Following on from the initial creation of the nRF Multi-Network, I’ve had a bit more time to re-think the design. I was starting to think of having each nRF request for nearby nRFs, exchange nRFs addresses that each one can see and then have them build a routing table of the possible paths to take to each other.
However that’s a bit too much effort, so instead I’ve decided that if an nRF wants to forward a packet, that it should forward it to all nRFs and the nRFs that receive it should forward it onto all others and so on. For the moment I wanted to test out checking for neighbouring nRFs so they would just reply back to the sender only.
Just a quick simple project I’ve been meaning to do which is to have a nRF24 Wifi module connected to my 12V battery that’s charged by a solar panel and is running the siren of my alarm system. The nRF reader module will connect via USB to the PC and we’ll output the voltage to a file so we can parse the file anyway we like.
For the TX side, we use an ATtiny84 and as we’ve seen before, we use an analog voltage reference of 1.1V and I’ll use a 1M / 39K resistor divider which allows us to measure up to 28V (not that I expect it to go that high). We’ll send the voltage once a second.
Today we’ll be briefly looking at an old Philips 29″ CRT TV (29SX8771/75R). We’ve had this TV for at least 10+ years and I’ve been told that it cost $2,000 back in the day. The image has degraded a bit so instead of throwing this out, we’ll take it apart and harvest the parts.
Quite a lot of screws later and we’re in; there was 4 speakers which I took out for parts.
It’s very dusty, you can see the main board and the CRT with electron gun. I discharged the CRT by hooking up the plug coming from the CRT (anode) to ground.
Today we’ll be looking the Cisco WAP200 Wireless-G AP with PoE and RangeBooster which as it implies is a 802.11g access point that can be powered from PoE.
Four screws later and we’re in.
We’ve got the processor, flash, RAM and an add-on Wifi card.
A few things worth mentioning which isn’t really standard – the Wifi add-on card is soldered on to the socket so it doesn’t come out. The external shield of both crystals are soldered down. Both antennas are directly soldered to the Wifi add-on card.
From Part 3, we tested out the INCT on a gigabit network and made some changes to the resistor and capacitors, now that the analog switch has arrived we can try switching between the RX/TX lines to the peak detector. The reason we need an analog switch is that we need to isolate the RX-/+ and TX-/+ lines from each other.
An analog switch is just that, a switch which joins two wires together that is activated by driving the control pin either high or low. I went with the M74HC4066 which has minimum supply voltage of 2V and a quiescent supply current maximum of 1uA at 25C which is pretty good. We also need to keep an eye on the resistance of the switches when they are on, as it will never be 0 ohms, the maximum of 170 ohms should be ok.
When re-testing my current configuration without the analog switch I found that if I removed the one of the two network cables the LED would go off as expected but then a short time later it would flicker a little bit. Checking the scope I found some spikes were occurring and if I reversed the RX-/+ lines the issue went away – so I guess there is a specific way that lines should be connected. Also I found that you shouldn’t leave any control pin floating as that cause the analog switch to draw more current (about 100-200uA).
Whilst waiting for the analog switch to arrive as mentioned in Part 2, I decided to try out a gigabit network switch.
The waveform is faster than the 100Mbit switch but is essentially the same however this time when I connected the diode and 0.1uF capacitor (to form a peak detector) and tried copying data over the network it would slow down and when connecting the 47 ohm resistor the network would drop out.
After reducing the capacitor to 10pF it doesn’t drop out anymore and playing around with the resistor I found that a 330 ohm resistor worked well and I also added a 4.7M ohm resistor on the cap. The problem that I found is that the resistor divider’s input (yellow) and output of the peak detector (blue) is a bit noisy, on the left we have how it looks normally and on the right when the 330 ohm resistor is applied, as you can see there isn’t much difference between them – about 100mV average drop.
I had to adjust the resistor divider to be very close to the peak detector output so this is another problem how to adjust the resistor divider to suit the network switch – we can use a digital pot to adjust it automatically however there certain points in the waveform that could incorrectly trigger the LED.
Just a quick teardown today, we have the Kingston 60GB SSDNow300 SSD with model code SV300S37A/60G.
One problem with this SSD are the screws, they are the security torx screws which I didn’t have any screwdrivers available for. Somehow using a small flathead I was able to break off the middle pin so I can now access it using it regular torx screwdriver.
We’re in and there isn’t a whole lot, just 8x NAND chips and the memory controller.
Following on from Part 1, we looked at building our In-line Network Cable Tracer and I was hoping to do a PCB for it however when I connect another INCT to the network cable at the other end, it sometimes didn’t work.
I started playing around with the circuit, removing the capacitor, adding a resistor between AIN1 and AIN0 and tried to use an opamp with it too – none worked well.
But now that I have an oscilloscope I can actually check how the signal looks, something that the multimeter can’t ever help us with (I removed my earth pin on the oscilloscope plug to make sure nothing blows up). We can see that there is just about -1V and +1V to give 2.38V peak to peak.
It’s been a while since I’ve done a teardown, I have a few modems/routers in the ‘waiting to teardown pile’. Today we have the Netcomm NB5 Rev2 ADSL Modem Router.
The PCB has a date code of 0713 so 13th week of 2007. You can tell it’s fairly old because it has an AC input instead of a DC input. They’ve put the SMD parts very nicely grouped together on the bottom left and not a lot of chips as you would expect for an old modem.
I recently bought a Rigol DS1052E and have been playing around with the different functionality it has. Eventually I came to the point that I’d like to see what frequency does to various components and wanted to build a voltage controller oscillator (VCO).
By using 2 opamps and some passive components we can build a VCO as shown above. But I wanted to make one that was simpler so I looked at the ATtiny25.
ATtiny25 Clockout as VCO
The ATtiny25 has a clock out fuse bit which when enabled, outputs the system clock of the ATtiny but the downside is that you can only run on the internal oscillators when you enable that option. As we do with the SATVL, we can use the 16MHz PLL for our system clock and by modifying the OSCCAL value we can underclock and overclock our ATtiny25 – this is the one VCO method we could use. Download ATtiny25_Clockout
Underclocking and overclocking the ATtiny25 – 10MHz minimum to 33MHz maximum.
ATtiny25 Timer1 as VCO
But what if we don’t want to overclock our ATtiny25? I didn’t know but you can actually use the PLL as a clock source for Timer1 and we can keep our system clock at 1MHz.
AdvanceVGA – Play your GBA on the big screen! Swap out the LCD for our board, solder some wires, connect 5V USB and VGA and you’re ready to go.
GBxCart RW allows you to backup GB/GBC/GBA ROMs, save or restore game saves and re-write supported flash carts. Mini RW option available for GB/GBC only.
Wireless Gameboy Controller – Use your Gameboy, mGB, GBC, GBA, GBA SP, GB Micro, NDS and NDS Lite as a wireless controller on Windows, Linux, Raspberry Pi, etc, and on your NES, SNES, N64, Gamecube and Wii.
