I think this mod really needs a bit of added thinking to assure safety and prevent things from burning up.

To calculate resistor for LED Resistor:
RL = (70V - VF(diode) - VF(LED)) / IF(LED)

Assumptions:
Coil voltage really = 70V
Rectifier VF = 0.8V (typical for a lightly loaded 1N400x rectifier).
LED VF = 2.1V. Varies with LED color and specific LED manufacturer.
LED IF - assuming using typical value of 10mA

(70 - 0.8 - 2.1)/0.01 = 6710 ohms, closest standard value = 6.8K
Power dissipated by 6.8K resistor = (6800)(0.01)(0.01) = 0.68W (assuming 100% duty cycle).
A 1/2 watt resistor will burn up if left on too long. I try to never exceed 25% of rated power so you want a 5W resistor there. Wattage can be lower if you can *guarantee* the duty cycle is low (how long the LED is on / how long the LED is off). Even with a low duty cycle then you never want to go below a 100% load rating so you need a 1W resistor. But from a safety point of view - you should not assume anything other than a 100% duty cycle (e.g. shorted driver transistor).

If you use a 4.7K resistor (as shown in photo) -- LED IF will be 14mA. Might be OK but it depends on your specific LED. I try to never exceed 10mA for most LEDs.
Power dissipated by a 4.7K resistor = 0.94W! That puppy will be burning up if your solenoid voltage is really 70V and the LEDs were left energized.

So -- what is the typical period of time that the trolls are left hoisted up (hold voltage on coil)?
And can you guarantee to coil driving transistor will be shorted?

Edit -- forgot there are two LEDs in series so this does drop the power a tad:

RL = (70V - VF(diode) - (2 * VF(LED))) / IF(LED) = (70 - 0.8 - 2*2.1)/0.01 = 6500
Power of RL = 0.65W

Less than 5% change in power dissipation so not a lot of difference if driving one or two LED's at the solenoid voltage.

4.7K would be right if the voltage is 50V. But that resistor would be dissipating (4700)(0.01)(0.01) or 0.47 watts. That is still way too high for a 1/2 watt resistor. The 1/2W rating is assuming the resistor is going to operate at maximum temperature which you don't want to do. Never exceed more than 50% of rated power and I always shoot for about 10-20% max.

Note that the 50V is an approx unregulated voltage and is largely dependent on his local mains voltage levels. If his solenoid voltage measures significantly higher than a larger resistance may be required.

Yes - it will work. But what size resistor (wattage) did you use? And can you hold onto the resistor with your fingers the entire time the troll is held up?

Correction to "that's the only correction":
4.7K but rated at 1W *minimum*. At 1W, it will still be running very hot if left energized for any amount of time.
Would be better off if you used a 2W resistor for lower temperature.

Instructions just call out generic "Red LED" but doesn't get into specifics on sizing. Based on drill bit size specified (7/64"). Way too small for popular T1-3/4 (5mm) LEDs. My guess, these would be size T1 LEDs at about 3mm diameter.

Per eyeball pair:
Qty 1 resistor, value 4.7K ohms at 2W. RMO2-4.7K
Qty 2 LED, Red, Size T1 = Avago part number HLMP1700
Qty 1 Blocking diode, part number 1N4004 (anything 1N4001 and up).

Working up an inexpensive kit for these but the resistors are backordered.
Also out of heat shrink. Looking at:
M23053/5-102-0 for leads, maybe size 103... will try both.
M23053/5-105-0 for over all
Wasn't planning on including the wire, though. I don't normally stock wire and heat shrink.

I don't like the idea of a 2W resistor being heat shrinked (cut off airflow). Looking at ideas for mounting near solenoid rather than inline splice.

2mm isn't a standard size. Who has 2mm?

... sigh.... obsolete according to other resellers.
Will check direct.

Or maybe use the 22mm version?

There's also 2mm tower style LEDs where only the tip sticks through the opening.

Current Limiting Diodes (CLD) never really got popular as you can do the same thing with cheaper components (simple FET and resistor) or just a resistor since minor variations are acceptable.

On the plus side - CLDs can give you a much more stable current regardless of differences in source (solenoid) voltages... not that anybody's eyes could notice the slight deviation in LED brightness from one machine to another.

But - these have too many disadvantages to them:
CLDs in a low input voltage & low current application work well but when you start increasing the input voltage & output current, they become less attractive. I haven't researched them for all variations but most CLDs that can survive this high solenoid voltage cannot put out enough current.
You would probably have to parallel multiple CLDs to get sufficient current for the LEDs in this application (I would shoot for 10mA).

CLD's are a also linear circuit which means you get lots of heat - just like the current limiting resistor method.

You still need the blocking diode (1N400x or similar).

And using a pair of CLDs such as the E562 per LED would cost about 50x more per LED with no real gain in functionality.

Ya know - Barakwins1 may be onto something with post #33.
Why go complicated with the large resistor and only turn on the LEDs when the head is up.
If the head is down and the LEDs are still on... who's gonna know?
Just use a low voltage with a much smaller resistor and leave the LED on full time.