Hi guys, I've been trying to get in contact with Stephen, but he hasn't logged into PL for a while, and the captcha on his website is broken, so I thought I'd ask here.
I have build the harp to the plans exactly, EXCEPT I used different light sensors. The light sensors are the same manu, same package, pretty much the same thing except mine say "high sensitivity".
The problem I am having is with the output of opamp.
Here is what you are meant to get (The red trace)
Here is what I get:
Which looks pretty close, but at the output of the whole opamp, you are supposed to get:
I get:
However, the weird part. Even with the sensors completely sheilded from ANY light at all, I still get this on the output of that stage (Notice how there are only 9 "dips" now):
Is there anyone here who could help me in setting up the sensors/opamp correctly?
Just reading the 2 datasheets for the light sensor, are there any obvious problems that would cause the TSL275 to operate differently to the TSL12S?
"The TSL12S, TSL13S, and TSL14S are cost-optimized, highly integrated light-to-voltage optical sensors, each
combining a photodiode and a transimpedance amplifier (feedback resistor = 80 MΩ, 20 MΩ, and 5 MΩ
respectively) on a single monolithic integrated circuit. The photodiode active area is 0.5 mm × 0.5 mm and the
sensors respond to light in the range of 320 nm to 1050 nm. Output voltage is linear with light intensity"
And:
"The TSL257 is a high-sensitivity low-noise light-to-voltage optical converter that combines a photodiode and
a transimpedance amplifier on a single monolithic CMOS integrated circuit. Output voltage is directly
proportional to light intensity (irradiance) on the photodiode. The TSL257 has a transimpedance gain of 320 MΩ.
The device has improved offset voltage stability and low power consumption"
Creaky Old Award Winning Bastard Technologist Infinitus Excellentia Ion Laser Dominatus
Join Date
May 2007
Location
A lab with some dripping water on the floor.
Posts
9,902
Shobley licensed the design to some one. I don't know if he can even comment on it.
First Educated Guesses from looking at the pics. I'd need to see the schematic, and layout before I commented further. Transimpedence converters (photodiode current to voltage circuits) are a notorious witch to work with if you want any bandwidth at all. For a first try, other then the modulation in DARK mode, your looking very good.
I'd beef up the filter caps. Looks like your getting bleedthrough on the supply rails or ground from some laser or galvo modulation some where. The photocell current into the opamp on the photochip is nanoamps to a microamp or so, thus any large variation in supply voltage is likely to bleed through. Keep the wires to the photo opamp short and shielded, and have bypass caps on one both rails of the opamps. When it says 320 megohms, the internal gain is huge, and noise pickup by the chip is a problem. Usually you need a shield can/ground plane and circuit careful layout around the photochip in the end.
The opamp peaking on the square wave is normal with photosensors, your seeing risetime issues from the area of the photocell having a lot of capacitance. High sensitivity usually means bigger area and or missing external compensation. The TI/TAOS integrated photosensors need some care in their use, some of which is not on the data sheets. As your signal magnitudes are well out of the noise, I'd say ignore it.
The white noise on top of the square wave is either something like RF pickup from a cell phone or switching psu or the signal is strong and is clipping.
As Bill Benner has reminded us often at LEMs, a galvo pair doing a long jump can dump 4-8 amps peak into the ground wiring of a circuit, if your DC ground had a resistance of 1 ohm between your projector circuits and the PSU, you could just have bumped your ground 4 volts above where it should be. So use "star" or " home run" wiring between your big loads and your sensitive circuits, and I bet some of that noise goes away. This means a lot of extra ground wires,but it can be worth it.
See attached. Keep all the parts in the drawing right at the 257, and consider putting a can around it.
What you mentioned was exactly the problem Steve. The peaks in the dark were caused by the galvo's drawing peaks from the power supply. The sensors were also wayyyy too sensitive, so I changed the gain on the first opamp and brought it down, and I'm getting useable signals. The harp works now, however I think there may be an error with the 555 schematic he put in the manual. I believe the capacitor and pot should be reversed, should it not? In any case, I bypassed the 555 and it works
Bugger. I had a quick look at your CRO shots over breakfast, not they are gone due to license issues.
Correct me if I am wrong, but did the original op-amp output signal swing +/- while yours only swung +ve? I did think earlier your op-amp was missing the -ve supply.
Feel free to ignore me, it's been a loooong day in the office....
The first opamp is swinging +/-, it's just not all that visible on the scope trace.
Whats really odd however, is the microcontroller is setup to trigger an interrupt when the pin goes LOW, which is connected to the output of that opamp stage directly. However, when there is no signals showin gon the oscope, it sends out a heap of midi data, so I turn up the gain pot until it starts showing signals, and the uC starts to respond. However, it only seems to detect the pulse when my hand is in the beam, even though the scope trace is showing the output dropping LOW frequently ..
Creaky Old Award Winning Bastard Technologist Infinitus Excellentia Ion Laser Dominatus
Join Date
May 2007
Location
A lab with some dripping water on the floor.
Posts
9,902
Yeah,there are errors in the 555 schematic just looking at it.
The potentiometer should not have its wiper tied to the 555, the wiper should be tied to the top of the pot and the 555 tied there as well. The 555 needs a variable resistor, not a variable voltage.
That is just error one. Its almost like he is using the 555 as a buffer, not a schmidt, monostable etc Anyways, its way wrong in the way its drawn.
Pin 3 probably needs a pullup resistor to clean up the output, or he is using the active pullup in the micro, or had a cmos 7555 or something.
I'd watch pulling the 555s or microcontrollers input much below ground, that just screams eventual substrate lockup in the chip.
I've wasted too much time looking for proof of OBLs death, so I need to get to work. I'll look at the whole schematic tomorrow if you find me a working link or scan it in. If you can post what your working from I'll take a look at it.
Hey Steve, don't worry about it too much, I've got the laser harp working pretty much as it should now
Still hasn't got a case yet, but will get to that shortly.
I haven't really learnt much of the song, and had to play it slowly because I blew up the differential line balancer, meaning my galvo's can only move at 1/2 the scan angle. I only have like 5mm between beams, so I had to watch where my fingers were going
This is a frameless laser harp, beams terminate right on the roof, and the sensors pick up the reflection on the bottom of your hand (Green box).
[yt]NHsk5Mbtfd8[/yt]
Gonna have some serious fun with this. Even considering building another one so a friend and I can them at the same time
Opamp/laser harp help please?
Reply With Quote
Problem solved
