Edit:
First question Solved!
Next question post #5
My latest project involves controlling the dimming of this power supply: HLG-240H-C2100B Datasheet
I am unsure of whats needed to provide the "additive PWM" shown.
Can I just connect a mosfet across the dim+ and dim-?
Just wanna be sure to contain the magic smoke.
"Normally" to get 10V PWM you'd use 10V power supply and a transistor as a driver/switch [u]like this[/u].
But, I'm pretty sure 10V and the pull-up resistor (RL in the attached schematic) is built-in so all you should need is the transistor and base-resistor. (This a low current application so almost any NPN transistor should work along with a 1K - 10K base resistor.)
Note that the transistor is an inverter so you'll have to invert your logic (100% PWM is fully-dim and 0% PWM is fully-on).
There are two reasons for deducing that 10V is built-in... With nothing connected to the dimming circuit you get full-brightness. And, it can be dimmed by simply adding a variable resistor.
Your right I measure 10v across dim+/dim- and am currently using a resistor for dimming.
So a bjt with the collector connected to dim+, the emitter connected to dim- and a resistor between the base and an arduino pwm pin is all thats needed? TBH this was my first thought but i figured it couldn't be that easy.
Maybe safer to use an optocoupler.
Opto LED with (1k) CL resistor driven from an Arduino PWM pin.
Opto transistor just across the DIM terminals of the LED driver (no other parts).
Make sure you get polarity right (collector to DIM+).
Leo..
That sounds prudent. I think I'll run out to fultons before they close.(Yes we still have a real electronics supply store in town. 
)
First Karma to @DVDdoug & @Wawa Their suggestions worked perfectly Thanks 
Next I would like to monitor secondary circuit operation without cutting into the circuit.
My first thought was to use a LDR, but that would require drilling the case which I would rather not.
Then I started looking for a non-contact dc current sensor, but the ones I've found so far are rather expensive. Any ideas on how to solve this?
It seems dc current sensing is harder than I thought, at least the sensors I've found so far aren't cheap.
I need a inexpensive non invasive way to determine if a led or power supply has failed.
Any thoughts would be appreciated
An ACS712 breakout board rated for the secondary LED current could be an easy solution.
Remember that you're measuring PWM, so sync measurements with the PWM pulse or average many readings.
Leo..
I was really hoping to not cut into the circuit.
This may be a little far out but....because you don't want to break the circuit you might try this:
The inputs to the ACS712 simply brings a current carrying conductor in close proximity to the hall cell. If you don't need accuracy, you might be able to run the wire directly over the chip (instead through the provided terminals) and still get a GO / NoGO indication. It might only work at higher brightness or maybe not at all. If someone has a ACS712 they might be kind enough to perform a quick test.
Interesting... yes I only need to confirm circuit operation no need to know the actual current.
Have one on hand so I will test this after work tonight. Thanks
Unfortunately that didn't work. I did not get any response at all. 
You can use a clamp on current sensor on the AC side - you will still have to open the wire as only one conductor should go through the sensor, not both.
As you're using PWM, this kind of AC sensor should also work on the DC side - until you reach 100% duty cycle the current does fluctuate all the time, and that such sensors should be able to pick up. Assuming they can handle the frequency; PWM from an Arduino is typically ~480 Hz or ~960 Hz while those sensors are designed for 50-60Hz.
Hutkikz:
Unfortunately that didn't work. I did not get any response at all.
You will have to use only one conductor, not both. Otherwise the magnetic fields cancel out one another...
I am planning to monitor input power as well. monitoring secondary power is to determine if the power supply or one of the leds burns out.
I am thinking that since the dimming circuit in the power supply can take a variable voltage, a variable resistance, or pwm input. (see pic in OP) that there is a good chance that the pwm from the arduino is not directly passed through to control the led's but is used only to set the dimming level within the power supply itself. unfortunately I do not own a scope to test this.
You will have to use only one conductor, not both.
Yes, since the led's are in a series circuit I could not get it around both leads anyway 
I also tried this using my 6 amp vacuum with no response there either.
Not sure what the application is, and what/why you're trying to do these things, but PWM turns the LED driver fully on (100%) and fully off (0%) in a very fast rhythm with a fixed ratio. Because it happens so fast, your eyes see it as a dim light. Resistance dimming, with a pot connected to the dimmer, most likely just changes the output current of that LED driver. Voltage falls into place when you connect LEDs to it.
When there is nothing connected to the driver (LEDs are open circuit/broken), the driver outputs a higher voltage. Maybe you can detect that. Must be done optically, because there can't be a galvanic link between LED and DIM terminals of the LED driver.
Leo..
From what I understand you can simply put a CT (Current Transformer) in the primary side. This simply involves running one and only one wire from the primary feed through the CT. You can wrap several turns if you need more sensitive solution. This will let you know if it is on or off and approximately how much power is being output. Looking at the data sheet the easiest way to control the output try using a DAC (Digital to Analog Converter) you can purchase a chip or do it with PWM. This response is to help you get started in solving your problem, not solve it for you.
Good Luck & Have Fun!
Gil
From what I understand you can simply put a CT (Current Transformer) in the primary side.
The primary side is monitored at the outlet.
The application is grow lights and we want to know if any power supplies or LED's burn out which will adversely affect the plants.
Light sensors may also work for that.
wvmarle:
Light sensors may also work for that.
This was my first suggestion but it would require drilling a hole and my friend is adamant about not modifying the light in any way.
TBH I did not think it would be a problem to accommodate his wishes. But at this point if I can't find a better solution I'll have to tell him it's either a $20 sensor(cheapest I've found that I think might work Honeywell CSDA1BA ) or we'll have to drill a hole/cut a wire.
Since each light has 3 circuits $60 in sensors is not likely to fly but c'est la vie(such is life).
No need to modify the light for that - just place it in a strategic position away from the light, somewhere in the plant's canopy or so. Or tie it with a cable tie onto the light, that'll do as well.
