They're all 16 chip boards. I misunderstood your logic as entering the frequency yielded a specific hashrate. Yes, the frequency seems to correspond with the number of chips on board. When I switch the boards out they yield the same hashrates so something is going on with the power supply or operating system I'm not sure which yet. My step up transformer that runs my 220V BGA station went down so I had the chips placed professionally at Mastek Innerstep, which by the way if anyone wants professional chip placement I already paid for the stencil I think they want $25 a board for 16 chips but the stencil would work for less chips obviously. PM me for contact info.
I don't have a lot of USB cables so I wan't sure if one was messed up or not. I'm just shooting darts for the difference between my two sets of boards. I figured mabey high data transfer coupled with lead resistance. It doesn't matter, like you said I used a 6 foot PS3 controller cable and nothing changed.
If someone blows up a regulator please post details
In the mean time I'll keep stepping up a 16 chip board to blow one up on purpose and report back. My better thought is to lift the inductor used for the overcurrent protection and take AC and DC measurements there.
With hand or with a sensor, no matter how good is it, there is a fundamental problem of access to junction temperature. You can't measure temperature directly on the chip die, and that's the only one that matters. Temperature on the heatsink may be substantially lower than that on the junction if thermal coupling is poor between the chip and the heatsink. In this case the heatsink will stay cool but the chip will remain hot. Just imagine a heatsink that is not in physical contact with your chip. The same thing happens when it is in physical contact but for some reason thermal coupling isn't very good. Temperature difference also increases when the amount of dissipated power increases, so your estimation of junction temperature will be increasingly less accurate as the amount of power dissipated by chips gets higher. You could of course do all the math if you have thermal models of your system and you correctly estimate their parameters but well that's a very demanding job to do it right.
I've taken these measurements for several companies, you can't do anything about the inside of the chips to the best of my knowledge but most models suggest platinum with 4 wire yields at least +/- .1 C I think. You take a thin strand of platinum that's zero to at least an ice point reference and a span of 100. Then you thermal paste the platinum in between the heatsink and the chip, any exposed platinum must be insulated. Or you can buy a thin film prt also platinum that's already charted and hook it up to an Agilent meter and something like Benchlink.
You build a model of the metal instead of the system so you can reuse it. Really you can just weld (not solder) the tips of any RTD together as long as you have an idea of the composition, and do the same thing as long as you can generate ice point and 100C. It's probably only worth it if you're going to build a lot of boards. I have one set up to control the heat for an infrared reball station in my shed as well, it's pretty accurate. I favor my hand because it's easier. I was just curious how many days you've had them running at 35.
