Couldn't you just write a script with some control logic (ie. restart miner, send email) to monitor the outputs and act accordingly? (or make changes to the miner software, it's open source after all, but I'd rather use some external script)
Topic: BTCMiner - Open Source Bitcoin Miner for ZTEX FPGA Boards, 215 MH/s on LX150 - page 17. (Read 161727 times)
March 15, 2012, 02:05:21 AM
March 15, 2012, 01:59:40 AM
It's not that I have a heat issue, it's that I use passive heatsinks and they are cooled by the building ventilation, so come summer (when we open balcony door etc.) the airflow will stop intermittently and I need to limit the damage.
The *REALLY* strong case for temperature measuring control is that you can raise/lower the frequency forever and thus leave the chips TOTALLY unmanaged for longer periods at maximum "non damaging" speeds.
Using error control (specially one way like it's today, having to reboot the java process to up the frequency again) will wear the chips more and/or not get max hash rate.
What happened to me was that my gf opened the balcony door and one chip triggered the overheat thing. If I hadn't noticed the diode lighting up I would have lost 20% of my mining power!
Of course I could just use the -oh 0.9 parameter and let the chips error down to some low frequency, but temperature would let me leave the chips (balcony door open or not) and have them perform maximum depending on heat not error over years without worrying.
Also a shame that DeepBit doesn't support MH/s drop warning mails!
The *REALLY* strong case for temperature measuring control is that you can raise/lower the frequency forever and thus leave the chips TOTALLY unmanaged for longer periods at maximum "non damaging" speeds.
Using error control (specially one way like it's today, having to reboot the java process to up the frequency again) will wear the chips more and/or not get max hash rate.
What happened to me was that my gf opened the balcony door and one chip triggered the overheat thing. If I hadn't noticed the diode lighting up I would have lost 20% of my mining power!
Of course I could just use the -oh 0.9 parameter and let the chips error down to some low frequency, but temperature would let me leave the chips (balcony door open or not) and have them perform maximum depending on heat not error over years without worrying.
Also a shame that DeepBit doesn't support MH/s drop warning mails!
March 15, 2012, 12:09:27 AM
Hm, so they sell you a 160$ without adding a temperature sensor to it? I mean common, intel sells you a complete Atom motherboard with processor, GPU and everything for 80$!?!? I'm pretty sure theres a temp thingy inside the Spartan-6, it's just not connected on the ZTEX board?
My guess is if there would be a temperature sensor ZTEX would be using it. The ZTEX board is very well designed and the stock heatsink/fan is absolutely sufficient, mine barely gets warm. The firmware/software deals with heat in a very well thought through way (clocking down) that allows us to mine at maximum capacity of each board at an individual level (not each FPGA/board is identical). I think it's more us people worrying over something that is a non-issue.
For me personally, having temperature sensors or at least some sort of solution to remotely monitor the temperature is because I plan building a bigger mining rig. I'd like to have a way to know the temperature of my boards or in the room or in the cabinet, so that I can react to it if it gets too hot. By switching on a second aircon, shutting down some equipment, increasing airflow, etc. If I would be just running a few boards, I wouldn't even waste a minute thinking about all this.
For comparison, my Icarus boards get VERY hot for my taste (and I expressed my concerns in the Icarus thread about it and even went out to buy an IR thermometer) while the ZTEX boards 20cm next to them are barely warm.
As posted by CAcoin here, he has a few dozen of the ZTEX boards and has no heat issues at all.
March 14, 2012, 08:36:00 PM
Hm, so they sell you a 160$ without adding a temperature sensor to it? I mean common, intel sells you a complete Atom motherboard with processor, GPU and everything for 80$!?!? I'm pretty sure theres a temp thingy inside the Spartan-6, it's just not connected on the ZTEX board?
March 14, 2012, 08:09:01 PM
Putting analog circuits on an all-digital chip is more difficult than you think.
The particular semiconductor process chosen for FPGAs is an all-digital, high-speed process.
Adding analog elements like a thermal diode, its associated amplifier and an ADC on an all-digital semiconductor process is something like squaring the circle - at best, time-consuming, cumbersome and costly.
Just because Intel can do it, don't assume a fabless manufacturer like Xilinx can do it.
The particular semiconductor process chosen for FPGAs is an all-digital, high-speed process.
Adding analog elements like a thermal diode, its associated amplifier and an ADC on an all-digital semiconductor process is something like squaring the circle - at best, time-consuming, cumbersome and costly.
Just because Intel can do it, don't assume a fabless manufacturer like Xilinx can do it.
How about something like this, it's digital.
1 WIRE DS18b20 Digital Temperature Sensor
http://www.alibaba.com/product-gs/348880219/_ISO9001_2008_Most_Popular_1.html
Isn't this something the GPIO's are for on the ZTEX 1.15x?
- 8 GPIO's from EZ-USB FX2 and 8 GPIO's from FPGA, can also be used for LED's
-- // --
Quote from: BR0KK
Since i moved my VM to Parallels i got this error some times?
What could cause that?
What could cause that?
Slow USB / timouts / interruptions ?
March 14, 2012, 07:25:56 PM
Code:
[code]2012-03-14T17:32:47: bus-0-0: ztex_ufm1_15d3-04A32E1205: f=210,00MHz, errorRate=0,00%, maxErrorRate=0,39%, hashRate=210,0MH/s, submitted 15 new nonces, luckFactor=0,99
2012-03-14T17:32:47: bus-0-1: ztex_ufm1_15d3-04A32E1205: f=204,00MHz, errorRate=0,17%, maxErrorRate=0,77%, hashRate=203,7MH/s, submitted 7 new nonces, luckFactor=0,83
2012-03-14T17:32:47: bus-0-2: ztex_ufm1_15d3-04A32E1205: f=198,00MHz, errorRate=0,00%, maxErrorRate=0,29%, hashRate=198,0MH/s, submitted 19 new nonces, luckFactor=1,14
2012-03-14T17:32:47: bus-0-3: ztex_ufm1_15d3-04A32E1205: f=204,00MHz, errorRate=0,18%, maxErrorRate=0,86%, hashRate=203,6MH/s, submitted 12 new nonces, luckFactor=1,11
2012-03-14T17:32:47: bus-0-4: ztex_ufm1_15d3-04A32E1205: f=210,00MHz, errorRate=0,93%, maxErrorRate=2,12%, hashRate=208,1MH/s, submitted 10 new nonces, luckFactor=0,98
2012-03-14T17:32:47: bus-0-0: poll loop time: 21ms (USB: 0ms network: 21ms) getwork time: 552ms submit time: 440ms
2012-03-14T17:32:47: bus-0-1: poll loop time: 19ms (USB: 0ms network: 19ms) getwork time: 530ms submit time: 548ms
2012-03-14T17:32:47: bus-0-2: poll loop time: 25ms (USB: 0ms network: 24ms) getwork time: 553ms submit time: 488ms
2012-03-14T17:32:47: bus-0-3: poll loop time: 23ms (USB: 0ms network: 23ms) getwork time: 595ms submit time: 470ms
2012-03-14T17:32:47: bus-0-4: poll loop time: 22ms (USB: 0ms network: 21ms) getwork time: 568ms submit time: 535ms
2012-03-14T17:32:47: Total hash rate: 1023,3 MH/s
2012-03-14T17:32:47: Total submitted hash rate: 1030,9 MH/s
2012-03-14T17:32:47: --------
2012-03-14T17:32:48: bus-0-1: ztex_ufm1_15d3-04A32E1205: Submitting new nonce 7c96834e
2012-03-14T17:32:49: bus-0-3: ztex_ufm1_15d3-04A32E1205: Got new work
2012-03-14T17:32:50: bus-0-1: ztex_ufm1_15d3-04A32E1205: Got new work
2012-03-14T17:32:52: bus-0-2: ztex_ufm1_15d3-04A32E1205: Got new work
2012-03-14T17:32:57: bus-0-4: ztex_ufm1_15d3-04A32E1205: Got new work
2012-03-14T17:32:57: bus-0-0: ztex_ufm1_15d3-04A32E1205: Got new work
2012-03-14T17:32:58: bus-0-2: ztex_ufm1_15d3-04A32E1205: Submitting new nonce 48b712be
2012-03-14T17:33:01: bus-0-3: ztex_ufm1_15d3-04A32E1205: Got new work
2012-03-14T17:33:03: bus-0-4: ztex_ufm1_15d3-04A32E1205: Submitting new nonce 48ea5c24
2012-03-14T17:33:06: bus-0-0: ztex_ufm1_15d3-04A32E1205: Error: bus=bus-0 device=\\.\libusb0-0001--0x221a-0x0100: Read hash data: libusb0-dll:err [control_msg] sending control message failed, win error: Das Gerät erkennt den Befehl nicht.
: Disabling device
2012-03-14T17:33:06: bus-0-3: ztex_ufm1_15d3-04A32E1205: Error: bus=bus-0 device=\\.\libusb0-0004--0x221a-0x0100: Read hash data: libusb0-dll:err [control_msg] sending control message failed, win error: Das Gerät erkennt den Befehl nicht.
: Disabling device
2012-03-14T17:33:06: Stopped thread for bus bus-0-3
2012-03-14T17:33:06: Stopped thread for bus bus-0-0
2012-03-14T17:33:07: bus-0-1: ztex_ufm1_15d3-04A32E1205: Error: bus=bus-0 device=\\.\libusb0-0002--0x221a-0x0100: Read hash data: libusb0-dll:err [control_msg] sending control message failed, win error: Das Gerät erkennt den Befehl nicht.
: Disabling device
2012-03-14T17:33:07: Stopped thread for bus bus-0-1
2012-03-14T17:33:07: bus-0-2: ztex_ufm1_15d3-04A32E1205: Error: bus=bus-0 device=\\.\libusb0-0003--0x221a-0x0100: Read hash data: libusb0-dll:err [control_msg] sending control message failed, win error: Das Gerät erkennt den Befehl nicht.
: Disabling device
2012-03-14T17:33:07: bus-0-4: ztex_ufm1_15d3-04A32E1205: Error: bus=bus-0 device=\\.\libusb0-0005--0x221a-0x0100: Read hash data: libusb0-dll:err [control_msg] sending control message failed, win error: Das Gerät erkennt den Befehl nicht.
: Disabling device
2012-03-14T17:33:08: Stopped thread for bus bus-0-4
2012-03-14T17:33:08: Stopped thread for bus bus-0-2
2012-03-14T17:32:47: bus-0-1: ztex_ufm1_15d3-04A32E1205: f=204,00MHz, errorRate=0,17%, maxErrorRate=0,77%, hashRate=203,7MH/s, submitted 7 new nonces, luckFactor=0,83
2012-03-14T17:32:47: bus-0-2: ztex_ufm1_15d3-04A32E1205: f=198,00MHz, errorRate=0,00%, maxErrorRate=0,29%, hashRate=198,0MH/s, submitted 19 new nonces, luckFactor=1,14
2012-03-14T17:32:47: bus-0-3: ztex_ufm1_15d3-04A32E1205: f=204,00MHz, errorRate=0,18%, maxErrorRate=0,86%, hashRate=203,6MH/s, submitted 12 new nonces, luckFactor=1,11
2012-03-14T17:32:47: bus-0-4: ztex_ufm1_15d3-04A32E1205: f=210,00MHz, errorRate=0,93%, maxErrorRate=2,12%, hashRate=208,1MH/s, submitted 10 new nonces, luckFactor=0,98
2012-03-14T17:32:47: bus-0-0: poll loop time: 21ms (USB: 0ms network: 21ms) getwork time: 552ms submit time: 440ms
2012-03-14T17:32:47: bus-0-1: poll loop time: 19ms (USB: 0ms network: 19ms) getwork time: 530ms submit time: 548ms
2012-03-14T17:32:47: bus-0-2: poll loop time: 25ms (USB: 0ms network: 24ms) getwork time: 553ms submit time: 488ms
2012-03-14T17:32:47: bus-0-3: poll loop time: 23ms (USB: 0ms network: 23ms) getwork time: 595ms submit time: 470ms
2012-03-14T17:32:47: bus-0-4: poll loop time: 22ms (USB: 0ms network: 21ms) getwork time: 568ms submit time: 535ms
2012-03-14T17:32:47: Total hash rate: 1023,3 MH/s
2012-03-14T17:32:47: Total submitted hash rate: 1030,9 MH/s
2012-03-14T17:32:47: --------
2012-03-14T17:32:48: bus-0-1: ztex_ufm1_15d3-04A32E1205: Submitting new nonce 7c96834e
2012-03-14T17:32:49: bus-0-3: ztex_ufm1_15d3-04A32E1205: Got new work
2012-03-14T17:32:50: bus-0-1: ztex_ufm1_15d3-04A32E1205: Got new work
2012-03-14T17:32:52: bus-0-2: ztex_ufm1_15d3-04A32E1205: Got new work
2012-03-14T17:32:57: bus-0-4: ztex_ufm1_15d3-04A32E1205: Got new work
2012-03-14T17:32:57: bus-0-0: ztex_ufm1_15d3-04A32E1205: Got new work
2012-03-14T17:32:58: bus-0-2: ztex_ufm1_15d3-04A32E1205: Submitting new nonce 48b712be
2012-03-14T17:33:01: bus-0-3: ztex_ufm1_15d3-04A32E1205: Got new work
2012-03-14T17:33:03: bus-0-4: ztex_ufm1_15d3-04A32E1205: Submitting new nonce 48ea5c24
2012-03-14T17:33:06: bus-0-0: ztex_ufm1_15d3-04A32E1205: Error: bus=bus-0 device=\\.\libusb0-0001--0x221a-0x0100: Read hash data: libusb0-dll:err [control_msg] sending control message failed, win error: Das Gerät erkennt den Befehl nicht.
: Disabling device
2012-03-14T17:33:06: bus-0-3: ztex_ufm1_15d3-04A32E1205: Error: bus=bus-0 device=\\.\libusb0-0004--0x221a-0x0100: Read hash data: libusb0-dll:err [control_msg] sending control message failed, win error: Das Gerät erkennt den Befehl nicht.
: Disabling device
2012-03-14T17:33:06: Stopped thread for bus bus-0-3
2012-03-14T17:33:06: Stopped thread for bus bus-0-0
2012-03-14T17:33:07: bus-0-1: ztex_ufm1_15d3-04A32E1205: Error: bus=bus-0 device=\\.\libusb0-0002--0x221a-0x0100: Read hash data: libusb0-dll:err [control_msg] sending control message failed, win error: Das Gerät erkennt den Befehl nicht.
: Disabling device
2012-03-14T17:33:07: Stopped thread for bus bus-0-1
2012-03-14T17:33:07: bus-0-2: ztex_ufm1_15d3-04A32E1205: Error: bus=bus-0 device=\\.\libusb0-0003--0x221a-0x0100: Read hash data: libusb0-dll:err [control_msg] sending control message failed, win error: Das Gerät erkennt den Befehl nicht.
: Disabling device
2012-03-14T17:33:07: bus-0-4: ztex_ufm1_15d3-04A32E1205: Error: bus=bus-0 device=\\.\libusb0-0005--0x221a-0x0100: Read hash data: libusb0-dll:err [control_msg] sending control message failed, win error: Das Gerät erkennt den Befehl nicht.
: Disabling device
2012-03-14T17:33:08: Stopped thread for bus bus-0-4
2012-03-14T17:33:08: Stopped thread for bus bus-0-2
Since i moved my VM to Parallels i got this error some times?
What could cause that?
[/code]
March 14, 2012, 04:24:10 PM
Yes, but then the chip will be really hot, hot enough to burn your fingers on the heatsink, and that's not good for longevity.
I don't mind hashing slower (I'm talking 100-200 range here), if it allows me to hash longer.
ZTEX, does the chip use 8W at 100MHz too? Or will the chip use like 4W and be half as warm?
I'm really just looking for settings to be able to passively cool the thing without having to worry.
It's a shame there is no temperature reading on the chip!
I don't mind hashing slower (I'm talking 100-200 range here), if it allows me to hash longer.
ZTEX, does the chip use 8W at 100MHz too? Or will the chip use like 4W and be half as warm?
I'm really just looking for settings to be able to passively cool the thing without having to worry.
It's a shame there is no temperature reading on the chip!
Ok, i get you. I trust ZTEXs experience together with my extra fans.
March 14, 2012, 03:31:11 PM
ZTEX, does the chip use 8W at 100MHz too?
No. Power consumption of a VLSI chip is the sum the static power (i.e. leakage), which is almost negligible, and dynamic
power, which is (by and large) proportional to the clock frequency.
Or will the chip use like 4W and be half as warm?
Yes.
It's a shame there is no temperature reading on the chip!
Putting analog circuits on an all-digital chip is more difficult than you think.
The particular semiconductor process chosen for FPGAs is an all-digital, high-speed process.
Adding analog elements like a thermal diode, its associated amplifier and an ADC on an all-digital semiconductor process is something like squaring the circle - at best, time-consuming, cumbersome and costly.
Just because Intel can do it, don't assume a fabless manufacturer like Xilinx can do it.
March 14, 2012, 02:02:31 PM
Yes, but then the chip will be really hot, hot enough to burn your fingers on the heatsink, and that's not good for longevity.
I don't mind hashing slower (I'm talking 100-200 range here), if it allows me to hash longer.
ZTEX, does the chip use 8W at 100MHz too? Or will the chip use like 4W and be half as warm?
I'm really just looking for settings to be able to passively cool the thing without having to worry.
It's a shame there is no temperature reading on the chip!
I don't mind hashing slower (I'm talking 100-200 range here), if it allows me to hash longer.
ZTEX, does the chip use 8W at 100MHz too? Or will the chip use like 4W and be half as warm?
I'm really just looking for settings to be able to passively cool the thing without having to worry.
It's a shame there is no temperature reading on the chip!
March 14, 2012, 01:35:46 PM
Then can I get a parameter for specifying the initial frequency and another to disable initial step-up of frequency. You can cap the initial frequency parameter to max 200 since I'm going to use this to _lower_ the frequency for passive cooling during summer.
Hmmm i think the board can do this by itself ?! Sounds to me like reinventing the wheel or am i missing something ?
March 14, 2012, 11:40:15 AM
Then can I get a parameter for specifying the initial frequency and another to disable initial step-up of frequency. You can cap the initial frequency parameter to max 200 since I'm going to use this to _lower_ the frequency for passive cooling during summer.
Excellent idea, I second that.
Being able to specify the start frequency and the maximum frequency would be ideal for those of us who:
- have inadequate air conditioning
- want to make sure the FPGA has a long lifetime
- want to install a smaller cooled due to space constraints (i.e., putting multiple vertically arranged boards into an enclosure)
March 14, 2012, 08:50:07 AM
Then can I get a parameter for specifying the initial frequency and another to disable initial step-up of frequency. You can cap the initial frequency parameter to max 200 since I'm going to use this to _lower_ the frequency for passive cooling during summer.
March 14, 2012, 05:20:28 AM
Hi, can you gather temperature from the chip? Could we get a temperature/frequenzy controller as a complement to the error/frequenzy controller for passive cooling mining? My dream would be a -t parameter which keeps the chip at that temperature!
No, chip temperature cannot be gathered.
March 13, 2012, 04:44:05 AM
Hi, can you gather temperature from the chip? Could we get a temperature/frequenzy controller as a complement to the error/frequenzy controller for passive cooling mining? My dream would be a -t parameter which keeps the chip at that temperature!
March 06, 2012, 03:46:13 AM
Code:
2012-03-06T08:53:13: bus-0-0: ztex_ufm1_15d3-04A3469722: Error: bus=bus-0 device=\\.\libusb0-0003--0x221a-0x0100: Read hash data: libusb0-dll:err [control_msg] sending control message failed, win error: The device does not recognize the command.
: Disabling device
: Disabling device
That are USB connection errors. Typically caused by the power supply (loose contacts, instabilities if the PSU becomes to hot / output current to high) or by the USB hardware (cables, hubs, devices that draw to much USB power). A potential equalization problem is possible too.
March 05, 2012, 09:45:24 PM
I have received another 2 units so now I have a "cluster" of 3 ZTEX 1.15x. Had it running over night on my new cheapo 10 port USB hub (since I only plugged in 3 boards I didn't use the additional USB 5V 2.0A power supply that came with the hub).
It worked well from 21:12 to 08:53 this morning, but then the following happened:
Then the second one followed at 10:29 when I plugged in my Icarus to the same USB hub.
For the last one I didn't have to wait long:
Could that be a USB power issue or because the cheap USB hub interrupts when I plug/unplug other devices?
For the last two units connecting/disconnecting USB cables (connected to Icarus) seems to be the culprit, but I don't know what happened for the first unit at 8:53, I wasn't even near my computer at that time as far as I know (I may have logged in, but I definitely didn't touch the USB port/1.15x boards).
It worked well from 21:12 to 08:53 this morning, but then the following happened:
Code:
2012-03-06T08:53:13: bus-0-0: ztex_ufm1_15d3-04A3469722: Error: bus=bus-0 device=\\.\libusb0-0003--0x221a-0x0100: Read hash data: libusb0-dll:err [control_msg] sending control message failed, win error: The device does not recognize the command.
: Disabling device
: Disabling device
Then the second one followed at 10:29 when I plugged in my Icarus to the same USB hub.
Code:
2012-03-06T10:29:46: bus-0-0: ztex_ufm1_15d3-04A32E00E9: Error: bus=bus-0 device=\\.\libusb0-0002--0x221a-0x0100: Read hash data: libusb0-dll:err [control_msg] sending control message failed, win error: The device does not recognize the command.
: Disabling device
: Disabling device
For the last one I didn't have to wait long:
Code:
2012-03-06T10:31:32: bus-0-0: ztex_ufm1_15d3-04A346CEC7: Error: bus=bus-0 device=\\.\libusb0-0001--0x221a-0x0100: Read hash data: libusb0-dll:err [control_msg] sending control message failed, win error: The device does not recognize the command.
: Disabling device
2012-03-06T10:31:32: Stopped thread for bus bus-0-0
: Disabling device
2012-03-06T10:31:32: Stopped thread for bus bus-0-0
Could that be a USB power issue or because the cheap USB hub interrupts when I plug/unplug other devices?
For the last two units connecting/disconnecting USB cables (connected to Icarus) seems to be the culprit, but I don't know what happened for the first unit at 8:53, I wasn't even near my computer at that time as far as I know (I may have logged in, but I definitely didn't touch the USB port/1.15x boards).
March 05, 2012, 05:39:44 AM
Is that possible or something that can't be done?
Another UI is possible, but not planned.
March 04, 2012, 12:24:15 PM
I like it the way it is.
March 04, 2012, 11:50:10 AM
Hi,
I'm using five boards in cluster mode atm. They work fine so far
Would be nice if BTC Miner gets a new interface --> i don't have the right words in english right now. i mean it should look like cgminer or some thing like that. Displaying every information like it is now can be confusing?
Is that possible or something that can't be done?
Would like to be able to use cgminer ore different miners with ztex boards aswell:)
I'm using five boards in cluster mode atm. They work fine so far
Would be nice if BTC Miner gets a new interface --> i don't have the right words in english right now. i mean it should look like cgminer or some thing like that. Displaying every information like it is now can be confusing
?Is that possible or something that can't be done?
Would like to be able to use cgminer ore different miners with ztex boards aswell:)
February 28, 2012, 09:51:26 PM
Possibly the pool's response time?
What happens if you change to a underutilized pool for testing?
What happens if you change to a underutilized pool for testing?
Jump to: