BM1384 Pod Miner plus trade-in/recycling - an interest and feasibility poll

[sidehack]

I won't be building ethernet directly into anything anytime soon. Maybe if I end up building a rack miner (probably using 6 or 8 TypeZero boards) it'll have built-in ethernet, but that'd be because of an internal Pi-like controller tying everything together not because I put an ethernet-capable microcontroller on anything.

[2112]

(and probably also code work, since the guy who's supposed to be helping hasn't gotten back to me in 3 weeks)

I handed it to Novak for firmware but he never got around to it and then he left for a freakin' sweet rocket scientist job

I understand your problems, but let us help you.

Maybe somebody who's reading those threads has some experience working, or desire to work, ith those microcontrollers.

Originally, you had in mind megaAVR with assembly-level coding.

Later, you've mentioned some NXP part (presumably programmable in C).

Let us know what exact part numbers you have in mind for your design. Post somewhere or link to the code you may already have on your mind. People here may be able to help you. It doesn't cost you much. We (all of us, readers) may be able to constructively help you.

I, personally, am not making any promises to deliver on some deadline. I'm mostly curious. But I do have significant experience and I'm certainly able to help with various design decisions.

But in order for us to help you, we have to know what microcontroller choices you've considered and what are your preferred choices for the applicable toolchain.

[sidehack]

Well I haven't made any design or toolchain choices. All that was Novak's job, so I'm trusting his decisions and picking up where he left off. I do actually have degrees in embedded and programming so I'm capable, just out of practice. And either way, it's not happening for a BM1384 pod.

[2112]

Well I haven't made any design or toolchain choices. All that was Novak's job, so I'm trusting his decisions and picking up where he left off. I do actually have degrees in embedded and programming so I'm capable, just out of practice. And either way, it's not happening for a BM1384 pod.

I'm not buying your answer.

You must have looked over his shoulder. Did he paid for his toolchain or used a free (limited) one? What was his debugging box that he used?

I know that Atmel (now acquired by Microchip) offers Windows Atmel Studio 7 unlimited for free covering AVR, AVR32 and ARM32 (raw metal, not Linux).

You can post here or on any other thread of your choice. It is fairly obvious that the microcontroller and toolchain choice will be brain bandwidth (wetware) limited, so please post what were the tools that you've used and are comfortable with. (Plural "you", meaning "sidehack" & "Novak" & whoever else works with singular you "sidehack).

[WorkHorse78]

What is the overhead to use a USB controller in the Pod Miner?

Well part of the advantage is I don't have to write a driver for it, which is good since I'm already doing everything else (and probably also code work, since the guy who's supposed to be helping hasn't gotten back to me in 3 weeks) for a couple other projects right now - on top of manufacturing and hosting. If someone wants to port over S5 code to work with a USB/UART controller, this guy would hook right up to it - except you wouldn't have a temp sensor since that ties in through I2C on the 18-pin header. I had a board design about Christmas with all that stuff integrated, and I handed it to Novak for firmware but he never got around to it and then he left for a freakin' sweet rocket scientist job so it's not gonna get done in any time frame where the BM1384 is even really viable as a hobby miner, not when I have A3218 and BitFury projects already underway.

I was hoping the usability would be just as good as the Compac.  Could find a chip that supports both USB and Ethernet but adding a microcontroller can be heavy with programming and finding one that is easy to work with(Software version 7.0?).  I could understand by keeping it simple and the cost down by not having 3-5 different chips to support it that the Beaglebone could already do.

I'm sorry that you lost the resource of someone who could do firmware and software.

Why are the Compac's selling out?

A3218 and BitFury projects already underway?  Should we see a different version of this that uses less than 50watts?  With Next-Gen Chips?

One of my computers has issues with USB miners (U3, nanofury stick, compac, you name it), I would generally prefer miners to use ethernet connections.  Not sure if I'm in the minority or majority on this one...

Some computer mainboards either use cheap USB chipsets and hope that they work or the driver/firmware support may not be 99%.   Everyone only needs to use USB for just a mouse and keyboard right?

...And either way, it's not happening for a BM1384 pod.

I'm not buying your answer.

So, Novak and I are working on a proof-of-concept device which we think has potential.

October 8th of 2015... Without Novak working on this idea, I'm pretty sure that it's dead unless if someone else can make up for 5 months of development?

[sidehack]

I've been saying it for at least a month now, but there will be no mass-produced standalone BM1384 pod. If I had one in December, maybe January it could have worked. But I didn't. So I won't.

Compacs are selling out because people want them. They're sold out and no longer available from me because I'm tired of building them. I don't know if anyone could estimate how much manual work goes into assembling one of them, but when you make them with pulled chips the amount of manual work gets tripled. Now consider how much work would go into it if it had 8 pulled chips. Now consider how much of the price would actually be profit, from which I might actually get paid a bit for that work. See why it's not worth it?

Novak chose an LPC11u23 ARM, which is USB-capable and has a variety of ports and ADCs. It's in the same family as the ARM on the Avalon Nano and Ava6 control board (which I believe is LPC11u14? Going from memory on all these part numbers). I've worked with 8051 programmed from Windows, but not ARM programmed from Linux. He had a Linux toolchain set up, and before he left I believe he shifted it over to one of the general shop machines rather than his own box. I figure on utilizing the USB bootloader supported by the chip, so I can write code to it without requiring extra hardware - that also makes firmware updates in the future more possible.

So, I reckon if someone is pretty good with ARM, and maybe also c for some cgminer drivers, and has time to work on it in the next couple weeks, let me know.

[sidehack]

So at some point I mentioned a BM1384 Pod prototype that worked off an S5 controller. Well today I finally got around to putting it together and guess what, it works. The same frustration with power that plagued the first prototype some time ago was still present, but this time I actually figured out the problem and solved it. So. The thing's currently sitting at 350MHz. And having that problem solved takes away some hesitation to future designs with the same power setup.

It's quitting time now, so I'll put up a picture tomorrow. But for now, it's mining at Eligius and should average 154GH.

I still have a bit of testing to do; there's a node voltage display circuit built in that I haven't populated yet so I don't know if it works. I figured I'd get base functionality first.

I can put together about 8 more of these; if people actually want them maybe I'll do something about it.

Tell you what, this has actually been a pretty good week for prototypes. But the rest of that story will have to wait for another day.

[EDIT] And of course as soon as I say all that, I poke the wrong thing and now it's down again. Grumble grumble fix it tomorrow.

[gt_addict]

I'm interested aswell

[sidehack]

Well, got it back to working after whatever happened last night. I think my poking around toasted the buck chip - well, not completely toasted. Screwed up the current trip level detection I think, because it would start but as soon as it tried to mine and detected an appreciable load the buck would reset.
So now I'm back to functional. However, the bigger problem I thought I solved is now a problem again and it's making me question how it was solved yesterday. I might have time to get back to it today but no guarantees. Hopefully I get it ironed out because I was really looking forward to using that power circuit on some other stuff for you guys.

If anyone's wondering, I'm using a pulled TPS53355 to generate about 2.6V across four nodes of BM1384. The '53 has an internal 5V LDO which drives internal switching circuitry and powers the switching FET gate drives. When the load increases, this 5V starts to sag. It's not a problem on any of the commercial miners I've looked at, and I've replicated some of those circuits in test but it still misbehaves on mine. I noticed some inordinate gate drive currents dragging it down, so a resistor inline with the boost capacitor solved some of that. Yesterday it started working beautifully after increasing the boost capacitor and 5V bypass cap, but now with the same parts in place it's still sagging so I'm a bit perplexed. Anyway it's running right now at about 175MHz because if I push it farther the 5V sags below 4V and the buck resets. I'm sure it's something very trivial that I've overlooked.

[NotFuzzyWarm]

<snip>
If anyone's wondering, I'm using a pulled TPS53355 to generate about 2.6V across four nodes of BM1384. The '53 has an internal 5V LDO which drives internal switching circuitry and powers the switching FET gate drives. When the load increases, this 5V starts to sag. It's not a problem on any of the commercial miners I've looked at, and I've replicated some of those circuits in test but it still misbehaves on mine. I noticed some inordinate gate drive currents dragging it down, so a resistor inline with the boost capacitor solved some of that. Yesterday it started working beautifully after increasing the boost capacitor and 5V bypass cap, but now with the same parts in place it's still sagging so I'm a bit perplexed. Anyway it's running right now at about 175MHz because if I push it farther the 5V sags below 4V and the buck resets. I'm sure it's something very trivial that I've overlooked.

pm me a pic of yer circuit. Have you looked at the gate drives with a scope? I'm guessing either the switching times are off or possibly oscillations.

And of course definitely interested in one

[sidehack]

It's a monolithic with internal FETs.

[CjMapope]

So at some point I mentioned a BM1384 Pod prototype that worked off an S5 controller. Well today I finally got around to putting it together and guess what, it works. The same frustration with power that plagued the first prototype some time ago was still present, but this time I actually figured out the problem and solved it. So. The thing's currently sitting at 350MHz. And having that problem solved takes away some hesitation to future designs with the same power setup.

It's quitting time now, so I'll put up a picture tomorrow. But for now, it's mining at Eligius and should average 154GH.

I still have a bit of testing to do; there's a node voltage display circuit built in that I haven't populated yet so I don't know if it works. I figured I'd get base functionality first.

I can put together about 8 more of these; if people actually want them maybe I'll do something about it.

Tell you what, this has actually been a pretty good week for prototypes. But the rest of that story will have to wait for another day.

[EDIT] And of course as soon as I say all that, I poke the wrong thing and now it's down again. Grumble grumble fix it tomorrow.

omg yay! where do i send my btc?!?! haha
no but mark me down for atleast 2, 1 will do if i can get it. or 3 or 4....
great job man! your keeping the home miners in the game!

[Mikestang]

I've been saying it for at least a month now, but there will be no mass-produced standalone BM1384 pod.

Well today I finally got around to putting it together and guess what, it works.

Is what you are working on here just for fun now?  I think it's really cool that you have a working unit, but I am unsure what the end game is now that there are no plans to develop it for production.  Thanks.

[sidehack]

I can put together about 8 more of these; if people actually want them maybe I'll do something about it.

I was really looking forward to using that power circuit on some other stuff for you guys.

There is no endgame. It's a one-off that I finally had time to mess with since I've been solidly busy pretty much every day for the last month and some. I just figured, if I can make it work right and I have enough parts to build half a dozen I'll see if folks want half a dozen. If nothing else it'd pay for the parts and be another dev step toward a product that simultaneously doesn't suck and is released at a reasonable timeline.

[Mikestang]

Got it, thanks for the explanation.  I bet you could auction off the ones you do make, I'm sure there will be a demand for the limited amount you produce.

[sidehack]

That's kinda what I'm thinking. I'm building one fresh now based on the properly working configuration from yesterday. I gotta cut out early though so I doubt I'll get to fully test it.

If I can isolate whatever accident made yesterday's work perfectly for several hours and I have screen shots to prove it, I'll probably work on a similar configuration using a newer chip sometime in the near future.

By the way, this board integrates the S5 temp sensor so that works too. Fan control, not so much.

[NotFuzzyWarm]

Fun things I've found that make circuits mysteriously work/not work:
probe loading
gentle breezes from cover off vs cover on - and not exactly related to local internal ambient temps
bright light when illuminating for close inspection

What a fun world eh?

[johnny5]

I can put together about 8 more of these; if people actually want them maybe I'll do something about it.

Tell you what, this has actually been a pretty good week for prototypes. But the rest of that story will have to wait for another day.

Would love to get one (or more) of these. I'm nearly certain people definitely actually want these