Printable View
Hi,
I'm building my own DAC controller - I put together a simple R-2R DAC driven by a 595 shift register.
The output stage is through the same op-amp config used in norm's laser show DAC. (Parallel port DAC)
I noticed that I was getting noise at the lower end of the cycle - this is +5 -> -5v.
Does anyone know if R-2R DACs are prone to this, or is there something wrong? Either way, what would you recommend to try to fix it? It looks way too noisy to send to a galvo.
Steve
OK I'll reply to myself...
:)
I think I'm entering the "noise zone" so I'll apply a liberal sprinkling of caps to the power supply and see what happens...
the output amplifier have enought power supply to output +/-5v (supply rails @ 8v minimum for 5v output)? or there is an output load? what is the op amp reference?
I'm using +/-15v to power the op-amp stage - these are the JFET TL082 opamps as used in norm's laser dac - I don't have the diagram to hand, but it's been posted here before.
I don't have a load on the output, running straight into my scope inputs (differential inputs, as the GND (center) is actually at +10v to my PC GND running the scope software)
ok this is perhaps before the op amp at the DAC output when the output fall near the ground but this is strange.
do you have the wiring diagram to help?
Wow! - someone trying to do the same thing as me...
I've been trying to put together a working laser harp for about 25 years:D
I built a really simple one a couple of years with a frame around it, but now I'm trying to build a scanner-based frameless harp.
The trace you see above is after the op-amp output stage which is what is shown in the attached diagram -
I tried to use the same DAC chip but I could not get the Arduino to talk SPI properly - so I built an R-2R ladder and controlled it from a 595 shift reg.
The output of the r-2r is a fairly clean 0-5v signal, with no distortion around 0v. So the noise you see on the scope is being added by the TL084 stage.
I cleaned up the r-2r output with a bypass cap (0.01uF) but this lead to spikes in the final AC output. It did not help with the noise so I removed it.
I reprogrammed the Arduino to output values in the 25-225 range and this avoided the noise area completely. I've lost resolution, but I'm going to open and close the harp with a digital potentiometer, so I don't think it matters. I might try scaling the final output back to -5 <-> +5v and see if the noise reappears.
regarding your laser harp project - did you get a working reflection trigger? Mine are all very difficult to calibrate.
Steve
this is strange because tl084 is good op amplifier and in this wiring (invert add) there's theorically no noise
about my laser harp yas i have a good sensor system and it's an open frame :)
here is a video take last week end:
http://www.dailymotion.com/relevance...rpe-laser-test
and free laser show :)
http://www.dailymotion.com/relevance...w-laser_events
Can you explain the sensor system? I have problems with the light reflecting from the ceiling of the room I am in. This interfers with the signal.
What is the power of your laser?
Steve
the laser is a 150mw dpss and the schematic need to be made under software to post it on the web but on paper you just need to follow my link "laser harp" at the bottom of this post.
you have the schematic in picture on the first page :)
about your problem when you have -5v at the output you need to have 0 volt at the output of the current to voltage converter. there is noise at this point?
If I understand you correctly then yes the DAC output is 0v and the TL082 output is -5v (approx).
The DAC output is not very noisy, and seems to be OK over the whole range. So the noise is being introduced in the '82 section.
Tonight I will plot the output of the first phase of the '82 and see if there is noise there.
Laser Harp: I've been looking at the schematic for your sensor - can you explain what the last stage is for? It has unmarked components / output.
Would an LM324 op-amp chip be OK for this circuit?
(What kind of photodiode did you use?
One sensitive to the laser frequency?)
(Also what is the component '103' - a 10K resistor?)
Steve
i use tl084 as i remember and yes 103 is the chip code for 10K resistor.
the last stage is a comparator to create interruption trig on the microcontroller.
the photodiode is just a standard wide area use in barcode laser scanner.
this system detect the variations of light and the laser need to be modulated to be detected by this sensor.
this is why the ambiant light is rejected (but not fluorescent light due to her frequency and fast response)
about your problem can you explain more your scope connexion? you can add a 10K ohms resistor between the ground and the output of the dac to make a fake load.
peraps your 5v rail is noisy and add noise by the invert add wiring of the last stage.
Thanks - I will try to build your sensor at some point.
Tonight I will try to load the DAC output, and also inspect the 5v being added in the last stage.
I will post the results here.
Thank you for your help.
Steve
Quick question - I noticed that your circuit uses nF (nanoFarads) - I cannot find a vendor in the US that deals in nF capacitors - just pF and uF.
Even converting the values into pF/uF I can't find matching caps over here.
Is there somewhere I can get them from?
yes from radiospares and farnell :)
but this is strange because they are standard values.
I goofed in my conversion...
They are common - you are right..
OK, just *three* more questions about your beam detector and I *promise* I'll leave you alone...
:D
Is the comparator 2nd (-) input a copy of the laser blanking signal?
What is the collector of the NPN transistor tied to? +5v or -5V?
Is the diode in the comparator an LED, or there for another reason?
don't worry about questions ;)
i not understand the first one.
for the second and third the collector is connected to the +5v logic through the led for test but in my system the led is replaced by the INT0 input of the microcontroller :)
OK thanks!
My first question was about the inverting input to the final op-amp.
See the picture, what is connected to 'X'?
+5v, -5v, or the laser blanking signal.
One more question - do you know why Babelfish French->English keeps translating 'harpe' as 'toothing-stone'?
It did it with your pages, and it did it with Sebastien Pamart's Harpe Laser pages...
the X point is at 5v :)
and the modulation is the signal expected at this point and not the laser modulation ;)
about the translation i don't know this is strange ^ ^
Thanks again - I'm soooo happy that there's someone out there that's achieved what I want to do. :D
Finally what do you think - Jean Michel Jarre's laser harp - does it really work, or is it just a stage prop?
I was at the Manchester concert when the beam failed, yet the sounds continued... :eek:
Very disappointing for me. :(
ah? the laser had a problem in concert?
about JMJ i think some concerts was in direct and other not.
In your diagam here:
http://thequark.free.fr/harpe/photos/5.jpg
Can you tell me the frequency of the laser blanking?
And how the circuit is tuned to work with this frequency? I assume it's something to do with the feedback caps on the op-amps...
Thanks.
Steve
hello :)
this is the part i need to tune before to give the schematics on the web.
but with theses value the sensor work well with 70-200Hz of blanking. but need to be tested :)
I managed to get your sensor built and working - thanks for posting the schematic.
Not quite as elegant as yours, but it's a good start!
Steve
great :)
and that work fine for your system? :)
Well....
I can get it to generate an interrupt on the Atmel processor, so it's doing it's job OK - that's as far as I've got
Do you think that an LM324 Quad op-amp would be OK for this circuit?
(I'm using a TL082 x2 at the moment)
Do you have any information on the photodiode you used?
The only thing I was able to find with a large surface area was the OPT211 chip - everything else was tiny.
you can use lm324 for this job but the TL084 is better :)
about the photodiode i not have information. this is a part from barcode laser scanner. but a photodiode with large area can work fine i think.
Photodiode - Can you tell me where you got it from? I'm struggling to find anything like the one in your pictures.
the one used on my system is from an ebay member who send these surplus part for free.
Mr Genesis,
I was looking at your board assembly and I couldn't help noticing that the two resistors on either side of the photodiode did not make it into the build.
Are you using the PD in a photovoltaic mode?
Steve
hello :)
this is a prototype and i need to make the final schematics. and i use some smd resistors on my boards perhaps on this one. this is a friend who have my harp and i can't verify this now :)
but the schematic poste above was used when i make the board ;)
Genesis,
I'm almost done with my harp...
I wanted to ask you where in the main control loop you output MIDI messages?
I have put my MIDI output routine just after the return signal and it works OK, but it does seem to slow down the beams a little...
Steve
Hello :)
about the midi routine she was on the main program and use bits to know if a string is played.
the bit (flag) is set by the timer routine who draw the stings.
in the main routine if the flag go off the "note off" is sent :)
what is the microcontroler you use? his frequency?
greetings and happy new year :)
I'm using an Arduino (atmel atmega 168) at 16Mhz.
The MIDI output routine is slowing down the harp strings - but I have a control loop with timing delays - I never thought about using a timer interrupt for this.
When you play a note you can see the harp "flash" very quickly as the delay of sending MIDI stops the next refresh.
I am using TAOS light to voltage sensors in place of the photodiode. I made some (accidental) modifications to your beam detector circuit recently and it started working really well. I shall post you my modified circuit at some point.
One final question - how do you deal with multiple note triggers? When I trigger the note I get 2-3 notes sounding on NOTE ON, and 2-3 sounding when I take my hand away.
ok :)
about the multi strings i use multi flags :)
each string use a proper flag to send each note on and note off.
for my midi routine i use the serial uart inside the atmel and note bit twidling. perhaps you loose time in bit twidling when you send the message.