I’m waiting for three 16x-16x risers to arrive in few days, so I think the best option is to try one of those in that middle 16x slot to see if it works.

If it’ll not work I’ll try to short it…but I need to find a guide on how to do it properly in the 16x slot (that thread shows the pic of a 1x slot shorted), to avoid some nasty problems :P

Btw, thank you again for your help, I’ll let you know as soon as the 16x-16x risers will arrive!

[quote name=“MrFeathers” post=“54521” timestamp=“1390529891”]
To fix the problem where one card mines super slow/not at all you have to boot up with the monitor attached to that card. Try booting with the monitor on different cards and see if it fixes the problem. As for the card not being recognized you may need to set pcie to 1x in the bios to enable all slots to work over 1x at the same time. Try these and let us know if they work.
[/quote]

Thank you for your help MrFeathers!
I fixed the problem of the card mining superslow doing as you suggested! Switching the monitor over the slow card (even after the boot) brings that card to the same normal Hashrate as the others!!

Regarding the other hint you gave me, I’ve looked for an option like that on the bios but I can’t find it…
Here the screen of my bios menu:
[IMG]http://i40.tinypic.com/27wx4rp.jpg[/img]
[IMG]http://i39.tinypic.com/2mhzlsp.jpg[/img]
[IMG]http://i41.tinypic.com/2qnp8qu.jpg[/img]

Can you explain me better how to force the PCIE to 1x?
Cause it will come back usefull to run 4 or 5 cards at the same time using all of the slots available.

Thank you again for you help, really appreciated :)

Hello to everyone!

First of all, I’m a noob in Mining and such, so I’m sorry if I’m asking something obvious for you.

That sayd, I’ve some problems with my mining rig, here’s my setup:

MB: Asrock 970 Extreme4
CPU: Sempron 145
PSU: Corsair AX860 Platinum
RAM: Gskill 1x4Gb 1333Mhz
SSD: Kingston 60Gb
GPU: 3x Sapphire R9 280x OC Dual-X 3GbGDDR5
Riser: 3x Powered Risers 1x-16x

I tested all the 3 cards and all the 3 risers in every combination, and all of the items are working good at 700-750 KH/s.

My problem is the following:

The PCIE4 slot, the 16x in the middle of the MB, is not recognizing any card if I use them with any of the risers -> D6 error on the Dr.Debug led on the MB, that is: “No console output devices are found”.

I tested the same slot, the PCIE4 slot, plugging in any of the cards directly, without risers, and it recognizes them normally.
I also tested the 3 16x slots together, plugging all of the cards directly in, without the risers, and all of 3 works fine.
I tested the other 2 16x slots of the MB with risers, and they recognize the cards normally.

So, why that single slot isn’t recognizing the 1x-16x risers? Should I need to buy a 16x-16x for that slot? There’s a way to fix it somehow?

And also, I’ve encountered another probem trying to avoid using that 16x slot, here the configurations I’ve tried so far:

Config A:
GPU 0 -> Riser plugged in PCIE 16x slot --> 700Kh/s
GPU 1 -> Riser plugged in PCIE 16x slot --> 700KH/s
GPU 2 -> Riser plugged in PCIE 1x slot --> 100KH/s

Config B:
GPU 2 -> Riser plugged in PCIE 1x slot --> 700KH/s
GPU 1 -> Riser plugged in PCIE 1x slot --> 700KH/s
GPU 0 -> Riser plugged in PCIE 16x slot --> 100KH/s

So, with both those configurations, all of the 3 cards are correctly recognized with the risers.
BUT, in both cases, the single card that has a different PCIE Interface compared to the other 2 cards (GPU 2 in config A & GPU 0 in config B) is having a really low hash rate for some reasons.

Why?
I think this could be a setup problem, cause if only the cards on the same slots interface are getting proper Hashrate it would be nosense to put more than 3 cards on the same MB.

Thank you in advance to anyone that will try to help me out, I hope I’ll be able to set it up properly to continue mining with all the 3 cards!

Jsnow

[quote name=“nimmneun” post=“51009” timestamp=“1389297512”]
Have you tried testing each card individually, by actually unplugging one card from the system? Thats where I’d start from.

And maybe start with lower settings like 1000mhz engine / 1250 mhz memory. I tested those with 3 different brands and they all gave me about 690Kh/s. The difference between 1250mhz and 1500mhz on memory is 0.5-3Kh/s at the most - so basically nothing =)
[/quote]

Hey,
I’ve did some tests as you suggested, testing 1 card at a time and setting them at 1000 clock and 1250 mem clock.

With both the cards that are at 1.2v from stock I can’t go any lower than 1.14v, anytime I try it gives me driver error…
With the card that has 1.163v from stock I can go down till 1.08v, as yours.

So, why can’t I go lower than 1.14v with those 2 cards? Could it be a driver problem? Which driver do you use?
Or what could be the problem?

And, using those 3 cards at, respectively, 1.14v + 1.14v + 1.08v, could I’ve problems using my PSU Corsair AX860? Or I can use it safely?

Thank you again for your help!

Thank you for your answers!

@Kojima18: at stock settings my GPUs temps are around 60�C, totally open case in a room with windows open all the time to get better airflow.
My setup is almost same as OP:
CPU - Sempron 145
MB - Asrock 970 Extreme4
SSD - Kingston 60Gb
Ram - 4Gb GSkill

I’ve never touched mem clock, tried to increase the engine from the 1020 stock value to 1050 and 1080 max, it goes well with respectively:
1020 - ~705KH/s
1050 - ~720KH/s
1080 - ~730KH/s
No problems at all when I OC without undervolting.

@nimmneum: I was thinking the same and I’m actually testing each card individually, will try to lower the memory and engine settings as you suggested.
I’ll edit this post with the results as soon as I finish testing them!

Thank you again for your patience :)

Ok I tested a bit more and something strange happens:

• I tried to undervolt both cards with Trixx, but even undervolting them by a tiny amount (from 1163 & 1200 to 1150) after a while the GPU 0 started to look strange in CGMiner: Hashrate jumped to 850KH/s on the GPU 0 line, but the total Hashpower amount was not respecting it, and after a while I got a driver error, the monitor went black for a while and went back to normal…but after the driver error restarting CGMiner causes the GPU 0 to stay at a ridiculously low Hashpower…
• I tried to undervolt only GPU 0, but got same as above…
• I tried to undervolt only GPU 1, but, again, same as above…

Just to mention, the GPU 0 is the one connected to the monitor during the tests.

So what could be the problem?
Thinking about the fact that I’m getting always the same problem to GPU 0, no matter which GPU I undervolt and how much I undervolt, could it be something else and not a card problem??

I hope that you can give me some tips and help to figure it out!
Thank you!

Thank you kojima18 for your answer.

I did some tests with 2 of my GPUs, and I’ve some questions:

• Does the CGMiner command --gpu-vddc works for those R9 280x cards? Cause I’ve tried it but apparently nothing happens.
• Does CGWatcher recognizes the voltage changes made with Trixx? Cause I’ve changed the voltange on Trixx but starting CGWatcher it shows the normal stock voltages.
• Does overclocking has sense in conjuction with undervolting?

The two GPUs tested are from stock:
1020 Core Clock
1500 Mem Clock
1200 vddc
&
1020 Core Clock
1500 Mem Clock
1163 vddc

I’m actually running them both at 1113 vddc, so undervolted just a bit, and everything seems fine, with no loss of performance.
If I go at 1100 or under I get a driver error and one of the GPUs looses half of its Hashpower, going to 300-400 KH/s, the other one stays normal…how that could be? Is this a problem of the card?

Thank you again!!

[quote name=“ChekaZ” post=“50713” timestamp=“1389188075”]
Hey Jsnow,

welcome to our Feathercoin Community! :)

If you like you can add yourself to our FeathercoinMap: [url=http://local.feathercoin.com]local.feathercoin.com[/url]
[/quote]

Thank you, I’ve just added myself, I’m from Rome :)

PS. I was able to choose just one label so I put forum member, but I also mine FTC, even if I switch with LTC and other alts.

[quote name=“nimmneun” post=“43479” timestamp=“1386862446”]
+rep - UV4Life! I’m running the same setup (gpu+mobo), except for the CPU (Athlon x2 260, Semprons were out of stock *lol)

I’d be curious about your watts + hashrate if you set Trixx [b]1.07v[/b] and [b]–gpu-engine 1000,1000 --gpu-memclock 1250,1250[/b]?
In theory this should be
With difficulty growing towards 5000 for LTC by end of the year (other altcoins will probably also seriously rise next year) power consumption will be quite a factor =)

edit4memclock fix =)
[/quote]

Hi there!
Firs of all, I’m in the mining world from christmas, so I’m totally noob!
That being sayd, I’m using almost the same setup as in the OP and I’d like to run 3x R9 280x with my Corsair AX860 80Plus Platinum PSU.

As far as I’ve read I could run those cards also using the defaults stock settings at 840W, right?
But I’d like to use the settings you have in your signature to save some power…I can achieve it by just changing my CGminer setup?

My actual cgminer setup is the following:
cgminer --scrypt -o *** -u *** -p *** --thread-concurrency 8192 --intensity 13 --worksize 256 -g 2

Should I change it that way? :
cgminer --scrypt -o *** -u *** -p *** --thread-concurrency 8192 --intensity 13 --worksize 256 -g 2 --gpu-engine 1080 --gpu-memclock 1500 --gpu-powertune -20

Should I also add a fan speed command and a target temp command? I’d like the cards to stay under 70�C.

I need to do something else or doing so I’m already achieving my goal?

Thank you very much in advance!

PS. I’ve the opportunity to place my rig in my office, avoiding any electricity cost. Buying the beta of CGRemote I’d be able to monitor my rig from home and also switch from LTC, FTC and other coins as I wish and at every moment? Cause If so I’m definetely going for it!

Hello everyone!
I’ve just joined this forum!
I found it on google looking for some guides and I decided to join.

You’ve done a great work, this forum has a lot of really helpfull threads, especially for who is approaching the mining world for the first time!

Jsnow