp?




Offline AAH

  • Sr. Member
  • ****
  • Join Date: Jul 2014
  • Location: Australia
  • Posts: 481
  • Kudos: 14
  • Blinky blinky blinky
    • I love blinky lights
Re: RGB Amplifier
« Reply #5 on: March 22, 2017, 10:17:57 PM »
  The 1k resistor needs to go from the incoming R,G and B terminals to the +5 if you're using a 7805 regulator. You also need to put in a capacitor on both the input and output of the regulator as they can oscillate and self destruct if there's no capacitor. A 47uF on both is probably all that is required. Your schematic doesn't show the 0V connection to the 7805 reg and the 4 +IN inputs of the op amp need a fixed voltage of something like 2.5V from 2 1k resistors in series. The 4 47R resistors from OUT to the gates of the mosfets aren't drawn either.
  Depending on the incoming voltage the 7805 can either be a good thing or a bad thing. For voltages over 12V it will reduce the voltage swing of the op amp and make the high to low transitions faster. For under 8V there will be insufficient voltage to power the regulator. As your using op amps and not CMOS logic you can probably get away without the regulator unless you use 1 of the high speed low voltage op amps that only handle 5V and not to 30V or more that most op amps can handle. If it wasn't for the requirement for an output pullup resistor the use of a quad comparator like the LM339 could be used. They switch a helluva lot faster than op amps but only pull down in the same way as the mosfets do. You need a low resistance pullup in order to overcome the mosfet gate capacitance.

Offline bowhunter3125

  • Full Member
  • ***
  • Join Date: Jan 2016
  • Location:
  • Posts: 233
  • Kudos: 0
Re: RGB Amplifier
« Reply #6 on: March 23, 2017, 10:01:19 AM »
Now i thought, sense the incoming RGB lines are essentially grounds, that they connect to the Negative inputs of the OP amp? I did not include the 47k or the 10k resistors between the OP and the Mosfets in my schematic, i know they have to be in there, just did not want to include them yet in case any major changes needed to be done. And i just looked, the OP482 can handle up to 15 volts input,, so yeah, i can eliminate the 7805 if i go with that Op Amp. The LM324 can handle 3-32 volts so wont need a regualtor with that one either.

Offline bowhunter3125

  • Full Member
  • ***
  • Join Date: Jan 2016
  • Location:
  • Posts: 233
  • Kudos: 0
Re: RGB Amplifier
« Reply #7 on: April 07, 2017, 10:18:24 PM »
It looks as though the 12 volt incoming needs to connect to each of the + input pins for the RGB lines, is this correct?

Offline AAH

  • Sr. Member
  • ****
  • Join Date: Jul 2014
  • Location: Australia
  • Posts: 481
  • Kudos: 14
  • Blinky blinky blinky
    • I love blinky lights
Re: RGB Amplifier
« Reply #8 on: April 08, 2017, 08:07:06 AM »
They need to be connected to something of about half supply voltage. Dividing the voltage in half by using 2 1k resistors between 12V and gnd will do the job. I'd draw a pic but i'm pretty much limited to 1 finger typing for forseeable future (or at least until I see doc on Wednesday).


 

Back to top