...snip....
{
"index":"1",
"chain_acn":"4",
"freq":"737,fan_num=1,fan1=5880,fan2=0,fan3=0,fan4=0,fan5=0,fan6=0,temp_num=3,temp1=0,temp2=0,temp3=35,temp2_1=0,temp2_2=0,temp2_3=46,temp_max=0,Device Hardware%=0.0000,no_matching_work=0,chain_acn1=4,chain_acn2=4,chain_acn3=4,chain_acs1= oooo,chain_acs2= oooo,chain_acs3= oooo,chain_hw1=0,chain_hw2=0,chain_hw3=0,chain_rate1=4.99,chain_rate2=5.97,chain_rate3=5.24|",
"fan":"5880",
"temp":"0",
"chain_acs":" oooo"
}
...snip...
...snip...
Ok, weird. temp1 and temp2_1 both go together. One is the PCB (first one, i think) the other is the ASIC. For some reason the temps are coming out empty. I figured it was their web UI because if anything is wrong in the numbers, the web UI goes blank. The firmware that came on it -- what frequency options exist? Beyond that, is the version1 firmware providing you the frequencies you expect? Also, for anyone reading, I just want to be clear this is on the latest batch minis with the factory firmware.. :-) Jason |
|
|
|
today I received 2 new z9 mini's and found they wont run batch 1 firmware correctly. When running the batch 1 it will not show the temps on more then 1 hashboard but if I use the current version from bitmain everything runs fine showing all the temp's. Tried all the resets and everything but sameissue. When the temps arentshowing the thing hashes finefull speed just no temps which cause fan to run full speed and fault light to be on. Anyone else have this issue on any mini shipped after nov 1st
Eh, that's probably their crappy web interface and nothing more... but it's possible they think they are being tricky.... the fault like thing is a little interesting, but probably still a simple fix if it's what I think it is. On your miner, go to these two URLs and PM me the output please: /cgi-bin/get_miner_status.cgi as well as the output from this /cgi-bin/get_status_api.cgi Jason |
|
|
|
|
Folk,
I know I have not provided a lot of update here recently, and I apologize. The reason is simple, I didn't have any news worth reporting.
I am now confident the release is imminent (a couple of more evenings) although I will scale it out slowly to ensure the infrastructure and miners are working properly.
The current status is this:
* Almost done with the infrastructure
* Testing various positive and negative infrastructure conditions such as server down, not found, license not found, etc.
* Licenses can be upgraded on the fly and clients will "magically" pick up the updates (have a license for 2, need 4, can do it on the fly, no need to change firmware)
* the dev fee version will be updated along with this one and will bring them to feature parity.
* and more...
I still have a reasonable amount of work to do, but the hard parts are behind me (I think....). I'm still in the process of updating the webUI with support for providing a messages to the user (for example "upgrade available") in the dashboard, displaying the licensing information and current counts,
per hashboard feature: This will be enabled in all next releases, both paid and developer fee supported.
For dev-fee users, I am working with a couple of the pools to optimize how the underlying aspects work -- the benefit there is that I will be able to greatly reduce the impact of a dev-fee on hash rate (should be close to 0 when all is said and done). It may be that this functionality will be pool specific in the near-term; I'll provide more detail here as it is available.
NiceHash: I've generated the traces requested by nicehash and submitted them for analysis. My eye does not see anything obvious as to the failure reasons, so I am reliant on them to provide information in order to try to add in support.
Fan Control: Hardware aspects like fan/voltage are the next items on my list. I've done a lot of pre-work on this and honestly things don't look great, but I am not out of ideas just yet.
Voltage Control: see fan control.
Mini: This will follow on the footsteps of the next release. I've been testing. I will likely have a lower dev fee rate for the mini.
Dev-fee: I am strongly considering lowering the dev-fee over time; I will be able to handle this with the infrastucture I've been building. What this means is the longer a node runs, the lower the dev-fee goes to some minimum amount.
Other: I'm sure I'm forgetting something(s)... more soon.
Thank you for your patience and support,
Jason
|
|
|
|
Any update on the paid version of the firmware without the dev fee? Any ETA on release? I would like to use the individual board clocking feature but do not want to pay the fee since I already bought the firmware.
@CrypticMiner - I'm working on it as quickly as I can, I swear. I'm beyond disappointed in myself with how long this is taking me. I keep running into road blocks and I simply cannot release until the main issues are resolved. I apologize. The upcoming features are more about the long-run of the units instead of the near-term performance -- tunings that can help ensure they are remain profitable as time progresses. I will release as soon as I can. I am not sitting idly by. Thank you, Jason |
|
|
|
Waiting for the z9 mini version, if ever !
Me too! Jason |
|
|
|
|
In other news, my miner that "Made it through DHL without customs hit" just received a $900 invoice about 45 days later.
Sigh.
My guess is any of us that thought these came through without the 25% hit will be getting bills in the future.
Jason
|
|
|
|
@efudd - just want to check that the latest firmware 2.0d works on the latest batch of z9? Thanks
@roymunn- yes, that is my understanding. I have not personally been able to test. Please provide feedback if you find out different from your testing. Thank you, Jason |
|
|
|
i know sometime is good to not say anything but i used this dev fee firmware and i dont see any profit other than making less than usual just because the fee. i went back to smartass111 free overcloked firmware and i have been running @650 just fine making way more than yours.
this dev fee is killing us, im following your work wich is awsome and i dont take anything off it. but since theres is a free FW and its working flawless i dont see why anyone should pay or use this dev fee fw just to kill their machines to not even make anything other than your dev fee.
@dohckiller - Thank you for your review and feedback. I am thrilled that you have not had any issues with smartass111's firmware and I do hear that you believe the development fee is "killing you". I would not want that to happen, so I do encourage you to use the other firmware and make "way more than [mine]". Your experience is the unusual one as I have had multiple users and farms with thousands of machines approach me after having issues with smartass's firmware modification. I know for a fact that his modification misses a couple of pieces in the frequency calculations and will throw a system off. Here is a list of the check-ins to my code repository for these modifications. The time I've put in is not trivial by any means. I personally value my time and have put significant effort into the current modifications and the upcoming modifications. I'd also like to point out that I respond near-instantly to almost every inquiry both public and private. The value I provide is more than just the firmware, but also the service and support. Date: Sat Nov 10 00:50:45 2018 -0500
Date: Fri Nov 9 00:44:33 2018 -0500
Date: Tue Nov 6 22:35:25 2018 -0500
Date: Tue Nov 6 21:53:50 2018 -0500
Date: Tue Nov 6 17:57:51 2018 -0500
Date: Sun Nov 4 01:10:25 2018 -0400
Date: Sun Nov 4 01:10:17 2018 -0400
Date: Sat Nov 3 01:23:08 2018 -0400
Date: Fri Nov 2 21:28:52 2018 -0400
Date: Fri Nov 2 01:38:06 2018 -0400
Date: Fri Nov 2 01:15:48 2018 -0400
Date: Thu Nov 1 15:51:31 2018 -0400
Date: Sun Oct 28 18:30:35 2018 -0400
Date: Thu Oct 25 13:00:42 2018 -0400
Date: Wed Oct 24 16:19:16 2018 -0400
Date: Wed Oct 24 15:57:50 2018 -0400
Date: Wed Oct 24 15:20:41 2018 -0400
Date: Wed Oct 24 14:49:28 2018 -0400
Date: Wed Oct 24 03:15:15 2018 -0400
Date: Wed Oct 24 00:36:36 2018 -0400
Date: Tue Oct 23 23:31:10 2018 -0400
Date: Tue Oct 23 23:21:51 2018 -0400
Date: Tue Oct 23 02:57:34 2018 -0400
Date: Tue Oct 23 01:33:11 2018 -0400
Date: Tue Oct 23 00:50:24 2018 -0400
Date: Mon Oct 22 22:26:09 2018 -0400
Date: Mon Oct 22 15:40:27 2018 -0400
Date: Mon Oct 22 15:20:47 2018 -0400
Date: Mon Oct 22 02:05:09 2018 -0400
Date: Sun Oct 21 16:08:25 2018 -0400
Date: Sun Oct 21 14:14:46 2018 -0400
Date: Sat Oct 20 15:24:51 2018 -0400
Date: Sat Oct 20 12:39:39 2018 -0400
Date: Fri Oct 19 20:49:18 2018 -0400
Date: Fri Oct 19 19:25:15 2018 -0400
Date: Thu Oct 18 22:55:13 2018 -0400
Date: Wed Oct 17 22:00:18 2018 -0400
Date: Wed Oct 17 01:09:55 2018 -0400
Date: Tue Oct 16 19:29:07 2018 -0400
Date: Mon Oct 15 02:57:36 2018 -0400
Date: Mon Oct 15 01:45:40 2018 -0400
Date: Sun Oct 14 20:52:57 2018 -0400
Date: Sun Oct 14 20:14:59 2018 -0400
Thank you, Jason |
|
|
|
....
Antminer-Z9-Mini-NAND-500M-201808311812 firmware
<option value=“500”>blance</option> Change to <option value=“700”>blance</option>？？？？？I restarted and restored the default settings. Is it where I am making mistakes?
Or do we set the configuration file to read-only? How can I set it up?
https://releases.broked.net/antminer-z9-mini-201805262047-500m.tar.gzJust install the batch1 firmware. -j |
|
|
|
...snip...
Some boards can not hold overclock. It is rare, but it happens. Your pictures are also small and blurred so I cannot make out a lot of detail. At least in your second picture the pool side average is 54 to 55. That's even with whatever dips you are seeing. It seems that your miner side average is 55 and your pool side average is 54 -- this looks good. Your pool side dips could be any number of reasons, including random network or pool issues. The real time is not important -- it is your average (and ultimately the pool side average and share submission rate) that is important. If you are staying in the ~55kSol range and your boards are not dropping, then you are in a good place. I have 1 board that I constantly have to reset as it goes to 0 -- it's on one system and does not happen on the others. One day it will be fine at 650, the next day it only wants 643. I cannot explain these things. Jason |
|
|
|
OP first post has a link but i am unable to download....
" Time out Error"
Is there another link or possible method to download?
Where are you trying to download from? What IP address? https://releases.broked.net/There is not anything wrong with the server. Thank you, Jason |
|
|
|
An important note is more important than frequency is your share rate -- which is more important than hash rate.
Share rates depending on difficulty and pool settings and can be different for the same coin depending on where you are mining. This makes the share rate a poor metric except for comparing machines connected to an identical pool. Share rate is what you get paid on. And yes, it varies by luck and such... within a pool, share rate is how I'd measure the effectiveness of a modification... so I stand by my statement.  Jason |
|
|
|
Hi Jason,
I new here and I'd like to OC my Z9.
What should I do??
Thanks in advanced.
Hello, version 2.0d is linked at the beginning of this thread. Thank you. Jason So give me the link fie 2.0d please.... - the server was down on accident for http but was running for https. I apologize. Jason |
|
|
|
Jason! When do we wait “ cooming soon “ with z9 mini? Hi Jason,
I new here and I'd like to OC my Z9.
What should I do??
Thanks in advanced.
Hello, version 2.0d is linked at the beginning of this thread. Thank you. Jason "soon". and not one moment before. Sorry folk, it is what it is -- I'm focused on getting the paid users update first. -j |
|
|
|
Hi Jason,
I new here and I'd like to OC my Z9.
What should I do??
Thanks in advanced.
Hello, version 2.0d is linked at the beginning of this thread. Thank you. Jason |
|
|
|
And very important benefit from voltage control
If i slightly undervolt, i can then overclock to higher frequencies. It has a max amp/power limitation.
My batch 2 could never hold stable anything above 650Mhz. I can do 737Mhz now.
Batch 1 owners should be able to go as high as 800Mhz.
An important note is more important than frequency is your share rate -- which is more important than hash rate. This is not to dismiss your statement on overclocking, as more should be "Better" -- but I agree that their voltage/frequency curves may not be ideal. As I told you in PM, bitmain has attempted to hide the software interfaces for I2C, which means communications with the PIC and voltage controller, as well as fans, are obscured at a minimum. I have some ideas and some pointers on breaking down that wall, but I have to lock down some other components of my work before I move to this more esoteric (yet valuable) work. Thank you, Jason |
|
|
|
I've started collecting data on what is going on with nicehash. I'm going to have to contact them because I do not see obvious explanations for the failures. extranonce is supported by this firmware, it just does not function properly sometimes. I have just submitted the traces to nicehash to try to get some assistance in understanding why this works sometimes, and fails sometimes. Thank you, Jason Maybe this information will be useful: https://www.nicehash.com/help/patches-for-extranoncesubscription?lang=enYeah, I've been through that, and the patches. The extranonce stuff actually appears to be working just fine unless there is something wrong in the calculations of the returned results. I have two traces that are exactly the same (as best I can tell) except for the work requested to be performed. nonce's the same, etc... one works, the other doesn't on NH. When I connect to some of their stratum, they provide extranonce information back, on some they don't.... I may do a set of tests tomorrow without #xnsub and see what happens. Here is an example -- one of NH's servers did provide me an extranonce setting: {"id":null,"method":"mining.set_extranonce","params":["0000908e00"]} Yet that format does not match their specification. It should have a second parameter option of '5' per their own documentation... I just don't know at this point in time. Thank you for providing the link! Jason |
|
|
|
alright, PM sent. I'm reasonably confident I can guess what you did, the software solution is the more challenging thing, as you suggest. Jason |
|
|
|
I've started collecting data on what is going on with nicehash. I'm going to have to contact them because I do not see obvious explanations for the failures. extranonce is supported by this firmware, it just does not function properly sometimes. I have just submitted the traces to nicehash to try to get some assistance in understanding why this works sometimes, and fails sometimes. Thank you, Jason |
|
|
|
The ban, I think, was due to API calls -- it has since been lifted and flypool is communicating with me on the concern now. Jason |
|
|
|
|
Show Posts
