Block Erupter USB - Overclocking/ hacking ?

[SRoulette]

Hello,
I will shortly be ordering a usb key in a group order.
I dont expect to make my money back, I actually want to use this device to mine altcoins

Has anyone managed to overclock this chip ? is there any room for hardware hacking ?
I have seen the upgraded heatsinks people are doing and it got me thinking what else is possible.

Thanks.

[1715147776]

1715147776

[1715147776]

1715147776

[Photon939]

It uses a fixed frequency crystal oscillator, you would have to desolder that and wire in an adjustable clock source to try any sort of overclocking.

[SRoulette]

It uses a fixed frequency crystal oscillator, you would have to desolder that and wire in an adjustable clock source to try any sort of overclocking.

Im game 
I shall wait for  it to arrive and see if I cant upgrade the crystal

[SteveB]

I don't actually own one of these as they are way overpriced even at the new lower price.
But every screenshot I have seen show lots of hardware errors when mining.
Isn't that a sign that they are already overclocked way to high?

[SRoulette]

I don't actually own one of these as they are way overpriced even at the new lower price.
But every screenshot I have seen show lots of hardware errors when mining.
Isn't that a sign that they are already overclocked way to high?

It could be a simple power supply issue as well, I plan to feed it power independently of the usb key.
"FOR SCIENCE"

[tom_o]

I don't actually own one of these as they are way overpriced even at the new lower price.
But every screenshot I have seen show lots of hardware errors when mining.
Isn't that a sign that they are already overclocked way to high?

It could be a simple power supply issue as well, I plan to feed it power independently of the usb key.
"FOR SCIENCE"

You can only mine SHA256 coins btw, not other alts. As for this - the chips themselves run on <2v so it may be a regulator limit and supplying more amperage on the USB socket/5v line may not actually help at all.

Anyway, would anyone with one already care to read off the crystal speed? Also the voltage regulator part number would be interesting, voltage reading even better, and maybe a current reading of what is being pulled over the USB. Then we could work out exactly how much headroom we have. An unlimited current supply on the 5v might help if the regulator has a decent amount of headroom.

[nubbins]

kano posted this a while ago:

[nubbins]

...although I feel like the new ones say Rev. 3.00 on the back...

[bobsmith652]

There is an 11 and a 12 MHz surface mount crystal on the Rev3.  The 11MHz is the 3rd device from the left at the top in that photo that has the flash whitewashing the markings out.  Not sure how that makes 333MHz...

Even i could match the wonky alignment on the crystals in the photo (the Rev3 shares this trait)

[tom_o]

Looks like someone has blanked out the power mosfet though, gonna have a search for third party images...

[Trongersoll]

it has been bugging me, anyone know what that "4R7" thing is? it is like the highest part on there.

[Photon939]

it has been bugging me, anyone know what that "4R7" thing is? it is like the highest part on their.

It's an inductor used in the power supply circuit for the ASIC chip

[bobsmith652]

Is it the one in the bottom right with 2 legs on one pad?  I don't think it is scrubbed - it's just not photogenic.  Didn't pick up with flash, so this one is with a torch.  I can try daylight on a windowsill if you can't see it in this. (click to open in new window, click twice more to zoom)



The 11MHz is actually 11.0592MHz.  The 12 is 12.000 I think it says KD905L on it under the speed.

[tom_o]

That 11(ish) one is connected directly to the USB interface chip so that one is far less likely to be the base for the clock speed. As for that power supply, looks to be this:
http://www.aosmd.com/pdfs/datasheet/AOZ1021AI.pdf

That part is rated at 3a output, the USB spec says a max of 2.5watts so as long as the chip is supplied at >1v there should be a bit of headroom left, especially once you take into consideration that some of the power draw is the buck conversion inefficiencies not the ASIC itself. Would definitely require a re-wired/modded USB hub though.

To work out the exact Vout to the ASIC either measure voltage out across pin 3 and the 'far' side of R1 whilst running (riskier), or measure the resistance values of R1+R2 whilst off and check against that datasheet. (pinouts and supporting circuitry p12/13, voltage programming table p7)

[bobsmith652]

That 11(ish) one is connected directly to the USB interface chip so that one is far less likely to be the base for the clock speed. As for that power supply, looks to be this:
http://www.aosmd.com/pdfs/datasheet/AOZ1021AI.pdf

That part is rated at 3a output, the USB spec says a max of 2.5watts so as long as the chip is supplied at >1v there should be a bit of headroom left, especially once you take into consideration that some of the power draw is the buck conversion inefficiencies not the ASIC itself. Would definitely require a re-wired/modded USB hub though.

To work out the exact Vout to the ASIC either measure voltage out across pin 3 and the 'far' side of R1 whilst running (riskier), or measure the resistance values of R1+R2 whilst off and check against that datasheet. (pinouts and supporting circuitry p12/13, voltage programming table p7)

By risky, do you mean I might touch the wrong thing by accident while trying to measure, or something else?

I will get my meter out when I get home.

As for the hub - I can festoon some extra wire around the insides between the power jack and the USB sockets on mine very easily.

Bob

[tom_o]

The resistance offered by the multimeter is usually 1 or 10megaohm so it shouldn't disrupt the circuit at all. Just yeah try not to touch the wrong thing/slip whilst measuring! If you did both measurements you could check the voltage sag compared to what it has been set to with the resistor values.

Also need to work out how the clocking works/clock multiplication ratio. It is probably  be the 11.0592MHz( x 30 ) going into "IC3" which I can't seem to find a datasheet for. 12mhz is normally used for USB but is connected to the ASIC directly...

And yeah that's what I meant on the hub, just be careful as there should be decoupling capacitors in there that would complicate things; make sure you're only reinforcing existing connections and not making new ones!

[bobsmith652]

I'll try to get both sets of measurements.

I was wondering if any of the test points at the end of the unit could be used to supply power.  I have seen a picture with a connector fitted there.  I will see if there is 5V on any of them.

I also found this:
http://www.asicminer.co/about.html
Stating that they can be run at a higher speeds with a 0.1V increase.

Bob

[tom_o]

I'll try to get both sets of measurements.

I was wondering if any of the test points at the end of the unit could be used to supply power.  I have seen a picture with a connector fitted there.  I will see if there is 5V on any of them.

I also found this:
http://www.asicminer.co/about.html
Stating that they can be run at a higher speeds with a 0.1V increase.

Bob

You'll definately need some better heatsinking btw. Definately look into getting something like these for both sides of the device, ASIC and VRM. Might need to cut them up on the component side though.
 http://dx.com/p/aluminum-cooler-heat-sink-chipset-for-ram-memory-w-type-8-piece-set-35816

 If they overclock enough I might buy some if ROI looks likely, was planning on mounting them on an old CPU cooler!

[bobsmith652]

The Erupter circuit board is on an aluminum substrate, and the coloured bit on the back is a screw-on aluminium heat spreader, with some kind of rubbery (I assume silicone) sheet between.

I already have some copper heat sinks on the way from China (ebay items 290928970602 and 390610210994 look like the ones).  If you order from DX, it will probably take 3 weeks - If you order from Chinese Ebayers, it usually takes 2 weeks (I have known 8 days).  Sticking 2 Erupters on a CPU heat sink would probably be the best.  You might get away without the fan, as the HSF combo is meant to take 65-95W

I like the look of these too http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=251195651419 the clip could touch the ASIC with the main flat on the underside heat spreader.

[bobsmith652]

To work out the exact Vout to the ASIC either measure voltage out across pin 3 and the 'far' side of R1 whilst running (riskier), or measure the resistance values of R1+R2 whilst off and check against that datasheet. (pinouts and supporting circuitry p12/13, voltage programming table p7)

R1 = 603R
R2 = 1058R

I make that 1.25V or 2.2V, depending on which way around the 2 would be.  They aren't K ohm though.

R3 = 7.45K
R4 = 3.36K

These give 1.16V with the formula, but I can't tell where the vias go as the other side of the circuit board is an aluminium plate, so I am not sure which of these 4 are the ones.  R3 and R4 are the only ones in the K ohm range out of the 4 nearby though.  I can see R1 is the one on pin 4, so it must be one of the 2 in the divider circuit.

Measuring the voltage from pin 3 to the opposite side of R1 (where the via is) gives 1.04V.


IDK WTF is going on, but then I am a software engineer who has to drive hardware, not design or debug it.