Hacking BFL Monarchs and servicing them while times are weird.

[lightfoot]

These units are very interesting. I'm impressed that BFL was able to produce such a decent unit (albeit hot-running and below performance expectations) given all of their failings.

Yeah they had a tough time building the damn things. Main fault in retrospecitve was building such a fucking big chip; it ran so hot they couldn't run them in series like ants did, and trying to build a power step down supply that could feed the chips wound up with FETs that ran almost as hot as the chip (inefficient). A series string of 15 smaller chips would have been amazing, but they wanted to do the stupid plug it in your computer form factor.

And they were insane. Ah well, no more domestic US miner manufacturers. Ants or nothing.

[1714105785]

1714105785

[1714105785]

1714105785

[1714105785]

1714105785

[mkj_14]

apologies if this has been covered before.

i have a number of BFL Monarch.  they power up and run fine at idle.  within 5 minutes of mining the temps at 90C and cut off.  any ideas where i should start?


Thanks

[DaveF]

apologies if this has been covered before.

i have a number of BFL Monarch.  they power up and run fine at idle.  within 5 minutes of mining the temps at 90C and cut off.  any ideas where i should start?


Thanks

Could be several things.

1) Cooling pump is dead so the fluid is not moving.
2) All the coolant leaked out (or was drained)
3) The cooling blocks are not solidly attached to the chips.

If you know someone with PC liquid cooling experience they can usually tell you in a few minutes if it's any of those. And for the most part #1 and #2 they are easily (if not cheaply) fixable. #3 is quick and simple but usually not the issue.

Did you get them used or have you had them since new? Did they ever work?

-Dave

[lightfoot]

Dave's got it. I'd check first to make sure the fan on the front of the radiator is hooked up and working, then check to see if the radiator is getting warm (indication of pumps running). Then check to make sure the fan on the back works, then let's go from there.

C

[mkj_14]

Hi,  thanks for the reply's.

i bought them used at a very god price for a bit of fun.

1 - work but at a lower rate. (500 ghz instead of 700 ghz)
1 - I've got one to work, however runs a little hot (68 - 78 C)
2 - overheat within 5 minute of mining.

unfortunately there no one in my area that every seen water-cooling.  plus i couldn't trust the quality of work as I'm in working in the middle east at the moment.



Cheers

[mkj_14]

Just out of interest.

Has anyone been able to get the monarch to work on Dash Coin?

[cbjarscoobs]

Just out of interest.

Has anyone been able to get the monarch to work on Dash Coin?

Monarchs hash SHA256 coins. Dash is a different algorithm, won’t work.

[cbjarscoobs]

apologies if this has been covered before.

i have a number of BFL Monarch.  they power up and run fine at idle.  within 5 minutes of mining the temps at 90C and cut off.  any ideas where i should start?


Thanks

I have same issue but mine cut off in 5 secs not 5 minutes.

[mkj_14]

hi,  does anyone know where the temperature sensor are, and if there replaceable?  my showing 150 C at start up........could be the problem me thinks.

[cbjarscoobs]

hi,  does anyone know where the temperature sensor are, and if there replaceable?  my showing 150 C at start up........could be the problem me thinks.

i would like to know as well

[lightfoot]

Sorry, been away for awhile. The temp sensors are labelled T1 and T2. They are bottom center of the board, right outbound of the bottom two screws by the ISL chips that control the 6 phase FETs on each side.

You could ignore them by setting the temp on BFGMiner.conf to 200 or so, but make sure the temps on the chips really are reasonable before you do this.

C

[cbjarscoobs]

Sorry, been away for awhile. The temp sensors are labelled T1 and T2. They are bottom center of the board, right outbound of the bottom two screws by the ISL chips that control the 6 phase FETs on each side.

You could ignore them by setting the temp on BFGMiner.conf to 200 or so, but make sure the temps on the chips really are reasonable before you do this.

C

Got a IR temp reader, and looks like I am getting to hot. Looking next to the pumps. Is there a way to test the pumps without powering the board?

[lightfoot]

Sure, the pumps just use normal 12v fan plugs. Use an adapter to connect a fan to a 12v power supply and fire them up. You should be able to hear the water running in the radiator if they are working.

Any evidence of water leaking (on the copper plates or bottom metal frame)?

[cbjarscoobs]

Sure, the pumps just use normal 12v fan plugs. Use an adapter to connect a fan to a 12v power supply and fire them up. You should be able to hear the water running in the radiator if they are working.

Any evidence of water leaking (on the copper plates or bottom metal frame)?

No water leaking evidence at all. I quickly tested one pump per your suggestion but got nada. Would I need to power both pumps at the same time?

[lightfoot]

No each pump is independant. Try the second one, I haven't seen both fail together very often.

[cbjarscoobs]

No each pump is independant. Try the second one, I haven't seen both fail together very often.

I got 2 monarchs and somehow all 4 pumps are lifeless. I decided to sacrifice one. Any thoughts on where my pumps on this one went wrong? The little bit of liquid that did leak out was pretty black in color.

I have some burn/brown marks in the  in the middle

https://i.imgur.com/advZOYnl.jpg

One pump looks to have some green inside?

https://i.imgur.com/CobqcsHl.jpg

Think I snapped off the two wires for the pump controller.

https://i.imgur.com/prgaeZfl.jpg

here are other pictures for others interested

https://i.imgur.com/4QRnopzm.jpg https://i.imgur.com/jxGPyO7m.jpg https://i.imgur.com/ZArTFV1m.jpg https://i.imgur.com/nTlVhAFm.jpg

https://i.imgur.com/D1OrSFVm.jpg https://i.imgur.com/0KlOhRmm.jpg https://i.imgur.com/7tIqYLKm.jpg https://i.imgur.com/fML0JOsm.jpg

[NotFuzzyWarm]

My guess is that the cooling systems failed from pure and simple corrosion. A result of the crappy cooler mfgr they used.
A side note to this - and owners of PC's using similar self-contained cold-plate/pump/heat exchangers:A while back I came across Intel's performance and design specs for their CPU coolers. One rather surprising thing I learned that that they all have a rather significant coolant loss rate - not from leaks, but from the fluid just slowly passing through the hose walls via osmosis. Up to 1/2 ounce per-year with ~ 0.2.-.3 oz/year as typical.

[BadgerHobbs]

I have a 750GH/s unit that when powered on, makes a few beeps and has loads of red lights flashing.

BFG miner is unable to detect it.

[cbjarscoobs]

I have a 750GH/s unit that when powered on, makes a few beeps and has loads of red lights flashing.

BFG miner is unable to detect it.

[...]

Do your pumps work? you can test the pumps without powering the board.

[lightfoot]

My guess is that the cooling systems failed from pure and simple corrosion. A result of the crappy cooler mfgr they used.
A side note to this - and owners of PC's using similar self-contained cold-plate/pump/heat exchangers:A while back I came across Intel's performance and design specs for their CPU coolers. One rather surprising thing I learned that that they all have a rather significant coolant loss rate - not from leaks, but from the fluid just slowly passing through the hose walls via osmosis. Up to 1/2 ounce per-year with ~ 0.2.-.3 oz/year as typical.

No, they're full of propolyne glycol and although they run under pressure they don't corrode. What kills them is the insane heat; and if the fan up front ever stops then they will overpressure and blow out the seals.

Water cooling is neat, but as we learned with toy trains everything is hell with water.