For Linux users

For Linux users

Arranging meeting - How and why (a reason to exist)

Attending: Tony, Les, Ollie, Aaron, Joe and Mike. (Nearly forgot James.)

Continuation of the discussion started on the mailing list about how to organise meetings.

The main points. Discussed but not yet agreed.
If an invite system is to be used, it would need to be a 'blanket' invite to everybody on a list.

We do not want to accidentally leave someone off and have them think there is no meeting or worse, that they have not been invited because we do not want them to come.

People should be able to join and leave the list themselves, so we do not have to allocate someone to construct the list.

Anyone on the list should be able to instigate an invite if they want a meeting to happen. This takes the burden off one person always having to remember to do it.

A method of cancelling should exist if there is no response to an invite.

A couple of us felt that the existing mailing list could handle this. It has been the one stable thing that has always been there since the group formed in 2004, while various other options we have tried have been and gone.

Currently, we get apologies on the list from people who are not going to attend. This does not give any indication of who is going to attend, or if there is even going to be a meeting.

Switch this around. An invite goes out, people respond to say they are attending and the meeting happens. No response and the invite is cancelled.
It may need fine tuning, but it covers the points discussed.

NOTE..  The above has not been agreed. There are plenty of other options available.

The second part, was the suggestion that we should make it known what was going to happen each meeting. This generated another discussion.

Main points. Again, not yet agreed.

Most felt that the current ad-hoc, laid back system works well. Trying to add structure was not seen as a positive move.

Currently, this means that we do not advertise what we are going to be doing, and people just turn up and get on with whatever they want to do.

However, it was suggested that the facility should be there to organise sessions if members want them.
Previous examples of this working well were mentioned :- Jon put on a Python tutorial for beginners, which stretched across several weeks, and was well received. Jon also did a video editing tutorial, again well received.  When we recorded a video of a computer being dismantled and reassembled, that was one of the best attended meeting we have had:-
http://blackpoollug.blogspot.co.uk/2013/07/blackpool-lug-dismantle-computer.html

Another point.

It appears that the group has evolved into what was described as a 'gentlemans club',  meaning that it is becoming a closed entity just for our own private use. "our" meaning the core members who regularly turn up.

Given that a user group is about the members that attend, and what they want, does the above 'gentlemans club' description mean that we are stagnating,? OR

Does it mean that because this is happening by default, we have a good setup, and this is actually what the active members want? 

I have heard from other groups who have driven up numbers, only to become unmanageable, and unsustainable as a large group, and subsequently split into smaller sub groups.

The LUG is a small group, is that what we want, or should we try to increase numbers?

Suggestions came up about things we could do to increase numbers, but again, it was not agreed that we should do any of the following:

Promoting the Makerspace.

Install fests.
to coincide with the new release of Ubuntu/mint. This is also a good time to promote the light distros to replaced the expiring XP.

repair fests (restart parties)
http://therestartproject.org/

I guess it all resolves down to our reason for existing as a group.
We have no mission statement or objectives. What is driving us, where are we going, will we know if or when we arrive?

Reading through all this again, there are a lot of questions, and it seems like the easiest option is to ignore it all, and wait to see what happens :)

But is that the best option?








Puppy Linux - Wireless and WPA Vs WEP

Geoff signed up to the LUG mailing list this week... Welcome Geoff.
 
"Hi Mike,      
Got your email address from a B'pool Ubuntu site. Think
we may have corresponded before. Anyway see you have meetings (with
biscuits!) in Blackpool, am hoping one of your members could help if I
attended the next one. I have a Satellite SP6000, I am trying to put
12.04LTS on it with no success over a number of weeks. When I tried
Mint mate it flashed a message about BIOS needing update. 
Cheers Geoff ."

I checked the Satellite spec online, and it came with 256M memory, so I replied to Geoff indicating that most of the major distros like Mint and Ubuntu 12.04 would struggle, and suggested he try Puppy Linux.

Geoff had no luck with Puppy either, and when he arrived at the meeting, the laptop had a clean install  of Windows XP on it.

It turns out that Geoff was trying Linux because XP would not connect to his wireless router . But neither would Linux. Geoff had updated the bios, clean install of XP, added an extra 512M memory, and installed the latest wireless card drivers from Toshiba.

All to no avail, because the wireless card in the laptop does not support WPA .

I added an extra wireless netgear router on WEP to our network and the laptop connected perfectly in XP and Linux.
I know WEP is insecure, but if that is all you have.........

Puppy is great for low spec machines, it even warned us about the WPA problem!

Puppy booting


Network config

Warning about WPA

Brightbox was on WPA - can't connect

Netgear on WEP


Connected!

Meeting 2014-03-08 not really a LUG meeting?

Attending:  Joe and me (Mike)


Joe uses Apple computers, and I prefer a Windows PC, so not really a Linux meeting this week.

Joe has a Virtual Private Server (VPS) and was working with Gary's Mod on a game server.

I decided to install Windows 7 on the Packard Bell imedia box pictured below  to use it as a media centre at home. Windows 7 comes with a free version windows media centre. Windows 8 does not.


After a clean install, about 130 updates downloaded, and took several hours to install. This was followed by more updates, and then service packs which took several more hours.

Media centre itself was straightforward to configure once the USB tuner problem was sorted out.
Basically, windows tried to install the tuner itself, but wrongly identified it, so it didn't work. I installed the driver from the disk, but each time windows did updates, it put the tuner back to none working. Eventually I identified the update that was doing this and disabled it. 
The power saving scheme that I set up didn't seem to work during testing until I realised a reboot was needed for the setting to work. 
Sound comes out of the TV speakers via the HDMI cable, but needed adjusting. At first, with the TV remote sound settings on normal, I couldn't hear anything because the computer volume settings needed to be turned up to maximum.
The hard disk makes too much noise when it is recording, I am going to try an acoustic jacket on the disk, but if that doesn't work, I will image it across onto a quieter disk.

Now a bit about Linux.
Some of you will know that I have made several attempts to use Mythbuntu as a media centre, and failed each time, even though I have had Donald expertly guiding me. 
The problem is probably this: I have been using Linux since the mid '90's, first Slackware, then on to Redhat, Mandrake, Mandriva, PClinuxOS, Knoppix, Puppy, Debian, Ubuntu, Mint ............. you get the picture, Distro hopping. 
I have never really stayed on one distro long enough to really get to know it enough to sort out problems when things goes wrong.

I wouldn't call myself a windows expert either, but I have always used it, and tend to recognise problems when I see them, and have an idea about what to do. 
I said earlier that windows media centre was easy to set up, but it had just as many problems to overcome as MythTV had, so the easy feeling must be down to my familiarity with windows.

The reason I bring this up now is that the media centre install highlights the issue.
The problems I experienced setting up media centre on windows more or less mirror the problems I experienced with Mythbuntu, but I was able to sort the problems on windows but not on Linux.

If I had stayed with Slackware from the outset instead of distro hopping, chances are I would have been able to sort out the MythTV problems on Linux too.





Meeting 2014-03-01 Logic probe nears completion

Attending: Tony, Elizabeth, Kieran, Joe, Mike, and welcome to a new visitor, Stuart, who has just moved to Blackpool.

Tony installed a LAMP stack into virtualbox with the intention of doing some web work on the likes of Drupal, Wordpress, Mambo, and/or Joomla.
Elizabeth has been dual booting Suse Linux and Linux Mint for a while now, and has come to the conclusion that Mint is 'The One'
Kieran was working on a game using Python, and Joe was working on a game server.
Stuart described his introduction to computing:
"Started in 86 after college doing old dos support with things like wordperfect and lotus123 and have been doing it for so long that I guess I am caught now.First computer a ZX81 in 79 Oooof those where the days :)"




The breadboard version of the logic probe working.

It looks a bit scary with all the jumper wires, but there are only a few components involved.
The next step is to assemble it into the barrel of a pen.


A logic probe is a useful piece of equipment to have when fault finding on digital circuits.
Here is a description of a logic probe on Wikipedia http://en.wikipedia.org/wiki/Logic_probe



Some very basic theory using this diagram.
The two jagged lines labeled R1 and R2 are resistors.
R1 will be connected to the positive supply, and R2 will be connected to the negative supply, 
which will provide Point A (R1) with the voltage of the circuit under test, which can be up to about 15Volts for the CMOS chip we are going to use later. Here we will use 12volts, and Point B (R2) will be 0volts.

With point C unconnected, current (electricity) will flow between point A and B through the resistors R1 and R2.

If point C is used as a probe and connected to 12volts, no current will flow through R1 between points A and C, because they are both at 12volts. This is described as there being no potential difference (PD) between point A and Point C.
However, there is a 12volt PD between point C and point B, and current will flow through R2.

If the connection is now reversed, and the probe (point C) is connected to 0volts, the opposite happens, and current flows through R1

Electricity always takes the path of least resistance 
When point C is connected either way around, it cuts out one of the resistors from the circuit.





Add two Light Emitting Diodes (LED) to the circuit, back to back as shown.
Now when the probe, point C is used, one LED will light up when connected to 12volts, and the other will light up when connected to 0volts.
LEDs are diodes, and will only pass current in one direction, which is why both LEDs do not light up at the same time. If the probe was connected to an output that pulsed high and low, then both LEDs would flash on and off, but in most cases, that would happen too fast for the eye to see.



This circuit can not be called a logic probe for a couple of reasons, but it does supply a basic understanding.

A third LED is needed to indicate pulsing outputs. This LED needs to be connected to additional circuitry to be able capture and latch a single transient pulse, or to stretch out high speed pulse trains long enough for the eye to see that pulses are happening. A CMOS 4001 integrated circuit will be used to achieve this. 

Also, attaching the probe should have as little effect as possible on the circuit under test, or in other words, attaching the probe should not cause the circuit under test to behave differently. This can be a very complex topic. Here is an article which covers oscilloscope probes called "understanding the probing problem" http://teledynelecroy.com/doc/probes-probing

Here is a full circuit for construction, provided by http://www.zen22142.zen.co.uk/Circuits/Testgear/lprobe.html

The two resistors and two LEDs previously described can be seen in this circuit, near the probe. The LEDs are marked HI and LO.
IC1, the CMOS 4001, costs a few pence.





The logic probe can display four output states, High, Low, Pulsing and tri-state (or high impedance). The tri-state output is a high impedance state in which the output pin has no value, it is not at logic 0 and not at logic high.

IC1a (Pinout pins 1,2,3) is wired as an inverter with a 2M2 feedback resistor. With the probe not connected to a logic circuit, the output of gate 1a, pin3 is fed back to the input, pins 1 and 2 via the 2M2 resistor. This gate will oscillate at a very high rate and resultant output voltage at pin 3 will be approximately half the supply voltage. The Hi and Lo logic indicator LED's are also connected to a potential divider consisting of the two 1k resistors. The voltage at the resistor junctions is also half supply voltage so with no input, no output LED's light representing the tri-state. 

A Hi or Lo logic condition at the probe input, will cause IC1a to rest in a permanent state indicated by either the Hi or Lo LED illuminating. 


With a fast oscillator or clock signal input both Hi and Lo LED's will light but appear quite dim. 


To increase brightness, the input signal is slowed down. This is achieved by gates IC1b and IC1c which are wired as a monostable. 
The time constant for the monostable is determined by the 100n capacitor and 4M7 resistor. A fast input pulse now continually triggers and re-triggers, the monostable, effectively slowing the input signal. The output of the monostable is inverted by IC1d, wired as an inverter, and increasing output current to the pulsing LED. 

http://electronicsclub.info/cmos.htm  supplies this info about CMOS chips
  • Supply: 3 to 15V, small fluctuations are tolerated.
  • Inputs have very high impedance (resistance), this is good because it means they will not affect the part of the circuit where they are connected. However, it also means that unconnected inputs can easily pick up electrical noise and rapidly change between high and low states in an unpredictable way. This is likely to make the IC behave erratically and it will significantly increase the supply current. To prevent problems all unused inputs MUST be connected to the supply (either +Vs or 0V), this applies even if that part of the IC is not being used in the circuit!
  • Outputs can sink and source only about 1mA if you wish to maintain the correct output voltage to drive CMOS inputs. If there is no need to drive any inputs the maximum current is about 5mA with a 6V supply, or 10mA with a 9V supply (just enough to light an LED). To switch larger currents you can connect a transistor.
  • Fan-out: one output can drive up to 50 inputs.
  • Gate propagation time: typically 30ns for a signal to travel through a gate with a 9V supply, it takes a longer time at lower supply voltages.
  • Frequency: up to 1MHz, above that the 74 series is a better choice.
  • Power consumption (of the IC itself) is very low, a few µW. It is much greater at high frequencies, a few mW at 1MHz for example.

Meetings - January 2014 overview

Attending: The usual suspects

I can't believe it is the middle of February already, I took a couple of weeks off over Christmas, and it has taken me ages to get back 'in the swing'. I didn't write up any of the January meetings, so this is a brief overview of what we got up to.

Work on the soldering/electronics area is finished.
The  workbench is up,  the extractor is installed and  general purpose tools are available, including temperature controlled iron, solder sucker, multimeter, usb oscilloscope,  hand tools, cutters, pliers, clamps, tweezers, jumper wire, croc clips, breadboards, etc........
Also, multiple parts bins and old computer boards to practice on.



The G4 powermac has been abandoned by Apple, and was no longer viable for use on the internet. (no up to date browser) 
Wiped the hard disk, and installed Debian for PPC. The installer is exactly the same as the PC installer, once you have learnt the 'magic' mac keypress to get the CD to boot.


Installed and working fine, but what I don't understand is the really low bogomips score, these scores are in the thousands on PCs. Maybe because PPC is RISC, it doesn't need to do millions of instructions per second (mips) to get things done.


Our 37" TV has been broken for a while. It started to switch itself off after a couple of hours of use, then the time it would stay on got shorter and shorter, until it would not stay on at all.

Donald and I decided to attempt to repair it. 
There is only two (PCB)  boards inside, a power board and a picture board, so a 50/50 guess as to where the problem is?
Given the way the problem developed, it was a fair guess that the psu board had overheated and eventually failed.
Donald researched the board and we ordered this:-




PSU board removed


New PSU in place


And here it is working again on the back wall.

And finally, the hard disk clock on the wall.