New R-Box Upgrade Kit - page 4. | Bitcointalksearch.org

philipma1957

legendary

Activity: 4116

Merit: 7849

'The right to privacy matters'

Quote from: sidehack on July 08, 2015, 10:13:26 PM

If I'm not able to fit S1 holes on it, they'd still fit on a redrilled S1 heatsink. So, anything you could mount a standard S1 board on would fit two of these side-by-side.
Yes, it'd have a PCIe for power and a USB.

I think I can make some nice heat sinks match this very well.

sidehack

legendary

Activity: 3318

Merit: 1848

Curmudgeonly hardware guy

If I'm not able to fit S1 holes on it, they'd still fit on a redrilled S1 heatsink. So, anything you could mount a standard S1 board on would fit two of these side-by-side.
Yes, it'd have a PCIe for power and a USB.

philipma1957

legendary

Activity: 4116

Merit: 7849

'The right to privacy matters'

Quote from: sidehack on July 08, 2015, 09:21:26 PM

Yes, the heatsink fans and box would be reused. You know, all the really heavy parts that suck to ship and also that I don't have the tools to make. Just like with the other large-scale boards we're making.

Someone requested New R-Box form factor and it's close enough to something I was already going to do that the idea seems appealing. If I wanted to make a new Gridseed, I'd be making a new U3.

Anyone saying "Make the S1 board instead" is not paying enough attention to everything said so far.

so this board will have a pcie jack on it ? along with a usb port?

If I have heatsinks, boxes, fans and an usb hub. along with a pc/controller like a rasp pi

I can jury rig it up correct?

I have an idea for cooling it. I have some long aluminum bars.

___________________________________________________

___________________________________________________

just like above I could do 4 boards attached to it.

sidehack

legendary

Activity: 3318

Merit: 1848

Curmudgeonly hardware guy

Yes, the heatsink fans and box would be reused. You know, all the really heavy parts that suck to ship and also that I don't have the tools to make. Just like with the other large-scale boards we're making.

Someone requested New R-Box form factor and it's close enough to something I was already going to do that the idea seems appealing. If I wanted to make a new Gridseed, I'd be making a new U3.

Anyone saying "Make the S1 board instead" is not paying enough attention to everything said so far.

CrazyGuy

legendary

Activity: 1973

Merit: 1007

I'd definitely be interested for myself, but you probably wouldn't get much more than 50 sales. My suggestion would be to make something that is compatible with as many chassis as possible, which may leave the board looking like swiss cheese. If you have to pick one compatible chassis, go with the S1. S1 is going to be your enthusiast board, with the highest ratio of forum users and tinkerers.

SEGMining

member

Activity: 88

Merit: 10

What about the New R-Box hardware exactly is reused? The heatsink and fans? The box?

I guess what I'm getting at, is, why wouldn't you just design the board with heatsink at that point? By reusing a not-so-popular piece of hardware, I would suspect that you're limiting the potential to mass produce it. Unless you were to design it such that it didn't require a heatsink and operated at ambient temps, like those bitfury kits. Which would be spectacular.

Anyway, If I could vote for which form factor you should reuse in this small miner market, I'd say try to make a Bitcoin only miner out of the old Gridseeds. There are an abundance of those in the wild from my understanding, since it was one of the first scrypt asics on the market. I personally, have about 150 left.

AJRGale

hero member

Activity: 767

Merit: 500

Well, i vote Yes, due to the fact i have some dead New-R-Box's.. or they get trashed/hacked/used in some bitcoin cultist ritual

if the S1 is the same/similar footprint as the NRB board, and people want to go that way, meh, i'd re-drill the sinks to match and see if it fits in the case

philipma1957

legendary

Activity: 4116

Merit: 7849

'The right to privacy matters'

Quote from: sidehack on July 08, 2015, 03:34:10 PM

Yeah, it'd definitely be geared toward people that already have 'em.

So I'm running one right now and it's giving me a reported stats of 40C on 110GH, which since these run about 1W/GH I can assume is at least 110W. That cool, I bet these wouldn't have much trouble cooling 150W or so. With that in mind, we could actually do a full 18-chip TypeZero (not implementing fan speed in hardware, but everything else) in a layout sized for this box. 18 chips would get you 300GH at 300MHz on about 120W, and then the only changes required for S1-style boards is PCB layout. If a new chip comes out, we can retool for the specific footprint and voltage changes, adjust the layout and BOOM we have a TypeZero Spec1. I think that could work. Price would probably be higher since the PCB is both bigger and in smaller quantity, but otherwise ought be not much difference. And if not enough people want them, well I just won't make 'em to sell.

you bring a lot of good points up. getting it to market fast seems very possible.

I have a question when you build the board for it the 18 chips would be placed so the rbox heat sinks and fans do the job?

If that is true and I take a few of these 18 chip boards to use in one of my many cases + heatsinks.

It seems like I would need to provide heatsinks + fans + hub + rasp pi and I could build a 6 board unit doing

1800gh and using 720 watts.

If the boards will use usb to a hub to a controller . my concern is heatsinks and fans.

I have lots of cases heatsinks and fans and hubs and rasp pi's.

If this is true I want 4 to 6 boards.

sidehack

legendary

Activity: 3318

Merit: 1848

Curmudgeonly hardware guy

Yeah, it'd definitely be geared toward people that already have 'em.

So I'm running one right now and it's giving me a reported stats of 40C on 110GH, which since these run about 1W/GH I can assume is at least 110W. That cool, I bet these wouldn't have much trouble cooling 150W or so. With that in mind, we could actually do a full 18-chip TypeZero (not implementing fan speed in hardware, but everything else) in a layout sized for this box. 18 chips would get you 300GH at 300MHz on about 120W, and then the only changes required for S1-style boards is PCB layout. If a new chip comes out, we can retool for the specific footprint and voltage changes, adjust the layout and BOOM we have a TypeZero Spec1. I think that could work. Price would probably be higher since the PCB is both bigger and in smaller quantity, but otherwise ought be not much difference. And if not enough people want them, well I just won't make 'em to sell.

notlist3d

legendary

Activity: 1456

Merit: 1000

Quote from: sidehack on July 08, 2015, 02:33:58 PM

Quote from: notlist3d on July 08, 2015, 02:13:52 PM

The biggest thing I don't like is this would make it where I need the R-box to upgrade. I'm guessing a lot of potential customers do not have a R-box.

I'm sure you will have some sales, but I think a product that does not require a R-box will have higher sales.

If you don't already have a New R-Box, no offense but you're sorta irrelevant to the discussion. Saying "a lot of potential customers don't have R-Box" is saying "a lot of potential customers aren't actually potential customers". I'll probably go ahead and design it, like I said it gets me to about 90% of what I wanted to design anyway. The question is going to be whether or not there's enough demand to try and mass-produce.

I meant no offense or to be irreleant. I personally would like to have one and be a support, as I appreciate and do like your work.

It's just the R-Box still has a decent price for it on sites such as ebay. Paying over 1/2 of upgrade kit's cost for me to acquire a R-box makes this version one that you really need to have the new R-Box to justify.

Either way best of luck with it.

-droid-

legendary

Activity: 1050

Merit: 1001

Quote from: TheRealSteve on July 08, 2015, 02:17:06 PM

Just as slight clarification:

	R-Box	New R-Box
	not it	it

oh damn, this puts me out then Sad

I have the smaller R-Boxes

sidehack

legendary

Activity: 3318

Merit: 1848

Curmudgeonly hardware guy

Quote from: notlist3d on July 08, 2015, 02:13:52 PM

The biggest thing I don't like is this would make it where I need the R-box to upgrade. I'm guessing a lot of potential customers do not have a R-box.

I'm sure you will have some sales, but I think a product that does not require a R-box will have higher sales.

If you don't already have a New R-Box, no offense but you're sorta irrelevant to the discussion. Saying "a lot of potential customers don't have R-Box" is saying "a lot of potential customers aren't actually potential customers". I'll probably go ahead and design it, like I said it gets me to about 90% of what I wanted to design anyway. The question is going to be whether or not there's enough demand to try and mass-produce.

Meech

hero member

Activity: 924

Merit: 1000

I found the original Rbox software a little buggy. Had to reset every few days. I don't know if that carried on with the new Rbox. Sure liked the simple design though. Price would be the factor.

HerbPean

legendary

Activity: 1638

Merit: 1005

Quote from: sidehack on July 08, 2015, 02:07:17 PM

This would be board only; you'd need an R-Box to upgrade.

The decision isn't whether to make this or the S1, but whether to make this in the dead time we're waiting to make the S1. We'd need to do a similar build anyway as an initial test, so the question is whether I should work on an S1-precursor with BM1384 that I design to completion but never sell, or if I should make an R-Box with BM1384 that I design to completion and then sell, and which gets me 90% of what I'd have needed for the S1-precursor anyway. The way I figure, if I'm going to design the thing anyway, what's wrong with trying to make it sellable?

By which I mean, I design and prototype this inside a month. It's then sellable. I take two more months to shift the design to a new-gen chip for TypeZero and it's then sellable. Or I design and prototype this inside a month, do not sell it, then spend two more months with pretty good hardware on my shelf not helping anyone while I redesign for a new-gen chip for TypeZero and then sell it. The TypeZero timeline won't really be affected much at all.

Make a lot of sens.

The only question is when will Bitmain release a new hardware and if it is with a new chip. ( should be this month .... but still who knows)

For me, the R-Box board doesn't interest me much since I have none R-Box. (This is only me)

If both the bitcoin community and yourself (with the design) will benefit more with that path. I don't see why we would try to convince you to not do it

TheRealSteve

hero member

Activity: 686

Merit: 500

FUN > ROI

Just as slight clarification:

	R-Box	New R-Box
	not it	it

wlefever

legendary

Activity: 1173

Merit: 1001

Dang, all I've got are the original R-Boxes sitting at the house. I sold off my "New" R-Box models.

notlist3d

legendary

Activity: 1456

Merit: 1000

The biggest thing I don't like is this would make it where I need the R-box to upgrade. I'm guessing a lot of potential customers do not have a R-box.

I'm sure you will have some sales, but I think a product that does not require a R-box will have higher sales.

-droid-

legendary

Activity: 1050

Merit: 1001

Makes sense. R-Box it is then :-)

sidehack

legendary

Activity: 3318

Merit: 1848

Curmudgeonly hardware guy

This would be board only; you'd need an R-Box to upgrade.

The decision isn't whether to make this or the S1, but whether to make this in the dead time we're waiting to make the S1. We'd need to do a similar build anyway as an initial test, so the question is whether I should work on an S1-precursor with BM1384 that I design to completion but never sell, or if I should make an R-Box with BM1384 that I design to completion and then sell, and which gets me 90% of what I'd have needed for the S1-precursor anyway. The way I figure, if I'm going to design the thing anyway, what's wrong with trying to make it sellable?

By which I mean, I design and prototype this inside a month. It's then sellable. I take two more months to shift the design to a new-gen chip for TypeZero and it's then sellable. Or I design and prototype this inside a month, do not sell it, then spend two more months with pretty good hardware on my shelf not helping anyone while I redesign for a new-gen chip for TypeZero and then sell it. The TypeZero timeline won't really be affected much at all.

HerbPean

legendary

Activity: 1638

Merit: 1005

I would prioritize S1 and S3 board first.

At least for me.

Topic: New R-Box Upgrade Kit - page 4. (Read 4808 times)