Ilegator

A lot of people are suggesting that once the 5000 series starts shipping, prices for the 3000 will fall, and it's a reasonable prediction, but I wouldn't be so sure. Looks at what's going on with RTX 2080 and 2080Ti pricing right now. Used cards sell for as much if not more than the RTX 3080 MSRP. The new Ryzen's are not cheap, and looks like there will no 5700x, so personally, I don't think we will see any fire sales on the 3700x parts anytime soon. If you see a good deal now, I'd say go for it!

I would if I were you, that's a great deal!

yep buy it. the prices are through the roof sadly for the new ryzen series... 

Well the 6 core 12  thread 5600X will cost 300€/£/$
 
To OP - I'd buy that.

Unless you open it up and/or poke around inside it while it's plugged into the wall then absolutely no. Even the charged electrolytic capacitors aren't very dangerous when the device isn't plugged in.

Contrary to what anyone will tell you, the amps are not the killer. Technically the amps are what kills you, since amperage is the measurment of the actual flow of electrons, but just because something can deliver a shitton of amps does not mean it's dangerous to a person. Skin and flesh has a resistance to the flow of electricity, the epidermis (outermost skin layer) is especially good at this and can provide several million ohms of resistance. Ohms law (V/I=R) says that at a given voltage, the resistance determines the amperage. Thus, touching a car battery fully charged to 12 volts with a typical skin resistance of 100k ohms, you would only have 120µA flowing through your body, more than 100 times smaller than the 20 or so milliamps needed directly across the heart to cause ventricular fibrillation. Keep in mind that for an electric current passing through your body to be seriously harmful or lethal, it needs to pass vital organs or your nervous system/brain. For this reason, arm to arm contacts are most dangerous, that's where you touch one polarity with each hand so the current passes through your chest.

Excluding extraordinary circumstances, no voltage below about 30 volt can possibly be dangerous to a person and you need to get upwards of 60-100 volts before there's ANY kind of danger.
The only time where voltages lower than about 100 volts can be dangerous or lethal is when both of the points of contact bypass the skin, as flesh and tissue have much much lower resistances.

When contacting the skin, voltages over about 50 volts can cause slight discomfort or pain, voltages over 100 volts can be very painful and are potentially dangerous and voltages of about 200 volts or above are potentially lethal. Any voltage over 100 volts should be treated with respect, as if it could kill you, to prevent any accidents. Voltages over about 200-300 volts should be treated with extreme caution as these could potentially be instantly lethal, depending on the way you contact them.

But back to the subject at hand, a charged electrolytic capacitor, even those big ones inside a PC power supply used for smoothing rectified mains voltages simply does not contain enough energy to kill a person.
You'll get a hell of a shock if you bridge the connection with your finger, you'll get a burn mark and you'll be feeling it for hours, but you won't be dead. This is because the current drops off too fast as the capacitor discharges through you.
An analogy would be that flicking your finger through a candle flame won't burn you because the contact time isn't enough, there's not enough exposure time to heat up the skin to give you any burns.

Heck yeah it could. If you threw it hard at someone's head, it'd probably knock them out cold and cause a severe haemorrhage. 

Do that^ and also check the back of the motherboard for damage. Mine has damage on the back so as a result, its stuck in with 2 slots working, and those in single channel (luckily the RAM overclocks to  2200MHz).
The i7 3770 by far.

it will only cause problems if its touching the contacts of the caps = the body/can isnt part of the circuit.

4-pin fans get 12V all the time and use a signal to determine how long the motors turn on and off. 
3-pin fans get a variable voltage (between 5V and 12V) to control the fan speed. 
 
If your motherboard's BIOS is set up for PWM fan control, it will be sending that pulse to the 4th pin instead of lowering the voltage.  The 3-pin fans will basically get the full voltage all the time. 
So make sure the fan settings in your BIOS are set for voltage control rather than PWM. 

I thought I'd share a short tale today about fixing a dead graphics card!
 
A few months ago, I bought a box of graphics card for the sole purpose of getting an EVGA GTX 295 for a display model (what can I say? It's a neat looking, beefy dual-GPU card and I thought it would be neat to display).
 
Turns out that all four of the cards included in the box were dead. While I didn't really need any of them to work, it was kind of a bummer that I wouldn't get a conveniently usable GPU out of the deal...
 
UNTIL TODAY!
 
Yesterday I took another look at one of the dead GPUs, an XFX 8600GT 256MB, and I noticed that the a number of capacitors had blown. With the hope that the capacitors were the only issue, I scheduled a trip to a local electronic component store, Lee's Electronics in Vancouver, purchased a number of replacement capacitors and attempted to replace the bad ol' capacitors with some brand spankin' new ones.
 
For the people who might be contemplating doing something like this, it's important to note two things about replacing capacitors, and the person replacing them:
 
1) Capacitors can be replaced with another capacitor as long as the capacitance (measured in micro-Farads or uF) is an identical amount, and the rated voltage of the new capacitor is equal to OR greater than the old capacitor. While people say you CAN use a higher capacitance in some cases, it depends on what the capacitor is being used for exactly, so for someone (like me) who is not sure, using an identical value is the safest option.
2) I am pretty amateur/bad at soldering, and even I managed to remove and replace the bad capacitors (though perhaps not... perfectly. More on that shortly.)
 
While I initially struggled to get the first capacitor out of the spot it was soldered into, I eventually managed to get a technique down of heating up the solder that was holding the capacitor legs in place while holding and gently rocking the capacitor back and forth until the legs slid out of the holes they were pushed into. The first try must have taken me a whole 5-10 minutes just to take out the capacitor, not to mention another 5 minutes of trying to seat a new one in place. Eventually though, I figured out a solid methodology:
 
1) Heat up the solder attached to the legs from the backside of the graphics card
2) Grab the capacitor on the front side of the card and rock it back and forth, pulling gently on it to pull the legs through the holes on the PCB.
3) After pulling out the old capacitor, take the replacement one and trim the legs down to a more manageable size (I trimmed them down from 3-ish centimetres to about 0.5 centimetres)
4) If there's leftover solder in the holes where the old capacitor legs used to be, heat the solder up and push the legs of the new capacitor through (making sure the positive and negative ends of the capacitor are in the right spots). You may need to rock the capacitor back and forth as you push the legs through, heating the solder to make sure it doesn't cause the legs to stick.
5) Once you push the capacitors all the way through, apply flux to the metal legs sticking through the PCB. I used a flux pen that I also bought at Lee's Electronics.
6) Finally, apply some solder the capacitor legs to secure them to the graphics card PCB.
Done!
 
I have a bit of a confession to make. I mentioned earlier that I didn't quite perfectly perform the capacitor replacement, and that's because I accidentally burnt the tip of my finger when I brushed it against my very-hot soldering iron. That, uh, sucked.
 
But! Did it work? Was there a point to spending $5 in capacitors, burning the tip of my finger and possibly inhaling some maybe-not-good-for-me-fumes because I'm not the smartest cookie in terms of workbench placement?
 
The answer is Yes! The once-dead 8600GT has risen once again to deliver pixels anew, booting into the BIOS and Windows without breaking a sweat!
 
So, that's it! Hope you enjoyed the brief story/guide of a dumb-ass electronics amateur fixing a graphics card with blown capacitors! Hopefully this is useful to others who might be thinking about trying this on one of their dead graphics cards, or at the very least they find it mildly amusing!
 

Haswell is improved for single thread and power efficiency, and at identical clocks it is quite tough for the I7 3770. But once the latter runs with turbo up to 4.3ghz (at 100mhz fsb) the Haswell is the one looking old, well, on >=30% lower wattage. 

You appear to have better granularity than me for this one.  My statements were fairly rough.

I don’t know.  I haven’t done ram overclocking. I understand manual ram overclocking is a famous pita.  There’s supposed to be an unofficial app that helps. Called dram[something]. I’m not sure how much. It’s apparently sort of a questionnaire where you type in a bunch of data about your ram and it spits out what you should set your stuff to in various places.  There are several descriptions.  I’ve never used it myself.

The 3770 will beat the i5 in more cpu-demanding titles where threadcount matters, and if partial-overclock is used it does beat the Haswell in almost every way possible. Also, the Haswell-system doesn't need much power with the stated gpu, 400-450W should be fine if the psu is compatible with c1/c2. Cooling doesn't need to be great to get this system running.

So you can bench it all first hand...
I'd go for the rtx2060 Super, it usually beats the 1070/1070ti in newer titles and it will be tough (albeit not impossible) to find a 1070ti or 1080 at a reasonable price.

That CPU shouldn't bottleneck any of the cards you have listed, I would go either 1070 or 1080, really just depends on the games you play and at what resolution/refresh rate

I would say the i5 myself as it’s got the least future, though it’s kind of a tossup between the 3770k and the i5 at this point both of them are limited life.
 
As for GPUs the current conventional wisdom is if you already have something that works OK, sit on it and wait because new GPUs are being teased to be announced soonish.

I don't think it makes sense to compare a R7 2700 with such old processors from Intel.

The 2700 is better in single-threaded performance too, unless you compare it at stock speeds to a heavily overclocked Haswell/Ivy Bridge.
It's the obvious choice here imo, unless you can get your hands on a 3600, which would be even better.

I'd go with the Ryzen 2700 over an Ivy Bridge i7 or Haswell i5. I realize this is the Ryzen 3600 review, but it has the 2700X on the charts, which is going to be identical (if not within 1fps) of the non-X: https://www.guru3d.com/articles_pages/amd_ryzen_5_3600_review,23.html
 
From here you can appoximate from the i5-6600K on the charts, which will be about your best case scenario with the Ivy Bridge i7 and Haswell i5.

would be interesting to see if it clocks well regardless of the pin used. I would anticipate, at worst, slight voltage problems, but it would be cool if it was business as usual in this CPU.

First, I tried if it worked without these two pins, but it just would not post. I tried inserting two pins from an old AMD CPU into the MB socket so it made contact with the CPU, but it woulnd't post neither. 
It took me a while to get all the necessary items to fix it: soldering station (with flux, tin, etc.), 0,3mm copper wire, an AM4 socket motherboard to test it...
I followed the same steps than this video: 
 
Here are two pictures of the fixed pins:

 
 

First, I tried if it worked without these two pins, but it just would not post. I tried inserting two pins from an old AMD CPU into the MB socket so it made contact with the CPU, but it woulnd't post neither. 
It took me a while to get all the necessary items to fix it: soldering station (with flux, tin, etc.), 0,3mm copper wire, an AM4 socket motherboard to test it...
I followed the same steps than this video: 
 
Here are two pictures of the fixed pins:

 
 

I enjoy reading this thread ? please continue ? 

Yeah this is an odd board.  I was learning more than anything over the last few minutes LOL.