Topic: GekkoScience BM1384 Project Development Discussion - page 77. (Read 146665 times)

The project for a (ultra) low power stickminer is great.

 For example, if BM1384 is 0,3 W per GH, it's can be ~5 W for 15-20 GH :-). USB 2 power supply can go up to 10 W, then the maximum will be 33-40 GH/s, perhaps 50 GH (with higher voltage).
The important thing for personnal use and learning/hobby is that it must be cheap and sold per individualy or in pairs (but not by 500 units ;-)).

Thanks for the job :-).

Im working in x Ray.  Im a biotechnologie engineer so if you want Send me a pcb I will image it.

Cheer
Valkir

Got a friend with an X Ray machine?

First Protos yam 

That's unfortunate - though at least you know the basic design should work, so unless the PCBs didn't come out right after all, at least odds are it's in the soldering somewhere.  If you're going to get a microscope, also have a look for an endoscope (some have endoscope attachments, no experience with those).  Makes it easier to get shallow angle views.

Let us know if your comfort burger fund needs replenishing

Well, so far the prototypes are tickin' me off. I have two fully populated and neither works properly. One seems to be powering the chip properly, but the chip's never really lighting up. It's getting serial data but not responding at all.
The other, I honestly don't know what's up. Sometimes it enumerates in cgminer and starts to hash around 300MH/s, usually craps out after a few seconds but a couple times it's run for minutes and once or twice I briefly saw the hashrate over a few GH/s. If you look at the 1BURGER stats on Eligius you'll see a slight lump recently; most of that is from a U2 I was using to test the powered hub setup but the lone share about 19:00 was one single 128-diff that got submitted while the thing was temporarily approximately working. So TECHNICALLY a success. A stupid one.

I'll keep working over the weekend and see if I can't get things ironed out. Those BM1384 are really hard to get stuck down right, and since the pads don't wrap at all there's no good way to inspect them for proper solder joints. I need to get a friggin' microscope or something.

sidehack

Our human heating systems are a 3 loop system with a small (3-5 gal) expansion tank operating at 1 bar +5psi/-0psi gauge. Basically the system runs at atmospheric pressure with the +5psi in the system plumbing being caused by fluidic resistance in the plumbing. One loop continuously circulates fluid thru the miner(s)/immersion tank contributing heat to the expansion tank. The second loop pump is driven by a room thermostat (on/off) i.e. it runs when the thermostat demands heat and flows heated fluid into the PEX in the floor or into a heating coil mounted in the plenum of a forced air furnace when the fan in running, and the third loop pump is driven by an adjustable temp sensor on the tank which circulates fluid thru PEX buried outside the building (we like to put it underneath the driveway and sidewalks to minimize snow shoveling in the winter, one could use it to preheat the water in a hot tub also). Typically the system is sized @ 1/8 BTU capacity of an identical system with a restive element boiler. This sizing minimizes the third loop pump operation. All the pumps are "run of the mill" hydronic heating system circulators. At the high point in the system we install a automatic "burper". All PEX in the system is of the oxygen diffusion barrier type (this minimizes rust contamination/creation from cast iron pumps, fittings etc.). A safety sensor on the expansion tank will disrupt the power to the miner(s) if tank temperature becomes excessive. The system is tweaked to maintain a ~40C expansion tank temperature with the automatically resetting high limit miner shutdown set at ~60C. Careful attention is paid to tubing sizing to minimize fluidic resistance and maximize flow rates for optimal heat transfer in all zones of the system. Because we operate the system a lower temperature and pressure than a typical restive element heated system the system typically needs larger sized piping/tubing to transfer the heat quantity desired. Oh yeah and the fluid is a 50/50 mixture of distilled water and green antifreeze.

It's amazing how few miners it actually takes to heat a structure. One of the installations has 16 manifolded C1's heating the slab floor of a reasonably well insulated 40'x30' pole barn.

Well enough of this off topic rant.

Can't wait to get my hands on one of your boards.

Is the "nice hardware guy" named tophack maybe?   

Thanks for your efforts to describe and be transparent about what you are doing, including your less than successful stops along the way. It's quite refreshing.

AJRGale - Many of the server PSU's have a Vout adjustment and can easily be tweaked +- 1V and probably a lot more than that depending on the particular supply in question.

Most of the buck converters I've worked with want at least a Vin = Vout + .8V with a more typical minimum of Vin = Vout + 1.2V.
This is where SEPIC (Single-ended primary-inductor converter) and/or CUK converters come into their own with a wider Vin range at the expense of component count and circuit complexity. Where Vin can be either above or below Vout within an acceptable design range. Typically, these converters (DC-DC) have a practical maximum current design limit owing to massiveness of the inductors at high current outputs.

Several of the chip manufacturers have single chip SEPIC/CUK controllers on the market. But they still require a butt load of external power components. Linear, Fairchild, and TI to name but a few.

http://en.wikipedia.org/wiki/Single-ended_primary-inductor_converter
http://en.wikipedia.org/wiki/%C4%86uk_converter

The buck in any PSU worth anything is running a transformer-based forward converter behind an active PFC boost converter. The active PFC uses an inductor to push regulated input current uphill into a bulk storage capacitor, adjusting on a cycle-by-cycle basis to keep the driven current proportional to the input voltage (so it appears to your power lines to be a resistive load with a power factor of 1), and adjusting the I/V ratio between every mains cycle based on how much energy needs to be pushed into the bulk cap to keep the voltage relatively constant even if output power requirements fluctuate.

What this means is, the buck part is always taking from a constant voltage and shifting down to a constant voltage. The part that works better with an increased mains voltage is the PFC boost circuit. They're usually taking the mains up to something like 380VDC, which it takes a lot less push to get that from 340Vpeak rectified mains than it does from 170Vpeak. Switching duty cycle is reduced, and more importantly the overall current is reduced, which decreases your conduction losses (which are proportional to current squared).

If you had something outputting a 350-400VDC bus you could wire it straight into the bulk cap, bypass the PFC entirely, and probably run your PSU at about 95% efficient.

Yes, the S5 has 30 chips. This was discussed extensively in the first few pages of the S5 thread. It gives 0.8Vcore average per chip so they operate up to 400MHz typical.

If you're warming ambient conditions for humans while outside conditions change, you might want an adjustable thermostat. Something that could restart cgminer occasionally with changing parameters to alter the heat generation from your miners (and calculating optimally profitable hashrates for the desired heat output) would be pretty slick.

Additionally, I got the Compac test PCBs in about 7PM. There's exactly one UPS driver on our route that has realized we're not here at 9AM, and the other fifteen that seem to cycle in and out appear at exactly 9AM to leave door tags every single time. And then they don't come back by on their way to the depot at the end of the day (which, by the way, we're one mile from and they drive past us anyway). I should request we get put on the same route as the tractor store next door that gets daily deliveries in the afternoons. Anyway if we ever want to actually receive our stuff we have to wait until the day's end when the truck has driven past our building probably several times without stopping and pick it up from the depot. So that's why it's midnight and I'm still working on the prototype instead of having it going ages ago. I'm also getting annoyed because when you couple the stupidly laggy browser with the junk Novak dumped in my keyboard today it's really freakin' hard to type anything. I will not be posting pictures tonight. Y'all just gotta wait and remember that I'm the curmudgeonly hardware guy, not the nice one.

16V Jap Caps with a low ESR. Good stuff maynard.

Ya' know I looked at a S5 miner stats page as I was writing that post and saw 32. Just went back and looked at the same page still open in my browser and sure as Sh*% it indicates 30 chips. Must be going senile.

Ya' got a point on the remote S/W adjustment, although once corn-figured it'll likely never get changed.
One of the beauties of immersion cooling is the ability to R&R components while the rest of the machine is up and running. Hence the inquiry for edge card connectors. Hot swappable hash boards/PSU's.
You're spot on about bucks liking a narrow Vin-Vout. Why I'm always amazed that many of the server PSU's we use run more efficiently off of 240VAC than 120VAC.

the question is, how many designers worked out that the system may be running a lower voltage then the 12v out of the PCIe power plugs?
some how i don't thing my miners will be happy with less then 11v.. i have no variable voltage supply to test though.. i could rig up 7v out of my PSU with some random PCIe plug i cut off a dead psu :-?

unless you know a simple voltage dropper that could handle many amps, and be able to be slapped together on a breadboard?

none the less, a simple mosfet and resistor (well you know what i mean, get it form 12v to 5v/3.3v/1.2v/etc.) is cheap enough to keep 12v "clean" without spending another $5-10/per Vin on components to step up/down the voltage in.

come to think of it, im going to tare down the old "new r-box" and poke'n'prod them voltage lines, and maybe throw a rs232 input on the chips tx lines. (watch me let out that blue magic smoke on something), i really want to see where its failure point is (bad solder on one chip?).. also i want to get messy with all that white ceramic thermal compound... i ment, *ehem*

A well-designed miner will run with an input voltage below 12V. Depends on the topology, but the Spec1 should be designed to work off a Vin probably 8V or so. Spec2 may not, since its power requirements are different. A buck regulator also works more efficiently with a smaller Vin-Vout difference so running at a lower Vin is actually better (to a point; reduced Vin means increased Iin which can cause increased wire losses). For a well-designed miner.

Car battery, bad idea, they are more like capacitors then batteries, they are designed for large amounts of current at a small time scale.
SLA batteries on the other hand are OK, just don't let them discharge less then 25% of the stored capacity, they don't last long there after (think UPS's been discharged before power gets back online. you basically need to replace them)

Running miners off batteries can be done, i have looked into running one of my "new r-box" off solar + batteries, but i need a lot of batteries to keep up with 1000Wh (10 hours of 100W pull) and need many more solar panels to charge and run the miner in the short time there is sunlight..

and also voltage drop from discharging cells may not work with miners, the miners have a small voltage tolerance 12v+/-5% and cells drop to 9V at ~75% charge



Speaking of using the miners as heaters, im using my 2 miners atm to warm my lounge room, im thinking of throwing in a temperature sensor on the fans so they both keep a certain level of heat out, throw a potentiometer on it to change the heat sensitivity, so drop it for winter, raise it up for summer

Almost all of your questions have been asked and answered more than once already. Most of them are addressed in the first post.

The chip I'm looking over for the TypeZero power systems should take in over 12V easily. Limitations might come from accessory components (management rail power, input capacitors, etc) but I'll probably make it work off at least 16V safely.

Also, the S5 has 30 chips not 32. And with software-adjustable clocking, you don't have to take the machine apart and put it back together to make changes, which is probably good for an industrial or immersion install.

At the hash board level, clock & volt adjustment is about all I need.
Hell, jumpers or dip switches would work for me to set clock and CoreV.
Call me old fashioned but I like dip switches I can look at versus S/W.
What else can one do at the hash board level?

How close to having proto boards made for the Spec1 are you?
BTW, I got a dozen S1's that are turned off right now. Will be interested in 24 Spec1's.
What'ca figure the hash rate and power consumption will be on the Spec1's?
The Spec1's would be easily waterblockable (is that even a word?) with the waterblocks from syscooling, me thinks.

So the Spec2 will be similar to the S5 hash board (32 chips) with a Prisma formfactor and a USB interface?
Guessing the Spec2 board will draw ~240W @ 350MHz clock?

Do the DC-DC's you're thinking about using take a higher Vin than 12.2Vdc?
Just thinking that if the DC-DC's would live with a Vin range of 12.2 to 15.0 one could run them off a 12V lead-acid cell.

Hey has anybody out there tried running a miner straight off a 12V deep cycle or car battery (@ 12.75V) ?

Dinner time, l8r folks.

Dev's gonna depend on feature set. Right now we're looking to refit the Prisma formfactor with a 30-chip board that'll be software clock and voltable. The primary goal is the Spec1 board to fit on S1 chassis, but once that design is done it won't take much to extend it to the Spec2 Prisma refit board. If you want higher density I can probably work on condensing it with more chips per node and a higher current power system. The easiest thing will be a straight string topology, but for your needs the ability to alter the operating points (core voltage and clock) could be worked into an adjustable thermostat.

I do like the idea of a miner-based heating system.

For waterblocks, we've had good luck using the MX-4 thermal compound as it is electrically inert and has a thermal transfer co-efficient 6-8 times that of Arctic Silver.

Now you have me drooling over a Prisma 1.0/2.0 refit board. And we would be VERY interested in a board like that that didn't use a pre-historic propriety controller.

Yeah that's what I anticipated you'd say about socketing. Don't know till you ask.

Just outta' curiosity, guestimate - ball park - off the top of your head - a number I can't/won't hold you too, the design/dev cost of a high density Prisma style board using the BM1384's with edge card connectors and a matching backplane with DPS-2000 style PSU connectors. Dimensionality is unimportant as we can build a tank around anything. [email protected] Don't spend a ton of time on this just curious how many zeros are to the left of the decimal point a design like that would have.

We had/have 3 installations of "solid state" hot water boilers operating for comfort heating (at a profit) all this last winter in and around my local area in northern MN. All in all, an extremely successful test/prototyping and are working on a final design. Over designed one installation to the point that when it was -20F they had windows open dissipating the excess heat with everybody in tee-shirts.

Again, before I spent too much time doing board design I'd want some kind of a commitment from Bitmain for chip supply. Or maybe a licensing agreement to manufacture chips . . . . .
Just sayin'

One issue with a board-cooled string design is that each node in the string is at a different ground potential. If there's any short through the heatsink from the ground plane of one node to another, the best case is your PSU trips out and the worst case is stuff catches on fire.
For an anecdotal reliability comparison, the Prisma used board-cooled string and the S5 uses top-cooled string. Not many S5 catch on fire. As much as it'd be easier to manufacture and probably easier to install, I don't like the idea of a board-cooled string because it also makes it easier to fail.

I was thinking of mentioning to you that we're looking into a Prisma refit board with the BM1384 - same size and stock power level. I know you were playing with immersion cooling on those several months ago.

Socketing seems to be an unlikely way to go. Given that (with the exception of BitFury R1 and R2 chips) nobody has ever made a pin-compatible next-gen chip, given that sockets for the fairly customized footprint may not exist, given that if they do (or can) exist they'll cost more than the chips themselves and the chips themselves would cost more than the PCB and every other part on it, it's probably not feasible.

My primary interest is still to build things geared toward individual miners. If one of the big guys building industrial stuff wants to repurpose for immersion cooling, that's fine. But unless the big guy specifically commissions a board design and pays all the dev costs, I'm probably not going to do anything specifically for his benefit.

Been outta' the loop for awhile, long time no talk.
I hope you have better luck getting chips outta' Bitmain.
We've wanted to buy some for quite a while for a similar project (high chip density board for immersion cooling) but Bitmain never seems to want to release/sell them in quantity.
Like you (and others) we were hoping that the BE300 would come to fruition. Alas I fear that's not going to be the case in the near term.

In our immersion cooling world, a high hashing-chip density board consuming 1Kw-2Kw with edge card connectors mating to a backplane for power and USB interface, designed to dissipate the heat through the pad with all the components on the opposite side would be PRIMO. A design like that would lend itself to air (heatsink) or waterblock cooling solutions as well. If the backplane had an appropriate number of connectors like your DPS-2000 adapter one could keep the power and ground planes short and thick.

Another concept might be to socket the hash chips assuming that the BM1384 and BM1386 remain pin compatible.

Anyway, you'se guys are the brilliant ones and I'm sure whatever you come up with will be fricking amazing.

BTW, I have contacts inside 3M that may be able to apply a little external corporate pressure to prod Bitmain to release/sell chips to a board manufacturer that makes an immersion coolable board cuz' 3M, obviously, wants to sell their Novec engineered fluids, adhesives, and sealants. Let me know if I should make a couple of calls to "grease the wheels of motion".