I tried to put up a post last night can't find it now so if this is a duplicate my apologies.
I have the basic ryzen 7 with 16 gig ram version of the 5675, I ordered the liquid cooler from dell. First getting the fan heatsink off was difficult as most builders tend to do too much heatsink compound. I finally was able to get it off with some crc and careful nudging on the side of the cpu. I installed the radiator and fan at the back of the case and moved the fan that was there to the front of the case. I was looking for a usb power cable for the pump but this one apparently takes all power from the fan header. The cable on the pump is a three pin while the plug is a four pin. I took the cpu out of the socket gave it a spritz of crc then put it back locked the lever (yes I do know how you tell how it plugs in) put on the pump block checked to be sure I didn't unplug anything or jar the memory sticks and closed it all up...Nothing...no case lights no fans. push the green button on the psu it lights up and the lights on the keyboard light up so this is not power supply. I'm used to boards with beep codes does this not have that?
So we seem to have two or three possible scenarios here...1-CPU may not be seated properly, 2-Pump block is not working and thus internal safeties are preventing the machine from starting because it doesn't think the cpu has cooling, 3-something is monitoring pin 4 on the cpu fan plug and since the pump doesn't use the fourth pin again is triggering the safety situation for no cpu cooling. 3 seems a little improbable since this is supposed to be the correct cooling unit for this machine. I noticed when I pick up the cooler that it really feels light like there isn't any liquid inside it but i don't see any leaks on it. I wish there was some way to test the pump on this if I had a bench board I could probably test it that way but don't currently have that. New house will have shop with that for sure.
I appreciate any help I can get please.....
I guess mostly I'm looking for information on whether this board tests the cooling before boot and if there are supposed to be any post beeps. I've been building machines for a lot of years this isn't my first rodeo but I am sort of new to dell desktops have have a lot of the laptops but first desktop. Help? Am I missing something obvious?
Well, I guess I would put everything back the way it was originally (ie, no liquid cooler, etc.) and see if motherboard is still operational.
You released flea-power before you started, right?
I hope that CRC is the plastic-safe kind.
I've never heard of using CRC to remove a heat-sink. Do you think some thermal compound washed onto board or inside socket? I've heard some Thermal Compound is electrically conductive and will cause a short. I don't think the stuff I use is conductive, but I'm still really careful with it.
I haven't messed with AMD-CPUs is some time (since Athlon64-x2 days) ... are the new ones fragile or something? If this was Intel, I would say 99% sure it's NOT the CPU (they are pretty durable).
If all back to original config and its still dead ... might be time to give Dell a call.
Using solvent was a poor idea -- usually, it's simply necessary to run the system up to temperature, which heats up the thermal compound and allows you to gently slide off the heatsink.
That said, return the system to its default setting (it's ok to leave out the heatsink paste for this test). If the system won't power up, the mainboard is the first suspect. CPUs are generally robust and likely the CPU was not damaged - it's more likely you shorted out the mainboard by applying solvent (the heatsink paste is electrically conductive -- particularly when dissolved in a solvent, which will hasten its migration where it should not go).
Yes tonight I will reinstall the fan and see what happens, I would hope that dell tested this refurb cooler before it was sent out to a customer but who knows. I have been building AMD systems since the 90's AMD are robust, there were zero bent pins. The heat sink was pretty virulently stuck to the cpu I have used CRC for many many years and never once had it cause an issue I learned a long time ago to keep the stuff very thin too much produces the stuck issue it just has to be thin and cover the cpu a plastic credit card makes a fantastic spreader it's what I've been using for years.
I understand that you probably deal with people who don't know what they're doing...but I assure you that after building something on the order of a hundred machines and repairing many more that is not what you are dealing with here. I've been using crc for years never had an issue and if you don't know what "spritz" means it means very little. The machine was warm when I was trying to get the sink off and it was still quite stuck. I'm sorry, I don't buy the you shorted it out stuff...dude that comes close to insulting.
I still haven't been given an answer about moron proofing on this motherboard. Does it have safeties built in that prevent the machine from starting if the cpu fan is not running? What about beep codes? Or are diagnostic beeps entirely gone from pre-built machines these days?
I did go back and make sure I didn't unplug the power switch as that's all up there in that area, it was plugged in. Just to ask, is this pump setup entirely powered by the fan header? If there's supposed to be something that goes to a usb header (like my corsair unit) don't see that or a plug for it on the cooler.
The beep codes prevail IF the system can progress far enough into POST to activate them, and if there's enough functionality left on the mainboard to allow them to sound.
Sorry if this adds to the insult your feeling, but it's clear that if you've really built that many systems, something slipped here -- perhaps it was taking things too casually. Just because you'd never seen damage before from doing something dangerous, doesn't mean the foolhardiness of using a solvent near the CPU socket wouldn't cause damage.
After all, the space shuttle flew many times with the same O-ring damage that eventually destroyed the Challenger -- eventually, you run out of luck and must depend on prudent practice to carry the day.
Your best path forward is to replace the mainboard - which will almost certainly solve the problem -- unless you fried the CPU with the board by your engineering malpractice -- which is possible but unlikely.
1. Yes tonight I will reinstall the fan and see what happens,
2. I would hope that dell tested this refurb cooler before it was sent out to a customer but who knows.
3. I have been building AMD systems since the 90's AMD are robust, there were zero bent pins. The heat sink was pretty virulently stuck to the cpu I have used CRC for many many years and never once had it cause an issue
1. Yes, return machine to it's original working config
2. I would not have bought a "refurbished cooler". A "working pull" maybe, but I would have to be pretty desperate when you can get new Corsair 'Asetek clones all day brand new.
3. If you say so. Anyway, as long as the CRC was the "plastic safe kind" it was likely OK (as long as it was dry before the machine was powered up again). Anyway, my concern was mainly about where the old thermal-compound got washed-to or ended-up.
1. I understand that you probably deal with people who don't know what they're doing..
2. I don't buy the you shorted it out stuff...
3. dude that comes close to insulting.
4. I still haven't been given an answer about moron proofing on this motherboard. Does it have safeties built in that prevent the machine from starting if the cpu fan is not running? What about beep codes? Or are diagnostic beeps entirely gone from pre-built machines these days?
5. Just to ask, is this pump setup entirely powered by the fan header? If there's supposed to be something that goes to a usb header (like my corsair unit) don't see that or a plug for it on the cooler.
1. Exactly. We don't know your skill level (and you did not state it in original post).
2. Shorts are always a risk. As is conductive thermal-compound, as is static-electricity, as is flea-power, as is ... many other things.
3. Careful buddy. We both took the time to reply to try to help you (even though your request is beyond the technical scope of this forum).
4. It should still BIOS-Post ... even without ANY cooler attached. Definitely with original fanned-cooler attached.
5. We can't see your new cooler nor did you provide a link. Usually, yes, the pump is powered-by and monitored by one cable. The other cable is usually for the PWM fan on the radiator. All the Dell (Asetek clone) liquid-coolers I've ever seen only had the two cables.