Question Want to replace proprietary fans in a workstation with standard ones, they appear to run at a very high CFM which I can't find replacements for

Cyber_Akuma

Distinguished
Oct 5, 2002
451
12
18,785
I had asked about this system before, but now I realized some new possible issue with attempting to replace the fans in my system.

They are a Dell Precision T3610 and a T5810 system and they have proprietary fans, there are three fans in the front of the case all mounted in a metal cage like this:

View: https://i.imgur.com/F9J6hq5.jpg


I am planning to replace them with standard ones (adapters exist) as well as use splitters so that I will have PMW control over all the fans instead of just connecting the new fans to SATA power and having them running full blast 24/7.

I looked at the existing fans and saw that they were listed at a ridiculous 0.95A. I figured either they were just old or poor quality and since all the fans I was going to add together would not add up to that much I would be fine. But then I looked closer and saw that it also is listed as having an insane 93CFM.

Here is a picture of one of them: View: https://i.imgur.com/Pf4fkFf.jpg


Model numbers:
PVA092J12M-P
V67MK-A00

The second model number seems to bring up more results. However, I can't find any specs on them. One site claimed that they ran at an also ridiculous 6100RPM, but it's stats weren't very reliable (it claimed they are 90mm when they are 92mm. Another also claims they are 32mm deep instead of the standard 25mm, which appears to be true according to my measurements)

Now I am starting to worry that if I replace these fans, my cooling performance will actually go down, not up. The whole point I was doing this was because the system (especially around the CPU and PCIe slots area) runs pretty hot and I wanted to add additional fans as well as use the PMW signal from the existing ones to control them, but since the slots are proprietary I would need to convert them to standard fans first.... but if these are some special heavy duty fans that no standard fans can compete with then it will end up giving me WORSE cooling performance.

The system lists the three fans as "RAM Left Fan" "RAM Right Fan" and "PCIe Fan", it becomes obvious why when looking into the system: View: https://i.imgur.com/84fexnn.jpg


There is a shroud that covers both sides of the RAM that the first two fans are directly in front of. There are cut-outs on the inner side of the shrouds which I assume is so the CPU can get some of the airflow being funneled into the RAM slots too.

The third one just has some vents that look designed to make the air kinda sorta blow in the general direction o the PCIe slots area, no idea how effective that one would be.

Anyone have any information these? Are they really 6100RPM or have a CFM of 93? No other 92mm replacement fan I can find comes even CLOSE to that kind of output in either RPM or CFM. I know that RFM does not matter as much if the fan's design is more efficient so it can push more air at a lower RPM, but the best I have been able to find is 30-50CFM, which is considerably lower than the 93CFM listed on the fan itself.
 

Paperdoc

Polypheme
Ambassador
It is unlikely you will find "standard" computer case fans that perform as these Foxconn units do. I doubt the 6100 rpm spec, but the 12 VDC, 0.95A, 93CFM all sound right.

I don't know why you PRESUME these are made to the "standard" PWM fan design for computer cases. The simple fact that they have FOUR wires to the connector is NOT a guarantee that they are that design. Do you KNOW that the SPEEDS of the originals do change with workload, or do they run full speed all the time? Do you KNOW which wire in their cable is for which electrical signal from the power source?
 

Cyber_Akuma

Distinguished
Oct 5, 2002
451
12
18,785
The simple fact that they have FOUR wires to the connector is NOT a guarantee that they are that design.

They are just the same 4 wires standard fans use but with a proprietary connector and the wires swapped around, they even use the same wire colors. I have tried connecting them manually to standard fan sockets and the sockets to standard fans and they work. They literally just used a proprietary shorter connector and changed the pinout while not changing what any of the wires actually do.

The only exception to this is the CPU cooler (and HDD cooler, but that is optional and I have never seen a system with one) which has a FIVE pin port. Apparently you can STILL adapt that to standard 4 pins just fine, but that 5th pin is apparently somehow used to sense the fan and will throw an error every boot that cannot be disabled of a CPU fan error if you try connecting a standard fan to that. (Why they need a 5th pin when all the other 4 pin ports are ALSO able to detect if a fan is not connected I have no idea, most likely Dell being Dell)

Do you KNOW that the SPEEDS of the originals do change with workload, or do they run full speed all the time?

Yes, not only can I both hear and feel it, but there are options in the bios to adjust their speeds and the BIOS has a built in diagnostic program that can test every aspect of the system, including the CPU and all three case fans seperately, you can very clearly see, hear, and feel them spin up and push far more air during the test when it ramps them up to max.

Do you KNOW which wire in their cable is for which electrical signal from the power source?

Like I mentioned, they are the same as standard 4 pin fans, just with the wires switched around and a thinner connector. I was able to connect various different standard fans to the port just fine using an adapter, an adapter that just simply re-wires which pin is which and uses a standard connector on the other end, as well as manually wiring up the stock fans to a power source using jumper wires and getting them to work.

Yes, I KNOW that they are just using the same 4 wires stock fans use and I KNOW that they are PWM controlled.

... the issue is that they appear to be significantly far more heavy-duty than most standard fans in terms of how much air they push.
 

Paperdoc

Polypheme
Ambassador
I found a site FrozenCPU.com

https://www.frozencpu.com/cat/l1/g36/Fans.html

From the listings for 12 VDC PWM fans in 92mm size I found

Delta 92 x 92 x 25 mm, 103 CFM, 1.0 A, 4800 RPM, ball bearings
https://www.frozencpu.com/products/...C0912D-PWM_Bare_Lead_Wire.html?tl=g36c365s940

Cooljag Everflow 92 x 92 x 38 mm, 110 CFM, 1.2 A, 4800 RPM, ball bearings
https://www.frozencpu.com/products/..._38mm_PWM_Fan_R129238BUAF.html?tl=g36c365s939

Delta 92 x 92 x 38 mm, 160 CFM, 2.1 A, 6000 RPM, bearings not specified, requires adapter sold on that page
https://www.frozencpu.com/products/...Fan_-_16022_CFM_AFC0912DE.html?tl=g36c365s939

Evercool 92 x 92 x 38 mm, 117 CFM, 1.2 A, 5000 RPM, bearings not specified
https://www.frozencpu.com/products/...Speed_PWM_Fan_EC9238H12BP.html?tl=g36c365s939

Either of those first two might suit your needs.
 

Cyber_Akuma

Distinguished
Oct 5, 2002
451
12
18,785
Hmm, the problem with those is their power draw. A big reason I wanted to replace the fans was to use ones that draw less power so I could add additional fans to the header. Normally PC fan headers handle up to 1 amp (though who knows with this proprietary system) and considering that the stock fans are 0.95 amps, I am guessing it's likely a 1A header. The fans listed there go from 1 Amp to 2.5 Amps. Not only would there be no room to add additional fans, but they pull MORE current than the stock ones, one of the more than double.

At this rate I might have to build some custom cables that just simply extract the PWM signal from the headers while powering the additional fans off of SATA, issue is I have no idea where to get the connectors for the Dell fan headers. I know they are some type of connector standard, just have no idea what it's called, and it's obviously no the standard used in PC fans.
 

Paperdoc

Polypheme
Ambassador
OK, suggestions for those issues.

Power draw. You want to add more fans to the three already there (not clear whether you are keeping or replacing the old ones). I gather each of the old fans is plugged into its own mobo header now. Rather than custom-creating a new wiring scheme to accommodate added standard fans, try a standard fan HUB. This is one type of device to allow you to connect multiple fans to a single mobo standard header. What is special about a HUB is that, in addition to one cable to connect to a mobo host header and a number of output ports for fans to plug into, it has a third type of connection. It MUST be connected to a power output from the PSU (either 4-pin Molex of SATA power). The Hub gets ALL power for its fans from the PSU directly and draws none from the host header, so it does not overload the header. It can only work with a header that does supply the PWM control signal, and only with PWM type fans. That is because it replicates a PWM type header only. It supplies all its fans with power from the PSU in the form of a fixed +12 VDC line on Pin #2 (in a standard pin layout) and takes the PWM control signal from Pin #4 of the host header and simply connects that to all its fans. This does no overload the host header's PWM signal line. There is still a power limit, but higher. The max CURRENT the Hub can draw from a SATA power output of the PSU is 4.5 A on the 12 VDC lines. This is a limit of the contacts in the SATA connector, not a limit of the PSU output. So, for each Hub you install and connect to one mobo fan header, you can afford to load it to max 4.5 A for the fans on that one Hub. This does all you wanted with your plan to custom-fabricate connections.

Connectors. You appear to have figured out exactly which wires of the old fans do which functions, so you do know HOW to connect those fans' wires to a standard mobo 4-pin fan header. Consider doing a wholesale connector swap. When you buy additional 4-pin fans, also buy (maybe scavenge from junk at a used parts shop?) several standard 4-pin female fan connectors. For each old fan, cut off the connector a couple inches from the end of the wires and solder onto those wires the new connector with the right pinouts. Now, examine closely the connector on the end of the cable from one of the new HUBS to a host header. The key item here is that you do NOT need to connect from the Hub to Pin #2 of the mobo host header - that is the +12 VDC power supply line from the header that is NOT used. The standard pinout is: Pin 1 Ground, Pin 2 +12 VDC, Pin 3 fan speed signal, Pin 4 PWM signal. The two ridges down the side of the connector to fit around the tongue beside the header male pins are next to pins 1 and 3 - Pin 4 is outside their space. The Hub will send back to the host header the speed of only the ONE fan plugged into its marked output socket via Pin 3 of the header, and will ignore the speed of all other fans on that Hub. So, wire the connectors snipped off the old fans properly to the wires into the new Hub, and now you can plug those Hubs into the non-standard headers of your mobo to pick up the PWM signal and return a speed signal. Plus, you have already converted the old fans to be able to plug into the standard ports of the new Hubs you bought, right alongside the new fans.

HUBS may look like collections of cable "arms", like a small circuit board, or like a box with ports along the side. ALL will have those THREE types of connections including the one to a PSU power output. Examples

https://www.amazon.com/XMSJSIY-Adapter-Splitter-Desktop-Computer/dp/B093ST9W5L/ref=sr_1_69?crid=1UW8VL8MI0896&keywords=fan+splitter&qid=1679619806&sprefix=fan+splitter,aps,109&sr=8-69

Fits 4-pin Molex power from PSU
https://www.amazon.com/Electop-Chassis-Cooling-Molex-Controller/dp/B071KNT7FW/ref=sr_1_20?crid=3B6QATBQFJ0DV&keywords=fan+hub&nav_sdd=aps&qid=1679619944&refinements=p_n_feature_keywords_three_browse-bin:5462063011&rnid=5462059011&s=pc&sprefix=fan+hub&sr=1-20

Fits SATA power from PSU
https://www.amazon.com/Chassis-Fan-Header-Controller-Dedicated-Multiple/dp/B08Q367BXM/ref=sr_1_21?crid=3B6QATBQFJ0DV&keywords=fan+hub&nav_sdd=aps&qid=1679620041&refinements=p_n_feature_keywords_three_browse-bin:5462063011&rnid=5462059011&s=pc&sprefix=fan+hub&sr=1-21

https://www.amazon.com/DEEPCOOL-FH-10-Integrated-Occupying-Motherboard/dp/B077YHLDSP/ref=sr_1_3?crid=2G28Y5R4ZIRKT&keywords=Deepcool+fan+hub&qid=1679620280&sprefix=deepcool+fan+hub,aps,97&sr=8-3
 
Last edited:
  • Like
Reactions: Cyber_Akuma

Cyber_Akuma

Distinguished
Oct 5, 2002
451
12
18,785
Power draw. You want to add more fans to the three already there (not clear whether you are keeping or replacing the old ones). I gather each of the old fans is plugged into its own mobo header now.

Correct, yeah. Each of the three stock fans have their own header, and are three of the exact same fan in a row that all are connected by rubber plugs into a metal casing that screws into the front of the case:

F9J6hq5.jpg


Rather than custom-creating a new wiring scheme to accommodate added standard fans, try a standard fan HUB.

The problem with that is no fan hub would support the proprietary connector. If I wanted to connect the fan headers to any type of hub, splitter, controller, or anything I would have to use an adapter to convert the proprietary Dell port to a standard fan port. And once I do that, I can no longer connect the stock fans to that header anymore. I have not found any sort of adapter to go from standard back to the Dell connector.

On top of that, since the fans are much higher CFW than standard fans I can't replace them with a standard fan either, and all 3rd party fans that have the same CFW are also 0.75A to 2.5Amps (not to mention very expensive). The stock fans are 0.95 Amps and those ports normally max out at 1 Amp so using the stock fans would leave no room to add more than one fan to the same header.

Basically I can't convert the headers to standard because standard fans are much lower CFW, and I can't add more fans to the existing headers because the stock fans take up too much power.

I could convert the stock fans to use a standard connector themselves and then use adapters on all three headers, but the problem is most hubs and splitters seem to take power completely from the SATA port and ONLY use the header for PWM, so I am worried that the SATA port would then be overloaded if all the fans were powered off of it. (Also, I would not want all the stock fans to run off the same signal, the system actually controls the PWN of each fan separately).

I was thinking of using something like that first product you linked along with a DIY cable to just get the PWM signal to the additional fans, something like this:

Q94Sa57.jpg


Would this work or am I wiring that wrong? I am going to have to find what male and female connectors Dell uses (supposedly some kind of thin/slimline molex standard?) in order to make such a cable though.

(I was personally hoping to not have to rely on SATA power at all, but once I saw the fans were high-CFM I knew I would not be able to just replace all of them with standard fans that are lower amperage)
 

Paperdoc

Polypheme
Ambassador
You miss what I was trying to explain. IF you do BOTH the steps I talked about, you would swap connectors so that:
(a) ALL fans, new and old, have STANDARD 4-pin female fan connectors so they all will fit to outputs from any Hub.
(b) ALL Hubs have the non-standard Dell female connectors on the ends of their cables to mobo headers, so they DO connect there properly to get a PWM signal from each header separately, and to return to each header the speed of one of the Hub's fans.

In that arrangement, each Hub would draw from its PSU direct connection the max amp total of only the fans plugged into that hub, and would not draw any significant power from its host header. The limit on power to each hub , as I said, is max 4.5 A, and that is imposed by the CONTACTS in the SATA power connector IF you are using that power output from the PSU. The limit for the max power available from ALL SATA output connectors on one set of wires from the PSU is larger than that, but that limit is stated on the PSU itself. Of course, any other SATA-powered device on those same wires (not the same connector) would have to be taken into account.

What I don't know is: how many SATA power output connectors does your PSU provide? If only one, you DO have a dilemma, no matter how you re-wire. IF you also have 4-pin Molex power outputs from the PSU, those can provide the +12 VDC source you need for some of the Hubs. Further, the Molex connector design can provide even higher max Amps than a SATA can. A related option: IF you want to use only Hubs with SATA power inputs, but have too few SATA connectors and DO have an unused Molex power output, you CAN get an adapter to convert one 4-pin Molex output into two or more SATA outputs. Then EACH of the new SATA outputs is limited to 4.5 A max current, but total load on two or three of these from a single Molex is still OK.

We have not talked yet about possible limit of the PSU itself. Normally there IS a limit on max Watts or Amps it can provide on all its 12 VDC outputs combined. Some of that is already committed to mobo and HDD uses. So how much is available to ALL these fans you plan?

Something I noted in your diagram. In the group of three added fans you had one as 4-pin and the other two "3 or 4 pin". There is some misunderstanding of how the 3-pin system impacts speed control and ability to read speed signals. In BOTH fan designs the speed signal generated by a fan is returned to the host header on Pin #3. In a 4-pin PWM type of fan, Pin #4 is to deliver to the fan the PWM speed control signal that a 3-pin fan does not use. Further, in the 4-pin system the power supply from Pin #2 is always 12 VDC and speed control is accomplished within the fan by use of the PWM signal. To control the speed of an older 3-pin fan the power supply on Pin #2 MUST vary, from 12 VDC for full speed down to about 5 VDC for min speed without stalling. So, when you plug a 3-pin older fan into Pins 1-3 of a 4-pin header that is using the new 4-pin signal set, that fan gets full12 VDC power from Pin #2 all the time, and always runs full speed. That is why you cannot connect a 3-pin fan to any common HUB and expect to control the fan's speed.

Pin #4 has NOTHING to do with sending the SPEED signal back from the fan. That signal is on Pin #3. When you connect to a header more than one fan using a Splitter or Hub, the connection of that speed signal from Pin #3 of the male output connectors to the host header is made for only ONE output, and no others.
 
Last edited:
  • Like
Reactions: Cyber_Akuma

Cyber_Akuma

Distinguished
Oct 5, 2002
451
12
18,785
You miss what I was trying to explain. IF you do BOTH the steps I talked about, you would swap connectors so that:
(a) ALL fans, new and old, have STANDARD 4-pin female fan connectors so they all will fit to outputs from any Hub.
(b) ALL Hubs have the non-standard Dell female connectors on the ends of their cables to mobo headers, so they DO connect there properly to get a PWM signal from each header separately, and to return to each header the speed of one of the Hub's fans.

Yes get that, but using such a hub would power the fans completely off the SATA port and only use the fan header for the PWM signal and RPM reporting right? I only wanted to power the additional fans from a hub or splitter cable, not the stock ones too since their draw is high.

What I don't know is: how many SATA power output connectors does your PSU provide? If only one, you DO have a dilemma, no matter how you re-wire. IF you also have 4-pin Molex power outputs from the PSU, those can provide the +12 VDC source you need for some of the Hubs.

The PSU has no molex. It has four SATA connectors, at least two will be in use. Originally all four were going to end up in use but I am likely going to have one free now.

We have not talked yet about possible limit of the PSU itself. Normally there IS a limit on max Watts or Amps it can provide on all its 12 VDC outputs combined. Some of that is already committed to mobo and HDD uses. So how much is available to ALL these fans you plan?

I have no idea how the PSU splits the power across anything. It has two 10-pin CPU power connectors, one of which is in use, an 8-pin PCI-e power connector that is in use, and a 24 pin connector that goes to the motherboard and splits into four SATA connectors, one of which is a mini-SATA designed for slimline optical drives.

Something I noted in your diagram. In the group of three added fans you had one as 4-pin and the other two "3 or 4 pin". There is some misunderstanding of how the 3-pin system impacts speed control and ability to read speed signals. In BOTH fan designs the speed signal generated by a fan is returned to the host header on Pin #3. In a 4-pin PWM type of fan, Pin #4 is to deliver to the fan the PWM speed control signal that a 3-pin fan does not use. Further, in the 4-pin system the power supply from Pin #2 is always 12 VDC and speed control is accomplished within the fan by use of the PWM signal. To control the speed of an older 3-pin fan the power supply on Pin #2 MUST vary, from 12 VDC for full speed down to about 5 VDC for min speed without stalling. So, when you plug a 3-pin older fan into Pins 1-3 of a 4-pin header that is using the new 4-pin signal set, that fan gets full12 VDC power from Pin #2 all the time, and always runs full speed. That is why you cannot connect a 3-pin fan to any common HUB and expect to control the fan's speed.

Yeah, that "4 or 3 pin" thing was the manufacturer of that specific SATA cable claiming so. I am only going to be using 4-pin fans so hopefully that will not be a problem. With the cable I wanted to make only the blue wire will be connected to that SATA splitter anyway.... the issue though is finding the connectors to make said cable. It should be easy to find connectors for standard 4-pin fans, but I have no idea where to find the connectors or the specific name of them for the Dell fans/headers. I know they are SOME standard that Dell just re-purposed for their own and not completely custom, but I have no idea what it is or where I can buy some that don't require a purchase of hundreds or thousands from an industrial supplier. (also wire of the proper gauge and colors. I know the colors don't make a difference but I would like to have the proper colors so I don't lose track of which wire does what since while the Dell fans use the same wires and even the same colors, they are connected to different pins)

Pin #4 has NOTHING to do with sending the SPEED signal back from the fan. That signal is on Pin #3. When you connect to a header more than one fan using a Splitter or Hub, the connection of that speed signal from Pin #3 of the male output connectors to the host header is made for only ONE output, and no others.

Yeah, I figured that's how it works. That's why I was confused why that SATA splitter seemed to have the blue and yellow wire going to multiple connections instead of just the blue. But I will only be connecting the blue wire from my DIY cable to the "motherboard header connection" of that splitter so that should work as long as I only use 4 pin fans right?
 

Paperdoc

Polypheme
Ambassador
So, you have three older fans with non-standard connectors on them plugged into mobo headers that match them for pinouts, etc., and you want to leave that unchanged. Then you want to add more fans possibly (but maybe not - depends on the fans you can get) with "standard" 4-pin connections and pinouts on them as supplied. Since the three mobo fan headers actually control each existing fan potentially differently you MIGHT want the added fans to be controlled slightly differently by getting separate PWM signals from each of those headers. That would mean doing all of this in triplicate, maybe. OK.

Focus on ONE header. What you want is to tap into the PWM signal line from it to feed to the PWM signal input line of a new fan. Then you want to connect power from a single PSU SATA power output to this new fan, bearing in mind that only the Ground and +12 VDC lines are to be connected to this source. The PWM signal is coming from the tap into one header, and the speed signal of the new fan cannot be connected to anything - all mobo headers already are receiving speed signals.

Now, IF your intent is that ALL of the newly added fans are guided by the SAME PWM signal from just one header, then the diagram you showed above is quite correct (allowing for the all-4-pin-fan plan). The main "trick" then is tapping into that header's PWM signal line. Assuming you already KNOW which line that is, the simplest way would be to strip off a little bit of that wire's insulation. Go the the input connector for that Splitter you show (the one that would go to a mobo header normally), cut into the wire for Pin #4, bare that cut end, wrap it around the stripped spot of the old fans' lead, solder lightly and wrap tape around. Now you have one line connecting the PWM signal from one header to the outputs for all fans of the Splitter, and that does not overload the PWM signal line of the mobo header. Then the POWER for all the new fans is supplied from the single SATA power output connector that feeds the Splitter. As long as the total max CURRENT of the fans on that Splitter does not exceed 4.5 A, you are safe.

Among the four items I linked above, three use max currents of 1.0 to 1.2 A, so three such on one Splitter and SATA feed is OK. The fourth one has too high a max current and it ALSO is a 5-pin design requiring an adapter for use with any other header type and unknown pinout specs, so a good one to avoid!
 
Last edited:

Cyber_Akuma

Distinguished
Oct 5, 2002
451
12
18,785
ince the three mobo fan headers actually control each existing fan potentially differently you MIGHT want the added fans to be controlled slightly differently by getting separate PWM signals from each of those headers. That would mean doing all of this in triplicate, maybe. OK.

Huh? No, I am not trying the combine the PWM signals somehow.

Basically, the case only has three intake fans in the front, that's it. No top, bottom, rear, or side fans. It also has a single CPU fan.

The case fans are labelled "RAM Left Fan" "RAM Right Fan" and "PCIe Fan". It's pretty obvious what they do, but I suspect that the two RAM fans also supply air to the CPU since the vents that co over the RAM/Fans have a cut-out area that would blow towards the CPU.

Anyway, I am going to add a 80mm fan to the rear to help exhaust more hot air (I notice that the rear of the case tends to have hot air pool there) and was planning to connect that fan to the PWM signal from one of the RAM fans.

I also wanted to install either a single 120mm, a 120mm and two 60mm, or a single 180mm (depends on spacing of my GPU and PSU) on the side to blow towards the PCIe cards (which include the GPU and a M.2 to PCIe adapter) that I am planning to control by the "PCIe Fan" PMW signal. Though this fan will likely also blow towards the CPU a little too just by where it will be located.

The computer looks like this (not a picture of my specific system), but close. main difference is I will likely install the M.2 adapter in that PCIe slot above the GPU:
84fexnn.jpg


And with these are the measurements I took of the side panel, I can definitely put in a single 120mm fan, and MIGHT be able to add in two 60mm fans (not sure if I should bother with those though) below it like this:

iOdUlif.jpg


CYWPhzX.jpg


Depends on if those 60mm fans can clear the PSU (it might be a right fit).

Also depends on if the GPU I am planning to install will even fit, the height is what will be the issue. I MIGHT have to remove that black "handle" part from the side panel. If that's the case, then I might just be able to fit a single 180mm in there then (again, depends if it will clear the PSU).

Yes, of course I am going to cut holes in the side panel for the fans.

Focus on ONE header. What you want is to tap into the PWM signal line from it to feed to the PWM signal input line of a new fan. Then you want to connect power from a single PSU SATA power output to this new fan, bearing in mind that only the Ground and +12 VDC lines are to be connected to this source. The PWM signal is coming from the tap into one header, and the speed signal of the new fan cannot be connected to anything - all mobo headers already are receiving speed signals.

I guess it would also work out to connect all of the additional fans, rear and side, to a single header too. Not sure what would be a better idea since one of them is dedicated to cooling the PCIe slots which is what I want the side fans for, but the rear fan is just for general exhaust/CPU exhaust.

And yes, that's the idea. the additional fans will be powered by SATA which will only be supplying the 12V and ground pins, and the PWM signal from the headers, with the RPM signal not connected since I will just have it go by the RPM of it's stock fans.

Now, IF your intent is that ALL of the newly added fans are guided by the SAME PWM signal from just one header, then the diagram you showed above is quite correct (allowing for the all-4-pin-fan plan).

I was planning on building two of those cables for the two sets of fans I wanted to control, and using two sets of those SATA cables as well. Although it would still work out to just have one control aoo of them too, but I figured it would be better to have the side fans controlled by the PCIe fan and the rear fan controlled by one of the RAM/CPU fans. I suppose I could also just wire the rear fan to the CPU fan header instead of one of the RAM fan headers. Not sure what would be a better idea for the rear fan. (The CPU header is actually 5 pin, no idea what the 5th pin is for but the other 4 pins are again the standard fan wires so I can still get just a normal PWM signal from it)

Assuming you already KNOW which line that is, the simplest way would be to strip off a little bit of that wire's insulation. Go the the input connector for that Splitter you show (the one that would go to a mobo header normally), cut into the wire for Pin #4, bare that cut end, wrap it around the stripped spot of the old fans' lead, solder lightly and wrap tape around. Now you have one line connecting the PWM signal from one header to the outputs for all fans of the Splitter, and that does not overload the PWM signal line of the mobo header.

Yeah, I know I can just strip the existing wires to connect the PWM line, but if I can find the connectors that were used to basically make a miniature extension cable that also splits out the PWM line I would rather go that route as it's a lot less messy, and easy to reverse if this doesn't work out by just unplugging the cable I made and plugging the unmodified fans back into the header.

Then the POWER for all the new fans is supplied from the single SATA power output connector that feeds the Splitter. As long as the total max CURRENT of the fans on that Splitter does not exceed 4.5 A, you are safe.

I can't imagine powering just the fans off of SATA would come close to that. Even if I use a SATA splitter to power two sets of those SATA splitter cables off of one SATA port, the fans I am planning to use are all rated at something around 0.05A to 0.08A at most. All of them together would not even hit 0.5A.
 

Paperdoc

Polypheme
Ambassador
Oh, big difference from what I thought we started with. Originally I thought you wanted to find new fans (either as added or as replacements for old) that have the SAME high air flow performance, and hence need the high amps. But if you are keeping the three old fans as they are and connected as they are, and simply adding about three more common 4-pin fan designs that each consume max 0.1 A, you surely CAN power all of the added units from a single SATA output from the PSU. In fact, since you have some limits on how many SATA output connectors are free to use, you could get a SATA power Splitter like this

https://www.amazon.com/CRJ-15-Pin-Power-Y-Splitter-Sleeved/dp/B07BMFG6J5/ref=sr_1_8?crid=D9A8N8IDT1WN&keywords=sata+power+splitter&qid=1679759791&sprefix=SATA+power+splitter,aps,105&sr=8-8

and feed two Hubs from one SATA source. Then rig your two Hubs as you say, using different PWM signal sources for the two.

I see what you mean about creating some removable plug-in adapters to access the PWM signal from the existing mobo headers. Ideally for one of those you'd make a type of Splitter with one female connector to plug into the mobo non-standard header, and two output arms with at least one of them fitted with a male non-standard connector exactly like the mobo header so you can plug in your un-altered original fan. The you use the other output to access the PWM signal from one line. The problem here is where to get those male and female connectors that match the mobo's non-standard headers?

As an alternative, how about making the single wire spliced onto the PWM line of an existing old fan connector into an insulated socket for ONE wire, and a matching connector on the end of the wire to the Hub. Something like these connectors

https://www.amazon.com/REDGOOSE-Connectors-Waterproof-Disconnect-Assortment/dp/B0BRH6JQMX/ref=sr_1_5?content-id=amzn1.sym.ea945d40-8e84-42be-ad5c-249b9bca6a40:amzn1.sym.ea945d40-8e84-42be-ad5c-249b9bca6a40&crid=24HB5BZ5T8JIK&keywords=Terminals&pd_rd_r=9bdeff44-380b-4eb9-b0d8-f6eca55bdfc7&pd_rd_w=bpXwW&pd_rd_wg=eefDH&pf_rd_p=ea945d40-8e84-42be-ad5c-249b9bca6a40&pf_rd_r=5T513G2QTA205VRCDPAJ&pid=zEC7aMN&qid=1679760315&s=industrial&sprefix=wire+,aps,152&sr=1-5

NOT necessarily these ones, but similar anyway. I'd suggest ones where the part connected to the fan cable as the source of PWM signal be one that has its contacts surrounded by a shroud for insulation, and the one on the cable feeding to the Hub has a protruding bare contact that inserts into the other. Both should have insulating sleeves so no wires are exposed when they are connected. The mod to each old fan's wires would be permanent, but they are simply not used if you decide not to connect according to the plan. If you choose, connecting the half of this to the wires from the Hub could be done similarly. That is, do NOT cut off the 4-hole female connector on the end of the Hub's feed cable. Just bare a bit of the correct wire from Pin #4 , wrap with a short piece of new wire, solder and tape exactly as you do for the connection to the old fan's wire. This gets you a wire to crimp on the mating connector with the protruding end. If unused, just tape over that.
 

Cyber_Akuma

Distinguished
Oct 5, 2002
451
12
18,785
Oh, big difference from what I thought we started with. Originally I thought you wanted to find new fans (either as added or as replacements for old) that have the SAME high air flow performance, and hence need the high amps.

Originally the idea WAS to replace them with new fans. At first I was going to replace them with standard fans, until I noticed the high CFM. Then I made this topic asking for replacement high-CFM fans, but when I saw that it could cost $75-100 for a set and they would still take up as much amps anyway then I saw no point in replacing the stock fans anymore.

But if you are keeping the three old fans as they are and connected as they are, and simply adding about three more common 4-pin fan designs that each consume max 0.1 A, you surely CAN power all of the added units from a single SATA output from the PSU. In fact, since you have some limits on how many SATA output connectors are free to use, you could get a SATA power Splitter like this

https://www.amazon.com/CRJ-15-Pin-Power-Y-Splitter-Sleeved/dp/B07BMFG6J5/ref=sr_1_8?crid=D9A8N8IDT1WN&keywords=sata+power+splitter&qid=1679759791&sprefix=SATA+power+splitter,aps,105&sr=8-8

and feed two Hubs from one SATA source. Then rig your two Hubs as you say, using different PWM signal sources for the two.

Yeah, that's what I was planning. Normally I would never use a molded cable like that since they are a fire hazard, but these would be so low power draw I doubt it would make a difference. Also it's very VERY hard to find a proper non-molded cable, 99% of them are molded. Nobody makes quality stuff anymore.... or no online store seems to carry stuff that is of quality anymore. Back when I needed to connect two SSDs to a single SATA port it took forever to find a splitter that wasn't molded, and all the good ones seem to be 4-way splitters, nobody makes a good 2-way splitter.

As an alternative, how about making the single wire spliced onto the PWM line of an existing old fan connector into an insulated socket for ONE wire, and a matching connector on the end of the wire to the Hub. Something like these connectors

I don't understand what you mean. I would still need both a male and female version of the connector. Even if I harvested the female connector off the fans, I would still need a male connector from somewhere. I am trying not to modify the fans themselves if I can avoid it so it would be easy to undo if this does not work out as well as replace them if they ever die. The less I have to jury-rig components in place of the original hardware, especially if it's not easy to reverse, the better.
 

Paperdoc

Polypheme
Ambassador
Maybe a better explanation of my idea. For each Hub you are connecting to a mobo header for the purpose of getting that header's PWM signal you need some way of tapping into that signal. The original suggestion was simply bare a short length of the correct (identified by you) wire near the end of the cable from your OLD fan at that header. Go to the cable from the new Hub that is intended to plug into a mobo standard 4-pin fan header and snip off the connector. Bare the end of the wire for that connector's Pin #4, wrap it around the bared piece of the old fan's wire, solder and tape. Now you have the PWM signal from that mobo header fed to both the old fan and the new Hub. BUT these are permanent mods hard to undo.

Alternative basically is to do this with a small single-wire connector pair at the mobo header end of the old fan's wires so that you can choose at any time to use this connection or not, without really changing anything about the feed cable to the old fan. For that purpose I suggested arranging the connector pair so that the SOURCE side of it - at the old fan's wire - is the half that is well protected by an insulating sleeve. Further, in following the concept of making a connection offering easy choices, I suggested that at the HUB cable's end you do NOT snip off its connector. Use that same technique of stripping off a small length of the Pin #4 wire and wrap-splicing a new wire onto it. Then mount the other half of the single-wire connector pair on the other end of that new wire. Then you can plug that into the source socket part at the mobo header if you wish for the new operating plan, or NOT do that and return easily to the pre-mod conditions. These DO make small mods to the wires of both your old fan and the Hub involved, but they are modest changes that do not impose a permanent change to your system.
 

Cyber_Akuma

Distinguished
Oct 5, 2002
451
12
18,785
Yeah, I wasn't looking to snip wires, especially since I might have to split some of the fans. I would rather not hard-wire the fan to anything. I want it to all use connectors.

I have to admit, I am COMPLETELY lost on that second paragraph. What's wrong with just making a male-to-female extension cable with two PWM wires?
 

Paperdoc

Polypheme
Ambassador
That extension cable you suggest will need to have one female connector to match your mobo header for its input end, and another of the male type to plug your existing fan into, then a third standard 4-pin male output connector to feed the Hub. Where can you get those non-standard male / female connectors for TWO such custom adapters?
 

Cyber_Akuma

Distinguished
Oct 5, 2002
451
12
18,785
That extension cable you suggest will need to have one female connector to match your mobo header for its input end, and another of the male type to plug your existing fan into, then a third standard 4-pin male output connector to feed the Hub. Where can you get those non-standard male / female connectors for TWO such custom adapters?

Apparently the connectors themselves ARE a standard, just not a standard normally used for PC fans:

I found the female versions at least:
https://www.molex.com/molex/products/part-detail/crimp_housings/0050579404

The crimp connectors for them:
https://www.molex.com/molex/products/part-detail/crimp_terminals/0016021114

And the crimp connectors for the male connectors:
https://www.molex.com/molex/search/partSearch?pQuery=&sType=s&query=70021

But I am still trying to figure out what is the male equivalent of those first two. The site lists dozens of options for compatible mating connectors, not sure which one I am supposed to go with. Also will need to see how to crimp those connectors as I have never done that before without using the official tool that costs like $300-500.
 

Paperdoc

Polypheme
Ambassador
You may have some difficulty finding a place to sell you only a few of those items. I expect they are sold mostly by large wholesalers who supply to manufacturers. Anyway, the the FEMALE connectors I see are sold as empty housings and separate crimp terminals. Some males are sold as complete connectors with exposed pins on the back for connections. That's because many of them are used as mobo headers soldered into a circuit board. I'm guessing you can find the empty bodies and separate male pins to construct male connectors for wire ends. The males that mate are listed on that first page of the female bodies under the tab "Mates With / Use With". For both males and females you need to match the size ordered to the exact dimensions of what you have. Then there are options for things like type of metal coating on the contacts, etc.

For crimping the female metal terminals onto a bared wire end you certainly CAN do that by hand.If you look closely there are small metal "wings" open at the back end. You place the wire between them then bend those wings closed and simply squeeze them tight with needle nose pliers. Check for secure holding that won't let loose. SOME people would put a tiny blob of solder on that, not necessary. Most often when you see them up close you find that this little metal terminal has a couple of little metal "barbs" along the side that point back to where the wire is. So after your wire is crimped into that, you slide the whole terminal into a hole in the BACK edge of the connector body until it stops at the front. Those little barbs stick into the interior of the hole so you cannot pull it back out, and it won't push out when you plug the connector in. The expensive special tool is for fast assembly of LOTS of connectors.
 

Cyber_Akuma

Distinguished
Oct 5, 2002
451
12
18,785
Those links I posted have a "Buy Now" link that lists several popular electronics suppliers like Mouser, Digi-Key, Arrow, Newark, etc nearly all of which state "No Minimum Order"

And yeah, I saw both male connectors and male headers, several kinds of each, need to figure out which is the best one. I saw the "Mates with" section too as I mentioned, but it lists like a dozen different kinds so it's a little confusing which one I should select.
 

Paperdoc

Polypheme
Ambassador
Yeah, I guess you just have to wade through the options and make choices. I believe you are just making your own custom version of SPLITTERS so that both an existing older printer and a new connection to a Hub can be made from each mobo header. For that purpose ALL FOUR wires from the header need to go to the one non-standard output for the old printer (identical pinout positions), but for the other output to the Hub only the Ground and PWM lines are connected. On the standard 4-pin male for that, those are Pin 1 Ground) and Pin 4 (PWM).