[QUOTE=Chryseus;46100479][t]http://u.cubeupload.com/Chryseus/tVq5AH.png[/t]
Now for the fun part!
In case you haven't been keeping up it's a 24 bit audio ADC with USB isolation (up to 5kV), which I'm
going to use with my ECG, and perhaps a narrow band SDR.[/QUOTE]
Enjoy your only 16-bit-or-so resolution :rolleyes:
[QUOTE=DrDevil;46102951]Enjoy your only 16-bit-or-so resolution :rolleyes:[/QUOTE]
Im confused? how is it 16 bit when its a 24bit audio ADC? I know you can sacrifice bits for bandwidth with the atmega328 and others (is that correct?) is that whats going on here?
[QUOTE=Subby;46103125]Im confused? how is it 16 bit when its a 24bit audio ADC? I know you can sacrifice bits for bandwidth with the atmega328 and others (is that correct?) is that whats going on here?[/QUOTE]
If your signal is too noisy, you will loose bits, since you can't distinguish signal features from noise anymore. At 24-bits your signal has to be very noise free to make use of all your resolution.
[QUOTE=DrDevil;46103215]If your signal is too noisy, you will loose bits, since you can't distinguish signal features from noise anymore. At 24-bits your signal has to be very noise free to make use of all your resolution.[/QUOTE]
Doesn't small-amplitude noise kind of get integrated out by a delta sigma converter?
[editline]blah [/editline]
24 bits is a lot though :v:
Does anyone know of a good, cheap crimping tool for crimping 2.54mm pin header wires? I just can't fucking stand soldering them as it always looks so sloppy and is weak. Also, on the same note, why are some crimping tools so expensive? To me it just looks like regular pliers with some grooves machined into them.
[QUOTE=alexaz;46104377]Does anyone know of a good, cheap crimping tool for crimping 2.54mm pin header wires? I just can't fucking stand soldering them as it always looks so sloppy and is weak. Also, on the same note, why are some crimping tools so expensive? To me it just looks like regular pliers with some grooves machined into them.[/QUOTE]
They are expensive because there is such a little market for them, manufacturers are having to choose between:
Make them really good and expensive, then they sell less but make enough money
Make them really crap and cheap then they sell more to make enough money
I drew a font for my multimeter! 5x3 pixels per character.
[img]http://i.imgur.com/BaOPBow.png[/img]
The stuff at the top is for the status bar at the top of the screen. Here it is in actual size:
[img]http://i.imgur.com/JW1ZaMV.png[/img]
Feel free to use it, anyone looking for a small font! It's standard ASCII layout, the stuff at the top is in the non-drawable section of ASCII.
[editline]29th September 2014[/editline]
Here's one for the multimeter readout
[img]http://i.imgur.com/DcVe8Rv.png[/img]
[img]http://i.imgur.com/WIRX8Es.png[/img]
[QUOTE=r0b0tsquid;46101537]Looked up that ADC you used, it looks really nice, farnell has it for like £2.76 as well! I see that you've connected your signal to both inputs of the ADC, are you going to average the two results to get a better SNR? And is that better than having multiple input signals, like a full-on ECG in a hospital?[/QUOTE]
I don't think it samples both channels at the same time, or if it does the samples are buffered since the output data alternates between the left and right channel.
Funny enough RS is selling the ADC for £1.70 at the moment.
[QUOTE=DrDevil;46103215]If your signal is too noisy, you will loose bits, since you can't distinguish signal features from noise anymore. At 24-bits your signal has to be very noise free to make use of all your resolution.[/QUOTE]
It's hard to say how many ENOB I'll get without testing, the noise floor of the ADC is quite low with the SNR around 3dB to 5dB which isn't that great but should do once I give it some serious software filtering.
[img]http://i.imgur.com/peaSR2o.png[/img]
Do these icons look fairly self-explanatory?
Can I get a suggestion for a 1.427GHz directional antenna design?
[QUOTE=r0b0tsquid;46110037][img]http://i.imgur.com/peaSR2o.png[/img]
Do these icons look fairly self-explanatory?[/QUOTE]
From left to right, "Please wind me up", "Caution, arcing multimeter", "Fuck it, it probably approximates to a sine wave" and finally "give up and watch TV"
[QUOTE=metallics;46111127]From left to right, "Please wind me up", "Caution, arcing multimeter", "Fuck it, it probably approximates to a sine wave" and finally "give up and watch TV"[/QUOTE]
Bingo.
I've decided on 2N7000 MOSFAT logic level transistors, and a 74HC595 shift register to drive my 7-segment LED display.
Making the PCB tomorrow. I'll have to see where I can get the transistors and shift register from.
Need to design the PCB now.
Anyone see anything immediately stupid with my idea?
I'm just going to hook the outputs from the shift register straight into the gates of the transistors.
[QUOTE=nikomo;46114338]I've decided on 2N7000 MOSFAT logic level transistors, and a 74HC595 shift register to drive my 7-segment LED display.
Making the PCB tomorrow. I'll have to see where I can get the transistors and shift register from.
Need to design the PCB now.
Anyone see anything immediately stupid with my idea?
I'm just going to hook the outputs from the shift register straight into the gates of the transistors.[/QUOTE]
You don't need separate MOSFETs to drive LEDs with a 74HC595. Look at the datasheet, it'll tell you the maximum current per pin.
7-segment LED display, common anode.
MOSFET comes after each segment's cathode, breaks the circuit.
Apply power to MOSFET gate, segment lights up.
[QUOTE=nikomo;46117587]7-segment LED display, common anode.
MOSFET comes after each segment's cathode, breaks the circuit.
Apply power to MOSFET gate, segment lights up.[/QUOTE]
I know.
Or, you could just have anode -> LED -> resistor -> 595, and just invert the byte before you send it to the shift register.
The external MOSFETs are completely unnecessary.
Well if he's doing it just to get a feel for how they work then all well and good, but yeah, totally not needed.
[QUOTE=Zero-Point;46119344]Well if he's doing it just to get a feel for how they work then all well and good, but yeah, totally not needed.[/QUOTE]
If he was making the display out of those 10W Cree LEDs or something and using some IRF540s instead of lil 2N7000s, that would be awesome. For this though, noooo :v:
From the way he worded it (stating that it was common anode, as though that matters when CMOS sinks and sources equally well) I think he might be using them for inversion... Which is cool, but an AVR can invert a register in a single cycle, so really not needed either. Hey, you could also put a single MOSFET inverter on the 595's data line! 1 transistor instead of 8, and save that single cycle! Efficiency :D
A 100w 7 segment display sounds cool actually, I need to find a reason to build that.
[editline]1st October 2014[/editline]
Nikomo I think it's really cool that you're making your own PCB, it took me ages to take that step. Keep us posted please :)
RIP tip.
1994-2014
[IMG]http://i11.photobucket.com/albums/a166/ballsandy/CGS_0946.jpg[/IMG]
[IMG]http://i11.photobucket.com/albums/a166/ballsandy/CGS_0948.jpg[/IMG]
I'll have to buy a new one tomorrow.
I used MOSFETs because I'm not entirely sure I can power the thing how I'm powering it, so it's a bit more flexible.
Also, it'll be the school covering the components, sooooooooooooo, shh
But it simplifies the design a lot.
I'll have a look once I get to school and get my two other boards done.
[QUOTE=pentium;46119817]RIP tip.
1994-2014
[IMG]http://i11.photobucket.com/albums/a166/ballsandy/CGS_0946.jpg[/IMG]
[IMG]http://i11.photobucket.com/albums/a166/ballsandy/CGS_0948.jpg[/IMG]
I'll have to buy a new one tomorrow.[/QUOTE]
Is this what you're soldering with? You do realize that those are only for really high power stuff like thick cables. They're really unpractical for PCB soldering since they can fry your circuits if you're not extremely careful.
I ended up deciding against the straight-to 595 method, if I didn't have something running on the Pi that set the registers all to 1, the LEDs would all be lit up if the board was hooked up, and I don't really want to do that.
Also, fire inspection tomorrow, which means I had to spend the entire day cleaning the workshop, instead of making my PCBs, and tomorrow we've got another teacher talking about microcontrollers, so I won't get to do it then either.
Oh well. Better than sending the design to some group order thing and waiting a month for it to be made and delivered.
Also don't have the MOSFETs yet, but the school has them, and I found the shift register.
[t]http://i.imgur.com/ZFutiV3.png[/t]
[media]http://www.youtube.com/watch?v=4ALTBRkzcaY[/media]
It's so cute :-3
[QUOTE=DrDevil;46095767]I've designed my first board in weeks:
[t]http://a.pomf.se/vnxbpd.png[/t]
It's a PWM LED strip driver based on an ATTiny.[/QUOTE]
Is that an ATTiny 84? What'd you use for getting the voltage down?
[QUOTE=DrDevil;46121342]Is this what you're soldering with? You do realize that those are only for really high power stuff like thick cables. They're really unpractical for PCB soldering since they can fry your circuits if you're not extremely careful.[/QUOTE]
I'm using the gun to do repairs on a 1953 television set. Not a single PCB in this baby!
[IMG]http://i11.photobucket.com/albums/a166/ballsandy/CGS_0905.jpg[/IMG]
My usual Weller pencil iron is just not powerful enough to handle some of the massive joints on this thing. Also yes. Half the chassis is made out of solid copper.
[QUOTE=Gulen;46122959]Is that an ATTiny 84? What'd you use for getting the voltage down?[/QUOTE]
[URL=http://www.digikey.com/product-detail/en/ATTINY20-XU/ATTINY20-XU-ND/2357328]ATTiny 20[/url]
and
[URL=http://www.digikey.com/product-detail/en/AP2210K-5.0TRG1/AP2210K-5.0TRG1DICT-ND/4505292]generic linear regulator[/url]
I just picked the cheapest one that came in a sot23-5 package, since I accidentally finished the PCB before realizing that the regulator I've previously chosen is 7V max. Oops!
[QUOTE=nikomo;46122089]I ended up deciding against the straight-to 595 method, if I didn't have something running on the Pi that set the registers all to 1, the LEDs would all be lit up if the board was hooked up, and I don't really want to do that.
Also, fire inspection tomorrow, which means I had to spend the entire day cleaning the workshop, instead of making my PCBs, and tomorrow we've got another teacher talking about microcontrollers, so I won't get to do it then either.
Oh well. Better than sending the design to some group order thing and waiting a month for it to be made and delivered.
Also don't have the MOSFETs yet, but the school has them, and I found the shift register.
[t]http://i.imgur.com/ZFutiV3.png[/t][/QUOTE]
You could have solved the default-on problem by putting a pullup resistor on the 595's OE# pin, which means the outputs will float (not be connected to vcc or ground) until your code starts and grounds that pin.
Good luck with your project though!
[editline]1st October 2014[/editline]
[img]http://i.imgur.com/tUKjQz7.png[/img]
It's always handy to read the [url=http://www.nxp.com/documents/data_sheet/74HC_HCT595.pdf]datasheet.[/url]
If you go with the MOSFETs, you mustn't forget to [B]connect OE to ground[/B].
You will get some very strange behaviour otherwise - because MOSFETs have such high input impedance, if you leave the gate floating, they will switch on and off just when you touch them. Also, your circuit wouldn't work :')
Also instead of going straight to a PCB you should consider prototyping your idea (possibly on a breadboard) first.
Or at the very least a decent simulator.
[t]http://u.cubeupload.com/Chryseus/ipr9vZ.png[/t]
Any ideas on what to do with the empty space ?
What the fuck I just noticed one side of C3 is floating when it should be connected to ground and DRC doesn't pick it up, I think I'll stick with Altium in future.
