What's new
Carbonite

South Africa's Top Online Tech Classifieds!
Register a free account today to become a member! (No Under 18's)
Home of C.U.D.

Input required please.

Status
Not open for further replies

ginax570

New Member
Rating - 0%
0   0   0
Joined
Feb 28, 2020
Messages
19
Reaction score
3
Points
185
Location
Durban, Pinetown
Good Day to you all friends. My name is Gina and i am new here. I am not really new to the PC world but when it comes to Ryzen i am still a newbie. Please Help me on this one.

So firstly here are my specs:

Asrock b450 pro4 updated to the latest bios: ASRock B450 Pro4
Ryzen 5 2600
256gb hp 920 nvme ssd
16gb gskill trident Z ddr4 3200mhz CL 16
RX 570 4gb overclocked to 1480mhz on the core and 2100mhz (power limit slider maxed out and voltage -18 on msi afterburner. On full load the gpu does not go beyond 67 degrees with an aggressive fan curve)


After reading a lot and researching, i managed to get my r5 to 4.05ghz stable on 1.35v. It is also stable on 4.15ghz but the voltage required is a little bit insane 1.425v so i left it at 4.05.

My problem is this motherboard does not have any LCC or LLC settings so sometimes i get vdroop. If i set my voltage to 1.35, it drops to around 1.32v on full load and my oc crashes. So i read on the internet, a way to counter this is to set a higher voltage, which i did. I set my voltage to 1.375v and when it drops it drops to 1.35v and my overclock is stable. Is there anything else i can do guys to counter vdroop without having to raise voltage?
I am trying to go for 4.2ghz as i saw on hardware unboxed and 4.15ghz is not really a problem but 4.2ghz seems to crash due to vdroop. Atleast on 4.15ghz the vdroop is not as severe so it does not crash. If you could point me in the right direction about vdroop, i would really appreciate it. I do not want to go above 1.425v in order to counter vdroop.
 
Didn't see you mention it, what PSU do you have? It could be a factor.
 
Good Day to you all friends. My name is Gina and i am new here. I am not really new to the PC world but when it comes to Ryzen i am still a newbie. Please Help me on this one.

So firstly here are my specs:

Asrock b450 pro4 updated to the latest bios: ASRock B450 Pro4
Ryzen 5 2600
256gb hp 920 nvme ssd
16gb gskill trident Z ddr4 3200mhz CL 16
RX 570 4gb overclocked to 1480mhz on the core and 2100mhz (power limit slider maxed out and voltage -18 on msi afterburner. On full load the gpu does not go beyond 67 degrees with an aggressive fan curve)


After reading a lot and researching, i managed to get my r5 to 4.05ghz stable on 1.35v. It is also stable on 4.15ghz but the voltage required is a little bit insane 1.425v so i left it at 4.05.

My problem is this motherboard does not have any LCC or LLC settings so sometimes i get vdroop. If i set my voltage to 1.35, it drops to around 1.32v on full load and my oc crashes. So i read on the internet, a way to counter this is to set a higher voltage, which i did. I set my voltage to 1.375v and when it drops it drops to 1.35v and my overclock is stable. Is there anything else i can do guys to counter vdroop without having to raise voltage?
I am trying to go for 4.2ghz as i saw on hardware unboxed and 4.15ghz is not really a problem but 4.2ghz seems to crash due to vdroop. Atleast on 4.15ghz the vdroop is not as severe so it does not crash. If you could point me in the right direction about vdroop, i would really appreciate it. I do not want to go above 1.425v in order to counter vdroop.

Hi Gina, welcome to the forum. With regards to LLC (load line calibration) I've learned over time to not touch it due to some vcore spikes it would sometimes generate. These spikes can only be seen with an oscilloscope, so it won't be seen using a multimeter. I did some tests on my previous mobo, the Asus C7H to see what difference the LLC setting makes and what would happen to the vcore if left on Auto by measuring the vcore on the mobo probelt with a multimeter. With the LLC setting at it's highest there was very little vdroop but still there was. When left on Auto there was no, and I mean zero vdroop.

If I may ask, what do you use to check the vcore while under load? My reason for asking is that when I check with hwinfo64 the CPU CORE VOLTAGE (SV12 TFN) do drop about 0.025 or there about but when I check with cpuz, there is basically no vdroop. When I did my tests with the multimeter I also found that the vcore given by CPUz is the most accurate. I was in a discussion regarding this on OCN, from where I learned that the CPU CORE VOLTAGE (SV12 TFN) given by Hwinfo64 drop as it shows what is actually going into each core, not really what is going into the processor.

As for overcoming vdroop, as things stand it would only be possible by raising the vcore in the bios up to where the vcore under load is similar to where you know the cpu is stable. Now with the Zen+ cores, you could possibly go higher on the vcore without damaging it, but the 4.05ghz you sitting with is already a good start. You could try 4.1ghz using 1.3875vcore and see it it is stable, if not then drop the speed to 4.075 and see if it is stable.

Lastly, what do you use to check stability? Most people would say nothing but Prime95 with small fft's. Well I've tried it and it generate a hell of a lot of heat. Currently I'm using the blender benchmark as it also generate a fair amount of heat and with your chip will run wel over an hour before complete.

Hope you come right.
 
Hi Gina, welcome to the forum. With regards to LLC (load line calibration) I've learned over time to not touch it due to some vcore spikes it would sometimes generate. These spikes can only be seen with an oscilloscope, so it won't be seen using a multimeter. I did some tests on my previous mobo, the Asus C7H to see what difference the LLC setting makes and what would happen to the vcore if left on Auto by measuring the vcore on the mobo probelt with a multimeter. With the LLC setting at it's highest there was very little vdroop but still there was. When left on Auto there was no, and I mean zero vdroop.

If I may ask, what do you use to check the vcore while under load? My reason for asking is that when I check with hwinfo64 the CPU CORE VOLTAGE (SV12 TFN) do drop about 0.025 or there about but when I check with cpuz, there is basically no vdroop. When I did my tests with the multimeter I also found that the vcore given by CPUz is the most accurate. I was in a discussion regarding this on OCN, from where I learned that the CPU CORE VOLTAGE (SV12 TFN) given by Hwinfo64 drop as it shows what is actually going into each core, not really what is going into the processor.

As for overcoming vdroop, as things stand it would only be possible by raising the vcore in the bios up to where the vcore under load is similar to where you know the cpu is stable. Now with the Zen+ cores, you could possibly go higher on the vcore without damaging it, but the 4.05ghz you sitting with is already a good start. You could try 4.1ghz using 1.3875vcore and see it it is stable, if not then drop the speed to 4.075 and see if it is stable.

Lastly, what do you use to check stability? Most people would say nothing but Prime95 with small fft's. Well I've tried it and it generate a hell of a lot of heat. Currently I'm using the blender benchmark as it also generate a fair amount of heat and with your chip will run wel over an hour before complete.

Hope you come right.
Thank you so much. I really appreciate the detailed response. For checking vdrop, i was using hwinfor, i also have cpu z and with cpu z, the vcore tends to stay fixed at around 1.35 or 1.37v where i set it minus about 0.025v which i think is normal. Then for stability i was using ryzen master and prime 95 but the cpu gets so hot around 80 degrees. It becomes very hot that i have to put all my case fans to max to have it drop to around 70s. As for checking vcore under load, i was mostly using hwinfor, I will try cpu z and see what happens. I will mess around with the bios and fiddle around with clock speeds. Do you think i can squeeze a few mhz by touching the base frequency? or should i just be happy with 4.05- 4.1ghz and call it a day?
 
I dont see anyone asking this why you trying to get more juice?
 
Well i just want the max performance my cpu is able to provide even though the rx 570 is not stressing the cpu at all.
But why? I understand max performance but what are you trying to run? games/app/dev?
 
Thank you so much. I really appreciate the detailed response. For checking vdrop, i was using hwinfor, i also have cpu z and with cpu z, the vcore tends to stay fixed at around 1.35 or 1.37v where i set it minus about 0.025v which i think is normal. Then for stability i was using ryzen master and prime 95 but the cpu gets so hot around 80 degrees. It becomes very hot that i have to put all my case fans to max to have it drop to around 70s. As for checking vcore under load, i was mostly using hwinfor, I will try cpu z and see what happens. I will mess around with the bios and fiddle around with clock speeds. Do you think i can squeeze a few mhz by touching the base frequency? or should i just be happy with 4.05- 4.1ghz and call it a day?

In all honesty, I do not know the 2600 that well. I had a 2700X which was able to do 4.2ghz, but most of these Zen+ cpu's max out at around 4.0 to 4.1ghz which is why I said you should try 4.1ghz using 1.3875vcore.
 
In all honesty, I do not know the 2600 that well. I had a 2700X which was able to do 4.2ghz, but most of these Zen+ cpu's max out at around 4.0 to 4.1ghz which is why I said you should try 4.1ghz using 1.3875vcore.
thanks man i think i will leave it at 4.1ghz. I will update with results after stress testing.
 
thanks man i think i will leave it at 4.1ghz. I will update with results after stress testing.
Sweet, looking forward to your findings.

Sent from my SM-G960F using Tapatalk
 
Status
Not open for further replies

Users who are viewing this thread

Latest posts

Back
Top Bottom