Pages:
Author

Topic: GridSeed 5-chip USB miner voltage mod - page 33. (Read 157001 times)

newbie
Activity: 2
Merit: 0
April 23, 2014, 04:56:45 PM
Hi Guys,

I am getting some Gridseeds and will be trying the 47kohm resistor mod. Which software should I use to get max overclock (like the approx 490 k/h I've seen screenshots of)? I am planning on using Linux, either raspberry Pi or a standard motherboard (max 10 Gridseed miners connected to Windows is not enough....or is there a way of connecting more?).
Can someone point me to a tarball I can download? I am experienced with CG miner, so CG miner would be ideal, but am open to trying other software out. I found the 3355 Gridseed on Github but had trouble cloning it, I guess a tarball would be much easier.
Also, if I used a standard motherboard with a dual core or quad core processor would I be able to connect like 100 or more to a motherboard, or is the processor irrelevant (I'm thinking of the CPU switching between hundreds of devices so would that be causing a burden, to the 700 mhz "slow" CPU on a raspberry PI, and maybe could be overcome with a fast CPU, rather than buying lots of raspberry PIs)
sr. member
Activity: 378
Merit: 250
April 23, 2014, 04:47:30 PM
wolfey - how many amps are your pods pulling with the 49.9k resistor? Or watts?  I'm looking for actual measurements if you have an easy to use tool like a clamp on ammeter or a watt meter.  If anyone is using the 47k resistor it would be nice to know the power usage too so we know if there's any difference in power consumption. Actual measurements please.  We could list the mod, freq, resistor used, watts or amps........  If anyone is using a different resistor and has tools to measure the please post your info too.  It would be interesting to know the volts too, it's nominal 12 but my guess is there's a lot of variance power supply quality / loading and wire gauge used etc.  It'll help determine how gridseeds we can put on a power supply and power cost per mod.

some of my measurement results:
vmod1 975MHz to 1000MHz 11.86v, 1.07a, 12.69 watts, 400 kH/s  24hr results
no mod dual mode 750MHz both 11.86v, 4.23a, 50.17 watts, 8.5 Gh/s, 300 kH/s 24hr results
vmod1 dual mode 850MHz both 11.86v, 5.50a, 65.23 watts, 10 Gh/s, 350 kH/s 12hr results   



Hi, I really don't want to go tearing into my pods right now to take measurements. They're mining so good! I can't bear to interrupt them!  Grin
But I can concur with your vmod1 readings. Very close to what I found when doing my mods. 1. something Amps at 12.something to maybe 14 to 16 or so Watts sounds about right for up to 1250MHz with my 49.9k resistor mod. Just going by memory from weeks ago but I know these figures are very close.
I wouldn't ever push these miners harder in Scrypt mode. I see your other figures are for dual mode. I have no experience with those as I've never run my miners in SHA mode.
By the way, as I've mentioned before and at least one other has concurred, 50k ohm is okay to use too. I wouldn't go any higher than that but it's your stuff so.... Wink

One other fact. All 12V DC power supplies output at 13.8V nominal. No power supplies (except maybe broken, poorly made or incorrectly altered ones - and car chargers) output 12V DC! NONE! In automotive electrical/charging systems, in all 12V DC systems, 12.8 to 13.2V is the bottom level for a 12V battery although they get down to 10 or 11 volts at times due to improper maintenance, wear and tear, age and or out charging system characteristics.
All of them charge and keep the voltage under load at 13.80 to 14.20 VDC.

Also, keep in mind that those LED switch mode power supplies, and those bricks, although regulated, may not have proper output filtering due to use of cheap low value/quality capacitors and other components, poor quiescent and high ripple currents which sensitive equipment like our miners do not like! That's why some of them end up releasing magic black smoke and unfortunately in the process, ruin your miners too.
sr. member
Activity: 378
Merit: 250
April 23, 2014, 04:30:19 PM
aaarrgh !

Minerd-sandor111.exe - system Error
The program can't start because libcurl.dll is missing from your computer. Try reinstalling the program to fix this problem.

off on an internet libcurl hunt now.


If you have any instances of cpuminer or cgminer etc. on your HD, you can simply do a search and find that dll. I had the same problem, searched then moved one of the several instances of that dll to the proper folder then minerd fired up and ran. Wink
I can send you a copy if you like.
Just shoot me a PM and I'll hand it over Wink
full member
Activity: 140
Merit: 100
April 23, 2014, 03:52:47 PM
wolfey - how many amps are your pods pulling with the 49.9k resistor? Or watts?  I'm looking for actual measurements if you have an easy to use tool like a clamp on ammeter or a watt meter.  If anyone is using the 47k resistor it would be nice to know the power usage too so we know if there's any difference in power consumption. Actual measurements please.  We could list the mod, freq, resistor used, watts or amps........  If anyone is using a different resistor and has tools to measure the please post your info too.  It would be interesting to know the volts too, it's nominal 12 but my guess is there's a lot of variance power supply quality / loading and wire gauge used etc.  It'll help determine how gridseeds we can put on a power supply and power cost per mod.

some of my measurement results:
vmod1 975MHz to 1000MHz 11.86v, 1.07a, 12.69 watts, 400 kH/s  24hr results
no mod dual mode 750MHz both 11.86v, 4.23a, 50.17 watts, 8.5 Gh/s, 300 kH/s 24hr results
vmod1 dual mode 850MHz both 11.86v, 5.50a, 65.23 watts, 10 Gh/s, 350 kH/s 12hr results   

hero member
Activity: 616
Merit: 500
April 23, 2014, 03:43:16 PM
i already downloaded the one from crypoblog, then i downloaded sandor111 latest file - which was just the .exe file. the one from cryptoblog had a libcurl_4.dll in it not just a libcurl.dll. i found a libcurl.dll easy enough now its going. pool now shows 720! before the autooverlcok has kicked in. nice!

Great work sandor, but i really miss the summary at top of the miner app - like in bfgminer, and cgminer. how easy would it be to add in Smiley? if it wasnt for the whole complie on windows lack, id knock somehtign up myself. trying to compile it on windows puts me off though.

I need to start learning ncurses. Smiley
member
Activity: 84
Merit: 10
April 23, 2014, 03:42:20 PM

Hi Cbutters,

Wire 1 is to be attached to LOWER pad of the Cut-off R46 bridge...As I pointed in previous pictures/posts...it is the SAME as LOWER row soldering points of the VID resistors...

So you CAN ATTACH to ANY...or ALL of them...as they are electrically connected together... you can see the traces from R46...or measure that...

The trim pot is intended to be a TEST Bench...Read my second sentence in the my previous post...to ADJUST voltage smoothly...It IS a ZiG_Mod for testing/experimenting...

The original first proposal of ZiG_Mod is fixed resistor...Equivalent to R139 ("R52) replacement...but without the need of actual de-soldering...  

I have been running without top heatsink at the early stage of modding...anything above 1000MHz was causing Gridseed chips to become hotter and untouchable...This was the time when I modded the thermal solution and posted here...like end of March...

After that I have them running with copper heatsinks on each chip...RAM/VID type...10mm x 10mm ... Phobya thermal pads... no problems even @1250MHZ...but they are getting HOT...above 1150-1175MHz...

ZiG  


Thanks for the clarification! I think I might try wiring up trim pots up on all of mine if it works good in testing... I like the idea of variable voltage to be able to fine tune so as to be as efficient as possible at around 450-480 hashrate. With that we could find a stable voltage then slowly move down from there. Plust trim pots are quite inexpensive as well. I'll report back once I receive my gridseeds and do some testing. I will post pics / guide if it works well.

i considering using a pot but didnt have right one on hand either too big reisitance or too small. didnt have right range.
Used wiht caution - as always when modding - it fine. decided on your MAX resistance you want, set it before mounting, use a mulitmeter to check it. MARK the where max is. then adjust downwards only. mark which way increases and decease the pot too. for thoose wiht ladies in the house a dab of nial polish works really well for marking the max limit. it holds longer and doesnt come off easy, plus its more noticedable and different from any other markings you might add to it while adjusting it. i use the wifes reddish ones and a tooth pick to mark my pots - for other non grid mods or tinkering ive done. Sparte multimeter connected to the pot to check the resistance helps great too. analogue is good here as it responds without delay, like lower and mid range digital mulitmeter suffer from. (ill always keep and use both my analogue meters)
member
Activity: 84
Merit: 10
April 23, 2014, 03:33:59 PM
stock pads are crap and are meant for stock speeds and stock heat. overclocking will produce more heat FACT. ok if your not using sha people SAY that dont get that hot. but damn touch thoose things without the top heatsink on they are HOT!!!!

Right now the heat pads on top of the chips ARE NOT THE PROBLEM. read that again. allow it to digest a bit. beath and now continue.

The gridseed chips have THERMAL VIAs to conduct the heat from the chip TO THE BOTTOM HEATSINK! yes thats right the bottom heatsink. SO to really improve your thermal conduciviy you need to be more concerned about the thermal vias, which are located on the underside of the PCB! remove the screww in rises - yeah thoose things that the screw went into. NOW USING SOMETHING PLASTIC VERY CAREFULLY lever the pcb away from the bottom heat sink. i say carefully as it can and probably is stuck there, from use and the stock thermal pad gettting warm and cooling etc. DO NOT USE METAL.
Now stop re read thoose last two sentences let it sink in. and continue - trust me been there done that, seriuosly screwed up my phone! dont risk it! Metal can and 99.999999% will damage the PCB especially the solder shield underneath it and maybe even create a short you didnt want.

Now once youve removed it toss the stock heat pad - if you want - check the thermal vias, they might need reworking, mine did, i had solder oozing out of one and a noticeable air pocket on the other - i heard the air being pushed out as i was handling it! so check your thermal vias. Highly recommend appling a very generous amount of thermal paste to the thermal vias. this will make sure you have good thermal conducivity there as the paste heats itll flow into the vias creating a good thermal conduct.

Now if your going the route of thermal pad no problem cut it out apply to the underside surface of the PCB and reassemble - take note of the milled out portions in the heatsink and there corresponding alignment to the PCB.
IF your going to cover the whole surface in thermal paste, then cover the serveral exposed copper surfaces - com3, reset switch and a few other, wiht electical insulation tape. you dont want thoose to be short on the heat sink, especially the reset switch!
Apply your paste and reassemble again taking note of the cut outs.

Sorry no photos at the mo. its late and i aint stripping my grid down now, but tomorrow i plan on doing so so hopefully - kids permitting- i can get it stripped and do a few photos.

Or you can wait a few post and see what wolfy has to say - prob no need to change the pads thermal paste crap etc etc.

Pads, or paste your choice id recommend adding paste to the vias though either way. just use care and caution when separating the PCB from the bottom heatsink.
sr. member
Activity: 339
Merit: 250
Vice versa is not a meal.
April 23, 2014, 03:31:36 PM
I did the 49.9k resistor mod, but I'm having an issue:



For some reason, although the Gridseeds are running at 7.65 Mh/s collectively, I"m only getting around 3.6 Mh/s effective. It's like this no matter what pool I use. Is this normal, and if not, any idea on how I can fix it?

Can yopmeone join in here and tell what the actual problem is ?
I have to say that i face the same problem with the bfgminer.
On Ubuntu when i use cgminer, i have to restart cgminer very ofter as there plopps an libusb error. On BFGminer verything runs as it should but the effective hashrate is extremely lower...



I did the 49.9k resistor mod, but I'm having an issue:

For some reason, although the Gridseeds are running at 7.65 Mh/s collectively, I"m only getting around 3.6 Mh/s effective. It's like this no matter what pool I use. Is this normal, and if not, any idea on how I can fix it?

Can yopmeone join in here and tell what the actual problem is ?
I have to say that i face the same problem with the bfgminer.
On Ubuntu when i use cgminer, i have to restart cgminer very ofter as there plopps an libusb error. On BFGminer verything runs as it should but the effective hashrate is extremely lower...


Try the cpuminer in my signature, it should work without issues.

wow guys you are genious.!
cpuminer was a little a pain in the ass to configure the script and run it as root and so on... but! It works perfectly.
Cpuminer even adjusts pods to their own specific best frequency, what a charm i love it.!

Thanks a lot all.
sr. member
Activity: 294
Merit: 250
April 23, 2014, 03:17:40 PM

Hi Cbutters,

Wire 1 is to be attached to LOWER pad of the Cut-off R46 bridge...As I pointed in previous pictures/posts...it is the SAME as LOWER row soldering points of the VID resistors...

So you CAN ATTACH to ANY...or ALL of them...as they are electrically connected together... you can see the traces from R46...or measure that...

The trim pot is intended to be a TEST Bench...Read my second sentence in the my previous post...to ADJUST voltage smoothly...It IS a ZiG_Mod for testing/experimenting...

The original first proposal of ZiG_Mod is fixed resistor...Equivalent to R139 ("R52) replacement...but without the need of actual de-soldering...  

I have been running without top heatsink at the early stage of modding...anything above 1000MHz was causing Gridseed chips to become hotter and untouchable...This was the time when I modded the thermal solution and posted here...like end of March...

After that I have them running with copper heatsinks on each chip...RAM/VID type...10mm x 10mm ... Phobya thermal pads... no problems even @1250MHZ...but they are getting HOT...above 1150-1175MHz...

ZiG  


Thanks for the clarification! I think I might try wiring up trim pots up on all of mine if it works good in testing... I like the idea of variable voltage to be able to fine tune so as to be as efficient as possible at around 450-480 hashrate. With that we could find a stable voltage then slowly move down from there. Plust trim pots are quite inexpensive as well. I'll report back once I receive my gridseeds and do some testing. I will post pics / guide if it works well.
sr. member
Activity: 378
Merit: 250
April 23, 2014, 03:12:47 PM

Zigman where are you we need your input  on this problem one of the boys are havin.

I would also like to see some more confirms of the zigmod before attempting it in the next few days.

Zig... in your diagram for the pentiometer, I don't understand why wire 1 is connecting suddenly to 3 different points? (and not even the points you describe in the original Zigmod.)

Is the pentiometer to trim the voltage and to be used in conjunction with the original zigmod you posted? or is this a variant?

Also... is it safe to mine / test voltages without the cap on these things for short period of times for testing?

Hi Cbutters,

Wire 1 is to be attached to LOWER pad of the Cut-off R46 bridge...As I pointed in previous pictures/posts...it is the SAME as LOWER row soldering points of the VID resistors...

So you CAN ATTACH to ANY...or ALL of them...as they are electrically connected together... you can see the traces from R46...or measure that...

The trim pot is intended to be a TEST Bench...Read my second sentence in the my previous post...to ADJUST voltage smoothly...It IS a ZiG_Mod for testing/experimenting...

The original first proposal of ZiG_Mod is fixed resistor...Equivalent to R139 ("R52) replacement...but without the need of actual de-soldering...  

I have been running without top heatsink at the early stage of modding...anything above 1000MHz was causing Gridseed chips to become hotter and untouchable...This was the time when I modded the thermal solution and posted here...like end of March...

After that I have them running with copper heatsinks on each chip...RAM/VID type...10mm x 10mm ... Phobya thermal pads... no problems even @1250MHZ...but they are getting HOT...above 1150-1175MHz...

ZiG  

Why tell others to experiment? Use a trim pot? That can be dangerous if not destructive to the miner if they accidentally over adjust it.
Which is it going to be Zig? Fixed resistor mod or tinker toy mod?
No problems with stock pads either! They work just fine on my overclocked miners. Wink
member
Activity: 84
Merit: 10
April 23, 2014, 03:11:10 PM
i already downloaded the one from crypoblog, then i downloaded sandor111 latest file - which was just the .exe file. the one from cryptoblog had a libcurl_4.dll in it not just a libcurl.dll. i found a libcurl.dll easy enough now its going. pool now shows 720! before the autooverlcok has kicked in. nice!

Great work sandor, but i really miss the summary at top of the miner app - like in bfgminer, and cgminer. how easy would it be to add in Smiley? if it wasnt for the whole complie on windows lack, id knock somehtign up myself. trying to compile it on windows puts me off though.
sr. member
Activity: 378
Merit: 250
April 23, 2014, 03:09:55 PM
Guys, I have been running Cgminer For 4 days Non-stop so far, I have been using Cgminer since the first day I started mining with GPU's I do not have to restart a single time. BFGminer looks weird Look at all the hashrate differences, where is the Frequency? are they all the same or not  - Almost like the pools hashrate averages.

PM me if you want to get cgminer working I will try to work you through it...

Anyways guys guess what? I did Vmod3 with the 49.9kohm  I have done 3 pods so far, I used a soldering iron and the 0402 SMD resistors... Wow what a Rush it is when you get the 1st one done DAMMM... Its like I am operating on an Ants knee cap.... 3rd One guess what happen!!..... here you can learn from me now, so I plugged it in and it would not accept ANY shares at all... So I put it back on the operating table and realised the 0402 resistor was not contacting the 2 pads correctly possibly a bridge .... anyway I re-tried it ... AGAIN still no accepted ... So it goes under the operating table again and this time I decided to take the resistor fully off... and plug it in and run it to see if it is doing the exact same problem this would atleast tell me it is definitley the R52 pads & resistor not connected properly .. anyway so I removed the resistor and plugged it in with empty r52 pads... and too my releif it did not accept shares .. so I now know its definitely that and I didn't burn something else along the way... So I decided to use a brand new resistor and plugged it in BOOOOOOOOOOOOM.... 3rd time lucky this time I connected it nicely and it is accepting shares...

3 Pods MANY to go :\ ...

I definitley see ZIGmod being easier as you do not have to remove the resistor... but now that someone states it not working... makes me afraid to even bother with it....

Zigman where are you we need your input  on this problem one of the boys are havin.

Hi Grtaah...

I am on the road...traveling...but will try to give you a hand...

Here is a picture of mine test bench...the guinea pig...:
 
SNIP


I can recommend you to solder 2 WIRES...marked in WHITE...thin like from old PC speakers...or fan... an inch or so long...

Attach...Solder 20K Trimm Pot...if you have one...Solder One END pin to Wire 1...Middle or Middle AND the other end bridged together to Wire 2...

ADJUST somewhere in the middle ...like 10K each side...Before soldering...

If you don't have 20K Trim Pot...just start with regular 1/8W axial R = 5K...and SWAP them as you go with HIGHER value...let say 7-8K...10K...12K...15K etc...of course - SWAP at the end of the wires...NOT at the board ... Wink

VERY IMPORTANT is to MEASURE the DVDD+ Voltage, supplied to the Gridseeds...so you know where are you standing ...the point to measure is MARKED in the upper corner of the picture to the left of 12V input connector and BIG capacitor to the left of it...It is C42 on the board...big yellow SMD capacitor...DVDD+ is on the INNER side...marked + on the board...I encircled it...GROUND @ 12V connector is the easier for me...or the otherside ...marked "-" on the same C42 capacitor...

Measure ...and write back...
Should be in 1.45V/1.50V...to 1.65V/1.75V range...

Within that range...It should be hashing...

ZiG

EDIT...:

IGNORE the GREEN bridge...it is from the first original posted picture of Sandor111...that I am reusing...

That's intersting if not weird. Doesn't match your original zigmod description at all......leaves us all wondering / guessing. Nice...  Huh

If you can read schematics...I posted one...and follow my previous posts and pictures carefully ...with intend NOT to bash ...but understand ...you will see that this is the SAME thing...

The only weirdness is in your mind, Wolfey... Wink

No bashing, just fact. You are confusing people. That doesn't help anyone, now does it....
hero member
Activity: 616
Merit: 500
April 23, 2014, 03:04:22 PM
aaarrgh !

Minerd-sandor111.exe - system Error
The program can't start because libcurl.dll is missing from your computer. Try reinstalling the program to fix this problem.

off on an internet libcurl hunt now.


Here is the full package: https://www.dropbox.com/s/ttqa9p851siz8oi/minerd-gc3355.zip
full member
Activity: 140
Merit: 100
April 23, 2014, 03:01:47 PM
amix - get the whole package here: http://cryptomining-blog.com/2113-new-cpuminer-fork-for-gridseed-asics-with-hashrate-reporting-and-other-useful-features/ it's sandor's source compiled for windows with all the dll's etc.
member
Activity: 84
Merit: 10
April 23, 2014, 02:54:29 PM
aaarrgh !

Minerd-sandor111.exe - system Error
The program can't start because libcurl.dll is missing from your computer. Try reinstalling the program to fix this problem.

off on an internet libcurl hunt now.
member
Activity: 84
Merit: 10
April 23, 2014, 02:49:51 PM
sandor111 in the config file for cpuminer where we stat the indivudaula frequenncy for each grid, can we reference wiht serial number as well or just come ports?
Also do we need to specify each com port or is there a setting like all that scans each port for a grid?
sr. member
Activity: 252
Merit: 254
April 23, 2014, 02:42:29 PM
I tried the "simple" instructions that are listed in just about every readme/build.txt and I also tried these http://pastebin.com/p2cRgHxd ...no luck with either.  Compiling this stuff in Windows is a real b*tch.  I'm going to attempt it again after work on a clean system, the one I've been using is full of all kinds of crap and really needs to be wiped.

I've got it to work, but I can't recall the exact steps I took... all I know is that it was a mess.

BTW, and sorry for off-topic, I've just compiled an optimized build for Windows (disable FIFO & flow control) + pool hashrate display.
https://www.dropbox.com/s/7z1cj75mhvojjah/minerd-sandor111.exe
This seems to push the pool hashrate beyond of what the chip is hashing at, or maybe it's just variance, but there is definitely some improvement.

I'm testing this build out right now.  Getting better results with it than the previous builds (including the one from cryto blog).  The first couple builds would show 989KH/s client side.  But while clevermining pool showed similar rates, manicminer pool showed half that.  With this newest build, I'm getting consistent hashrate reporting both clientside and poolside for both pools I've tried.
ZiG
sr. member
Activity: 406
Merit: 250
April 23, 2014, 02:29:28 PM
Guys, I have been running Cgminer For 4 days Non-stop so far, I have been using Cgminer since the first day I started mining with GPU's I do not have to restart a single time. BFGminer looks weird Look at all the hashrate differences, where is the Frequency? are they all the same or not  - Almost like the pools hashrate averages.

PM me if you want to get cgminer working I will try to work you through it...

Anyways guys guess what? I did Vmod3 with the 49.9kohm  I have done 3 pods so far, I used a soldering iron and the 0402 SMD resistors... Wow what a Rush it is when you get the 1st one done DAMMM... Its like I am operating on an Ants knee cap.... 3rd One guess what happen!!..... here you can learn from me now, so I plugged it in and it would not accept ANY shares at all... So I put it back on the operating table and realised the 0402 resistor was not contacting the 2 pads correctly possibly a bridge .... anyway I re-tried it ... AGAIN still no accepted ... So it goes under the operating table again and this time I decided to take the resistor fully off... and plug it in and run it to see if it is doing the exact same problem this would atleast tell me it is definitley the R52 pads & resistor not connected properly .. anyway so I removed the resistor and plugged it in with empty r52 pads... and too my releif it did not accept shares .. so I now know its definitely that and I didn't burn something else along the way... So I decided to use a brand new resistor and plugged it in BOOOOOOOOOOOOM.... 3rd time lucky this time I connected it nicely and it is accepting shares...

3 Pods MANY to go :\ ...

I definitley see ZIGmod being easier as you do not have to remove the resistor... but now that someone states it not working... makes me afraid to even bother with it....

Zigman where are you we need your input  on this problem one of the boys are havin.

Hi Grtaah...

I am on the road...traveling...but will try to give you a hand...

Here is a picture of mine test bench...the guinea pig...:
 
SNIP


I can recommend you to solder 2 WIRES...marked in WHITE...thin like from old PC speakers...or fan... an inch or so long...

Attach...Solder 20K Trimm Pot...if you have one...Solder One END pin to Wire 1...Middle or Middle AND the other end bridged together to Wire 2...

ADJUST somewhere in the middle ...like 10K each side...Before soldering...

If you don't have 20K Trim Pot...just start with regular 1/8W axial R = 5K...and SWAP them as you go with HIGHER value...let say 7-8K...10K...12K...15K etc...of course - SWAP at the end of the wires...NOT at the board ... Wink

VERY IMPORTANT is to MEASURE the DVDD+ Voltage, supplied to the Gridseeds...so you know where are you standing ...the point to measure is MARKED in the upper corner of the picture to the left of 12V input connector and BIG capacitor to the left of it...It is C42 on the board...big yellow SMD capacitor...DVDD+ is on the INNER side...marked + on the board...I encircled it...GROUND @ 12V connector is the easier for me...or the otherside ...marked "-" on the same C42 capacitor...

Measure ...and write back...
Should be in 1.45V/1.50V...to 1.65V/1.75V range...

Within that range...It should be hashing...

ZiG

EDIT...:

IGNORE the GREEN bridge...it is from the first original posted picture of Sandor111...that I am reusing...

That's intersting if not weird. Doesn't match your original zigmod description at all......leaves us all wondering / guessing. Nice...  Huh

If you can read schematics...I posted one...and follow my previous posts and pictures carefully ...with intend NOT to bash ...but understand ...you will see that this is the SAME thing...

The only weirdness is in your mind, Wolfey... Wink
ZiG
sr. member
Activity: 406
Merit: 250
April 23, 2014, 02:02:39 PM

Zigman where are you we need your input  on this problem one of the boys are havin.

I would also like to see some more confirms of the zigmod before attempting it in the next few days.

Zig... in your diagram for the pentiometer, I don't understand why wire 1 is connecting suddenly to 3 different points? (and not even the points you describe in the original Zigmod.)

Is the pentiometer to trim the voltage and to be used in conjunction with the original zigmod you posted? or is this a variant?

Also... is it safe to mine / test voltages without the cap on these things for short period of times for testing?

Hi Cbutters,

Wire 1 is to be attached to LOWER pad of the Cut-off R46 bridge...As I pointed in previous pictures/posts...it is the SAME as LOWER row soldering points of the VID resistors...

So you CAN ATTACH to ANY...or ALL of them...as they are electrically connected together... you can see the traces from R46...or measure that...

The trim pot is intended to be a TEST Bench...Read my second sentence in the my previous post...to ADJUST voltage smoothly...It IS a ZiG_Mod for testing/experimenting...

The original first proposal of ZiG_Mod is fixed resistor...Equivalent to R139 ("R52) replacement...but without the need of actual de-soldering...  

I have been running without top heatsink at the early stage of modding...anything above 1000MHz was causing Gridseed chips to become hotter and untouchable...This was the time when I modded the thermal solution and posted here...like end of March...

After that I have them running with copper heatsinks on each chip...RAM/VID type...10mm x 10mm ... Phobya thermal pads... no problems even @1250MHZ...but they are getting HOT...above 1150-1175MHz...

ZiG  
hero member
Activity: 616
Merit: 500
April 23, 2014, 01:53:16 PM
I tried the "simple" instructions that are listed in just about every readme/build.txt and I also tried these http://pastebin.com/p2cRgHxd ...no luck with either.  Compiling this stuff in Windows is a real b*tch.  I'm going to attempt it again after work on a clean system, the one I've been using is full of all kinds of crap and really needs to be wiped.

I've got it to work, but I can't recall the exact steps I took... all I know is that it was a mess.

BTW, and sorry for off-topic, I've just compiled an optimized build for Windows (disable FIFO & flow control) + pool hashrate display.
https://www.dropbox.com/s/7z1cj75mhvojjah/minerd-sandor111.exe
This seems to push the pool hashrate beyond of what the chip is hashing at, or maybe it's just variance, but there is definitely some improvement.
Pages:
Jump to: