|
I`ve never bothered with stuff like p95, Just faff about until it stays below 70 on the highest load i can find, like playing stalker while watching a film on the on die graphics, that sort of thing, if it gets erratic, I take it back a notch. right at the moment I am messing around with a 4770k in a rog formula, now its chilling out at 4.2 on a very small stock. the graphics card is oc`d, but I am looking at getting water on it as well, I`d be putting together my water setup now, but I am un-ironically dealing with a flood.
|
#
?
Sep 6, 2014 12:51
|
|
|
|
| # ? May 25, 2024 13:39 |
|
|
Factory Factory posted:The reason why we set limits for 24/7 overclocking is to guarantee a minimum length of time that the chip will survive. 72 C at 1.38V Sandy/1.3V-1.35V (Intel has not said) Ivy+Haswell is what you need for the chip to run 24/7 for three years. Regarding the lifespan mentioned: do you mean if it's in a constant fixed voltage or if it's using an offset voltage for light loads?
|
|
#
?
Sep 6, 2014 22:14
|
|
|
Titor posted:Regarding the lifespan mentioned: do you mean if it's in a constant fixed voltage or if it's using an offset voltage for light loads? At a constant full load and with overclocking overriding Turbo Boost logic, there isn't a difference between those things. The chip just runs at the steady-state voltage, whether that's specified manually or via chip logic plus offset. E: When I say 24/7 for three years, I mean IntelBurnTest/LINPACK 24 hours a day for 1095 days.
|
|
#
?
Sep 6, 2014 22:25
|
|
|
I know its not linear, but its figures such as those where I get very little concern about hitting mid 80's with synthetics and 1.36 volts (for example). Of course I say this with my PC dead in the water but I'll chastise myself if it turns out to be the CPU
|
|
#
?
Sep 7, 2014 03:46
|
|
|
With a G10 bracket, does anyone think it could be plausible for me to make a single-piece VRAM heatsink? My idea is to take off the stock heatsink but leave the thermal pads on the vram. Then I'd take a single 4" x 3" copper sheet and drill four holes for the G10 screws to pass through, and then sandwich the sheet between the PCB and the water cooler. Would that allow the water cooler to essentially "suck up" all the heat from the VRAM and GPU at the same time? I'm wondering if the copper would radiate efficiently.
|
|
#
?
Sep 9, 2014 01:12
|
|
|
Zero VGS posted:With a G10 bracket, does anyone think it could be plausible for me to make a single-piece VRAM heatsink? My idea is to take off the stock heatsink but leave the thermal pads on the vram. Then I'd take a single 4" x 3" copper sheet and drill four holes for the G10 screws to pass through, and then sandwich the sheet between the PCB and the water cooler. Mount it to the card, not the g10 bracket. Also this is essential what the MSi 780ti does stock and it works great. It has a flat plate covering the ram and VRM's
|
|
#
?
Sep 9, 2014 01:38
|
|
|
Don Lapre posted:Mount it to the card, not the g10 bracket. Also this is essential what the MSi 780ti does stock and it works great. It has a flat plate covering the ram and VRM's Sorry I probably didn't type clearly, that's what I was trying to say. I'm describing just a flat piece of copper with four holes, so when I pass the G10 screws through it, the G10 is clamping the copper plate against both the GPU and the thermal tape of all the heatsinks.
|
|
#
?
Sep 9, 2014 01:48
|
|
|
Zero VGS posted:Sorry I probably didn't type clearly, that's what I was trying to say. Why would you put a copper plate between the gpu and the g10?
|
|
#
?
Sep 9, 2014 02:31
|
|
|
Don Lapre posted:Why would you put a copper plate between the gpu and the g10? Cooper plate is lying flat directly on the GPU and the VRAM thermal tape, then the water cooler (clamped by the G10 comes down on onto the copper plate so it's sandwiched. The plate conducts the heat from the vram and the GPU to the AIO cooler plate directly above.
|
|
#
?
Sep 9, 2014 03:00
|
|
|
What videocard are you cooling? For all of the recent Nvidia cards (starting with the 600 series) you DO NOT need to worry about cooling the ram even a little bit. Just the passive cooling from case airflow passing over the bare ram packages is more than enough. I can't say for sure about AMDs cards, but I'd guess they're probably similar.
|
|
#
?
Sep 9, 2014 03:07
|
|
|
I was actually planning on it and bought some cheap heat sink core off ebay but I didn't end up using it on the ram. I used pre made copper heatsinks for the vrms which was my real concern. There is quite a bit of space under the g10 though so you're pretty open to do a lot craig588 posted:What videocard are you cooling? For all of the recent Nvidia cards (starting with the 600 series) you DO NOT need to worry about cooling the ram even a little bit. Just the passive cooling from case airflow passing over the bare ram packages is more than enough. I can't say for sure about AMDs cards, but I'd guess they're probably similar. I'd agree with the ram but the vrms may be a problem. The water block on the GPU is able to basically cool the rest of the card... until it can't. The top set of pictures is under maximum game loads, clearly showing the whole card is cooler. However the bottom set of pictures is running synthetics and it produces heat faster than the small fan remove it compared to the factory cooler. The concern to me is that this tipping point of sorts is pretty unknown depending on the individual card layout and design. This is why I put heatsinks on the vrms with the g10 setup, even if I'm likely okay I can't really know since my 780ti doesn't have temperature readings on the vrms. 
|
|
#
?
Sep 9, 2014 03:26
|
|
|
Yeah, you do want to heatsink the VRMs. I ended up building a heatsink out of a plate and soldering fins to it similar to how the old Thermalright coolers were constructed. I tapped some holes in the plate and reused 3 of the factory mounting holes, springs, and screws. Even that got pretty hot (I think my cheapo IR thermometer was reading in the mid 60s on the top of the heatsink) so I added a fan to it and it's very cool now. I'm still using a H100 on my 680 which is way overkill. I'm able to run the radiator passively and it only gets into the mid 50s even with synthetic loads, games are usually mid to high 40s. That's with extensive power limit modifications in the bios and a 1.21 voltage limit too. The H100 is a waste of money for a videocard, I only used it because I had it laying around.
|
|
#
?
Sep 9, 2014 04:55
|
|
|
So a few years back when I upgraded I went with a 2600K on an MSI P67A-GD65 and now I'm getting around to overclocking it to wrestle some extra life out of the dumb thing, but I'm not sure what I'm seeing and/or how to proceed and could use some insights: How do I set an upper ratio limit in the EFI/BIOS that leaves Speedstep/Power Management intact? I'm monitoring via MSI's Control Center application: -Without any OC the chip will idle at x16 and ramp up to x34 as necessary. Great! -If I overclock via Control Center, I can set an x42 Ratio without any additional voltage changes and things are stable. Now it will idle from x16-x42. Great! -If I change the CPU Ratio to x42 in the EFI/BIOS, then it appears to completely disable the power management, leaving the thing running only at x42, even when everything is idle. Am I missing something here?
|
|
#
?
Sep 11, 2014 20:00
|
|
|
craig588 posted:Yeah, you do want to heatsink the VRMs. I ended up building a heatsink out of a plate and soldering fins to it similar to how the old Thermalright coolers were constructed. I tapped some holes in the plate and reused 3 of the factory mounting holes, springs, and screws. Even that got pretty hot (I think my cheapo IR thermometer was reading in the mid 60s on the top of the heatsink) so I added a fan to it and it's very cool now. I'm still using a H100 on my 680 which is way overkill. I'm able to run the radiator passively and it only gets into the mid 50s even with synthetic loads, games are usually mid to high 40s. That's with extensive power limit modifications in the bios and a 1.21 voltage limit too. The H100 is a waste of money for a videocard, I only used it because I had it laying around. oops I misread "ram" as vrms
|
|
#
?
Sep 11, 2014 21:02
|
|
|
Cuntpunch posted:So a few years back when I upgraded I went with a 2600K on an MSI P67A-GD65 and now I'm getting around to overclocking it to wrestle some extra life out of the dumb thing, but I'm not sure what I'm seeing and/or how to proceed and could use some insights: I can't answer your question because I have the same problem, I only OC if I'm keeping throttling and power saving features and I'm only able to do this using the motherboard software, loving with multipliers in bios disable everything throttling related so it's either lovely software or FSB oc.
|
|
#
?
Sep 11, 2014 22:46
|
|
|
crossposting from the build a PC thread:WhiskeyJuvenile posted:PCPartPicker part list / Price breakdown by merchant I'm looking to get these parts. Will I see a significant benefit from switching from the Phanteks to a full water setup? Thinking of throwing a bunch of EK crap at it: switching the 770 to a reference board and putting a full coverage block on it, a supremacy block on the CPU, and a 280 radiator at the top of the case (case literature references dual 120 radiator, but says it can have two 140 fans at the top, so I figure a 280 radiator would fit). Throw in a reservoir and pump and I'm done. Yes, $500 is a bit more than $80. Am I going to see significant enough gains, or is this all just basically pimping my ride?
|
|
#
?
Sep 12, 2014 22:36
|
|
|
The only gains you're going to see are in terms of noise. A 770 isn't especially limited by heat, especially not one with an open air cooler. You might get .1 to .2 extra Ghz out of the CPU, but it's in the extreme range of diminishing returns. An extra 400$ would be better spent switching to a 780ti or one of the new Haswell-Es, depending on what you mostly use your PC for.
craig588 fucked around with this message at 22:52 on Sep 12, 2014 |
|
#
?
Sep 12, 2014 22:50
|
|
|
No, once you do the initial overclock to around 4.5ghz you will not notice much beyond that, dropping hundred of bucks to get that last few hundred mhz. A cheap water cooler may or may not be better than the air cooler. If you wanna spend tons of money for some reason, do this: Buy two used R9 290 or two GTX 770 with reference coolers, SLI them in that dinky case, nab a higher wattage PSU, and then limit the framerate to whatever your monitor runs at, I'm assuming 60 fps. Each card will only have to work itself around half as hard at half the temps/noise to produce the same framerate, and if you wanna run things at full blast you can always wear headphones or put the PC in another room or build a sound-isolating foam box that allows airflow through. In any event, my plan doubles your gaming performance for $500 (or gives you a fairly quiet system with potential to crank it later), your plan gets you maybe 5% performance increase and a very quiet PC. Figure out your priorities.
|
|
#
?
Sep 12, 2014 22:57
|
|
|
What should I be able to overclock a G3258 with a 212 Evo to? I know I'm getting extremely greedy not being happy with 4.4GHz out of a 3.2GHz chip, but I can't seem to get it past 4.4GHz reliably while running P95. The only change I've made besides the multiplier is the CPU voltage, which is now at 1.3V.
|
|
#
?
Sep 15, 2014 16:30
|
|
|
You're already within the should range at 1.3 volts. Are you hitting thermal limits? 1.3 is considered moderate, you could push 1.350 before people start getting uncomfortable. I know the text goes red and angry on that motherboard when you go above 1.300, I'm not 100% on that motherboard, but it's not like crazy or too much or something. Also you can increase vrin (to say, 1.900) to see if that's your wall. And vring (starting at something like 1.250). Also you can lower uncore frequency if its trying to match 1:1. If you have a vice I'd delid before I'd spend any money on a cooler though.
|
|
#
?
Sep 15, 2014 16:45
|
|
|
Awesome, I'll try that tonight. I wasn't familiar with "delidding" until I Googled it - that's probably a little further than I want to go with it, but with these chips selling for near $50 now, maybe I could give it shot. Even if I mess it up, it's not the end of the world.
|
|
#
?
Sep 15, 2014 17:01
|
|
|
Delidding isn't supposed to do much on Devil's Canyon, is it? DC and the Pentium AE are supposed to be as-good-as.
|
|
#
?
Sep 15, 2014 17:05
|
|
|
Factory Factory posted:Delidding isn't supposed to do much on Devil's Canyon, is it? DC and the Pentium AE are supposed to be as-good-as. I've read 10-20 degrees (a bit skeptical on the 20 side). I only searched briefly, but it appears the temperature gains are very close to before. What's interesting is that it seems people are able to get more speed for less voltage out of the refresh chips in general after delidding over the (delidded) haswell counterparts. Maybe suggests something else is going on besides "fixing the tim" which frankly looks almost identical pre refresh and DC. Not saying that it is identical, but maybe the real change was a more mature manufacturing process to produce consistent cpu's. Kind of feels that way since the early 4790k's sucked just as bad as 4770k's. I'm just talking out of my rear end though. Radbot posted:Awesome, I'll try that tonight. I wasn't familiar with "delidding" until I Googled it - that's probably a little further than I want to go with it, but with these chips selling for near $50 now, maybe I could give it shot. Even if I mess it up, it's not the end of the world. The idea of it made me grimace, until I saw a 10 second video of it. Considering what's going on it's actually pretty easy. The actual nerve wracking part for me turned out to be cleaning off the die itself rather than slamming the cpu with wood in a vice. But in retrospect I was probably over worrying about it, as long as you don't pet your cat with it I'm sure it's fine. If I get another microcenter bundle I'll definitely be delidding it just for the sake of data on it. 1gnoirents fucked around with this message at 17:26 on Sep 15, 2014 |
|
#
?
Sep 15, 2014 17:18
|
|
|
The change in the TIM was changing the compound (likely from a zinc oxide in silicone to an aluminum oxide), which is good for roughly a doubling of thermal transfer. DC vs. 4770K is a fairly consistent 10 C delta (10-16 in AnandTech's testing) isoclock/isovolt, vs. delidding giving a 10-20 C difference. Because there's still a glue gap, delidding would give a benefit in theory, but if it's 10-20 C on top of DC's 10+ C for what is otherwise the same CPU with a slightly different FIVR, "surprised" would barely cover how I would feel.
|
|
#
?
Sep 15, 2014 17:31
|
|
|
Factory Factory posted:The change in the TIM was changing the compound (likely from a zinc oxide in silicone to an aluminum oxide), which is good for roughly a doubling of thermal transfer. DC vs. 4770K is a fairly consistent 10 C delta (10-16 in AnandTech's testing) isoclock/isovolt, vs. delidding giving a 10-20 C difference. Because there's still a glue gap, delidding would give a benefit in theory, but if it's 10-20 C on top of DC's 10+ C for what is otherwise the same CPU with a slightly different FIVR, "surprised" would barely cover how I would feel. It's hard to find much consolidated information about it as opposed to pre refresh Haswell stuff but anytime I come across someone posting results it seems to be in the 10-15* range. I didn't have time to read this yet but the second graph suggests a 16 degree drop http://www.kitguru.net/gaming/uncategorized/anton-shilov/intels-devils-canyon-chips-ngptim-is-still-not-efficient-research/
|
|
#
?
Sep 15, 2014 17:41
|
|
|
I feel like throwing up my hands and giving up. 16 C with the Liquid Pro is nearly identical to the drop on a 4770K. Next time I buy thermal compound I'm just getting some Dow Corning TC 5022. This min-maxing gives me a headache.
|
|
#
?
Sep 15, 2014 17:48
|
|
|
It'd probably be pretty easy to use a few home depot nuts and bolts to fasten a cooler directly on the AE die, for $50 I wouldn't mind chancing it. Anything heavier than the stock air cooler could crack it though, I'd imagine.
|
|
#
?
Sep 15, 2014 18:13
|
|
|
The issue with the IHS is that it's too high off the die, not the TIM. Changing the TIM alone doesn't have much benefit, nor does removing the IHS:
|
|
#
?
Sep 15, 2014 18:24
|
|
|
Yes, that's why the 3770K and 4770K had temperature problems. But we're faced with the following facts: * The 4790K changed the TIM without fixing the gap. * Despite not fixing the gap, the 4790K has improved thermal performance over the 4770K comparable with the low end of results from delidding a 4770K. * Our best understanding of the 4790K is that it is the same silicon as a 4770K, not adjusted for improved heat characteristics. The only apparent differences are TIM type and FIVR adjustments for extreme overclocking. Intel basically brute-forced the gap with better TIM. What we're trying to figure out in a way that makes sense is what effect delidding a 4790K has over and above this. And the number's were getting from anecdotes are not helpful. On one hand, there's a link saying that Liquid Pro improves temps by 17 C on DC. On the other hand, there's a link saying that delidding a 4770K and using Liquid Pro improves temps by 16 C. Yet stock DC is already 8-16 C cooler than a 4770K by yet a third link. We don't have good enough data to say anything with weight behind it.
|
|
#
?
Sep 15, 2014 18:42
|
|
|
Yeah. I'm looking forward to making my own data. But, I think it's safe to say that L4 batch and on are at the very least more consistent and perform better per volt than pre-DC haswell. Perhaps they just figured out how to put out consistent silicon since there was such a ridiculous range of stability for pre-DC haswell chips. edit: I forgot to add, since lower voltage = lower temps... perhaps they didn't have to do anything else besides drop some food coloring in the TIM dispenser and replace the 7 with a 9 
1gnoirents fucked around with this message at 15:23 on Sep 16, 2014 |
|
#
?
Sep 15, 2014 19:52
|
|
|
Sh4 posted:I can't answer your question because I have the same problem, I only OC if I'm keeping throttling and power saving features and I'm only able to do this using the motherboard software, loving with multipliers in bios disable everything throttling related so it's either lovely software or FSB oc. Well this is disappointing, but I suppose 30 seconds to manually tweak it on every reboot isn't ~too~ bad for a free 800MHz overclock.
|
|
#
?
Sep 15, 2014 21:59
|
|
|
I've been trying to overclock a 3570K since yesterday, my first attempt at OCing a CPU since the before advent of 'K' chips, so I'm still coming to grips with how to go about it. I was using IBT to do initial tests, then I left AIDA64 to run after I went to bed. This was with a 45 multiplier, voltage offset of +0.02. It was still running fine in the morning (The Vcore maxed out a 1.32, and highest core temp was 75°C.), so I assumed this would mean that the overclock was pretty stable, as it was a few hours, but Planetside 2 crashed within literally 30 seconds, I relaunched to make sure it wasn't a one off, and to be sure, it crashed within 30 seconds again. I bumped up the offset to +0.025, and I was able to play for maybe 5-10mins before it once again crashed. I put the offset back to +0.02 and gave BF4 a crack, and all seemed to go well, but it too crashed in my 2nd round. I did notice that in PS2 with the offset at +0.02, my max Vcore was 1.27, and at +0.025 Vcore was 1.28, noticeably short of the 1.32 in AIDA64, and in BF4 with a +0.02 offset, the max Vcore was 1.304. I presume these shortfalls in Vcore are the causes of the instability. Is this Vdroop? I have CPU LLC set to 50%, should I leave put on auto, 0% or 100%? Should I keep raising the offset till the Vcore reaches 1.32 whilst playing Planetside 2/BF4? I'm not sure how much thermal overhead I have given my max CPU temp was 75°C in AIDA64, I'm afraid I could blow even further past that if I keep raising the voltage offset. Should I just settle for 4.4Ghz? Why is AIDA64 stable, and these presumably less demanding games crashing? The mobo is an Asrock Z77 Pro-4M, so a pretty low end Z77, so I figure that might be part of the equation. Any thoughts would be appreciated. Also, I actually don't play Planetside 2 really, I just was using it 'cos I heard is was hard on the CPU. Not playing Planetside 2 is actually a viable solution for me, but on the other hand, I play BF4 regularly. Kintamarama fucked around with this message at 08:39 on Sep 20, 2014 |
|
#
?
Sep 20, 2014 08:37
|
|
|
I dont know why it's reporting that way but I've never used an offset over manually setting a voltage (and just leaving c states on to drop voltage in idle) However I can tell you right now for validation purposes just push the vcore until you get synthetics to the low 80's, check stability in game. If you're simply on the edge it won't take a lot. I wouldn't go higher than 1.375 even though you can since this isn't 24/7 - and its unlikely you'd get there anyways before you hit an uncomfortable thermal limit.
|
|
#
?
Sep 20, 2014 10:19
|
|
|
Hey 1ngnoirents, this is the CPU I got during eBay's 15% off campaign that we briefly discussed in the parts-picker thread. Unfortunately this lower end board only has an offset option, I've looked up and down the bios pages and couldn't for the life of me find a way to manually set voltages. I guess I'll have to raise the offsets for now and see how it goes.
|
|
#
?
Sep 20, 2014 10:58
|
|
|
Kintamarama posted:I've been trying to overclock a 3570K since yesterday, my first attempt at OCing a CPU since the before advent of 'K' chips, so I'm still coming to grips with how to go about it. My guess would be the motherboard, I've had a lot of asrock and while they are good budget boards they are terrible for overclocking and keep overheating and lose oc capabilities over time
|
|
#
?
Sep 20, 2014 14:44
|
|
|
Kintamarama posted:I've been trying to overclock a 3570K since yesterday, my first attempt at OCing a CPU since the before advent of 'K' chips, so I'm still coming to grips with how to go about it. Planetside crashes all on its own too though
|
|
#
?
Sep 20, 2014 14:45
|
|
|
You're stressing the system more in PS2 since the graphics card is involved whereas it wouldn't be in AIDA64, along with the rapid voltage changes you'd see in games. If you're seeing too much vdroop to remain stable at full load you'd either need to compensate with LLC or a higher vcore offset to keep the CPU stable. I know gently caress-all about Asrock boards, but I assume that there's higher offset options and maybe a setting for LLC if you want to try that. If you're overclocking the GPU as well you should probably revert to stock while you're attempting to get the CPU stabilized.
|
|
#
?
Sep 20, 2014 19:13
|
|
|
Thanks for the input, you guys are awesome. I thought I'd add that the system is rock solid at 4.4ghz with the +.02 offset, my max Vcore in AIDA64, and PS2/BF4 was 1.304, so no shortfall. I've noticed in AIDA64's stress test, you can tick a box to stress the GPU too, so maybe I should add a tick to that box from now on. I'll keep tinkering, but I think that my wall might be 4.5ghz, and I feel as though my low end motherboard might be playing a roll in that, but I can live with 4.4Ghz, I'm not missing out on much if I can't get that last 0.1Ghz.
|
|
#
?
Sep 20, 2014 23:18
|
|
|
Kintamarama posted:Thanks for the input, you guys are awesome. I thought I'd add that the system is rock solid at 4.4ghz with the +.02 offset, my max Vcore in AIDA64, and PS2/BF4 was 1.304, so no shortfall. I've noticed in AIDA64's stress test, you can tick a box to stress the GPU too, so maybe I should add a tick to that box from now on. I'll keep tinkering, but I think that my wall might be 4.5ghz, and I feel as though my low end motherboard might be playing a roll in that, but I can live with 4.4Ghz, I'm not missing out on much if I can't get that last 0.1Ghz. You really arent missing out on much if you can't get another 100mhz more, and 1.30 is a good safe voltage especially for 3570k's. My general experience with voltages so far are - Once you hit a vcore wall (increased voltage does nothing) increase vrin (up to ~2.10 volts) - Cache voltage to 1.20 and uncore frequency manually set low, below boost levels, to start I generally don't do any more than that but those three numbers can take up a lot of time.
|
|
#
?
Sep 20, 2014 23:21
|
|
|
|
| # ? May 25, 2024 13:39 |
|
|
So, a few weeks ago I was tinkering with overclock settings but quickly found that my cooling setup wasn't adequate. I have a thing for really quiet PCs, and that's generally not the domain of the overclocker. I could put a bigger fan on the giant heat sink I had, but noise is public enemy #1 so liquid time! As a general piece of advice, don't buy XSPC stuff. I have 15m of ruined tubing and had a fun job cleaning out the blocks after the included stuff threw plasticizer all over the system, but that's slightly off point. The main thing I wanted to figure out was what was actually going on in my loop so I knew what to adjust. I have one 3x120mm rad for the CPU and a second 2x120mm for the GPU(s). However, as far as I could tell nobody makes a product that would let me just add remote temp sensors to the SMBus so I'd have to buy one of those awful bay mounted ones that I hate. Solution: Roll your own. Madhacker has an old guide on attaching a MAX1668 to your SMBus, but it required that your mobo already have an smbus header (mine doesn't) or you solder up some ram sticks (Not happening). However, this guide is quite old, and predates the now common PCI Express slots we have sitting unused in our motherboards. As it turns out, PCIe exposes smbus pins at the front of the connector so a quick breakout board would be easy to hack up. Yesterday my quick prototype PCB arrived, I soldered it up, and it works.  The next problem was that there aren't any existing sensors I could use, so I had to make some.  Sensors are just diode-coupled BC547 transistors. Cheap and easy to make.  I drilled out a g1/4 acrylic plug (The 5mm LED ones are perfect since the hole is already the right size, and it's pre-tapped. Just drill it the rest of the way out.  Add hot snot, done.  Right now I have two set up in easy to install places, right before and after the GPU waterblock. This lets me do interesting things like measure the thermal change across the waterblock. So what's the point? Ultimately I want to be able to make smart decisions about how to adjust the fans and pump. If we know the liquid going into a block is too hot, we can just crank up the fans on that one radiator. If we know the liquid is cool but the block is still getting too hot, we can increase the pump speed. These are decisions we can only make if the data is all consolidated into one place though, hence the need of temperature inputs. But we also need PWM fan headers, (more than number I have on my motherboard at least), so next is this:  This exposes 6 remote temp sensors, and 6 PWM fan headers. This is fine for my loop, but with a bit of tweaking I can support several more (36 of each I think?). But I don't know if anyone needs more than 6. If anyone's interested, I can upload eagle .brd files. I had the PCBs done at OSHPark... the dumb prototype 4 sensor version is about $5 for a set of 3, the new one was around $13. The 4 wire one should be easy to hand solder by anyone who's got a non-poo poo iron, it's just an SOIC package. The other uses a rather evil TQFN40 so you'll probably want an oven for that.
|
|
#
?
Sep 21, 2014 01:09
|
|