Signal generator question

[Gazmon]

Hi,

Apologies for the basic questions

Im building some hobby circuits with LED etc and other small voltage circuitry, my question is can I use my signal generator to make them work as well?

i.e 5v amplitude + hz for a simple led circuit etc

thanks
gazmon

[rstofer]

Which signal generator?  How much output current capability?  How much current required?

Were it me, I would have the signal generator drive a transistor and let the transistor blink the LED or whatever.

I'm assuming your signal generator does square wave output and it would be useful if the waveform could be offset such that it didn't extend below 0V.

[Gazmon]

It’s just a feeltech fy3200

24mz

[Zenith]

The FY3200 has a 50 ohm output impedance. The FY6800 manual, (I assume the FY6800 is similar to the FY3200), says it can tolerate having the outputs shorted for at least 60 seconds, so the FY3200 should be able to flash a LED with suitable ballast.

I wouldn't do it. I'd rather have the sig gen working into a transistor, (as mentioned by rstofer previously), which could easily drive one or more LEDs and is nice and cheap to fix if anything goes wrong.

But if you are using the FY3200 to drive monostables and such and are not requiring it to output into very much less than 50 ohms, that's what they are for.

[radiolistener]

if you are afraid of damaging generator with a low-resistance load, just connect your gen through 25-50 Ohm resistor.

But generator output cannot drive high current, it is designed for 50 Ohm. It means that if amplitude is set to 5 Vpp, the max current will be just 5 / 50 = 0.1 A. As other mentioned above, you can fix it by adding some transistor to control more powerful load.

[bostonman]

Sorry to dig up an old thread.

I'm curious about grounds. If the transistor is being powered by an external source (say a 9v battery), where does the ground on the BNC from the generator go?

[shapirus]

Sorry to dig up an old thread.

Always a good idea if it is to ask or answer a useful question on the topic that adds value for those who come via search later.

I'm curious about grounds. If the transistor is being powered by an external source (say a 9v battery), where does the ground on the BNC from the generator go?

It can technically go anywhere (as the generator's output is floating), but you generally want to connect it to the reference point of your circuit, aka its ground, which, in your case, will most likely be (guessing here -- it depends on your circuit) the negative terminal of the battery.

[bostonman]

I always get mixed up with multiple voltages and grounds.

Can't tying grounds cause the generator to rise above its potential? I was thinking the generator BNC was Earth ground (I don't have one handy to measure at the moment), so essentially it causes a ground loop.

[RoGeorge]

if you are afraid of damaging generator with a low-resistance load, just connect your gen through 25-50 Ohm resistor.

All 50\$\Omega\$ generators have an internal series 50\$\Omega\$ on their output, and can withstand to a short-circuit indefinitely.

To answer the OP question, I've used a few times a Rigol DG4102 AWG as a two channels voltage source.  DG4102 AWG is a model that can generate as well 0Hz (DC).  Of course, good as a voltage source only for very small current loads.

Be careful when turning the amplitude knob when on DC!  It's very easy to put too much voltage, or even worst, to turn the output into a negative voltage by mistake, when you try to reduce the voltage.

[shapirus]

I was thinking the generator BNC was Earth ground (I don't have one handy to measure at the moment), so essentially it causes a ground loop.

Yes it may be, I didn't think of it. You need to verify this, and if it's really mains earth-referenced, then your options of where you can connect it are limited, unless your circuit is floating.

[dobsonr741]

FY2300 is quite a big jump up if doing hobby circuits with LEDs, etc. If I would be doing it I would start with a 555 with two pots and a CD4060 on a breadboard.

[bostonman]

Personally I planned to use a generator to blink a light so I could demonstrate to someone what frequency is and show it on a scope.

I thought 1Hz and then up to maybe 10Hz would be a nice demonstration, but a 555 needs too much tweaking to get that range.

[dobsonr741]

That's where the CD4060 comes in. Can divide the frequency by 16384. Can use two 4060s - and have 1 Hz on 555, and once a day on the 2nd 4060. Here's a good example: https://diyodemag.com/education/the_cd4060_14-stage_counter

[bostonman]

But I can't change frequency quickly by turning a pot, or, in this case, adjusting the generator.

[dobsonr741]

Both 555 and 4060 can happily take a pot to adjust the frequency.

[CamJam]

Hello,

Amateur circuit enthusiast here.

Thanks to everyone who has contributed to this post so far! I was searching the web to find out if I could drive/modulate an LED through a square wave signal generated by a function generator (FG) rather than driving a transistor.

From what I gather, as long as the FG can supply the appropriate voltage and current that the LED is rated for, the FG can be used. (Please correct me if I am wrong.)

Other than protecting the FG from possible harm due to the LED drawing too much current, are there additional benefits to driving a transistor to modulate an LED rather than using a FG? Does it allow higher frequency modulation of the LED? Better LED performance (i.e., more intensity/brighter pulses)?

Apologies if my questions are mundane!

[BennoG]

A function generator can drive 1 small led, if you want to drive more leds you need power amplification (mosfet / transistor)
So that would be a very expensive led driver for a single led.
Most if not all FG's are build to deliver only a couple of mA.  A single led takes about 10-20 mA a COB can take up to 3000 mA

Benno

[radiolistener]

can I use my signal generator to make them work as well?

i.e 5v amplitude + hz for a simple led circuit etc

it depends on how much current is required for your LED.

Lab signal generator has 50 Ω output impedance, it means that for 5 Vpk it can drive max I = 5 / 50 = 0.1 Amps. Is it enough for your LED? If not,  then you're needs to use some driver which can provide more power.

Also note that RMS power for AC signal will be less than for DC with the same voltage. The max output power on 50 Ω port for sine wave with 5 Vpk amplitude will be P = 5^2 / (2*50) = 25 / 100 = 0.25 W.

So, if 0.25W is enough power for your LED, then you can drive it with 5 Vpk sine wave directly from signal generator, if you want more power, then you're needs power amplifier.

[CamJam]

Thank you @Bennog and @radiolistener,

I was originally trying to use a RIGOL DG952 function generator (50 Ω out) to modulate a 10 mW UV LED requiring a 7.2 V forward voltage and 100 mA... which I now understand my the DG952 can drive the LED directly!

Now I want to upgrade my UV LED to a 18 mW model requiring 7 V forward and 180 mA current... the RIGOL won't suffice for this one as 7/50 = 140 mA.

This is where the transistor driver shines as I can power my LED with DC power supply and current limit it to 180 mA and then use the transistor connected to the DG952 to pulse drive the LED to 10 kHz (I think). (See p. 8 of attached .pdf).

Thanks 

[RoGeorge]

I was originally trying to use a RIGOL DG952 function generator (50 Ω out) to modulate a 10 mW UV LED requiring a 7.2 V forward voltage and 100 mA... which I now understand my the DG952 can drive the LED directly!

Now I want to upgrade my UV LED to a 18 mW model requiring 7 V forward and 180 mA current... the RIGOL won't suffice for this one as 7/50 = 140 mA.

Rigol DG952, accordong to the specs, https://www.tequipment.net/Rigol/DG952/Function-Generator/#description , can do

Max Amplitude Vpp @ 50 ohms   10 V

which means, internally it has a 20Vpp in series with a 50 ohms resistor, or else said, +/-10Vdc in series with 50ohms.  Since the LED needs DC, the max you can set in the generator is +10V, but there is already an internal 50 ohms in series, inside the generator.

Your 10mW LED requires about 7V, so if you connect it directly to the generator, the max current through the UV LED can be no more than (10V-7V)/50\$\Omega\$ = 60mA.  So, the generator alone can not drive the 10mW LED at its max power.

Another trick you can do, is to connect the LED between the CH1 and CH2, and synchronize the 2 channels to output pulses of opposed polarity:  CH1 between 0/+10V, CH2 pulses between 0/-10V.  Yhe LED is connected between the center of the BNC connector at each channel, with anode at the middle pin of the BNC out at CH1, and the cathode at the middle pin of the BNC out from the second channel.  This will give you max 120mA, so still not enough to drive the 18mW to its max power.

If you want to drive the LED to the max, you will need a transistor and a power source, even for the 10mW LED, let alone for the 18mW one.
- Beware at the LED polarity, they got damaged easily if you connect them in reverse.
- Beware of the LED with 3 pins (with optical power photodiode included), they got damages if the optical feedback diode is neglected.  For those 3 pins "LEDs" you'll need a dedicated control loop if you want the max power without damaging the LED.

- Most of all, beware of the LED power, it can do permanent damage to the eyes, yours or others.

[CamJam]

Thank you for the informative reply, RoGeorge!

To reassure you a little, I wear UV blocking face shields, a lab coat, and gloves while operating any UV LEDs.

Thank you for correcting my original interpretation that I could run the 10 mW LED with the function generator and confirming that I need a transistor driver to properly modulate the 18 mW LED 

- Beware at the LED polarity, they got damaged easily if you connect them in reverse.

Appreciate the reminder, I am aware of the polarity issues, I always double check with a multimeter!

- Beware of the LED with 3 pins (with optical power photodiode included), they got damages if the optical feedback diode is neglected.  For those 3 pins "LEDs" you'll need a dedicated control loop if you want the max power without damaging the LED.

The LEDs I am using are surface mount LEDs with 3 pads, an anode, cathode, and heat sink. Thankfully no need for a control loop  (unless the heat sink is the "optical feedback diode"). If you have any insight on how to properly accommodate the heat sink pad of the LED I am all ears. (will look elsewhere on the forum for this info as well.)

I am currently planning on just having the heat sink pad soldered on a relatively large area of copper/metal to help with heat dissipation. 

[RoGeorge]

There are many types of LEDs, hard to say without the part number.  Are they already soldered on the PCB?  Usually, it's an aluminium substrate PCB.  If not already soldered, then follow the manufacturer recommendation for the radiator.  Should be in the application notes, or in the datasheet.  If you just have 1-2 LEDs for one prototype only, might be possible to use a consumer LED light-bulb, as a donor for the PCB/radiator plate.

Try to attach a picture of your PCB and/or LED, or else it would be too generic to talk about all the possibilities.

[CamJam]

Hi RoGeorge,

I am using a E275-10 LED from InternationalLight (it was some of the more powerful surface mount UV LEDs I could find at the time). Just need one LED for my enclosure/project.

The datasheet only mentions that the LED should be operated at -30 to 60°C. Tested one out today (powered the LED with DC power supply 6.5 V, current limited it to 180 mA) with a MOSFET transistor driver (IRL2703PbF model, applied 2.5 V at the gate).

It worked (i.e., the LED was emitting UV), but got bloody hot (leftover flux on pads of LED smoked).

I attached the datasheet (had to scan it as what is provided on the website is not the same).

Thanks for your insight!

Cam

[RoGeorge]

You still didn't show on what is the LED soldered onto.

What kind of PCB (printed circuit board) is the LED soldered on?  Did you use Aluminium substrate PCB, or normal FR4?  What thickness is the copper clad, and what copper surface did the pads have?

That type of LEDs can not work without radiator.  Get some Al substrate PCB from a defective light bulb.  If you don't have Al PCB at hand, you can find defective white light bulbs at any EE waste basket, or sacrifice a working light bulb you might have at home.  Remove the plastic cap of the light bulb (there are videos online of how to remove the plastic cap without damaging the interior), and retrieve the Al PCB.

The Al plate to retrieve from the light bulb looks something like this (without the wires):



Snip away one of the white LEDs, preferably one from the middle of the plate.  Do not try to desolder the white LED, that won't be easy (unless you use a preheater).  Snip the white LED with a pair of pliers, then clean the pads one by one, and solder your UV LED instead + some wire on the other side end of the copper clad.

Use the biggest soldering-tip you have, and preferably use a preheater, the Al underneath is a big heat sink, normal soldering iron won't be able to melt the solder.  Do not increase the temperature on the soldering iron, but instead preheat the Al plate at 10-150*C or so.  That will help a lot when soldering your LED.  I wouldn't use any more the UV LED that smoked the flux, because overheating shorten their lifetime a lot.  Use a new one if you have.

Then, test how much is it heating inside your enclosure.  If the plate gets so hot that you can't keep keep a finger pressed on the back of the Al plate, then it's probably too hot, more than the 60*C allowed by the datasheet.  Make sure there is good air circulation, eventually poke some holes to your enclosure, for natural convection to lower the temperature.  Without air circulation, any radiator will eventually heat too much, no matter how big the radiator.