Picked up a DX order today for some LED strip lighting stuff for my sister.
Decided to dig right into the PSU. 12v 6a ~$20 "weather proof" outdoor model. It's not actually for outdoor use but the "indoor" or chassi mount models have the connecting blocks outside the case which is a potential shock hazard. Even though it's gonna be mounted hidden I decided to put in the extra $3 to alleviate that risk somewhat.
[img]http://puu.sh/2wZsz[/img]
[img]http://puu.sh/2wZtO[/img]
[img]http://puu.sh/2wZuu[/img]
[img]http://puu.sh/2wZuT[/img]
[img]http://puu.sh/2wZvh[/img]
[img]http://puu.sh/2wZx7[/img]
[img]http://puu.sh/2wZzO[/img]
After 10-15 mins at 2.9 amps load with the cover on. The temperature was pretty evenly 40C and only two hotspots at about 50-60c on some chunky resistors at the input and at the output.
Main heatsunk devices where at about 35-40c. Voltage regulation was excellent with only a 70mV drop from 12.2V no load to 12.13V 2.9A load.
I have yet to test it with the led strip or at full load as my current sink isn't beefy enough but it's already exceeded expectations.
My new ringtone:
[media]http://www.youtube.com/watch?v=rp8hvyjZWHs[/media]
[QUOTE=Chryseus;40220112]I'll go through it as I see it and point out any obvious problems.
op-amp B senses the voltage drop across the current shunt R1 and amplifies it by 1 + R5 / R4 giving a feedback voltage of I * R1 * Av.
op-amp A controls Q2 adjusting the current to the load given by the above equation, the gate drive voltage must be able to swing close to the input voltage since ID is dependant on Vgs - Vt which in practice means op-amp A should have a supply voltage equal or greater than the input (there are other options).
op-amp D is a standard voltage regulator however as with op-amp A the gate drive must be at least close to the input voltage.
The circuit is also not quite stable and will be effected by noise in the power supply or reference voltages, possibly in worse case causing it to oscillate.
R3 and R2 are not needed due to the internal output circuit of the op-amps which prevents the transistors getting stuck on, if your directly driving a mosfet it's a good idea to add a small series resistor (< 100 ohm) to help isolate the op-amp from the gate capacitance which can cause instability.
Simulation of the circuit in LTSpice showed some issues so I thought fuck it and redesigned it to use one transistor, this does not include the current amplifier but you can add that yourself easy enough.
Feel free to ask questions.
[t]http://u.cubeupload.com/Chryseus/MZnXOw.png[/t]
[url=https://mega.co.nz/#!LxhTzQYS!PE9EwGASfWCoY4ywBnbsMhmfaoFrPb5zzE_5VU6_ljE]LTSpice file[/url]
Eh I forgot the capacitors, I suggest 10nF to 100nF between the output of each op-amp and the non-inverting feedback.[/QUOTE]
I built this on the breadboard, with the exception that there is a 100nF cap across the output and non-inverting input across the voltage regualting op-amp and added the current sense amplifier. The problem after adding the current sense amplifier the ripple jumps from a 20 mV to...:
[t]http://i.imgur.com/z1C0xPK.jpg[/t]
EDIT: Whoops, the cap was across the output and inverting input. Changed it. Now when the current sense reference is 0V, the waveform looks like this:
[t]http://i.imgur.com/DLbvb4y.jpg[/t]
And if the cap is shorted, the voltage is (like its supposed to be) 0 V
[QUOTE=alexaz;40269899]I built this on the breadboard, with the exception that there is a 100nF cap across the output and non-inverting input across the voltage regualting op-amp and added the current sense amplifier. The problem after adding the current sense amplifier the ripple jumps from a 20 mV to...:
[/QUOTE]
Congratulations you've built an oscillator!
Could you draw how you connected the current sense amplifier up, also how much gain, did you add any capacitors, does it work without the amplifier, etc
I'm going to dig out my breadboard, I recently built a current regulator myself and have had nothing but problems.
The current sense amp is connected like here:
[t]http://www.radio-electronics.com/info/circuits/opamp_basics/op-amp_basic_non_inv.gif[/t]
R1 = 1K, R2 = 10K, so the gain is 11. There are no caps in the circuit, except one 100 nF across the output and non-inverting input of the voltage regualting op-amp. Circuit fine without the current regulation part.
EDIT: I just fucked something up and wow i'm lost. I'll to figure out what i'm doing right now and then post again lol
try add a 100nF cap in parallel with R2 and see if that helps at all.
What op-amp are you using?
I've just discovered TL071 will not work at all in a current regulator circuit no matter what you do.
I never said i'm using a TL071, or did I? I'm using LM324. Now i rebuilt the circuit, and changed the 100 nF to 1 uF and added 1 uF on the output. The ripple is now ~10 mv. The problems are gone. For now. Thanks.
After hearing the news of making graphene supercaps with a DVD burner, I'm willing to build one. I've been looking at material dielectric values to find the highest, to maximize capacitance.
You know nature is teasing you when most of the high value dielectrics are flammable.
Makes me sad that recently i destroyed my disc drive. I wonder if a camera flash would be enough to reduce the graphite oxide to graphene.
[QUOTE=alexaz;40281370]Makes me sad that recently i destroyed my disc drive. I wonder if a camera flash would be enough to reduce the graphite oxide to graphene.[/QUOTE]
You may be able to get a hold of some IR LEDs or a IR laser within Lightscribe's range:
[URL="http://en.wikipedia.org/wiki/LightScribe#Mode_of_operation"]http://en.wikipedia.org/wiki/LightScribe#Mode_of_operation[/URL]
[QUOTE=LoneWolf_Recon;40280535]You know nature is teasing you when most of the high value dielectrics are flammable.[/QUOTE]
Unexpected fireworks :v:
[QUOTE=alexaz;40281370]Makes me sad that recently i destroyed my disc drive. I wonder if a camera flash would be enough to reduce the graphite oxide to graphene.[/QUOTE]
Not even close to the intensity required - the light from the camera flash is spread out too much, you'd need to focus it to a point.
There's a lot of smartphones with lcd panels that has high res resolution out there, but seem to be very few "bare" panels that are for sale that has standard connection interface or a converter. Does anyone has any suggestions for a LCD panel that's around 4-6 inches and has a resolution of 720P or higher? Possibly uses a standard connection interface such as LVDS?
[QUOTE=LoneWolf_Recon;40280535]After hearing the news of making graphene supercaps with a DVD burner, I'm willing to build one. I've been looking at material dielectric values to find the highest, to maximize capacitance.
You know nature is teasing you when most of the high value dielectrics are flammable.[/QUOTE]
Don't you rather need an electrolyte?
[QUOTE=B!N4RY;40284193]There's a lot of smartphones with lcd panels that has high res resolution out there, but seem to be very few "bare" panels that are for sale that has standard connection interface or a converter. Does anyone has any suggestions for a LCD panel that's around 4-6 inches and has a resolution of 720P or higher? Possibly uses a standard connection interface such as LVDS?[/QUOTE]
A large part of them either use LVDS or [URL="http://en.wikipedia.org/wiki/Display_Serial_Interface"]DSI[/URL]
In general look for replacement 3rd party panels for popular smartphones.
[url]http://www.acsmaterial.com/product.asp?cid=27&id=16[/url]
300$? No thanks, no graphene for me...
[QUOTE=alexaz;40289319][url]http://www.acsmaterial.com/product.asp?cid=27&id=16[/url]
300$? No thanks, no graphene for me...[/QUOTE]
ebay
[QUOTE=DrLuckyLuke;40286369]Don't you rather need an electrolyte?[/QUOTE]
[URL="https://www.youtube.com/watch?v=G-AF7_Blj_k"]This[/URL] guy was using various materials including some form of electrolyte. I was looking into some thin, porous material as the separator and then use Glycerol Water (antifreeze) as the dielectric.
TBH, I dont understand why electrolytes are recommended for double layered caps to classify them as supercaps.
New toy!
[img]http://i.imgur.com/p0COdld.jpg[/img]
The future of embedded systems is looking bright, a colleague of mine on my schools robotics team sent this:
[URL="http://www.kickstarter.com/projects/435742530/udoo-android-linux-arduino-in-a-tiny-single-board"]http://www.kickstarter.com/projects/435742530/udoo-android-linux-arduino-in-a-tiny-single-board[/URL]
I'm feeling iffy about no heat dissipation on that first ARM, however this is looking to be one hell of a powerhouse.
[t]http://i.imgur.com/l3eYYHg.jpg[/t]
I have rebuilt the entire regulator thousands of times and now i'm using a LM311 just for the voltage regulation. Still getting oscillations/ripples as in figure one, but they tend to "ripple themselves out" in a few minutes, then come back again. I think what i need is hysteresis for the comparator, but i can't really use hysteresis because my reference isn't constant?
[media]http://www.youtube.com/watch?v=ZN9QiRa8RBg[/media]
Cypress PSoC5, it's just amazing!
It's basically a CPU (cortex M0), but it also has discrete logic (and analog) blocks that can be programmed and mapped to any pin you want. So it in a sense is like an FPGA in an MCU. I made this hello world breathing blink with this "code":
[img]http://i.imgur.com/0TZnJY8.png[/img]
And all the code the CPU was running was this:
[code]#include <device.h>
void main()
{
PWM_1_Start();
PWM_2_Start();
for(;;)
{
//nop nop nop
}
}[/code]
[QUOTE=alexaz;40316346][t]http://i.imgur.com/l3eYYHg.jpg[/t]
I have rebuilt the entire regulator thousands of times and now i'm using a LM311 just for the voltage regulation. Still getting oscillations/ripples as in figure one, but they tend to "ripple themselves out" in a few minutes, then come back again. I think what i need is hysteresis for the comparator, but i can't really use hysteresis because my reference isn't constant?[/QUOTE]
Ahh no wonder you're having problems I forgot to take the phase shift of the transistors into account resulting in the circuit turning into an oscillator around 1.2MHz where the phase reverses, being on a breadboard only makes the situation many times worse with the added parasitic effects creating additional poles.
One quick possible fix I can offer is to connect a 220nF or there about capacitor between the collector and base of both transistors, that should severely reduce the bandwidth making sure the closed loop gain falls to unity long before 1.2MHz where the op-amp naturally has a poor open loop gain and limited phase margin.
I'll pull out my board in the morning for some testing. :eng101:
I have an HP 16500C digital logic analyzer. Where can I sell something like that? It's valuable, but nobody's buying them on eBay, probably because it's such specialized equipment.
[QUOTE=Agent766;40364671]I have an HP 16500C digital logic analyzer. Where can I sell something like that? It's valuable, but nobody's buying them on eBay, probably because it's such specialized equipment.[/QUOTE]
It isn't really valuable or useful by today's standards, the reason nobody is buying is because it's overly huge for the performance.
That said, how much are you asking for it?
[QUOTE=Agent766;40364671]I have an HP 16500C digital logic analyzer. Where can I sell something like that? It's valuable, but nobody's buying them on eBay, probably because it's such specialized equipment.[/QUOTE]
[url]http://www.ebay.com/itm/HP-Hewlett-Packard-16500B-Logic-Analysis-System-w-16522A-16550A-Modules-/370795882162?pt=LH_DefaultDomain_0&hash=item56552826b2[/url]
[url]http://www.ebay.com/itm/HP-16500B-Logic-Analysis-System-/330908287708?pt=LH_DefaultDomain_0&hash=item4d0babc2dc[/url]
[url]http://www.ebay.com/itm/HP-Hewlett-Packard-16500B-Logic-Analysis-Analyzer-Used-Condition-/390432313508?pt=LH_DefaultDomain_0&hash=item5ae79450a4[/url]
[url]http://www.ebay.com/itm/HP-AGILENT-16500B-LOGIC-ANALYSIS-SYSTEM-w-16530A-CARD-/321109270489?pt=LH_DefaultDomain_0&hash=item4ac39a9fd9[/url]
[url]http://www.ebay.com/itm/HP-AGILENT-16500B-LOGIC-ANALYSIS-SYSTEM-w-16530A-CARD-/321109270489?pt=LH_DefaultDomain_0&hash=item4ac39a9fd9[/url]
[url]http://www.ebay.com/itm/HP-16500B-Logic-Analyzer-w-2-cards-16550A-3-cables-1817-/121096700999?pt=LH_DefaultDomain_0&hash=item1c31ed0447[/url]
[url]http://www.ebay.com/itm/HP-16500B-Logic-Analyzer-w-2-cards-16550A-3-cables-1817-/121096700999?pt=LH_DefaultDomain_0&hash=item1c31ed0447[/url]
[url]http://www.ebay.com/itm/HP-16500B-Logic-Analyzer-Digital-Oscilloscope-Mainframe-16530A-16531A-400MSa-s-/290846268570?pt=LH_DefaultDomain_0&hash=item43b7c9dc9a[/url]
[url]http://www.ebay.com/itm/HP-16500B-Logic-Analyzer-Digital-Oscilloscope-Mainframe-16530A-16531A-400MSa-s-/290846268570?pt=LH_DefaultDomain_0&hash=item43b7c9dc9a[/url]
There are lots on there at $200+, but like you said, they dont seem to be selling. I bet at a sub-$150 price, it would sell. At that price youre competing with less capable USB logic analyzers. Someone who isnt bothered by the size should jump on it.
Is it pretty complete?
It works fine and has the probes and everything. From what I understand, the expansion cards though are pretty valuable compared to the base system.
[editline]21st April 2013[/editline]
Oh, I guess I didn't list them.
16530A oscilloscope timebase (400 MS/s) card, a 16531A 2 channel oscilloscope card, 3 10431A probes and a 2 16510B 80-channel state/timing modules
You can never have too many.
Also, please take a [url="http://25.media.tumblr.com/tumblr_lxiq57hQKW1r6p97to1_500.jpg"]"Why can't I hold all these pcb's?"[/url] pic.
Sorry, you need to Log In to post a reply to this thread.
