I am not sure if I didn't missed something, but did anyone do a real OC? I mean OVER 275MHz?
I am currently doing 1.2TH at 300MHz @ 0.775-0.815V, did anybody tried more? how is the GH/W?
C1 at 100-275MHz has following chart, it should be similar as on other devices based on BM1382 chips - S3, U3 and S4.
100MHz @ 0.625V
(ASICs do not work below this voltage) : 192W =0.48W/GH
175MHz @ 0.650V : 356W =0.51W/GH
250MHz @ 0.720V : 650W =0.65W/GH
250MHz @ 0.760V : 662W =0.66W/GH
275MHz @ 0.760V : 809W =0.74W/GH
275MHz @ 0.800V : 900W =0.82W/GH
300MHz @ 0.815V : not yet measured
Hope anyone is interested.
Please note that best W/GH ratio can be reached by lowering input 12V voltage, such as 9V, but highest eff. power supply can be done only at high voltages (12V and up) which can also dismiss conducting losses between PSU and miner.
This measurement was done at miner, not at the wall.
Interesting. I may give your suggestion a test eventually. One thing I did notice with my unnamed cheap adjustable power supplies, since I flipped my rad fans so I could have the ports and wires inside a window, and I note it was raining when I did the flip, my power at the wall has gone from 860 watts to 880 watts. I check my voltage at the miner and it's 12.08vdc same as it had been when I set it up (setting the voltage at 12.15vdc no load), then I checked the wall voltage. The wall voltage is at 118.1. When I shut down the miner it comes back to 120.1vac. The miner is on one of two lines I ran to the living room for AC and my treadmill. I'm kicking myself for not running 12-2 Romex instead of 14-2. These two lines run to to a pair of half size circuit breakers. The room temperature has come up from mid-60's to high 70's. Outside temperature has come up 20° since the low power days of 860 watts at the wall (includes the pump). So, when it's raining and the indoor humidity is 56.5%, cooling is reduced (low mass air drawing away less heat) and efficiency drops. Also perhaps my cheap power supplies efficiencies drop in a warmer room and high humidity. I'm running the C1 on 4 PCI-E connectors. I had ordered another 4 of the connectors late and they arrived at my post office box the other day and I'm about to run into town for those. So, later today the C1 will be running on 8 PCI-E lines. Presently the leads to the 4 connectors are warm, dropping voltage. So, before I install the new leads I'll carefully check the power at the wall. If it's as you indicate, the reduced voltage drop between my supplies and the miner should cause an increase in power consumption - unless the efficiency of my power supplies suffer as the miner improves - the supplies are running at ˝ full load each. So, home with the PCI-E-6 connectors. The power supplies were seeing 118VAC, wattage 880 at the wall and output of the supplies was 12.09vdc. The wires from the supply to the C1 were warm. Only 4 of the 8 connectors on the C1 were occupied. Completed a pair, shut down the C1, added the 2 new to one side and fired up. At fire up wattage was 870 and after 20 minutes it was up to 889 watts at the wall. Finishing the second pair of wires, shut down, added the final pair of leads, now all 8 connectors occupied, and at fire up the wattage was 838 watts at the wall, 118VAC. After 20 minutes, wall wattage measurement is 856 watts, wall voltage 117.9VAC, power supplies voltages 12.10vdc. What this is telling me is that the supplies' voltages rose 10 millivolts due to decreased load as a result of reduced voltage drop in the lines from the supplies to the PCI-E connectors. The higher voltage at the PCI-E connectors allowed more efficient mining. I was somewhat remiss in failing to record the voltage drop between the supplies and the PCI-E connectors before doubling the wiring. The rise in the supplies' voltages would have been greater if the AC voltage at the wall had remained at 118VAC instead of dropping to 117.9VAC as the supplies' outputs to some degree track changes in wall AC. This was shown when wall AC voltage rose exceptionally high when the neighborhood shut down air conditioning on cool evenings. After an hour the wattage at the wall is varying between 854-857 watts. |
|
|
|
|
Odd. I ssh'd into the C1 some time back and changed the root password using the unix command passwd, logged out and logged back in using the new password. Today I went to log in and the password had reverted to admin while the GUI was still connected and unchanged.
|
|
|
|
I am not sure if I didn't missed something, but did anyone do a real OC? I mean OVER 275MHz?
I am currently doing 1.2TH at 300MHz @ 0.775-0.815V, did anybody tried more? how is the GH/W?
C1 at 100-275MHz has following chart, it should be similar as on other devices based on BM1382 chips - S3, U3 and S4.
100MHz @ 0.625V
(ASICs do not work below this voltage) : 192W =0.48W/GH
175MHz @ 0.650V : 356W =0.51W/GH
250MHz @ 0.720V : 650W =0.65W/GH
250MHz @ 0.760V : 662W =0.66W/GH
275MHz @ 0.760V : 809W =0.74W/GH
275MHz @ 0.800V : 900W =0.82W/GH
300MHz @ 0.815V : not yet measured
Hope anyone is interested.
Please note that best W/GH ratio can be reached by lowering input 12V voltage, such as 9V, but highest eff. power supply can be done only at high voltages (12V and up) which can also dismiss conducting losses between PSU and miner.
This measurement was done at miner, not at the wall.
Interesting. I may give your suggestion a test eventually. One thing I did notice with my unnamed cheap adjustable power supplies, since I flipped my rad fans so I could have the ports and wires inside a window, and I note it was raining when I did the flip, my power at the wall has gone from 860 watts to 880 watts. I check my voltage at the miner and it's 12.08vdc same as it had been when I set it up (setting the voltage at 12.15vdc no load), then I checked the wall voltage. The wall voltage is at 118.1. When I shut down the miner it comes back to 120.1vac. The miner is on one of two lines I ran to the living room for AC and my treadmill. I'm kicking myself for not running 12-2 Romex instead of 14-2. These two lines run to to a pair of half size circuit breakers. The room temperature has come up from mid-60's to high 70's. Outside temperature has come up 20° since the low power days of 860 watts at the wall (includes the pump). So, when it's raining and the indoor humidity is 56.5%, cooling is reduced (low mass air drawing away less heat) and efficiency drops. Also perhaps my cheap power supplies efficiencies drop in a warmer room and high humidity. I'm running the C1 on 4 PCI-E connectors. I had ordered another 4 of the connectors late and they arrived at my post office box the other day and I'm about to run into town for those. So, later today the C1 will be running on 8 PCI-E lines. Presently the leads to the 4 connectors are warm, dropping voltage. So, before I install the new leads I'll carefully check the power at the wall. If it's as you indicate, the reduced voltage drop between my supplies and the miner should cause an increase in power consumption - unless the efficiency of my power supplies suffer as the miner improves - the supplies are running at ˝ full load each. |
|
|
|
Interesting. 36.9°F outside. Inside is 63.9°. Maybe it's my connection or maybe it's the pool but my hashrate goes though a long slow rise then fall and so on. The temperature of the water out port of the water block is cool to the touch while the air pushed thru the C1 is quite warm on exit, warmer than the exiting water. I just unplugged the single fan so the two closest radiator fans to the ports run. I wonder if letting the C1 mounted fan run at 12v continually would be a good idea. Since disconnecting the single radiator fan the C1 mounted fan has started changing speed fairly often. The hashrate has come up but that might just be the observed cycle.
Odd. GUI Fan2 speed indication never changes from 1200 although I hear it increasing and decreasing.
If its annoying you can try plugging the fan into the other header that it can reach. You'll have to take the case off unfortunately (and all the PCI-E's) to get to it though. You sure you pump seems to be running okay? I can't think of a reason why the fluid exhaust would be cooler than the air exhaust. I think problem might be there is gap between waterblock and hash board.You could have a try to open C1 and make sure whether every screw on board is tight. Considering it. Just for practice today, since I have two S3+'s that work well tho one a little better than the other, I ripped apart the slightly slower S3+ and reworked the thermal paste. There was a wealth of thermal paste there unlike my S3's which I reworked. So my new thermal paste didn't make a great deal of difference. It is still the slower of the two S3+'s but it's been producing more shares than the faster most of the afternoon and evening. It may have just been the removal of accumulated dust on the controller board as much as the new thermal paste. I plan to rip the C1 down and replace the thermal paste but I used most of what I had and just ordered more. |
|
|
|
Thanks notlist3d and dogie,
I will wait for new firmware that Bitmain will have to post due to some broken miners that may have gotten shipped.
Also, because I have no energy and time left budgeted for this miner over this weekend.
My only frustration is is that despite known problems in shipped miners, Bitmain has not reacted promptly to provide easy solutions to problems like firmware.
My C1 is operational for now though I have to run it at lower speed and manage it with stupid alternatives e.g. saving miner configuration still does not make all pools operational, (it is random but limited to first pool dead in most cases).
I have to run cgminer from command shell using ssh to make it work.
Thanks for your advise Tupsu, will keep that in mind and that precious link.
regards
The problem you described appears to be a bad SD card which isn't going to be fixed with new firmware. I just read this entire thread. I have a solo setup and previously had an S1 on it with no issues and I cannot get the C1 to work. It works with other pools no problem but no matter what I have tried it always comes up "dead" on the C1 . The S1 for the URL I just had x.x.x.x:port and then it would add the http:// on front and work fine (mostly/bit flaky still).. the C1 adds the "stratum+tcp://" and even when I set it to http://*.*.*.*:port it still shows up as dead. Is there something different that needs to be done in the C1s? This quoted post is the only thing I found semi similar but even in different pool slots other than #1 it still is always dead. Everything else is taken care of ie node is fine and other miners including S1 connect fine and ip is in the rpcallow in the conf. Any ideas.. it is 11/26 firmware My own server has worked exactly 1 time in the last few days of messing with it.. then I changed another unrelated setting and reapplied and poof... hasn't worked again since. S1's work absolutely no problem. I even switch IPs between S1/C1 no matter what the C1 just shows up dead pool for no reason I can see. Pretty sure whatever the problem is it is the C1... This is absolutely pissing me off. It appears as it is C1 does not work with the standard BTC node solo mining. I tried CG 4.61 by itself and it asks for “--btc-address” and would not connect to a server until supplied. I think this is the reason it is not functional. I must have been seeing things when I thought it worked once. This is the last response from bitmain after multiple low comprehension emails. "Sorry to hear that. there may be some incompatibles with solo mining for C1." They keep telling me to use their pool and even told me I could user their "solo" setup which sounds like a complete joke... "suggest you consider to mine with antpool directly. We're using solo mining and will distribute the payment once blocks are found. stratum.antpool.com:3333, or solo.antpool.com:3333, solo.antpool.com:25 " Is there a way to supply arguments to firmware containing CG to actually fix my issue? I really expect to be able to connect to the standard client as a server.. seems pretty fundamental for btc mining equipment to be able to do. I understand everyone is very eager to say how stupid solo is.. I don't care, I am mostly using it to be involved / gambling and for heat... I never tried solo mining but about finding blocks: had a zTex, then USB Block Erupters, then a Blade, then my Jalapeno finally arrived, then I bought a Mercury. When I got the Mercury hashing it found a Block very quickly. That has been my only block in couple of years of mining. That includes the S3's, the S3+'s and now my C1. Seems to me finding a block is mysterious and rare. |
|
|
|
i have s3 miner which is connected to internet and i am able to access through default IP 192.168.1.99 but its not hashing..i have correct pool and the pool is connected..but in miner status page its not hashing its shows as zero...
i have tried to reset the machine but its same .Attached is the screenshot.
<snip> img </snip>
Go to Network tab then click edit on your connection then switch the protocol to DHCP and Save & Apply. You'll get a new IP address, so it will not be on 192.168.1.99 anymore, but will begin hashing straight away. yes i tried that but even then its not hashing..now the network have assigned it a value 192.168.1.100...  I don't see any ASICs registering at all. |
|
|
|
|
Clips made of cheap steel is the source of the rust.
|
|
|
|
after some repairing from my broken C1 i finally fixed it. it now runs STEADY on 1045 GH/S
What mods to original did you do may I ask? replaced the cooling block and all other cooling, which was totally destroyed from shipping (they shipped my C1 with water in the block which had made t he whole unit to Rust on the way to me so the miner was all wet with brown rust water everywhere. so i basically only replaced the cooling Take any photos of the rip down and repair? Particularly the quality of thermal paste application. |
|
|
|
after some repairing from my broken C1 i finally fixed it. it now runs STEADY on 1045 GH/S
What mods to original did you do may I ask? |
|
|
|
Interesting. 36.9°F outside. Inside is 63.9°. Maybe it's my connection or maybe it's the pool but my hashrate goes though a long slow rise then fall and so on. The temperature of the water out port of the water block is cool to the touch while the air pushed thru the C1 is quite warm on exit, warmer than the exiting water. I just unplugged the single fan so the two closest radiator fans to the ports run. I wonder if letting the C1 mounted fan run at 12v continually would be a good idea. Since disconnecting the single radiator fan the C1 mounted fan has started changing speed fairly often. The hashrate has come up but that might just be the observed cycle.
Odd. GUI Fan2 speed indication never changes from 1200 although I hear it increasing and decreasing.
If its annoying you can try plugging the fan into the other header that it can reach. You'll have to take the case off unfortunately (and all the PCI-E's) to get to it though. You sure you pump seems to be running okay? I can't think of a reason why the fluid exhaust would be cooler than the air exhaust. It's clear that drawing air in from outside, outside at 36.9°F, that coolant is going to chill. The miner was in a room at 63.9 beneath the window. Those temps were last night. I checked earlier this afternoon. Outside ~43°, inside 66°, air measured inside the rad exhausting into the room from outside, ~70°, air at exhaust of the C1, 99-100°. -------------- I'd link to see springs for the outlets that allow the hose to make a curve without collapsing. I probably want to put the fans on the other side of the rad on the same side as the hoses. In cold weather that would put the hoses inside while drawing in and warming outside air. I could always switch back in late spring. As it is I just rearranged a bit of foam from packing that I had taking up the remainder of the window width. It pressed against the hose and restricted a little. GH/s dropped into the low 900's. Hashrate has come back since cutting out a piece that compressed the hose a bit. --------------- My mistake. The fans are already on the side of the ports. I just need to flip them to reverse the rad. |
|
|
|
|
MrPark, the cap in the photo has been damaged by an impact. It was hit with something. You can see the slight dent in the cap can.
|
|
|
|
|
Odd. GUI Fan2 speed indication never changes from 1200 although I hear it increasing and decreasing.
|
|
|
|
|
Interesting. 36.9°F outside. Inside is 63.9°. Maybe it's my connection or maybe it's the pool but my hashrate goes though a long slow rise then fall and so on. The temperature of the water out port of the water block is cool to the touch while the air pushed thru the C1 is quite warm on exit, warmer than the exiting water. I just unplugged the single fan so the two closest radiator fans to the ports run. I wonder if letting the C1 mounted fan run at 12v continually would be a good idea. Since disconnecting the single radiator fan the C1 mounted fan has started changing speed fairly often. The hashrate has come up but that might just be the observed cycle.
|
|
|
|
Hmmm, seems like BITMAIN just used the controller and two hashing boards from the S4  ? and made the S5 out of it. no the chips are different. the s-5 is very efficient for power much better then the s-4. I did a review for the s-5 and it is very close to the sp20. 2 s-5's are 900 shipped to my door in the USA and 2 sp20's are 1000 shipped to my door in the USA. You can do a lot of comparisons of the two. For me in my home using about 10 pieces I give the sp20 a very small edge. For others the s-5 will be better. Just looked - an SP20 is $659, holiday special. How do you get $500 each? |
|
|
|
I am using the upgraded firmware. I am getting near 1000 only if I lower my freq to 243.75 and voltage at 0730. If I use the upgraded firmware settings of 250 freq with 0760 voltage, the first chain shows a few x's after a few minutes...
Thoughts?
What's your coolant? I know with the S3 I got from a Florida reseller that under performs, when it warms up I get an X. Are any of your hoses crimped? Could be the thermal compound isn't well spread on your water block. I'm using the stock firmware, stock wiring, distilled water coolant, and my pool shows 1071 GH/s. Ambient room temp 64°F. |
|
|
|
Am just about to fire mine up too. Have just done nmap. I've run the pump of course in getting the air out but about to apply power to the miner. I'm trying this with 4 PCIE-8 connectors not 8 only because I don't have the other 4 yet. If the wires show heat, excessive current, I'll shut down an S3 and get the connectors I need. ...
so, 10.45 amps sound right? P.F. 0.69, 857-860 Watts, 10.49 amps. |
|
|
|
|
Am just about to fire mine up too. Have just done nmap. I've run the pump of course in getting the air out but about to apply power to the miner. I'm trying this with 4 PCIE-8 connectors not 8 only because I don't have the other 4 yet. If the wires show heat, excessive current, I'll shut down an S3 and get the connectors I need. ...
|
|
|
|
Oh I didn't have any specific cooling liquid so I am using distilled water at the moment?
So no difficulty seeing air bubbles thru the clear tubing while getting the air out? Mine was suppose to arrive today but I screwed up and used my PO Box as a UPS shipping address. It's out for delivery but will head back to the office after customer hours so am told I can't get it until tomorrow. Was thinking of distilled water with food coloring but if the clear tubing stains that won't work well. Doggie's guide it very helpful if you have no water cooling experience, make sure you pay attention to the water flow around your setup the guide has pictures just follow them. I like having my cooling rig running from my psu not my C1 as I could run the pump and get the fluid sorted without any C1 issues. As far as air bubbles go, the main thing is read the guide filling the pump res while holding it above the C1 and Rad means you can get a lot of air out at the start, then you just need to cycle the power a few times as the res can empty out into the rig as the air comes out. I think I refilled mine 4 times or so ( As i was using Distilled water its not that messy with proper fluid I would use a filler bottle). Don't forget to rotate your rad, I rotated my C1 a few times to get some of the last bubbles out some were quite big. Oh Doggie mentioned we should use PC cooling fluid as the Distilled water has issues over the long term. I am ordering that as I write this. Oh and finally the air bubbles should end up in your res so as long and you rotate you main item's around, any air pockets will get back to the res/pump. good luck Thanks. Am sitting here waiting for delivery within the next few hours and the thought arises: where did I leave my Teflon tape last? I wonder if this is a positive displacement pump. A positive displacement pump might have two gears meshing in the center while fluid is forced around the gears outer sides. Or a positive displacement pump might be a flexible tube inside a semicircle and an inner circular drive wheel having rollers on the outside compressing the tube against semicircle forcing fluid in the direction of rotation. Or a positive displacement pump might be a pump with a piston and valves. A non-positive displacement pump would be like a jet pump. |
|
|
|
i kinda tested this out and i think its just the ui that is messed up.. the pool is still reporting a regular speed. Have you tried SSH'ing in, then commanding #/etc/init.d/cgminer stop, then, #/etc/init.d/cgminer start ? |
|
|
|
Oh I didn't have any specific cooling liquid so I am using distilled water at the moment?
You will have problems due to the mix of copper and aluminium in the loop. I have no idea over what timescale it will become a problem, but it would be advisable to add an additive or swap out to dedicated coolant which won't have the same problem. So the provided cooling radiator is copper? That's a surprise. Because water carries 50% more heat than ethylene glycol, the basic ingredient in most antifreeze coolant, I want to use distilled water (~.68 cents/gallon in Walmart) and perhaps rig up a PVC water filter with the guts from a quality water filter advertising removal of metal ions. Go with the coolant. What ever you save initially you will pay more in the long run if you don't have coolant or additives if you truly have to go with water. Well, the corrosion shown at a water block port was likely caused by the use of hard tap water. Distilled water alone would have slowed that way down. Distilled water with an ion filter it likely won't happen. We're talking about a 50% increase in heat carrying capability over ethylene glycol. When you use the phrase "in the long run" you must be addressing those getting free electric. e.g. padding their office expenses as for the rest of us profitability might not last long beyond recovering cost. Look at the Jalapeno. |
|
|
|
|
Show Posts
