Could very well be. The temp sensors were for some reason not handling rated voltage so a fix was applied but we didn't see the issue in bench testing, it showed up later on deployed units. That's probably all it is. An updated firmware can tell the difference between a bad temp sensor and an overheat so the flashing is different for the two conditions but yours probably doesn't have it. We were at least about 50 units in.

If it's flashing continuously it means overheating. Check that the fan is spinning.

Well I don't know about the protocol. VH is working on that. My job is the hardware, and I gotta say, all the hardware improvements to the chip are AWESOME. Like when I saw a spec sheet I was really confused about a lot of things because it was all so different, but when I saw 'em on a board it started making sense and now I'm genuinely impressed at how well thought-out these chips are. It'll mean a lot of design changes from the old stuff, but it also means the new designs will be a heck of a lot easier.

At least the footprint is the same, so the various iterations of stenciling and process changes we figured out for BM1387 will still apply.

Yes I did mean 1387. Had 1397 on the mind lately for, oh, no particular reason.

From 1387 to 1397 they use the same footprint but from what I can tell, entirely different pinouts, power levels, even the protocol has changed somewhat. It's definitely not a drop-in replacement.

That's a work in progress. We might be able to hit 300GH.

Also, the original idea mentioned in this thread has largely petered out. I was looking at a thing with a butt-ton of cheap 1397 running at bottom clock that I could put together quickly, but I think I'll move forward with new chips instead. The new Futurebit gear raises the stakes a little.

So, what advice do you have for improving the efficacy of the device?

Hey Kano - unless you start with a parts budget in the tens of thousands of dollars ON TOP OF what you had for R&D expenses, and don't take in orders until you've already paid for and assembled the first batch out of pocket on the hope that you'll be able to sell at least enough to make back your investment, you're gonna do a preorder first batch.

Also, making something with recent mining chips - I'm working on it. But that's not for jstefanop's thread.

Kinda what I figured. I'm probably the only one crazy enough to do PCBA in-house. Probably saves a bit of money but it sure ain't easy.

As a domestic manufacturer, I can confirm it's not easy to get full batch funding without some kind of pre-order. Very rarely have I been able to do a first batch of something without preorders, even though it sucks having to do that.

jstefanop, do you do production in-house or outsource to a fab?

BobLawblaw, you should have a chat with bitshopper

jstefanop, is the miner portion of this perhaps based around Bitfury RD6?

Guess I need to step up dev on my end. I'm way behind on R&D on account of manufacturing backlog and being shorthanded what with the plague we got going on. VH and I are working on some fun stuff, which includes having recently proved a USB connection like we use on NewPac and Terminus can exceed 4TH, but it takes more time than I have right now.

Then the problem is your controller. If you're running Windows it probably sucks. Might check through the logs here and see what other Windows builds and flags you might test with.

What's the voltage level set at? All 12 chips detected on both miners?

Retail clients all ask the same stupid questions over and over and you have to pay someone to answer them over and over.

Wholesale clients do their homework and ask stupid questions at most one time. Then pay you one giant lump of money. It's way easier on the backend. Even if the margins are lower up-front, you save on support staffing, logistics and headache.

Carefully clean as much of the black adhesive off as you can - probably will require some scraping; it's hard. Find what they call "thermal adhesive", not the forever-slick stuff used on CPUs and the like but one that sets like a glue, and use it to reattach.

USB units, eh? I might know a guy who could cut you a deal.

Nothing real to report, just doing some testing and figured I'd check to see if people would care if that test turned into gear.

May be something, may be nothing. Depends on markets and other projects.

If you have a NewPac, then you want the NewPac support thread and also NewPac frequency flags.

Page one, post one, FAQ section. Try this and see what happens.

cgminer.exe-o stratum+tcp://btc-us.f2pool.com:3333 -u kadet003.001 -p 0000000000000 x --gekko-newpacfreq 150 --gekko-newpac-boost

It's possible the erroneous 'x' in your command line is causing the error. There should only be one contiguous (no whitespace) section of text after '-p' and before the next command or you'll get a Command Not Recognized error.

It's also possible that having mis-spelled the frequency command (--gekko-newpac-freq) is your problem, and would cause the same type error.

Additionally and it's not significant to your problem, but for NewPacs you'll want the NewPac/R606 support thread. This one is for the previous generation of devices using an older chip.