The technical specifications and details can be seen here at fluke

But the highlights are:

• 6.5 digit resolution
• Basic V dc accuracy of up to 0.0024 %
• Graphical Dual display
• 100 uA to 10 A current range, with up to 100 pA resolution
• Wide ohms range from 10 ohm to 1 Gohm with up to 10 uOhm resolution
• 2 x 4 ohms 4-wire measurement technique
• Measure frequency and period
• Measures capacitance and temperature
• USB memory drive port for logging
• Communication interface IEE-488.2, Ethernet and RS 232 (including an RS232 to USB adaptor)
• Communication protocol as Fluke 884x or Fluke 45 emulation and Agilent 34401A emulation
• Trendplot paperless recorder mode, statistics, histogram
• CAT I 1000 V, CAT II 600 V
• Three-year warranty
   

Unboxing the package with the fluke 8846A meter



Front view with the only test leads there was include, so the package don’t include any 4-wire test leads so I have just made me own kelvin 4-wire test lead of some twin 0,75” good wires, 4 banana plugs and some kelvin clips that I got from EBay. I will not make a photo of them but you can see them in the other photos in this blog.



Rear view with the 3 interface ports, trigger port and banana port for test leads.



Measurement of an 10K ohm 0.01% test resistor
On left the Fluke 8846A with 4 wire test lead and on the right my Fluke 287



Measurement of an 1G ohm 5% test resistor (it is two 2Gohm resistor in parallel)
The Fluke 287 cannot measure this one because it is limit to 500Mohm but the Fluke 8846A can measure it without any issues.



For the DC volt measurement, I have an HP 6632B power supply there is also having the 4-wire output terminals so it will automatically adjust the output to compensate for the loss in the wires and there for the output will be very precis.
Here is a photo of the HP 6632B and the black box on the left is the output terminals.



HP 6632B output is set to 0.007 Volt DC, on the left is a Fluke 287 and on the right the Fluke 8846A



HP 6632B output is set to 0.502 Volt DC, on the left is a Fluke 287 and on the right the Fluke 8846A



HP 6632B output is set to 5.002 Volt DC, on the left is a Fluke 287 and on the right the Fluke 8846A



HP 6632B output is set to 12.001 Volt DC, on the left is a Fluke 287 and on the right the Fluke 8846A



For the A measurement in then as with the volt DC measurements I can use the HP 6632A power supply by setting the volt to the max 20V and then set the current limit function to a given value, then the test leads are then simple short circuit over the inputs of the meters.
HP 6632B output is current limit to 0.0022 A and this is the lowest value that I can set the power supply to.
On the left is a Fluke 287 and on the right the Fluke 8846A



HP 6632B output is current limit to 0.1025 A, on the left is a Fluke 287 and on the right the Fluke 8846A



HP 6632B output is current limit to 1.0011 A, on the left is a Fluke 287 and on the right the Fluke 8846A



HP 6632B output is current limit to 5.0019 A, on the left is a Fluke 287 and on the right the Fluke 8846A



To do AC volt measurement I then have this AC variable transformer 0-270VAC 2.43KVA and here in Denmark the power grid is 50Hz and 240 Volt AC



Measurement of the 50 Hz, on the left is a Fluke 287 and on the right the Fluke 8846A



Measurement of the 25 Volt AC, on the left is a Fluke 287 and on the right the Fluke 8846A



Measurement of the 100 Volt AC, on the left is a Fluke 287 and on the right the Fluke 8846A




Measurement of the 240 Volt AC, on the left is a Fluke 287 and on the right the Fluke 8846A



Measurement of a capacitor Tantalum 1,5uf 35V, the Fluke 287 shows it as 1,8uf and the photo show the Fluke 8846A



Measurement of a capacitor 68uf 400V, the Fluke 287 shows it as 64,4uf and the photo show the Fluke 8846A



Measurement of a capacitor 10000uf 25V, the Fluke 287 shows it as 12.0mf and the photo show the Fluke 8846A



Remote controlling the Fluke 8846A from a PC then there is 3 ports on the back and they are IEEE488, LAN and RS232.
I have tested all 3 ports and they works fine without any issues  but here I will show it via the RS232 com port because it is the easiest and echo’s the commands so it is readable.

To tell you what this output means I will then try to explain what I have done here line by line. The "x>" number at the beginning of most of the lines is just line numbers and the meter automatic count them up, so it is not something I have entered.

"System:remote"
This tells the meter to change into the remote command mode and the meter needs to be in this mode in order to response to other commands, this also disable the keys on the front of the meter.

"*IDN?"
This asks the meter to identify itself and the response from the meter is "FLUKE,8846A,xxxxxxx,08/02/10-11:53"
Note: The xxxxxx is the serial number that I have hidden.

"CONFigure:VOLTage:DC 10"
This tells the meter to change the measurement to DC Volt and with the range to 10 Volt

"Read?"
This simply tells the meter to take an measurement and return the value and it has responded with "+4.99488600E+00" and this is 4.99486 Volt DC

So this is how easy it is to control the meter from a PC and if I find the time to do it, then I will make a small application so I can get some smart live logging with cool graphs and a lot more ideas that I have…. But the issue here is the time to do it.


So now I am at the end of this long blog post and I must say that I am very happy with this new cool gear that I got it for free from Fluke, so thanks Fluke and I am sure that I will be very happy with this for years and it will give better and more precise measurements that I have been missing because until now I have only had handhold meters like my Fluke 287 and Fluke 789


Videos, I will upload some video clips and you can see them here

]]> Tooms http://www.tooms.dk/tblog/showblog1.asp?BlogID=201111082030033196 Blog Tue, 08 Nov 2011 20:30:03 +0200 http://www.tooms.dk/tblog/showblog1.asp?BlogID=201111082030033196

In my process of setting up my new servers i was missing the good old harddisk activity LED there was on the old PATA harddisk where you can connect an LED to each drive, so i have looked into how to do this on the SATA harddisks.
There is always a LED connector on the mainboard but that is just showing the activity for the controller chip on the mainboard and not for each harddisk,
So after some internet searching i then found two things. one, a lot of posts around the net there was trying to do the same thing but was unable to get it working and number two was that there was a pin 11 in the SATA power connecter there was labeled as "activity indication" on supported drives, so it seems that you just have to connect a LED to that pin.... how hard can that be.

I then researched and was reading a lot of posts on the net and then found the information, if you connect the SATA power cable to the connector on the harddisk then the PIN 10+11+12 is connected to the GND and because this PIN 11 there is the activity pin is pulled down to GND then it will not be able to drive a LED and show the activity, so it has to be floating and not connected to GND..... but that is not as easy as it sounds.

Here you can see the SATA power connector (the big one of the two) and what signals there is in the connector.



Here i have pulled the SATA power connector apart so you can see how it is connecting and note that each wire is having a connection to 3 pins and there for it is not very easy to just disconnect the pin 11 in the power connecter and then reconnect another wire for the LED and this power connector is the only one i have there can be pulled apart, all the other ones i have cannot be open up.
Another way is to disconnect the pin 11 on the harddisk drive and then connect a wire on the print board, but i don’t like that as that will break any warranty.



So i was looking for another way to do this and was thinking that there must be some sort of adaptor on Ebay that i can get for a low price and use to this and after some search time i then found this adaptor.

It is perfect because you can see each connection between the two connectors and easily modify it to the need i have.



Here you can see the cable connector, the adaptor and the harddisk drive as they will be connected.

Here I have drilled a hole in the connection of pin 11 between the two connector and there for it is no longer connected to the GND.
Then i have clean off the green stuff so i can connect the wires to PIN 4(GND), PIN 9(+5V) and PIN 11(activity)



Here i can connected the 3 wires to the adaptor and Black is GND, Red is +5V and White is activity.

Here is a diagram of how the LED is connected to PIN 9 and 11, only connect one LED and a 1K resistor as it is limit how much current that there is in the harddisk for the activity led.
So if  you like to connect more then use a optocoupler to make sure that there is not pulled to much current on PIN 11.


So now i have a flashing LED for the harddisk activity on each drive and i have not changed anything on the harddisk drives and/or the cabling so the warrant on the disks are good.


So if you’re getting the SATA adaptors from eBay like this one then it is not that hard to get the led to work and it is nice to now have a led for each harddisk drive.
]]> Tooms http://www.tooms.dk/tblog/showblog1.asp?BlogID=201110252324515723 Blog Tue, 25 Oct 2011 23:24:51 +0200 http://www.tooms.dk/tblog/showblog1.asp?BlogID=201110252324515723
This is blog post number 2 in my series about the build of my new home server setup and in this blog post I will look at what wattages and heat the parts in the server is having, so in this post i will look at the watt and heat for disk drives, interface cards and RAM blocks.



How much wattage are the SSD and HDD using?
So to better know how much the disk drives are using i then look at what drives i was having and found that there was 4 different hard disk where there was 2 of each model and then i have 2 of the SSD.
So by this i can then measure the use of 2 disks at the same time and then just divide the result with 2 and then have a bit more precise measurement then if i only have one drive. So the setup is this, one Fluke 789 meter is measuring the 12V and another Fluke 789 is measuring the 5V and a Fluke 177 to measuring the amps for the 12V wire and then a Fluke 287 to measuring the amps for the 5V wire.
On the photo below you can see the 4 meter for measuring the disk drives and on the right there is the Fluke 345 Power Quality Clamp Meter there is measuring the wattages on the 230V AC.

I have done the measurement by connection 2 disk drive to the system and then using the average function on the fluke meters by first have the drives doing nothing (idle) and measuring the average amps over 10 minutes and then i have done the same 10 minutes average measurement when the drives was at the highest load that i was able to do by using the IOmeter software to run a benchmark.

So it is not just marketing talk that the SSD's are using much less wattage then the old hard disk drives and one SSD can have  the same performance as many hard disk drives so with that the saving will be even better.
My old server is having 4x Seagate Barracude 7200.10 disks in a raid setup and that will give number around 20.8W in idle and 26,8W with high load for the disks, so if replaced by 3x SSD's there will use 1,7W idle and 4,7W with high load.... so this will give a saving at around 15-25W, This is a nice saving on the change from harddisk drives to SSD disk.
The only issue is the cost of the SSD, small GB size and limit lift time.




How much heat is there then coming from the disks?

So to make a comparable heat profile i have then power all the disk on for 4 hours in idle mode with just the power connector on and not any pc and then i run the disk at full load for 30 minutes to get them to heat up.
Then i have taken a thermal image with a Fluke TI25 thermal image camera of each pair of disks.
The image temperature color scale is set low point as 20 and high point as 65 degrades for each image to make sure colors and images are comparable.

HDD: Western Digital Caviar SE WD800 80GB 7200RPM 3.5"

The graph on the right shows the temperature across the to disks on the red line.
Seems to have okey temperature for 3,5" and the heat is coming from the main chip.


HDD: Western Digital Raptor WD360 36GB 10000RPM 3.5"

The graph on the right shows the temperature across the to disks on the red line.
The print board is have some hot spots, but for 10krpm disk i am a bit surprised that it is not hotter then this.


HDD: Seagate Barracude 7200.9 80GB 7200RPM 3.5"

The graph on the right shows the temperature across the to disks on the red line.
Wow that chip is getting very hot and the center of them is 72 degrades so if your using disks like this then your better make sure there is some type of air cooling across the disk.
This is when the disk is placed in free air on the floor so i can only guess that they will get even more hot when inside a computer case.


HDD: Seagate Momentus 7200.2 160GB 7200RPM 2.5"

The graph on the right shows the temperature across the to disks on the red line.
This is disks from a laptop and it is easy to see why laptops get so hot when they have hard disk drives installed.


SSD: Crucial M4 256GB 2.5"

The graph on the right shows the temperature across the to disks on the red line.
Wow, nice to see the SSD are so much cooler then the HDD's and if you look at the graph on the right you can see how low and cool the drives are in comparison to the other hard disk drives.

So this is super nice that i don't have the hard disk drives high heat in this new server and is just having the nice and cool SSD there is using alot less watt..

Wattages for RAM blocks.
By change the amount of memory on the mainboard and taking a number of reading with the Fluke 345 Power Quality Clamp Meter on the main 230V AC power, I have then calculated that 8GB(2x4GB) Corsair 8GB DDR3 1333Mhz XMS3 is using around 1,727W and 16GB(4x4GB) Corsair 8GB DDR3 1333Mhz XMS3 is using 3,454W, so from this it seems the ram blocks is not using much wattages and i think that the 8gb will cost more than the 16gb will in the long run because if there is only 8gb ram then the disk system will do a lot more swapping to disk and by this use extra watts.
so why not use 8gb blocks, it is because the 8gb blocks cost too much and if the prices was not that high then i will have installed 32GB(4x8GB) on the system, but this may come as a upgrade at a later time when the prices has come down, so for now it is only 16GB

Wattages for network cards.
So by doing the same as when taking measurement of the memory i have the cards changed around and taken a number of reading with the Fluke 345 Power Quality Clamp Meter on the main 230V AC power and from this i have found that the network cards cost around 2.5 to 4.5W per card and in the table below you can see the cards that i have tested with and how many watts they using.

Model	Interface	Ports	Total Watt	Watt per port
Intel PRO/1000 PT Desktop Adaptor	PCIe 1 Lane	1	2,272W	2,272W
Intel PRO/1000 MT Desktop Adaptor	PCI 32 bit	1	4,454W	4,454W
Intel I340-T4 Server Adaptor	PCIe 4 Lane	4	4,363W	1,090W

So because i have to make a Etherchannel of 3-4 ports between the server and the switch i will then use the Intel I340-T4 server adaptor as they have the lowest watt per port and i can use just one card and not have to use 4 cards, also it seems that Intel writes that is done a lot around the power management to keep the wattages down and be green and they list in the specs. That the card is using 4.3W and that is just what i have measured to use so it seems my readings for the cards are matching very well.

In the next post I will look at what power supply to use and the heat and cooling of the Mainboard, chipset, CPU, RAM and so on.

To see the other blog post in this series then click on the comment link and you will find a comment with links to each post in the series. ]]> Tooms http://www.tooms.dk/tblog/showblog1.asp?BlogID=201110091828348325 Blog Sun, 09 Oct 2011 23:52:34 +0200 http://www.tooms.dk/tblog/showblog1.asp?BlogID=201110091828348325 I am running my own home with VMware ESX and right now there is 7 VM’s, this server setup is old and cannot keep up with what I need, so it is time to look at a new setup there is better than the old one and if it can use less wattages then it will be a plus…

So to begin I then have a look at the setup I am running today to better see what I need the new one to do and where to improve the new one.

The old server is a single server with ESX 3.5 and is running 7 VM’s

Part	Description
CPU	AMD Athlon X2 4400+ 2.2ghz 2xcores and have a TDP = 89watt
Mainboard	ASUS A8N-VM CSM http://www.asus.com/Motherboards/AMD_Socket_939/A8NVM_CSM/
RAM	4 x 1GB Kingston Hyber-X Do to some bios issue then ESX can only see 2,8GB of the 4GB
Harddisk	4 x 250GB 7200RPM of mixed Seagate and Samsung harddisks In the Raid 5 this gives me a VMFS volume on 675GB
Controller	Intel SRCS16 – 6 SATA connections of 1.5 Gbits/s and 64mb cache With battery backup installed
Network	1 x Single port Intel gbit netcard 1 x dual port Intel gbit netcard
Other	1 x DVD drive
Fan controller	T-Balancer BigNG with a number of big silent fans
UPS	APC Smart-UPS 1500 with network interface

So do to some BIOS issue there is only 2,8GB ram for the ESX and VM’s to use and this then mean that it will swap a lot of memory out to the disk and is costing on performance.
For I/O’s on the disk system the highest number I have seen in ESX esxtop is around 3950 cmds/sec.,

Here is a analyze of the load time and what wattage it is using, the analyze is showing what time is use and wattage from the power on to the VM’s start to show idle time and is finish loading the stuff they do.

Time stamp	Watt used at that time	Description
00:00	124 - 126 watt	Power on and the post process
00:02	121 – 132 watt	ESX OS loading and VM’s starting
00:19	110 – 137 watt	All VM’s is started and starting to be alive and reply to ping
00:49	105 – 137 watt	Most VM’s is online and there services are working
01:19	109 – 130 watt	Starting to idle out and harddisk led is no on all time
01:40	105 – 112 watt	ESX and VM’s is now fully loaded the idle out

The power analyze is done with my Fluke 345 Power Quality Clamp Meter and is showing the Volt to be 233,1V AC RMS whit a power factor of 0.657PF

Here is some other wattage in others states

Power state	Watt
ESX 3.5 is load and no VM’s started	98 watt
In the BIOS screen	131 watt

So this is an old server and it is all too slow for me to be usable anymore and the wattage is use is all too high for what it is giving me of power and the only thing to say about it that it has been very stable and very silent design, In fact the server is so silent that I cannot hear it and it is below the background sound from outside the house.
So it serves me well and it is now time for a new setup.

So the new server setup will be something like this

• Low wattage and most be lower than the 105-130watt that the old system use and my goal is to be at max 80 watt and lower be better.
• Ultra low sound, I don’t like fan and harddisk noise so it has to be silent
• Low heat, the server is placed inside a closed rack without any cooling so it has to give away to low heat level as I can get it to do.
• More CPU power, This is not the biggest deal as I most of the time not has max the CPU out and most newer CPU are much more power full then the old AMD x2 4400 is, so here i will try see if I can use a CPU there is having a low wattages but still has the CPU power that I need for the VM’s.
• Memory has been a big issue with only 2,8GB, so the goal is a ESX server with 16GB or more
• Data volume, Harddisk is so last year and I think it will be all SSD to get the speed, low wattage, low heat and low noise.
  Data protection will be some raid level so the data is safe.
• Backup, some way to do backup of the servers, because today I don’t have this and the backup is power off the vm and manual copy.
• Maybe multi ESX servers for the same storage system, so I also can have a test ESX server to play with and the VM’s a storage on the same safe storage.
• Maybe using some of the newer smart things like thin provision, Snapshot, deduplication, host cache and other smart new things.

So what are the plan and the new design?

I've done some research and has now an idea for how this new setup will look like and it will most likely be two servers where one is the ESX server and the other one is a SAN/NAS server, yes I know it sounds a little wild but I think it will give me all the things I need and still be with a low wattages….. I hope but testing will show.
So why two servers, if I only build one new ESX server then I will not get many of the new options and just get a server there is the same as the old one but faster at a lower wattage.
So the idea is to build a server there is a SAN/NAS server with all the smart options like thin provision, Snapshot, deduplication, storage error alerting and many more things.

So now you’re thinking what SAN is he talking about and the answer to that is a free software call Nexentastor there is very cool project there is found here http://www.nexentastor.org/ and it is free for private use up to 18TB.
I have used this product for some years now and it is super cool and the best thing for me is the deduplication option there is allowing me to storage a lot more data on the SSD’s then if I just used the SSD in an ESX server.

So NexentaStor gives me this

• ZFS file system
• Multi GB read cache (the server will have 16gb ram and about 14gb is used for read cache)
• Online gzip compression
• Online deduplication
• Snapshot with almost no extra space cost(do to deduplication)
• RAW volume access via ISCSI or Fiber channel(need adaptor and switch)
• Share access via NFS and CIFS/SMB
• Backup software can access storage data via NDMP, ISCSI, Fiber channel, NFS, FTP, Rsync or CIFS/SMB
• Easy expanding volumes and shares.
• Cool performance monitoring tools.
• High performance I/O and Raid levels without the need for any high price raid controllers.
• Can bundle multi networks interfaces into one multi gbit interface.
• Support multi interface and VLANs
• On the support list at VMware as a supported storage system, in fact they use it on VMworld for the lab systems.
• And lot more.

And the ESX server will look like this

Mainboard
CPU 64bit Intel
16gb ram (Gives a lot of read cache)
1x 64gb ssd for ESXi OS and host cache
4 port Intel 1gbit network interface there is setup as a EtherChannel so it work as one 4gbit link and multi Vlan

Now to the server hardware….
In the research I have done I have found that I will go for the Intel 2100T CPU as it seems to have a low wattages and still have a lot of CPU power, so I think I will be good enough for what I need and for storage I will have to go for SSD’s as they are low noise, low wattage, high IO power and the only bad thing is the cost.
So I have been looking over of what hardware I already have and then order some of the other hardware parts so I can start doing some testing.

So here is the first setup.

Mainboard	Asus P8H67-M 6 sata port, 4 ram slot, onboard vga, onboard netcard, and other stuff
CPU	Intel I3-2100T LGA1155 2,5ghz 2xcores with HT and 3mb cache
RAM	16GB Corair XMS3 4*4gb 1333mhz DDR3
Harddisk OS	1x 146gb 2˝” notebook harddisk
Harddisk DATA	5x 3˝” harddisk of mixed 32gb and 80gb’s
PSU	300 fanless FSP-Group FSP300-60GNF

This setup is using around 35-60 watt and heat and noise is not low do to the harddisk noise and heat

This setup is using around 26-30 watt and there is no heat and noise

So now the watt is down at 21 watt and can go up to 26 watt under high load, so 21 watt server that is cool and a very good start but it will not be that low when I am finish because I have to add some extra netcards and so on, but with 2 that servers it will still only be around 42 watt and then there is some room for adding the extra netcards and so on and still be well under the wattages of the old server..

Here is a photo of the test setup on the floor as it looks right now, so it need a bit more work before it is finish.

Watt Meter
Before you ask how I am measuring the wattages and how precise that is, I can say that I am using a Fluke 345 Power Quality Clamp Meter so it cannot get much better than this and I also have a Fluke 435 Power Quality Analyzer and a Fluke 1735 Power Quality Analyzer but the last two are not very good on low power but I will use them when I have the whole setup put together, but a lot more on this in a later blog.

This blog is part 1 in my build and I will post a number of blog post as I go along with the build. ]]> Tooms http://www.tooms.dk/tblog/showblog1.asp?BlogID=201109241714328811 Blog Sat, 24 Sep 2011 18:24:10 +0200 http://www.tooms.dk/tblog/showblog1.asp?BlogID=201109241714328811

I have for some time been looking for a smart way to connect the test leads when testing as i don't like the old way where i had to turn off the main power and connect some extra test wires into each terminal and then turn on main power again, also when removing the test gear then i had to do the same.
so i was thinking that there must be a better way to do this and maybe some thing there can lock into each hole of the MCB's terminals, but i was unable to find a good way of doing this... so after a long time on Internet search I then found that a company call Electro PJP have made some smart magnet adaptors for 4mm banana connectors.

So I order a number of then from http://www.testoon.com/ to do some testing and see if they was up to any good and i must say they are very good and i like the way it is design, so they just work and i like them very much and they are super easy to connect.
For the size of the magnet in the them, i think that they are holding on good enough for the test lead to stay on but if you pull on the test lead then they will fall out, so i will maybe only use them when doing a short test and for longer testing where it is in when there is no one then i will use the old way as that is more safe.

Here are the details on them and where i got them
Brand : Electro PJP
Manufacturer Ref. : 606MG6.6-IEC3-IVx
Link to web shop: http://www.testoon.com/product-EN-4714-electro_pjp-0_606mg6_6_iec3_ivx_111-connectors_and_adapters.html

Price is 7,55 Euro ex. vat  (That is around 10,2$ ex.vat)
So the I think the price is very high and i don't understand why such a small piece of plastic with a magnet has a so high price tag, but i just need a small number of them so it is okay for me to pay the over cost to fix this problem.

I have also tested some other ones from Multi-Contact MC there is called XMA-7 but they was not very good and they was to big to get into the MCB's small holes and there for can i not use them, they also cost more (12,99 Euro + vat)

Conclusion

Positive

Negative

• Very very high price
• Hard to find webshop/shops there is having them
• Can fall of the connection, so not for longtime testing.

So the conclusion is that i love them allot as they fix a issue and does it in a good way, so if you have the same need for connection test gear like my Fluke 435 and others to the main power on MCB's and so on then i will say go buy the Electro PJP magnet adaptors for banana test lead as they they do the job and the only bad thing to say about them is why the hell do they cost so much and why has Fluke not included them with my test meters in the first place..

So i give them 4 banana out of 5 (only 4 because of the high price)


Thomas

Below is pictures of them in use

]]> Tooms http://www.tooms.dk/tblog/showblog1.asp?BlogID=201005222157282716 Blog Sat, 22 May 2010 21:57:28 +0200 http://www.tooms.dk/tblog/showblog1.asp?BlogID=201005222157282716 But this year the VMworld Europe is placed in Copenhagen at the Bella Center and that is only around 10 km from where i am living, so when it is so close it will be stupid of me not to go so i have now ask the boss at work and got the yes

So now there is payed for the access and i am ready to go for the VMworld Europe, that is super cool.

]]> Tooms http://www.tooms.dk/tblog/showblog1.asp?BlogID=201003082248570263 Blog Mon, 08 Mar 2010 22:48:57 +0200 http://www.tooms.dk/tblog/showblog1.asp?BlogID=201003082248570263

In the past i have made some electronic stuff for some minor projects but most of then ended with the lack of a easy and good way to make the PCB boards and i mostly used the prototype PCB board and then just connected a lot of wires, but at the end of each project it was not very good and it always look bad.

But some years ago i then had seen the blog posts and information around the Internet about how to print your own PCB design with a laser printer and making a PCB board out of them, but i never got around to try out this thing.
So some months ago when i read some pages about it again and was needing some PCB boards for a electronic meter project that i am working on, I was thinking that i will try to make some boards and I download the freeEagle PCB software and made the design and printed it on some Ink paper in the Laser printer and then put it on the copper board.
How the process of Laser printing the design and making the board is not some thing i will write about in this blog post as there is so many others guys on the Internet there has made some very good guides for that, so i will instead link to some of them at the end of this blog post and you can then read it from the good guides.

So i had then made the first PCB boards and was very happy with them and i then showed them to my good friend Kenneth Pallund and hi say "Tooms they look good but the items on the board most be horizontal or vertical and not all over the board as you fit.", hi is my electronic hero so as in the school, I then had to go home and do the homework again. :-)  so I then made another try where i placed each electronic item on the right way and used more time to make a better copper routing on the board and i most say hi was right and this time around it was much better and i guess over time i will learn to make it better and better.

Here is photo of some pcb boards from try number two and i am surprised how good and easy it was to make them and i even made a SMD adaptor board on the low left and on the lower right is a Fluke interface board for my Fluke Scope 123 but it was not working planed and i have not jet found the error on it.
The top PCB is a serial monitor board that i have made to monitor some RS232 serial traffic and meter it, but more on that later in this post.

The ruler size in centimeters

I have also made some double side PCB boards and they are also very good, it is not that hard to make it on the both sides, it just take more time to do and the hard part is to align the two sides, so I just made some extra 1mm hols outside the design and used them to aling the print on both sides and when finish it was spot on.
With two side boards you can make some very advanced designs and make some boards smaller then if they only was single side PCB.

As i small start project for the PCB design and some thing I was needing, I have made a small RS232 monitor board.
The RS232 serial monitor board that i have made is with some very easy to see LED's and some hooks to better connection my scope and/or logic analyzer when trying to understand some serial traffic.
The upper green LED row is when the signal i high and the orange LED row is when the signal is low.
There is a hook for each of the 8 signal lines and below them there is 8 GND hooks and there is a 9 pin female sub-d on the right side and 9 pin sub-d male connector on the left side


Here is a Video demo of the serial monitor board in use when my PC talking to a electric power meter there has a serial port but that is another blog post when i am ready with the software and that project.

I have made a small number of serial monitor boards so if you need one i can maybe sell some of them.



Here is some links to some informations about making PCB boards with laser printer.

Experimenting PCB board creation using a laser printer

HOWTO: An Illustrated Guide

Printed Circuit Boards for the Masses

How to make PCBs at home in 1 hour & WITHOUT special materials

Making PCB With Laser Printer

How-To: Etch a single sided PCB

Youtube.com - Laser print PCB

This is just some links i found via google, so try goolge on it and you will find many pages with alot of informations.
]]> Tooms http://www.tooms.dk/tblog/showblog1.asp?BlogID=201001261853097399 Blog Tue, 26 Jan 2010 18:53:09 +0200 http://www.tooms.dk/tblog/showblog1.asp?BlogID=201001261853097399

The bug was only seen on some versions of windows and they use AM/PM in the date format, this bug was the service see the config has changed every 60 sec. and then use cpu time to process the new config, even it never changed.

Thanks to Alan for reporting this.

It can be download from my software page.

]]> Tooms http://www.tooms.dk/tblog/showblog1.asp?BlogID=201001042306513681 Blog Mon, 04 Jan 2010 23:06:51 +0200 http://www.tooms.dk/tblog/showblog1.asp?BlogID=201001042306513681
This plug-in has long history there started some years back now, but now it is here and i am running this on high load mailsweeper with no issues at all and there has not been reported any bugs from the beta testes.
It is very stable and has processed many mails on my system with out any errors, so i think it has proven that it is now time for this release.

It is design for Mailsweeper.

• Works on 32bit and 64bit
• Fast and small
• Can use multi ClamD server (round robin)
• Log found Virus to a backend Database
• Nice webpage to show Top 10 virus stat.
• Nice GUI for setup

Note this is design for mailsweeper but in fact it is just a commandline app the can be used in many other MTA's like Hmailserver, MailMarshal and many others.
Maybe some day i will include a install guide for some of the other MTA's.

It can be download from my software page in the menu.

Hope you can use it.

]]> Tooms http://www.tooms.dk/tblog/showblog1.asp?BlogID=200912232325495989 Blog Thu, 24 Dec 2009 19:25:49 +0200 http://www.tooms.dk/tblog/showblog1.asp?BlogID=200912232325495989