Nice miner. Worked nicely.

Rented about 20 MHash.

Hashes delivered. Rigs seemed stable enough. No complaints. Will rent again. A+++

I have 1 million TOP. Will trade for 20,000 DOGE. Or pm your best offer

HI Crypthaus,

Where have all my coins gone? I had mined many coins, but now my balance is showing 0 and I cannot withdraw them.

When the lease expires, sometimes the customers pools aren't removed.

I just had 3 leased rigs finish their lease. The customer had set a main and a backup pool. However, when the lease expired, only one of the rigs started mining for me.

The other two continued mining on the customer pool. Leaserig had sent the addpool commands because my default pools were listed, but for some reason the removepool or switchpool commands hadn't taken effect.

I've examined all my cards (a variety of 7950s and 280x) and the auxillary power is not connected to the PCI-slot power. If you connect a second PSU to the aux inputs, then there is no risk of backfeed.

Obviously, i can't vouch for all cards and all VRM designs. However, I would expect that most VRM designs in use today could
be buillt so that backfeed cannot occur.

The most important thing to make sure if you are using multi-PSUs is that the ground connections are very securely connected together. In general, cards do directy connect the ground pins on the aux connectors to the ground pins on the PCI-e slot. This way, if you are not using risers, just connecting the 2ndry PSU to the aux connectors will result in a secure ground link.

The problem comes with powered risers. Theres a big danger with these, because they may separate the ground and power connections. If you connect the riser to 2ndry PSU, then the signal voltages will all be set with reference to the 2nd PSU, and if the ground voltage separate (e.g. when one PSU switches on) then this power can be conducted into the signal connectors and fry the motherboard. I killed several motherboards this way before I worked out what the problem was.

This problem is even bigger with the USB style powered risers, because they completely separate all the power, and only have a thin shield connecting the grounds - not like a ribbon cable which has about 10 ground connectors. If you are using USB style risers, you must ensure a very strong ground connection between the two PSUs. Easy way: Power all the risers directly from PSU 1 which is powering motherboard - Card aux power can come from other PSUs. Harder way: Make custom harnesses which connect the ground wires between 2 hard drive connectors. Use 2 or 3 of these to cross link the ground connections between the PSUs - don't trust a single wire linking the grounds on the big 24pin molex connectors.

When does leaserig actually issue a save command?

I was initially running with a custom .conf, but have switched to the default .conf file? I've also added the save command to the permitted command list, however, when I admin the pools from the admin console, no save command is issued, and a cgminer restart will re-load the original conf.

Is the command only issued when hiring starts and ends? Is it only customer that can issue save commands?

Agreed. You need new circuits installed. What you have installed now isn't good enough. I know it's not the answer you want, but realistically, there is no alternative.

I strongly recommend the use of 240 V outlets. PC PSUs operate more stable, more efficient and cooler when powered on 240 V.

In addition because of the lower amp requirement of the PSUs when operating at 240V it allows more flexibility in moving PCs from one circuit to another.

Got the risers from a Hong Kong vendor.

When I powered up, I saw sparks jump from the 1x slot and heard sizzling (both times). After that, though, there was nothing to see. No burned contacts, no burned traces that I could see. I could smell it though.

I've subsequently reused the "killer" risers, after doing a full electrical continuity test with a multimeter, and verifying no shorts or missing connections, and they work perfectly now that they are powered from the primary supply.

I can only guess that during power up, there was some sort of "ground bounce" problem, which caused the ground voltage on the PSUs to pull apart. I suspect that as the 2 PSU grounds were only tied together by a single piece of 18 G wire on the dual-PSU adaptor, that this was not enough to ensure full equipotential bonding, and that it was the ground current that fried stuff.

Perhaps, if all your PSUs are very strongly grounded together, then you could avoid this problem.

I can see why if the risers are on the same PSU as the mobo, this problem can be avoided. The cards will provide a very low impedance ground plane between aux power input and the 16x slot. This will provide a very strong cross-ground connection between the 2 PSUs. If the card and riser are on the same PSU, then any ground currents have to go through the PCI-e signal and shield wires.

Since I've moved all the risers onto the main PSU, I've not burned out any more mobos.

No. You must have power, there is no power connection through the USB cable. if you look at the PCB, the power pins are not connected at the motherboard end.

You must use a USB 3.0 cable, a USB 2.0 cable will NOT work as it does not have enough pins. You also need a non-standard type of cable with double A-male ends. Good luck trying to buy one of these cables separately.

The small socket contains a built in presence shorting link, which means that you do not need a jumper wire to activate the slot.

Issues:
The SMBus pins in the PCI-e slot are not connected. Some monitoring or diagnostic software may have difficulty working without these connections.

Also remember, that when powering the riser, you must power it from the same PSU as is powering the motherboard. You WILL fry the motherboard if you have 1 PSU powering the mobo, and a different PSU powering the riser.

This type of riser doesn't work very well at PCI-E v3 mode. If you have a mobo which uses PCI-E v3, you should select a lower speed in the BIOS, or you may get system instability (or card not detected).

The slot end of the connector and the USB plugs themselves are very loose. You need to use some hot-melt glue to keep the riser in the slot, and the USB cables attached.

There are a number of HK sellers on ebay whicih will post to the UK. I got mine in about 10 days, and paid £12 each inc postage.

I've been running a few mining rigs, which are basically just motherboards crammed with cards.

Due to heat problems, I've been looking to get some risers - and they finally arrived. I got 100 cm flexible USB cable risers, with SATA power. They look impressive 3 filter caps and a VRM on the riser board.

Anyway, the first thing I found was that no cards would be detected on the risers until I set the mobo BIOS to PCI-e v1.0.

So far, so good. So, one at a time, I transferred cards from mobo onto riser cables, and after a few false starts, I had 2 cards running on risers, and 1 on the mobo.

That's when I start to get adventurous, and connect a 4th card to a 1x PCI-e slot, and power it via a 2nd PSU. All the connections are double checked. The riser is in the slot the correct way round, etc.

Link the PSUs with a dual PSU adaptor, and power up.

*Sizzle* *Pop* *Crackle* <Burning smell> Cue sparks and smoke rising from the 1x PCI-e slot. The mobo is totally dead.

I recheck all the connections, and they are definitely correct.

Thinking I made a mistake, I decide to have another go.
This time, I connect directly a 1x PCI-e slot with a 4th card, after triple checking everything.

*Sizzle* *Crackle* Sparks. Another motherboard now dead.

I don't get it. I've built dozens of PCs in the past, and built a heap of mining rigs, albeit without risers. I've done dual PSU before.

What could have happened? What was cause sparks and sizzling and burn out a mobo. I've looked at the PCI-e riser design, and the power pins are not connected on the riser, so I don't see how a power surge or something could have come from the 2nd PSU back feeding the PCI-e slot.

Any ideas? Could it be a bad riser card? I've run a multimeter over it, and there don't appear to be any shorts or anything. I'm completely baffled.

WU/min and MH/s are the exact same thing, just slightly different units.

For Scrypt 1 WU = 655536 hashes. 1 WU/min = 1.09 kH/s (Different numbers are used in SHA-256)

Cgminer averages the WU and MH/s numbers differently, so they don't quite line up exactly on the display - but they are the exact same thing.

Share difficulty does not affect the WU number. However, share difficulty is measured in WU. So a share with 1024 difficulty needs on average 67,108,864 hashes to solve.

I've noticed that the EU server is mining totally different coins to the other servers (when it's working, it's down at the time of posting this message). The others are generally reasonably well synchronised, although there are slight differences in the exact timing of coin changes.



The lower color bar on the hash rate charts represents the server (blue = EU, Orange = USW, Red = Asia, Green = USE). The upper color bar represents the coin (or more correctly teh difficulty).

You can see that the EU server is mining different coins for much of the time (usually MOON, while the others are mining DOGE or LTC).

I'm wondering if this is deliberate, and whether it's H2O's attempt to mine some of the smaller coins by only directing a small proportion of hash to them.

Try it and see what happens in your configuration.
I reduced rejects from 6% to 2% by doing this. I lost 1% hash, but the reduction in rejects more than made up for this.

Going to Intensity 20 adds several seconds of latency to share processing. On very fast alt coins, this makes a big difference. People talk about choosing the correct server due to latency, etc. Intensity is the elephant in the room, adding more latency than anything else you could do.

Changing the intensity made more difference than moving my miners from a 28 kbps GPRS connection to a 60 Mbps fiber connection.

The latency was never much of a problem as I used to mine LTC with its long block time and stable difficulty period. The latency became much more noticable on middlecoin, because it spends a lot of time mining 30 or 60 second coins.

You could try www.leaserig.net

There's quite a bit of hash there. I'm hoping to add 5 MH myself, just waiting for my registration.

Please post feedback in this thread.

3 Mining rigs @ 1.6 MH/s available at www.leaserig.net

Rigs are offered on a best effort basis but do not have 24/7 support. I offer my rigs at a low price to reflect this.  However, rigs are very stable, and do not need rebooting, but are usually rebooted weekly as part of routine maintenance.

Refund policy
Any down time (due to rig or network failure) will be refunded pro rata at 2x the price paid, for up to 50% of time. Outages of more than 50% result in a complete refund.

Please send me a PM to claim.

Update 17/2/14: There was a brief outage last week due to a problem with my internet connection. The ISP has performed a repair, and I have been testing the connection for the last 72 hours, and it has had 100% stability.

For what's it worth, hashbros' stats are hopelessly broken as well. I was testing it, seeing as middlecoin's were broken in a new and unfamiliar way.

My "earnings" figures in hashbros change randomly every time I look at it. Right now, if the reading were accurate I would be getting an income of 0.001 BTC/Mh/day. The other problem is that the pool is simply too small to mine anything substantial like LOT.

You're right about the count(*), but other than the queries are identical.

Every DB I've ever used will elide the cross join in the first query into an inner join, when an appropriate equality exists in the where clause.

What is desperately needed here is indexing. shares.time should have an index, to avoid a table scan.

Doge coin has a rate of about 25 GH/s, and middlecoin doesn't mine it continuously. So, middlecoin is probably about 1/3 of doge.

42coin fluctuates in diff quite a lot, due to its very short difficulty adjustment, but has been averaging about 60. Also don't forget that block times are differernt, so difficulty isn't directly comparable and must be converted. Difficulty of 60 in doge is equivalent to difficulty of 42 in 42.

In short, 42coin has been averaging a hash rate of about 6-10 GH/s. Middlecoin would make a big dent to the profitability, but it wouldn't crash it.