Author |
Message |
cslammy
Joined: Apr 27, 2018 Posts: 206 Location: USA
Audio files: 1
|
Posted: Mon Oct 14, 2019 6:46 am Post subject:
|
|
|
Project_2501 wrote: | Frequency Central PSU is alive!
ut that will be amazing. I'm assuming the ic's are getting hot cause of a short somewhere. I did check my PSU is providing 12v/-12v. Hunt the continues! |
Master P:
I assume you are using this LMNC design:
https://www.lookmumnocomputer.com/cem-3340-diy-simple
From my own brief experiments with LMNC: He makes mistakes; we all do. I am tempted to joke around that he's too worried about his steampunk vibe and YT subscriber count to check his schemos, but really, I like LMNC a lot. He makes it cool to be a DIY nerd, and perhaps helps us collectively get dates, as long as you have a cool earring.
OK off the soapbox and on to your issue:
First question: did you socket the chips in your VCO? If not, you might consider doing so. It might even be worth rebuilding the circuit if you used stripboard or whatever, **always** socket chips for new designs IMO!
once you know it will work, then it's OK to not socket. Because: Troubleshooting hot chip issues is way harder when you can't pull the chip or at least lift some of its legs.
Second question: do you use a bench supply that limits current and tells you V and A in real time?
If not get one, absolutely, frigging **every DIY bench needs one**. Along with a DVM (which can be had cheap) and a scope (which can be had cheap) this is a 100% essential thing for your bench. Not taht cheap, but what is your time worth?
Something like this?
https://siglentna.com/power-supplies/spd3303c-series-programmable-dc-power-supply/
Get that out of the amazon box and then Set the current limit to 10mA and +/- 12V. 10mA should be fine for the LMNC circuit if it works.
power up the circuit again w your bench supply and see if you have a short (which will be apparent because the V on your psup will not go to 12V.) So you will see 10mA current source and say +2V instead of +12V. Yep--no bell bottoms for you, because you have a short! And since +12 only goes to +2, you know it's on the + rail!!!!
OK now pull the chips and power the circuit again. Do you still have a short? Yes?
Put in the chips (2 in this case) one at a time and see if you still have a short.
I always do this to help determine where the short is. If it's only one of the chips that isolates the issue. If you still have a short without the chips in at all, well, we know the IC's are not 100% of the problem.
OKlet's say it's the 3340 that's causing the short? First I would check the obvious....
The 3340 has specific power requirements, check out what fellow poster druid has to say:
https://electricdruid.net/cem3340-vco-voltage-controlled-oscillator-designs/
read the paragraph: "Another quirk is the supply voltage". Wit that in mind:
Is pin 3 being current limited correctly? Check for resistance between + on your power supply and pin 3. Is that the value you want?
Check the voltages at the chip (not at your psup) I assume you've already done that? So: What is the voltage seen at 3 and 16 relative to ground?
Check for shorts between 3340 pin 8, 4, and 10 and ground. Any of those would cause the chip to heat up I'd think.
OK let's say it's the TL072:
For the TL072, the most obvious thing I've found to heat up the chip (beyond putting it in backwards) is to not put a series resistor at output, then short out the output (say, a TS plug not plugged all the way in, so it sees ground). i don't see that on LMNC's design, and it's needed. So put a 1K resistor in series with those outputs. (I have been told the TL0xx family has output short protection, but in my expereience they still get hot if their output gets shorted)
Even if that's not the issue--it probably isn't, you need this. Is the output shorted?
OK that's a start. Try some of that and let us know. As my old tech boss said: anything can be fixed. _________________ Visit my AUDIODIWHY blog and website Last edited by cslammy on Tue Jan 14, 2020 7:35 am; edited 1 time in total |
|
Back to top
|
|
|
rockprada
Joined: Oct 24, 2019 Posts: 3 Location: Detroit, MI
|
Posted: Thu Oct 24, 2019 9:42 pm Post subject:
|
|
|
Project, do you have a picture or diagram of your board? I have just finished building the LMNC VCO based on his diagram and was able to get it to work *mostly*. The square wave is not very flat and PW pot will cause the sound to cut out if you are too high or low. However, I realized that my ICs were getting very hot when I first plugged it in. Turns out I made a simple mistake of not wiring the ground for the jacks correctly. I used a bottom pin instead of a side pin for ground. (I thought they were both ground) after changing the pin connection the ICs were not even warm and the saw and tri sound and look great.
My advice would be to triple check your board with the diagram you are using and make sure everything is grounded properly.
Now if anyone can explain why the PW pot would cause a cutoff at low and high values I'd be able to put this issue to bed. Currently adding the tri to sine based on the design in this thread. |
|
Back to top
|
|
|
Project_2501
Joined: Jan 31, 2018 Posts: 68 Location: Los Angeles
|
Posted: Mon Jan 13, 2020 6:30 pm Post subject:
|
|
|
Long time no see.
Moving from South California to North California took a lot out of me. 8 hours north and still in the same state..... just lovely.
Well news on the home front, sam posted a pcb (performance PCB, which In my peanut brain means TURBO!) recently. Going to order one soon.
Has a rotary switch for octaves! Just lovely!
https://www.lookmumnocomputer.com/1222-performance-vco |
|
Back to top
|
|
|
cslammy
Joined: Apr 27, 2018 Posts: 206 Location: USA
Audio files: 1
|
Posted: Mon Jan 13, 2020 6:53 pm Post subject:
|
|
|
Project_2501 wrote: | Long time no see.
Moving from South California to North California took a lot out of me. 8 hours north and still in the same state..... just lovely.
Well news on the home front, sam posted a pcb (performance PCB, which In my peanut brain means TURBO!) recently. Going to order one soon.
Has a rotary switch for octaves! Just lovely!
https://www.lookmumnocomputer.com/1222-performance-vco |
Hey project! I am in Nca as well. P-M me sometime--maybe we are neighbors?
As far as LMNC's new VCO design--I dunno.
I love the guy, who else can be as nerdy as we mostly are and still get a gazillion youtube hits, but really? And the punk earring thing? And The girlfriend? and the video game tie ins.
OK that's the dude, but what about his work?
I am into low parts count as much as anyone, but this design to me seems to cut a lot of corners--even by my standards, which are pretty low
The others can weigh in, maybe I am missing some of the schemos, but this looks to me like:
--Choose your octave switch resistors carefully because that's all you are counting on to make sure the octaves are really octaves. The resistors I think he wants you to use are going to need to be hand matched (all of them in any VCOs you build) and worse, sensitive to temp changes as well. Not great if you have bunch of VCOs that need to switch octaves in lockstep and stay in tune.
You will do much better with smd .1% 10K resistors or whatever, with good heat tolerance, which really aren't all that expensive, just harder to solder.
--I am not sure I am seeing everything but it looks like there is no accounting in this design for the various waveform outputs' dc offsets--you'll see a bias offset for each one. From my bench and online docs: The 3340 does not produce audio output that straddles 0V dc, and for most audio rigs you want that. Not a hard fix--to keep it low parts count you could cap-couple the outputs but it looks to me like for whatever reason he omits this from his design. Maybe his PA system inputs are is AC coupled? If you have certain VCFs VCAs etc downstream the bias offset in his design might make other modules' output sound really crappy. Depends on the module of course but the whole idea of modular is just that--modular. A VCOs output should work with other modules without making everything sound bad!
--He should use a poly cap for the main timing cap on a 3340. Others might disagree, but even in sunny CA USA this seemed ot me, on the bench anyway, to help my 3340 endeavors w/ overall tuning stability. And, I don't think you can find a poly cap in the size he specs here. (forgive me, I am not studying LMNCs design extensively, perhaps I am wrong here?)
--what format is he using? I am not sure what this PCB will fit behind. Not euro?
Turbo? I dunno. In general this design is built for speed--his. I think thats what he means. Low parts count, sure, it will work sorta, but overall, for me and again I will use cups and string for a phone if I think it would work--I'd want to do better-- For any studio type work I think the tuning issues are going to probably drive you crazy.
But then again what the hell do I know? I am not a youtube star. I don't have an earring. Don't tour and I dislike video games. I do however have a cool girlfriend. _________________ Visit my AUDIODIWHY blog and website |
|
Back to top
|
|
|
cslammy
Joined: Apr 27, 2018 Posts: 206 Location: USA
Audio files: 1
|
Posted: Tue Jan 14, 2020 7:30 am Post subject:
|
|
|
rockprada wrote: |
Now if anyone can explain why the PW pot would cause a cutoff at low and high values I'd be able to put this issue to bed. Currently adding the tri to sine based on the design in this thread. | \
My suggestion: Put the PWM audio signal on a scope.
If the waveform "goes away" at extreme settings that's your problem. It will look like a DC offset (a few volts) only, not audio. That's what's causing your problem, you've PWM'd so severely you aren't left with any audio.
If that's the case, you'll need to probably increase the current limiting resistor(s) that are used to condition the control signal for PW. Adjust this (make it bigger, probably) until it sounds the way you want.
If you post the schemo you are using I might be able to tell you the resistor I am talking about (again this assumes you can see the issue I mention on a scope) _________________ Visit my AUDIODIWHY blog and website |
|
Back to top
|
|
|
rockprada
Joined: Oct 24, 2019 Posts: 3 Location: Detroit, MI
|
Posted: Tue Mar 31, 2020 7:02 am Post subject:
|
|
|
cslammy wrote: |
If the waveform "goes away" at extreme settings that's your problem. It will look like a DC offset (a few volts) only, not audio. That's what's causing your problem, you've PWM'd so severely you aren't left with any audio.
|
I put the wave on the scope and when turning the pot i am getting my pulse range from about 2 to 7. Any lower of higher and the pulse is gone.
Stripboard layout is attached for review. Might not be the prettiest one out there...
Thanks for checking it out. I've got a lot more time on my hands with the current shutdown in Michigan.
|
|
Back to top
|
|
|
cslammy
Joined: Apr 27, 2018 Posts: 206 Location: USA
Audio files: 1
|
|
Back to top
|
|
|
rockprada
Joined: Oct 24, 2019 Posts: 3 Location: Detroit, MI
|
Posted: Sun Apr 05, 2020 12:48 pm Post subject:
|
|
|
So I found an updated actual schematic from lookmumnocomputer and it has a handful more resistors added in the PWM circuit. I was able to modify my layout to show what worked for me.
The biggest issue I found was that the original layout was taking in full +12V to the PW pot. This was causing an over-voltage amount being sent to the CEM3340 PWM pin. The PWM pin is looking for 0-5V to set the pulse width. I was going well beyond that. After some trial and error I found that the best resistor was 24kOhm to give me the full range when adjusting the 50kOhm potentiometer. (original schematic called for 100k)
This image is the fully tested version I have made. The only minor issue I have is that the PWM CV pot has some effect on the PW even when there is no CV coming in. It doesn't effect operation to the point that I would be frustrated with it. As long as the pot is fully CCW when no CV is plugged in you get the full pulse width with no cutout. If the pot is fully CW the pulse width will cut out when the PW pot is also fully CW.
|
|
Back to top
|
|
|
N-Bee
Joined: Oct 17, 2020 Posts: 1 Location: NYC
|
Posted: Sat Oct 17, 2020 10:11 pm Post subject:
|
|
|
I had a lot of trouble tuning Sam's design. I have found that the tracking is much better if you implement the Linear FM circuit as shown on the CEM3340 data sheet (pins 11,12,13). It is a very simple adjustment to boot! It now tracks near perfectly across 6 octaves for me. |
|
Back to top
|
|
|
|