LPF Donation via Stripe | LPF Donation - Other Methods
Links below open in new window
ArcticMyst Security by Avery
RF excited CO2 lasers
- Thread starter Hemlock_Mike
- Start date
- Joined
- Jan 7, 2007
- Messages
- 6,309
- Points
- 83
I think all the RF excited CO2's are sold but you need to ask hurrersciences. Chris will let you know what he has.
Mike
Mike
- Joined
- Mar 27, 2008
- Messages
- 478
- Points
- 28
Alright, well I've done a little bit of testing. Both boards have no ac signals going out to the laser or even on the large power transistor. I did some poking around and I the problem seems to lie with the crystal oscillator. The output on the oscillator on the board that I got from nightstick is around 48kHz, a far cry from the 27.12Mhz the laser needs. Furthermore I am not understanding the trace layout on this board. The output of the oscillator goes into a capacitor, and the other side of that capacitor doesn't go anywhere, there are no traces on either side of the solder pad and its not connected to ground. So either I am entirely wrong about something or there is an inner layer of traces on this pcb, in which case this whole thing gets much harder.
So I'm going to put in a request for a quote from the original manufacturer, though I would imagine it's not going to be reasonable for me. I did find the data sheet for the particular model.
http://www.gedlm.com/DataSheets/LM2000.pdf
The exact printing on the part itself is:
LM2000B
27.120M
GED
GED is the company that manufactures it, 27.120M being the frequency. LM2000 is the part number and the B denotes a frequency tolerance of +/- 50ppm.
I'm also going to ask Chris in case he has any spares. Has anyone heard of a crystal oscillator malfunctioning in this way? I don't know anything about these things so any insight would be super helpful.
So I'm going to put in a request for a quote from the original manufacturer, though I would imagine it's not going to be reasonable for me. I did find the data sheet for the particular model.
http://www.gedlm.com/DataSheets/LM2000.pdf
The exact printing on the part itself is:
LM2000B
27.120M
GED
GED is the company that manufactures it, 27.120M being the frequency. LM2000 is the part number and the B denotes a frequency tolerance of +/- 50ppm.
I'm also going to ask Chris in case he has any spares. Has anyone heard of a crystal oscillator malfunctioning in this way? I don't know anything about these things so any insight would be super helpful.
- Joined
- Jan 7, 2007
- Messages
- 6,309
- Points
- 83
48 kHz is way too low for even a crystal Osc. You are right that isn't a good reading to see. It sounds like a ground loop oscilation which isn't wanted here. Does the Osc have an enable pin to turn it on?
Mike
Mike
I would try to take out the oscillator from the board and try to power it alone, to be sure it's dead.
How critical is the frequency?
27Mhz oscillators are readily available:
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=300-8619-1-ND
but not 27.12Mhz
There is also the option to get a programmable oscillator, like this one:
http://www.sparkfun.com/commerce/product_info.php?products_id=9116
If you don't want to deal with pre-programming it I think you can ask them to program it to 27.12Mhz prior to shipping.
How critical is the frequency?
27Mhz oscillators are readily available:
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=300-8619-1-ND
but not 27.12Mhz
There is also the option to get a programmable oscillator, like this one:
http://www.sparkfun.com/commerce/product_info.php?products_id=9116
If you don't want to deal with pre-programming it I think you can ask them to program it to 27.12Mhz prior to shipping.
- Joined
- Mar 27, 2008
- Messages
- 478
- Points
- 28
The Osc pin 1, according to the data sheet, is used to turn the oscillator on. But the pin has no traces coming from it and the part is missing the (E/D) label that indicates the function is present on the device. So I believe that this particular oscillator does not have the on/off function.
I think I will look into the programmable oscillator thing, that sounds like a good place to start.
I think I will look into the programmable oscillator thing, that sounds like a good place to start.
- Joined
- Mar 27, 2008
- Messages
- 478
- Points
- 28
Also, if anybody has a KNOWN WORKING RF supply of this design, I would very much like to borrow/buy it to do some comparisons. If anyone has a spare working RF supply of a different design I would probably buy it too, if only to just to finally get this thing working once and for all.
- Joined
- Jan 7, 2007
- Messages
- 6,309
- Points
- 83
EH --
I think that there must be 3 versions of this driver which were sent out. I'm sure Chris tested these prior to shippment. I have two of the versions which work. I'm calling Lynn to get more info from him too. That tube has to light up.
The 27.12 MHz has to be pretty close for impedance matching. Your oscope readings don't make sense to me - that's way too far off.
Mike
Just a thought --- Because the high powered RF cloud around this system may mess up readings, try running it without the 28 volt input. This will reduce the crap in the area and may allow you to track the osc and driver circuits with the oscope.
I think that there must be 3 versions of this driver which were sent out. I'm sure Chris tested these prior to shippment. I have two of the versions which work. I'm calling Lynn to get more info from him too. That tube has to light up.
The 27.12 MHz has to be pretty close for impedance matching. Your oscope readings don't make sense to me - that's way too far off.
Mike
Just a thought --- Because the high powered RF cloud around this system may mess up readings, try running it without the 28 volt input. This will reduce the crap in the area and may allow you to track the osc and driver circuits with the oscope.
?
...
Guest
My best guess with the oscillator output is that you are seeing the pwm frequency, but not having set hand on such a supply (I am a much for direct dc injection kind of man) its hard to say. One thing you might try is looking at the enable pin (just take a voltmeter to it from the bottom of the board) and make sure that its within the operating spec for the board. Sometimes these things don't actually use the enable pin, but since the datasheet notes you need >2.4v on it to enable the output I would make sure theres actually 2.4v on that pin!
If you do think the chip is dead (one other thing to look into would be whatever it is driving is shorted, before buying a replacemnt I would unsolder this one and test it out of the circuit first) but mouser has a 3.3v smd one that will ship in 2 weeks http://www.mouser.com/Search/Refine.aspx?N=5115914+4294622267&FS=True, or it looks like you can custom order an 5v 8-pin one but it would be an expensive order!
http://www.io.com/~oe/OscFreqs.html#fsmc these guys have a 27.14MHz crystal, which is about 500ppm off from what you want--quite possible you could tune that out. You can also get 27.125MHz (which is withing the 50ppm tuning spec, so you should probably ok) and make an oscillator out of a simple not gate and be done with it. All you need is a not gate and maybe some capacitors to help keep things happy. see http://www.mpdigest.com/issue/Articles/2008/Mar/Crystek/Default.asp
I vote you screw your measly driver board and hook the laser up to this rf source http://cgi.ebay.com/RF-Plasma-Produ...5|66:2|65:12|39:1|240:1318|301:1|293:1|294:50
If you do think the chip is dead (one other thing to look into would be whatever it is driving is shorted, before buying a replacemnt I would unsolder this one and test it out of the circuit first) but mouser has a 3.3v smd one that will ship in 2 weeks http://www.mouser.com/Search/Refine.aspx?N=5115914+4294622267&FS=True, or it looks like you can custom order an 5v 8-pin one but it would be an expensive order!
http://www.io.com/~oe/OscFreqs.html#fsmc these guys have a 27.14MHz crystal, which is about 500ppm off from what you want--quite possible you could tune that out. You can also get 27.125MHz (which is withing the 50ppm tuning spec, so you should probably ok) and make an oscillator out of a simple not gate and be done with it. All you need is a not gate and maybe some capacitors to help keep things happy. see http://www.mpdigest.com/issue/Articles/2008/Mar/Crystek/Default.asp
I vote you screw your measly driver board and hook the laser up to this rf source http://cgi.ebay.com/RF-Plasma-Produ...5|66:2|65:12|39:1|240:1318|301:1|293:1|294:50
Would one of this be able to power a RF tube?
http://www.protechcbshop.com/fireinthywire.html
I got a co2 tube that looks exactly like the tubes from this thread but without any PSU. If and when I receive it I have to figure out a way to power it.
http://www.protechcbshop.com/fireinthywire.html
I got a co2 tube that looks exactly like the tubes from this thread but without any PSU. If and when I receive it I have to figure out a way to power it.
- Joined
- Mar 27, 2008
- Messages
- 478
- Points
- 28
Alright, so this is what we have so far:
Received laser, powered up, worked properly. Burned some things, shut it down. Came back the next day to the exact same setup, fired it up. Lased for a brief moment then stopped. RF OK light was not on, MOD OK light still working.
Tried different power supplies for +32V, +12V and +5V. Tried changing +32V input to as low as 24 volts and as high as 50 volts. All ground points are common, voltage is present on circuit board so it isn't a wiring issue.
Checked power transistor, and it's fine. Replaced all three IC's on the board with no change. Received an entirely new board from nightstick, same problem. Tried tuning the laser by compressing the coil on the head as per Chris's advice, no change.
Got oscilloscope. No signal present on laser, RF board output, or at power transistor.
NEW INFORMATION
Got an image of the ~50kHz reading from the oscillator:
The scale is 200mW/div and 10us/div
Interesting but useless info:
I was poking the probe around the middle IC, which is a 4 pin IC labeled
PY99AD
LM
393M
and I accidentally shorted the connection between pins 1 and 2. The oscope readout went crazy and the RF OK light suddenly came on. I did this three more times to confirm what I did and indeed the RF OK light comes on. I turned on the power to the laser thinking I might have just solved my problem, but sadly the tube does not light. I took this image of the oscope while the probe was shorting out pins 1 and 2:
The scale here is 5V/div and 0.2us/div!
So the signal present here was almost 45Vp-p and was at about 1MHz. Neither of the pins on their own have such a signal. But that STILL isn't the 27.12Mhz, all it does is turn the damn light on. I shorted out the IC again using a separate jumper and probed the power transistor to find no signal there, and I probed the oscillator to find the same old 50kHz signal there. Also, shorting pins 1 and 2 seem to turn on the RF OK light even when the PWM is not connected to 5V. So I'm thinking this mystery signal is merely two IC output pins cycling rapidly back and forth because they are each trying to drive opposite signals, or possibly some sort of loop of the IC shutting down due to the shorted pins. The Rf light comes on because the signal triggers whatever controls the LED.
So I will continue testing but I am finding out more about this board every time I do.
Important info:
I've taken a very close look at where the output of the oscillator goes. It goes immediatly through a filtering capacitor and from there into a can transistor which amplifies said signal. Starting at the output of the oscillator, the signal isn't getting very far. The The filtering capacitor connected in series with the output of the oscillator is NOT passing the 50kHz signal. So the signal isn't getting through that cap. This could be either because of a bad cap which is affecting the oscillator's output, or because the frequency from the bad oscillator is too low and the cap is filtering it out. I'm going to replace the capacitor first and we'll see what happens.
Received laser, powered up, worked properly. Burned some things, shut it down. Came back the next day to the exact same setup, fired it up. Lased for a brief moment then stopped. RF OK light was not on, MOD OK light still working.
Tried different power supplies for +32V, +12V and +5V. Tried changing +32V input to as low as 24 volts and as high as 50 volts. All ground points are common, voltage is present on circuit board so it isn't a wiring issue.
Checked power transistor, and it's fine. Replaced all three IC's on the board with no change. Received an entirely new board from nightstick, same problem. Tried tuning the laser by compressing the coil on the head as per Chris's advice, no change.
Got oscilloscope. No signal present on laser, RF board output, or at power transistor.
NEW INFORMATION
Got an image of the ~50kHz reading from the oscillator:
The scale is 200mW/div and 10us/div
Interesting but useless info:
I was poking the probe around the middle IC, which is a 4 pin IC labeled
PY99AD
LM
393M
and I accidentally shorted the connection between pins 1 and 2. The oscope readout went crazy and the RF OK light suddenly came on. I did this three more times to confirm what I did and indeed the RF OK light comes on. I turned on the power to the laser thinking I might have just solved my problem, but sadly the tube does not light. I took this image of the oscope while the probe was shorting out pins 1 and 2:
The scale here is 5V/div and 0.2us/div!
So the signal present here was almost 45Vp-p and was at about 1MHz. Neither of the pins on their own have such a signal. But that STILL isn't the 27.12Mhz, all it does is turn the damn light on. I shorted out the IC again using a separate jumper and probed the power transistor to find no signal there, and I probed the oscillator to find the same old 50kHz signal there. Also, shorting pins 1 and 2 seem to turn on the RF OK light even when the PWM is not connected to 5V. So I'm thinking this mystery signal is merely two IC output pins cycling rapidly back and forth because they are each trying to drive opposite signals, or possibly some sort of loop of the IC shutting down due to the shorted pins. The Rf light comes on because the signal triggers whatever controls the LED.
So I will continue testing but I am finding out more about this board every time I do.
Important info:
I've taken a very close look at where the output of the oscillator goes. It goes immediatly through a filtering capacitor and from there into a can transistor which amplifies said signal. Starting at the output of the oscillator, the signal isn't getting very far. The The filtering capacitor connected in series with the output of the oscillator is NOT passing the 50kHz signal. So the signal isn't getting through that cap. This could be either because of a bad cap which is affecting the oscillator's output, or because the frequency from the bad oscillator is too low and the cap is filtering it out. I'm going to replace the capacitor first and we'll see what happens.
?
...
Guest
STOP shorting those two pins! You should never ever see 50v across a lm393 its only even rated for a 35v max supply voltage and should probably be running off a +/-15v supply (or whatever you are feeding the laser). I don't know how there is 50v getting into it, but you are going to fry the poor chip (if you haven't done so already).
Also, on that note, I would look and see where the 50v is coming from, it really should not be in that part of the supply and if its getting into your supply rails it would explain why you fried the oscillator and probably any other silicon device on that bus (well maybe not the diodes)
Ok course the 50khz won't pass through a capacitor sized to pass 27MHz, you would need a capacitor 500 times as large to have the same impedance...
In any case, when you short pins 1/2 of the lm393, you are shorting the output to the non-inverting input. If there is a low impedance source on that input (well, lower than the output of the lm393 which is on the order of tens of ohms) you are shorting out the lm393 and feeding whatever is coming into the input to whatever the output of the opamp was feeding. If there is a high impedance source there (as is probably the case) you are telling it to ignore that input and just output the same voltage as is on the inverting input.
I would focus your efforts on the crystal oscillator, and see if you can't get something rigged to output something on the order of 27MHz into the supply. I don't know how good of a lab you are working in, but a simple function generator, pic IC with a 25mhz oscillator, radio transmitter for a rc car that runs at 25mhz just anything that will output a 25mhz signal to feed that preamplifier transistor.
Also, on that note, I would look and see where the 50v is coming from, it really should not be in that part of the supply and if its getting into your supply rails it would explain why you fried the oscillator and probably any other silicon device on that bus (well maybe not the diodes)
Ok course the 50khz won't pass through a capacitor sized to pass 27MHz, you would need a capacitor 500 times as large to have the same impedance...
In any case, when you short pins 1/2 of the lm393, you are shorting the output to the non-inverting input. If there is a low impedance source on that input (well, lower than the output of the lm393 which is on the order of tens of ohms) you are shorting out the lm393 and feeding whatever is coming into the input to whatever the output of the opamp was feeding. If there is a high impedance source there (as is probably the case) you are telling it to ignore that input and just output the same voltage as is on the inverting input.
I would focus your efforts on the crystal oscillator, and see if you can't get something rigged to output something on the order of 27MHz into the supply. I don't know how good of a lab you are working in, but a simple function generator, pic IC with a 25mhz oscillator, radio transmitter for a rc car that runs at 25mhz just anything that will output a 25mhz signal to feed that preamplifier transistor.
- Joined
- Mar 27, 2008
- Messages
- 478
- Points
- 28
I forgot to mention that the oscillator definately does not have the on/off function. The lead does not go anywhere, definately no traces leaving the pin.
I figured the 50v thing was not good at all. though it was 50p-p so only 25 volts were on the pins at any one time. My next move then will be to find a 27 MHz Source. My "lab" consists of my bedroom so I don't have any fancy equipment. I'm thinking I'll rip apart a 27mhz rc car controller. I should be able to just connect the antenna to the terminal the oscillator outputs to, right?
I figured the 50v thing was not good at all. though it was 50p-p so only 25 volts were on the pins at any one time. My next move then will be to find a 27 MHz Source. My "lab" consists of my bedroom so I don't have any fancy equipment. I'm thinking I'll rip apart a 27mhz rc car controller. I should be able to just connect the antenna to the terminal the oscillator outputs to, right?
- Joined
- Mar 27, 2008
- Messages
- 478
- Points
- 28
Pulled the oscillator out of the board and tested it, 50kHz output. I think we've finally found the problem after all this time. I'm going to pull the oscillator off my original board and see if it's any different.
EDIT: Oscillator on the original board is dead, no output at all. Tested the new one again and now it's operating at ~60kHz. Definitely toasted.
Time for a new one.
What would be really helpful to me now is for someone with the same supply to take an oscilloscope to the oscillator and let me know what to look specifically for. In particular I would like to know the p-p voltage output, including the max and min voltage, and the type of waveform outputted (sine wave, sawtooth, etc.). Again I don't know much about oscillators, maybe all of them are built to some sort of output signal standard, but I'd like to know the proper values regardless. The pin to test is opposite of pin 1, with the dot on the corner. You can't easily get at it, so instead meter this capacitor lead, it is directly connected to the output:
We're near the end of this journey I think.
EDIT: Oscillator on the original board is dead, no output at all. Tested the new one again and now it's operating at ~60kHz. Definitely toasted.
Time for a new one.
What would be really helpful to me now is for someone with the same supply to take an oscilloscope to the oscillator and let me know what to look specifically for. In particular I would like to know the p-p voltage output, including the max and min voltage, and the type of waveform outputted (sine wave, sawtooth, etc.). Again I don't know much about oscillators, maybe all of them are built to some sort of output signal standard, but I'd like to know the proper values regardless. The pin to test is opposite of pin 1, with the dot on the corner. You can't easily get at it, so instead meter this capacitor lead, it is directly connected to the output:
We're near the end of this journey I think.
- Joined
- Jan 7, 2007
- Messages
- 6,309
- Points
- 83
EH ---
Caution is advised --- You're about to become an electron like some of us here ;D Board level troubleshooting is an an art.
Mike
Caution is advised --- You're about to become an electron like some of us here ;D Board level troubleshooting is an an art.
Mike
- Joined
- Mar 27, 2008
- Messages
- 478
- Points
- 28
Updated the last post. Just need some readings from someone who has the same power supply design. I have to admit though, while I've been frustrated a few times, overall it was a positive experience figuring out the problem. Lets just hope this is all there is to it.
And Mike, I'm sorry to say, I'll never be an electron. I'm a photon at heart.
Daedronus, I have two boards and two dead oscillators. I'm going to try and fix both of them, and if I can get them both working I can sell you one to power your laser with. But I can't promise I'll be able to fix both of them so don't depend on me entirely!
And Mike, I'm sorry to say, I'll never be an electron. I'm a photon at heart.
Daedronus, I have two boards and two dead oscillators. I'm going to try and fix both of them, and if I can get them both working I can sell you one to power your laser with. But I can't promise I'll be able to fix both of them so don't depend on me entirely!
