If the 921's firmware continues to be tweaked/tuned, the power consumption of the 7nm A3206 chips could probably be cut down, making this miner profitable.

It would be nice if their fpga firmware was open source as it used to be on github a couple years ago. What is AVA9_P_SET_SS_PARA_EN? Why even bother posting a changelog?

I've tried emailing the contact listed on their downloads page (love@...) but it's returns a Delivery Failure Notification. Do they have a dev contact?

After running for 24 hours now, I'm thinking it is a power setting. Voltage Level Offset seems to behave as it should now.

---

So after ~34 hours of mining with the new firmware (and adjusting voltage offsets) I can say w/ certainty that the new MM reduces the hashrate. Of the two machines I installed it on, both lost 1.5-2TH/s. I've rolled back to the old firmware and the machines are back at their normal 20TH/s. If I had the machines locally I would test their power consumption with the new firmware running. I'm wondering if the loss of ~1.5TH/s came with a decent reduction in current; can anyone with new firmware running test this and report their findings?

New firmware was just released for 921 (and the 920). I wish they were more descriptive in the changelog. All that's noted is:

Got it working. The issue was resolved by updating to the latest MM921.mcs

Has anyone on this forum purchased an order of these? I'm wondering what their experience with Inno's shipping is like.

Mine is delayed as well due to "awaiting clearing agency review" (apparently in China). Is this the same for you two?

That looks like the correct kernel, but know that the kernel version and firmware version are two different things.
Firmware is listed below Kernel Version as "File System Version."

Did you see that Bitmain has a package to fix S9 firmware update https://shop.bitmain.com/support.htm?pid=007201611260753443104jm60Q6L0639
I've never had to use this but apparently it's an issue for the S9.


 

If you want to try and manually force the miner configuration (pool / worker / password) here are the steps. If something on the controller is corrupt then this may not work. This is only a last resort. I'm assuming you aren't familiar with nix shells from what you said in your first post.

You'll need to ssh into the miner. If you aren't working from a linux or mac, you'll need Cygwin or some other way of running a bash-like shell.
The username is root and the password is admin (unless you have changed the password)
*enter password*

Change to the config directory

Stop bmminer

Change permissions on bmminer.conf so we can edit it manually

Edit the file using vi

Use arrow buttons to go to where you edit text, press "i" and edit the text. If you get stuck using vi just google it—unfortunately there isn't any other editor installed. When you are done editing the pool and worker info, press escape, then colon and type "wq" (which mean write changes and quit vi) and enter.

Change permissions back to original state on bmminer.conf

Start bmminer


See if this gets your pools and workers back

My November 21-30 has just been marked with a UPS shipping label. So I've received both November orders (the other was the LTC payment batch).

What about my first question?

Did you just get these machines? Were they sold by Bitmain or secondhand/used?

You can try installing a new firmware image. Go to https://shop.bitmain.com/support.htm?pid=00720161126080548570Q2tDBXAH065D and depending in which S9 machine you have, download the firmware. If it's not a static 14TH unit, then download the one at the bottom Antminer-S9-all-201708151137-autofreq-user-Update2UBI-NF.tar.gz.

Go to the miner's page under System tab -> Upgrade and under "Flash new firmware image" uncheck "Keep settings" and select the tar.gz firmware you downloaded.
Give it 5-7 minutes and then see if you can enter your pool info.

Has anyone received a shipping notice for Nov 21-30 miners? I'm among the waitlist, just curious what others are hearing.

Look at the System Log tab, that might provide some clues. See what's happening around the time the hashrate is dropping to zero. Maybe there is something happening with dhcpd, etc.

You can ssh into it with username root and password admin.
I doubt the firmware is causing this, at least, if it was installed correctly.
I had an R4 controller fail and I replaced it with an S9 controller, flashing it with the R4 firmware and it works fine. If the image was installed properly, and your controller isn't broken, it should be working. Make sure you don't have anything preventing javascript from running on that page.

The "ant_data=;" you see in the miner configuration page is there because the header has an inline script that begins with The script is fetching cgi data from the config saved in /config/bmminer.conf. For some reason it is failing at that point.

@rs1x, How is your miner's pi connected to the internet? Is it connected directly to your router or does it go to an access point?
The Routes page is missing a gateway, and cgminer's api is telling you it cannot connect to a network. That's the issue; the cause could be many things.

Did Parallel miner sell you the pi? I'm wondering if you have the wrong version of OpenWrt is installed for your pi. You may just want to reflash the pi with the correct image from Canaan.

The sales rate depends largely on their (expected) stock quantity and the demand. Neither factors are publicly available. At the rate they are pumping out miners between now and January, one would expect the difficulty to escalate greatly. This may also be influencing sales.

This next bit is speculation:
History shows that as Bitmain starts producing miners that have better and better hashrates (also with lower failures) and then suddenly starts selling-off large quantities of batches which seem to be a regression towards earlier models' hashrates, they are dumping their stock to prepare for an improved model. 

What size CPU and RAM do I need to run a node of your branch? Would it be feasible to run on a VPS server (i.e. one with the space necessary to hold the blockchain)?

I'm resurrecting an old thread here. I'm curious if anyone with some pcb assembly experience has tried diagnosing this issue. I'm willing to test every inductor, capacitor, resistor, and chip lead with a multimeter—but how do I get this epoxy off cleanly? Is there anyone with feedback? Thanks.

The miner is consuming power because of its fans, which are probably still spinning at full speed due to the internal temp. Once the chips start cooling off, the SoC controller should regulate the fans to a lower speed accordingly (and even go into an idle state iirc). I would hesitate to immediately shut-down the miner before letting it cool off. Try giving it a little while before checking the power consumption after the internet goes out. If your internet is always intermittent, then your solution probably lies in finding a better ISP. That's just my two cents.