Not an EMI thread - Terminal slug cursor is PSU resistance based, not interference based

Everything about latency. Tips, testing methods, mouse lag, display lag, game engine lag, network lag, whole input lag chain, VSYNC OFF vs VSYNC ON, and more! Input Lag Articles on Blur Busters.
User avatar
r0ach
Posts: 95
Joined: 10 Oct 2023, 14:45

Not an EMI thread - Terminal slug cursor is PSU resistance based, not interference based

Post by r0ach » 31 Jan 2024, 15:09

So I've used all kinds of power supplies from beefy EVGA 1300w G2's, to ATX 3.0 Channelwells, to Corsair platinum SF600/SF750 series, to Seasonics, Superflower Leadex, etc. I ended up with a Superflower Leadex III gold as a backup unit while my Corsair was being RMA'd and even though this PSU is supposed to be ranked very highly, cursor movement is pretty much garbage on it.

Why is it garbage? The first likely suspect is that the higher end Corsairs use 16 AWG cables for things like PCIE and CPU while this turd Superflower uses 18 AWG. You can literally feel the increased resistance while moving the mouse. It comes in the form of both decreased granularity for cursor movement as well as just more disconnected/floatier cursor movement in general. I opened up one of the Superflower cables and see that they also use tinned copper as well.

18 AWG cables have 60% more resistance than 16 AWG, then tinned copper is 15% as conductive as bare copper on top of that. So if one brand of PSU uses 16 AWG + bare copper and another brand uses 18 AWG + tinned copper, you're looking at clown world levels of increased electrical resistance and your PC will feel unusable. A normie engineer specs things like 18 AWG as 'usable' for a Linux email server while not causing a fire. It's not actually usable for a gaming PC.

Next subject is the power cord itself. I have 18, 16, and 14 AWG cables and there's tremendous difference between which one you plug in to a PSU (one that doesn't actually need higher than 18 AWG for load). You see the same phenomenon here where you can feel cursor granularity increase each time you go to a bigger size. You will probably also see voltages slightly change in the BIOS as well.

The power cable subject is NOT as cut and dry as internal cables, though. I would say you ALWAYS want 16 AWG PCIE and CPU cables and should never go near companies using 18 AWG ones or tinned copper for that matter. Send them hate mail. If your current PSU uses this, your mouse movement will likely always be trash. Doubly so if it's tinned copper on top of that.

For the actual power cord, tossing a 14 AWG cable onto a PSU that uses an 18 AWG one by default can actually feel detrimental. Cursor movement will be extra slippery and wild. I didn't really see this kind of behavior on a 1300 watter that uses one by default, so doing it on a smaller PSU might be throwing some voltages out of whack. I believe in-cable capacitors probably cause resistance problems that board capacitors don't, so those should likely be avoided.

Before the inevitable "what PSU should I buy" posts come. Channelwell platform is used for most of the new ATX 3.0's and seem to have huge problems with causing coil whine. I had one that had LOUD coil whine just sitting at idle which I've never seen on any PSU before. So even though Corsair is usually one of the better companies for thick cables, this will likely affect tons of Corsair PSUs + their RME line using a different company also has the same problem, as well as pricey Seasonic Vertex. You can always buy and return these brands if they do have it and eventually they'll get the message and fix it.

FSP Hydro PTM X Pro don't seem to have the coil whine problem and have 16 AWG cables (at least on the 1000w unit, not sure about 850w), so that's really the only new age potentially safe PSU I know of. If their cables are tinned copper you would still probably be screwed with gimped mouse movement, though. Someone will need to email all these race to the bottom trash companies and force them to give up the info of what their cables are made of.

Bobo
Posts: 83
Joined: 05 Jun 2018, 11:44

Re: Not an EMI thread - Terminal slug cursor is PSU resistance based, not interference based

Post by Bobo » 31 Jan 2024, 18:17

as an entry level audiophile, i can say that i know a bit about power cables and power supplies.
when you compare the expensive high end pc power supplies sold by the likes of corsair, msi etc, with stuff that audiophiles use especially DIY stuff, the high end pc ones look like children`s toys.

if you`re looking into DIY`ing a proper power cable, i`d say go for solid core wire, non plated of course and with proper shielding, keep in mind that solid core is harder to bend so you need a bit more space behind your pc case, also make sure you use decent non plated connectors for the cable, you can find a lot of cheaper ones on aliexpress, if you`re not looking to spend a ton on a cable.

User avatar
r0ach
Posts: 95
Joined: 10 Oct 2023, 14:45

Re: Not an EMI thread - Terminal slug cursor is PSU resistance based, not interference based

Post by r0ach » 01 Feb 2024, 04:29

Bobo wrote:
31 Jan 2024, 18:17
as an entry level audiophile, i can say that i know a bit about power cables and power supplies.
when you compare the expensive high end pc power supplies sold by the likes of corsair, msi etc, with stuff that audiophiles use especially DIY stuff, the high end pc ones look like children`s toys.

if you`re looking into DIY`ing a proper power cable, i`d say go for solid core wire, non plated of course and with proper shielding, keep in mind that solid core is harder to bend so you need a bit more space behind your pc case, also make sure you use decent non plated connectors for the cable, you can find a lot of cheaper ones on aliexpress, if you`re not looking to spend a ton on a cable.
I’ve always looked for solid core stuff in the past instead of stranded but there’s a few problems with that. The first is that virtually nothing is marked to differentiate the two. The second is that solid core is probably unusable for internal cables due to flexibility issues. It would be usable for HDMI/Displayport and external power cables, but then comes the debate again of whether companies will have different failure rates between the two from bending.

If you get a single crack in a long cable it might effectively reduce your 26 AWG cable to the equivalent of 32 AWG or something. I don’t really know how likely these breaks are to occur. Probably nobody does as it depends on things like if the cables are packed in coils or 90 degree right angles, etc. So it’s not hard to see why companies have gone to stranded for due to potentially more reliability.

Sosiska
Posts: 1
Joined: 02 Feb 2024, 02:10

Re: Not an EMI thread - Terminal slug cursor is PSU resistance based, not interference based

Post by Sosiska » 02 Feb 2024, 02:17

As someone who is just starting to delve into the audiophile world, I've come to learn a thing or two about power cables and power supplies. When stacking up the premium PC power supplies from well-known brands like Corsair and MSI against the equipment favored by audiophiles, particularly the custom-built (DIY) options, the former can seem somewhat toy-like in comparison.

Mr1991
Posts: 151
Joined: 24 Nov 2019, 10:10

Re: Not an EMI thread - Terminal slug cursor is PSU resistance based, not interference based

Post by Mr1991 » 02 Feb 2024, 09:45

Sosiska wrote:
02 Feb 2024, 02:17
As someone who is just starting to delve into the audiophile world, I've come to learn a thing or two about power cables and power supplies. When stacking up the premium PC power supplies from well-known brands like Corsair and MSI against the equipment favored by audiophiles, particularly the custom-built (DIY) options, the former can seem somewhat toy-like in comparison.
Is there actually a noticeable difference, or is it just better on paper?

Also this is something I’ve noticed on my setup, I bought a slightly more expensive psu, I believe it’s a “Corsair 1200i”, when I bought it it was considered top at the time, but it does seem to have a bit more of a watered down/smoother feel to mouse movement, it’s hard to say wether it’s better or worse, I’ll admit I’ve had nothing but positives with this setup in terms of overclocking/smoothness, but feel I performed better competitively on my older setup with a bog standard motherboard and psu

User avatar
r0ach
Posts: 95
Joined: 10 Oct 2023, 14:45

Re: Not an EMI thread - Terminal slug cursor is PSU resistance based, not interference based

Post by r0ach » 02 Feb 2024, 10:13

Mr1991 wrote:
02 Feb 2024, 09:45
Also this is something I’ve noticed on my setup, I bought a slightly more expensive psu, I believe it’s a “Corsair 1200i”, when I bought it it was considered top at the time, but it does seem to have a bit more of a watered down/smoother feel to mouse movement, it’s hard to say wether it’s better or worse, I’ll admit I’ve had nothing but positives with this setup in terms of overclocking/smoothness, but feel I performed better competitively on my older setup with a bog standard motherboard and psu
The expensive, digital controller PSU might cause more subjective bogged down cursor movement similar to having your BIOS set to auto-regulate power phases instead of setting them on extreme, full phases with no regulation. I considered buying one of these expensive PSUs before thinking they would provide some kind of benefit like a solid, consistent 12.00v line instead of 12.09 or 11.95 or whatever, but load regulation isn't even an improvement on a normal PSU:

Image

Image

In fact, that 12v line looks only mediocre and the 5v line looks like it's not even good at all so what exactly is the point of these things? Anytime you introduce more complexity without focusing on achieving some specific goal, usually everything just suffers (like modern bloated monitor firmwares that feel less responsive and laggier than a barebones, fly by night China monitor even though they supposedly bench at 2.6ms input lag). So it's possible these expensive digital PSUs might be subjective input lag downgrades with really no benefit at all.

Mr1991
Posts: 151
Joined: 24 Nov 2019, 10:10

Re: Not an EMI thread - Terminal slug cursor is PSU resistance based, not interference based

Post by Mr1991 » 02 Feb 2024, 11:24

r0ach wrote:
02 Feb 2024, 10:13
Mr1991 wrote:
02 Feb 2024, 09:45
Also this is something I’ve noticed on my setup, I bought a slightly more expensive psu, I believe it’s a “Corsair 1200i”, when I bought it it was considered top at the time, but it does seem to have a bit more of a watered down/smoother feel to mouse movement, it’s hard to say wether it’s better or worse, I’ll admit I’ve had nothing but positives with this setup in terms of overclocking/smoothness, but feel I performed better competitively on my older setup with a bog standard motherboard and psu
The expensive, digital controller PSU might cause more subjective bogged down cursor movement similar to having your BIOS set to auto-regulate power phases instead of setting them on extreme, full phases with no regulation. I considered buying one of these expensive PSUs before thinking they would provide some kind of benefit like a solid, consistent 12.00v line instead of 12.09 or 11.95 or whatever, but load regulation isn't even an improvement on a normal PSU:

Image

Image

In fact, that 12v line looks only mediocre and the 5v line looks like it's not even good at all so what exactly is the point of these things? Anytime you introduce more complexity without focusing on achieving some specific goal, usually everything just suffers (like modern bloated monitor firmwares that feel less responsive and laggier than a barebones, fly by night China monitor even though they supposedly bench at 2.6ms input lag). So it's possible these expensive digital PSUs might be subjective input lag downgrades with really no benefit at all.
Yeah it makes sense, also on the topic of the plug/cable type I noticed the cable that came with this psu is extremely thick, so could relate to the guy who says he noticed a difference between the two of his

User avatar
r0ach
Posts: 95
Joined: 10 Oct 2023, 14:45

Re: Not an EMI thread - Terminal slug cursor is PSU resistance based, not interference based

Post by r0ach » 02 Feb 2024, 17:31

Mr1991 wrote:
02 Feb 2024, 11:24
Yeah it makes sense, also on the topic of the plug/cable type I noticed the cable that came with this psu is extremely thick, so could relate to the guy who says he noticed a difference between the two of his
I'm currently trying to work through a few theories of mine concerning things like:

- If very tight voltage regulation is beneficial or noticeable. I.E. a PSU that can hold a solid 12.0v vs a PSU that registers 12.17v or one that dips to 11.9v in use. I theorize that if your 5v rail is way above or below that it probably does affect things like a feeling of overshoot/undershoot. Probably similar effects on the 12v rail.

- If there's any negatives to PSUs with multiple primary capacitors (I believe most 1000-1300w units have this) instead of just one large one (like SF750) in terms of things that affect gaming like mouse movement, subjective latency, potential fluidity, etc

- If the single capacitor on smaller units like SF750 is actually large enough or not to deal with things like transients on big GPUs or if you're always in some type of state where the thing is struggling to stay filled up that could cause intermittent tearing, voltage related effects that disturb mouse movement, etc

- I think finding out what vendors are shipping bare copper cables instead of tinned copper is also important

Multi-big cap 1200w unit:
Image

Single-cap 750w unit:
Image

Mr1991
Posts: 151
Joined: 24 Nov 2019, 10:10

Re: Not an EMI thread - Terminal slug cursor is PSU resistance based, not interference based

Post by Mr1991 » 02 Feb 2024, 18:50

Interesting, I’ve just ordered a 125V power cable to test with this setup (both the monitor and psu are 250v) which should come tomorrow, saw another guy say he has better results with the 125, but I guess I’ll find out

User avatar
r0ach
Posts: 95
Joined: 10 Oct 2023, 14:45

Re: Not an EMI thread - Terminal slug cursor is PSU resistance based, not interference based

Post by r0ach » 03 Feb 2024, 08:47

Mr1991 wrote:
02 Feb 2024, 18:50
Interesting, I’ve just ordered a 125V power cable to test with this setup (both the monitor and psu are 250v) which should come tomorrow, saw another guy say he has better results with the 125, but I guess I’ll find out
I don’t know anything about that subject, but for PSU’s, the guy who runs HWBusters PSU reviews said something the other day like “smaller PSUs are able to outcompete the larger PSUs” in some kind of metrics (meaning ones that are like 1200-1300w), but I don’t know exactly what he was alluding to. I think he made the statement concerning SF850-L vs SF1000L.

I can tell you that cable length matters a lot. Obviously or you wouldn’t see companies shipping 25 feet 24AWG HDMI cables and then 28 and 30 AWG 6 foot ones (anything above 26 AWG sucks even at 6 feet though IMO). So on PSU’s with extra short cables like the SF600/SF750/SF850/SF1000 series, these have huge return rates on Amazon because people can’t get them to stretch far enough for their cases to work. Little do they know, instead of a negative, the short cables (they’re 16 AWG to boot and not 18 AWG junk), are probably actually a benefit.

Post Reply