I hereby certify that I am not simply asking someone else to design a footprint for me.
When I create a footprint for a through hole device to be soldered on a PCB, how much clearance should I allow for the lead? For instance, the data sheet for one device shows the leads to be 0.84 +or - .05. What size should I create the mounting holes on the PCB? Should I use the same clearance for all devices, or should they be a percentage of the lead?
I think it would be easier to help answer your question if you provided at least the mechanical specifications for the through hole device. Sometimes these calculations are easy and sometimes you might need to strike a compromise between large hole size for easy insertion versus adequate pad copper (OD-ID) and adequate spacing between pins.
I remember that through hole DIPs routinely needed leads to be formed (bent somewhat) before insertion into a pcb.
BTW I strongly encourage you to consider surface mount if that is an option. It is not particularly difficult!
The particular device Iām using, a current sensing transformer, is available only in through hole. There are only two leads on it and they do not need to be bent. But this device currently has to be pressed into place with more force than I like and it doesnāt seem like Iām getting a good solder joint. So, I want to redo the board with larger holes but, if thereās a recommended spec for clearance, Iād rather use it rather than just guessing.
I do have a surface mount resistor and an LED on this board and Iām pretty thoroughly convinced that Iām not very good with surface mounting stuff.
The current sense transformer is not likely to have very tight pin pitch, and the leads are also not likely to be very small; maybe AWG20 = 0.8 mm or a bit bigger? The other consideration is that they might not have a round cross section, or they may be solder dipped so that the diameter is not precise.
If you are using a board that is standard thickness = 0.062" = 1.6 mm, and you have plated thru holes, I would think that the hole diameter could exceed the pin diameter by 12 mils = 0.3 mm all around. Try to have your annular ring be generous such as at least 20 mils = 0.5 mm but this is not absolutely required.
Space between pad edges should be 6 mils = 0.15 mm bare minimum or 10 mils = 0.25 mm is easier to work with. If voltages pin-pin are more than 20V you might need larger spacing. The required spacing versus voltage also depends on cleanliness and operating humidity.
So if you have 0.8 mm diameter leads, you would then have 1.1 mm diameter holes and 2.1 mm diameter pads, if you have adequate pin pitch for that. If the pin pitch squeezes your annular ring you could go to an oval pad.
These numbers I am giving you are off the top of my head based on experience. There may be some standards for some of this, and I will happily bow to anyone quoting IPC standards if those apply for example.
Just to add: I am 72 years old. I started with electronics when I was 13, and I have always had some āessential tremorā which I inherited from my dad. I can remember struggling for maybe an hour to get a simple but hand-awkward solder joint done when I was a kid. To a large degree I have learned to work with my situation. Andā¦as an adult I find that if I want to have an easier time soldering it helps a lot if I have a beer beforehand. Nobody would call me a heavy drinker.
Having said all that, if we assume we are not discussing very tight devices, I donāt see that working with 0805 chips and SOICs is overall more difficult than working with DIPs and 1/4W resistors. But I am not saying that they are the same. Assembling through hole on a clean board may be a bit easier than doing that with SMT. But if you ever need to rework or desolder, then SMT can be significantly easier.
Thank you so very much! Iāll do as you suggest. BTW, I was using 0402 resistors and LEDs. They are too small! At 83 mu eyes arenāt what they used to be! 
Bart
You used term āclearanceā what in PCB design world is the distance between two copper things at PCB belonging to different nets. How much clearance is needed depends mainly on voltage difference between these two nets but also on expected pollution degree, humidity and even height above sea level where device will be used.
I think using the term āclearanceā make BobZ thinking about distance between pins so he said all of these:
I suppose you simply wrongly used term clearance thinking only of difference between pad diameter and hole diameter and not about distance between pads. But Iām not sure. I canāt find a definitive conclusion in your statements whether you mean real clearance or not. May be you are really having a problem with clearance because your pads are very close to each other so to have enough copper for soldering them you have to use very small holes.
I think it would be worth to solve this doubt and only you can do it.
One thing I would add . . . I have the luxury of having a Pace vacuum solder removal iron, this is great as long as it can suck the solder from the through hole, where the pin is a close fit in the hole this is made more difficult.
So donāt only think about fitting the device but also if you have to remove it.
I guess there are guidelines for this from the IPC, but it is not a simple number or ratio. A tight fit gives more mechanical strength, while a loose fit makes it easier to remove parts when needed. Filling up a plated through hole is also much stronger then a single sided PCB with no hole plating. Sometimes also āpress fitā connectors are used. These are just pressed into the PCB, and not soldered at all, and these need a well controlled diameter.
If you use test pins like below, then their springiness will keep them in the hole for easy soldering.
Those other tinned test pins, with a round pin and a tapered squarish piece to be inserted in the PCB are difficult to use by hobbyists.
I asked Chat GPT and got this 
Yes!! Unless you can be likee me and never make misteeks.
I do have a few 0402s but I hate to work with those. If I need to solder an 0402 I have about a 50% chance of losing it. Most of the engineers whom I have worked with seem to share the preference for 0603 as a minimum size for passive components.
Actually my default chip size for resistors and capacitors is 0603. But the footprint I use for these is a minimum size 0805. The smaller chip on the larger footprint allows more room for the soldering tip.
No, no! Cannot rely on the facts!! Rumors and speculation are all I want!! 
I indeed find that a 1-hole test pin is difficult to use. (Although I have not purchased much of the commercial parts.) Industrial operations can use a machine to press-fit a pin tightly into a hole but I am discussing assembly on my own bench. What I like to do is use an axial resistor lead as a small jumper. I save wire scraps including axial resistor leads that are 2 cm or longer. I use a piece of (maybe AWG24 or AWG26) wire soldered into 2 pcb holes spaced 2.5 - 4 mm apart. This requires a bit more space than a 1-hole test pin. But it is cheap and it is easy to bend the wire so that it does not fall out of the board before and during soldering.
Quoting an AI is problematic. I suggest we would refrain from that in the future.
Yes!!!
The hole size in the PCB should be slightly larger than the lead diameter to allow for easy insertion and soldering.
Artificial: maybe. Intelligent: certainly not!
Even I know the hole needs to be larger, unless I resort to my trusty hammer. 
@JohnRob 's reply is far more useful. 
Trick is I was having a laugh 
I even told you it was AI, if you find that problematic god knows what your gonna do in the future when nobody tells you. Oddly enough while the OP was being told life stories, quoted varies specs and receiving far more useful replyās he clicked my comment as a āsolutionā and buggered off with a smile on his face. ffs
Hi Bad Mouse,
I deliberately didnāt use the āQuoteā as my complaint is only with āAIā, not you. My almost automatic reaction to any āAIā reply is 
. It is not artificial and it is not intelligent. Someone, somewhere, wrote a bloody programme that spews out this garbage.
Perhaps, however on the other hand, useful information was posted and at least a few benefitted from it.
Consider, a good question in a forum helps everyone regardless of the source.
