Author |
Message |
MapacheRaper
Joined: Feb 15, 2018 Posts: 166 Location: Spain
|
Posted: Mon Jul 09, 2018 3:56 pm Post subject:
nicolas3141 simpler LFO troubleshooting |
|
|
***Newbie alert***
I adore the simplecity of the Nicolas3141 circuits and the simpler LFO is been my first try at veroboard. I have spend like 12h trying to debug the circuit with no avail.
When I connect the batteries only one LED stays on and does not blink. If I touch with the fingers the backside of the components to short it via my skin sometimes it starts to flipflop, erratically. The circuit:
I have built it with some discrepancies:
1-Im using 100nf for the timer cap. The build lists 1uF. SO it should work just outputting higher frequencys, right?
2-Im using a 100uF cap between the main +9/ -9 instead of 220uf, but I don´t think it is quite important.
3-The 9V bateries are 9.1V and 8.3V. Could this be the culprit?. Should I buy two new 9V?
4-Im not sure if I have fried the LM358. In fact, maybe I have fried 2 of them. Is there a fast way to check if the opamp is still working?
The conections are re-re-rechecked like 20 times and are correct. The solders are checked via tester and are correct.
Danke! |
|
Back to top
|
|
|
gabbagabi
Joined: Nov 29, 2008 Posts: 651 Location: Berlin by n8
Audio files: 23
|
|
Back to top
|
|
|
MapacheRaper
Joined: Feb 15, 2018 Posts: 166 Location: Spain
|
Posted: Wed Jul 11, 2018 4:53 pm Post subject:
|
|
|
I´m waiting for a chinese shipping to bring me the male breadboard cables so I can´t breadboard yet... Should be arrived 2-3 weeks ago, but you know chinese order always comes with some kind of surprise/lag or both.
I guess I´ll start with the version you post when the cables arrive. It can´t get more simple than that. One question: Why Nicolas is not using 100nF decoups in between opamp + - and ground but a 220mF between +9 and -9v?. I think is the first time I see it this way
Thanks GGabba! |
|
Back to top
|
|
|
gabbagabi
Joined: Nov 29, 2008 Posts: 651 Location: Berlin by n8
Audio files: 23
|
Posted: Wed Jul 11, 2018 10:39 pm Post subject:
|
|
|
on the breadboard i have switched to staples.
u can see it here, a bit http://electro-music.com/forwtopic.php?t=69028
As u sit in western Europe u may be able to get copper ones.
https://www.ecosia.org/images?q=staples+copper.
I guess the 220nF was what he just had laying around in the workshop, feel free to use 100nF.
Some are suggesting to use two 100nF Caps, one on each power-supply-pin to GND.
U would like to stay with the batteries or u plan to have a powersupply?
cu Last edited by gabbagabi on Thu Jul 12, 2018 9:05 am; edited 1 time in total |
|
Back to top
|
|
|
MapacheRaper
Joined: Feb 15, 2018 Posts: 166 Location: Spain
|
Posted: Thu Jul 12, 2018 12:55 am Post subject:
|
|
|
The staples idea is fantastic. I´ll check that. Thanks!
I plan to use it on a+-12v power supply in a eurorack case. The components of the power supply are about to arrive. When in does, I´ll check it with that voltage.
Your link is broken, but this is working. Cool finish!. Did you do the panels?
http://electro-music.com/forum/topic-69028.html |
|
Back to top
|
|
|
JovianPyx
Joined: Nov 20, 2007 Posts: 1988 Location: West Red Spot, Jupiter
Audio files: 224
|
Posted: Thu Jul 12, 2018 6:37 am Post subject:
|
|
|
While staples are nice and square and neat looking, they are not really the best idea for conductors. Staples are made of cheap steel and can rust over time which means your circuit will work until that happens. I know it's breadboard, but I've had breadboard projects stay on the breadboard for months - or more. And often power jumpers may stay in place even if the circuit is changed. Remember that fingers aren't perfectly dry and can leave moisture behind, and the moisture can rust the steel. Copper wire is a much better choice than staples.
Even worse is attempting to solder staples onto a board. I used to fix DEC minicomputers when DEC still existed. I remember one really frustrating day being called out for a "crashing" system. If you turned the computer off for an hour, then on, it would work for maybe a1/2 hour and crash. The problem turned out to be that some computer "engineer" had installed memory cards and used staples to program the address of the board. But solder will stick to steel only with a super hot iron and we just used basic pencils which worked fine with copper. Once I replaced the staples (which were showing signs of rust) with copper wire, the system booted up and never crashed again. When the system cooled off, things changed size from the difference in temperature and caused the bad solder on the staples to connect. Once the system heated up again, things moved around from heat and the staples disconnected.
YMMV, but this can happen and it's better not to tempt fate.
Looking through the thread, you mention the timing cap is 1uF instead of the 100 nF called out and that it would run at higher frequencies. A larger timing cap will make it work at lower frequencies. In fact, 10 times longer cycle time.
The cap across the batteries is not critical and 100 uF should work.
Batteries that measure as low as 8.3 v instead of 9 v means that battery is old, near dead and can't produce the current it ought to. I can't say for sure that this is a problem, but it's not good and should be replaced. I'd replace both with new ones OR use a mains type PSU.
Opamps are pretty hard to fry, it's possible, but not easy. I would look elsewhere for the problem. Besides miswiring, bad solder joints are one of the most common problems. I've built things that don't work, I then touch every connection with a hot soldering iron again and that often fixes it. So try to reflow all of the connections. Sometimes there can be crap on a lead or wire that prevents the solder from sticking. It's a good idea to always use flux when soldering, even if the solder has flux in it. The flux cleans and prepares the metal for soldering. When you use solder with flux in it, the flux boils away quickly and might not have time to clean the metal. I use paste flux, but there is also liquid. Flux will also help to solder to weird copper alloys that seem difficult to solder. Flux works like magic.
You can check things with a meter and they can show good, but if you're putting pressure on it with the probes, you might be causing it to connect for the test, but when you release the pressure the connection can open again. I'm thinking bad solder connections again, I've seen it so many times I can't count.
As g,gabba said, it is very wise to breadboard something before soldering. This becomes more important if you are substituting anything in the circuit. And I did read the note on the schematic that nothing is critical. However, you did change some things, and at least the timing cap being 10 times larger will cause the LFO to run 10 times slower than the circuit may have specified. You also mentioned that this cap change would increase the frequency - so your expectation was to see it run 10 times faster when in fact it will run 100 times slower than you expect it to which could lead one to think "it doesn't work". All "not critical" means is that it should still work, not that it will work properly. An LFO working at 1/10 the frequency stated is still working. Bad batteries might still work, but because of reduced current could cause it to run even slower.
Applying finger pressure to shunt things through the skin is not really a great technique, but can help discover things that aren't really connected that should be. Often what happens is the pressure or flexing is causing (again) bad solder connections to connect a little better (I said little) and the circuit may at least do something different. Also be very careful, you can still get a shock from 9 volt batteries - ever test one on your tongue? And two in series gives 18 volts. As a kid, I had a 24 volt Lionel train, I leaned on the track with one hand and got a surprise - so just be careful, it can scare the crap out of you when you don't expect it. 18 volts won't kill you, but it can still startle you. _________________ FPGA, dsPIC and Fatman Synth Stuff
Time flies like a banana. Fruit flies when you're having fun. BTW, Do these genes make my ass look fat? corruptio optimi pessima
|
|
Back to top
|
|
|
MapacheRaper
Joined: Feb 15, 2018 Posts: 166 Location: Spain
|
Posted: Thu Jul 12, 2018 9:12 am Post subject:
|
|
|
Thanks for your detailed comment, JovianPix.
The link g.gabba provides is to copper staples, which are the ones Im looking for while the duponts arrive.
In the circuit, the timer is 1uF, but Im using 100nF, so it should go 10x faster. Anyway, as I have 2u2 electrolitics and I prefer the LFO that works in the lower freqs I´ll try using two in series like -++- should produce 1u1 bipolar electroitic, isn´t?.
From somewhere:
Quote: | Short both the -ve terminals. The other two +ve terminals can be used as leads of a non-polar capacitor. |
This is for just testing as I have a bunch of mylar caps coming.
Im going to get two new 9v batteries and try it. I did the reflowing thing and the pressure quite times, but I´ll try the reflowing again. Is that or a faulty component
That´s totally cool to know that opamps are not that easy to fry. I guess for the high impedance of the pins, itsn't?. Whatever, I didn´t know and its relieving. Thanks again! |
|
Back to top
|
|
|
JovianPyx
Joined: Nov 20, 2007 Posts: 1988 Location: West Red Spot, Jupiter
Audio files: 224
|
Posted: Thu Jul 12, 2018 9:43 am Post subject:
|
|
|
MapacheRaper wrote: | Thanks for your detailed comment, JovianPix.
The link g.gabba provides is to copper staples, which are the ones Im looking for while the duponts arrive. |
Sorry, I didn't realize that copper staples were available at all. If indeed copper then they should be fine. I'm assuming these staples are not meant to work in a paper stapler.
Quote: | In the circuit, the timer is 1uF, but Im using 100nF, so it should go 10x faster. Anyway, as I have 2u2 electrolitics and I prefer the LFO that works in the lower freqs I´ll try using two in series like -++- should produce 1u1 bipolar electroitic, isn´t?. |
I must have misread that. Morning and coffee not taken full effect.
Quote: | From somewhere: Short both the -ve terminals. The other two +ve terminals can be used as leads of a non-polar capacitor. |
It makes no difference since regardless of which way, each capacitor deals with reversed polarity every other cycle. It's still very important to obey voltage ratings as too much can damage a cap. Under high current conditions, they can explode (such as backwards electrolytics in a PSU). This circuit won't ever cause that much current to pass through the timing capacitor.
Quote: | This is for just testing as I have a bunch of mylar caps coming.
Im going to get two new 9v batteries and try it. I did the reflowing thing and the pressure quite times, but I´ll try the reflowing again. Is that or a faulty component
That´s totally cool to know that opamps are not that easy to fry. I guess for the high impedance of the pins, itsn't?. Whatever, I didn´t know and its relieving. Thanks again! |
Relatively high output pin impedance is part of the reason. Depending on the opamp, there can be internal protection such as from carefully placed diodes or circuit bits that purposefully limit current.
I don't do vero or perf construction because it's a lot of whacky soldering. I prefer stripboards rather than make my own traces with wire and solder. It would seem to me that it's easier to generate solder bridges that way. I'd also look with a magnifier for such as even a hair fine conductor looks like a short to such circuits.
Flexing the board causing it to connect better to certain components and touching parts to change resistances doesn't tell you much because you don't really know how you're changing the circuit with your resistor of a finger. I've also had problems where I touch a meter or scope probe to it somewhere and all of a sudden it begins to work. Remove the probe and it dies, this is a sure sign that something is not connected with a near zero ohm soldered connection and that in reality it's a resistor that is low enough that if you parallel it with your finger, it starts to half-ass work. I've also seen soldered connections where it at first looks connected, but on magnified inspection shows that the solder became phobic of the thing it needs to connect (caused by crap on what you tried to solder - flux is your friend). I describe this as a reverse meniscus as can be seen with water against hydrophobic materials. Flux before soldering is the cure for that because it is caused by corrosion, oil, or other dirt on the lead or wire. It can also be caused by using cheap wire like from decommissioned telephone systems as some of that is a weird copper alloy that may contain metals like aluminum. It can solder, but may need high temperature. That wire was never meant to be soldered, it is meant to work on a punch down block which is a spring force connector. _________________ FPGA, dsPIC and Fatman Synth Stuff
Time flies like a banana. Fruit flies when you're having fun. BTW, Do these genes make my ass look fat? corruptio optimi pessima
|
|
Back to top
|
|
|
gabbagabi
Joined: Nov 29, 2008 Posts: 651 Location: Berlin by n8
Audio files: 23
|
Posted: Thu Jul 12, 2018 9:54 am Post subject:
|
|
|
staples in a professional microcomputer - nice!
i was using them once here in the school to teach how to solder, once the students had good results in soldering in the staples, we switched to copper bridges made of stripped cat6-cable then the results where almost perfect.
I have two TZVCO running on staples for a year now - it is an experiment how long they will last. But of course if u want quality u could not use them!
yes i made the panels, they are made of stuff called "Airex Forex" and simple color-ink-jet-print covered with transparent tape.
But i switched now to plywood and write what i need with a marker on it.
personally i would try to get the ciruit work with that 100nF and try later to increase the cap values, may u want to make them switchable between 100nF and 2,2µF?
As far as i know: by having two similar caps in series u not only half the capacitance but also the voltage rating, so having two 2,2µF/16V in series would result in 1,1µF/8V. Iam not shure about the polarity u mentioned.
edit: wow two posts while iam writing |
|
Back to top
|
|
|
MapacheRaper
Joined: Feb 15, 2018 Posts: 166 Location: Spain
|
Posted: Fri Jul 13, 2018 2:25 am Post subject:
|
|
|
So, I breadboarded the simple triangle that you posted using staples and it´s working at the first run!. Yeah!. Only 5min of building, so im loving the breadboarding thing
So, now I now the opamp is not fried and the 8v3 battery works for this circuit.
What´s the fault in the original veroboard/circuit I still don´t know but must be close. Thanks guys! |
|
Back to top
|
|
|
gabbagabi
Joined: Nov 29, 2008 Posts: 651 Location: Berlin by n8
Audio files: 23
|
Posted: Fri Jul 13, 2018 9:13 am Post subject:
|
|
|
ah, good news!
now u can test the original schematic.
and may u want to post a photo of ur stripboard? |
|
Back to top
|
|
|
MapacheRaper
Joined: Feb 15, 2018 Posts: 166 Location: Spain
|
Posted: Sat Jul 14, 2018 1:34 am Post subject:
|
|
|
Honestly, the crappiest looking building in the forum.
I have 5 different shots, but the attachment system is a bit strange... Last night I spent a good amount of time triying to debug but no avail. I have tried 2 full perfect 9v power sources, reflow all the conections check with the multi all the conections and possible bridges and test the separated components. Everything is OK, but it doesn´t work. Im about to desolder, trash the pcb and start a new one when I feel inspired. In the meanwhile there´s a lot of other builds waiting to happpen.
Your tri osc worked at the first run as I told you. Maybe I should keep breadboardind from there as you said
Description: |
|
Filesize: |
169.47 KB |
Viewed: |
214 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Filesize: |
169.47 KB |
Viewed: |
177 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Filesize: |
169.47 KB |
Viewed: |
184 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Filesize: |
169.47 KB |
Viewed: |
212 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
|
|
Back to top
|
|
|
gabbagabi
Joined: Nov 29, 2008 Posts: 651 Location: Berlin by n8
Audio files: 23
|
|
Back to top
|
|
|
AlanP
Joined: Mar 11, 2014 Posts: 746 Location: New Zealand
Audio files: 41
|
Posted: Sat Jul 14, 2018 3:42 am Post subject:
|
|
|
Those LEDs look tres' cool, even before they're turned on! |
|
Back to top
|
|
|
MapacheRaper
Joined: Feb 15, 2018 Posts: 166 Location: Spain
|
Posted: Sat Jul 14, 2018 8:56 am Post subject:
|
|
|
g.gabba wrote: | dont worry about the look, it will be better the more you solder
i cant see so much on the photo, but the dil socket looks already tierd, did u measured directly on the pins of the chip if the supply voltage is there - or on the copper trace?
just before u start new u could try to solder that what u have tested on breadboard
some common things i could say
- things are easier when u leave more space eg. one hole between every component or bridge and one hole on to the side of the chip
- if u use the upside down method it is good if u bend little feets or hooks on the end of the component legs, easyer to solder and makes better connection
- clean there where u break the board, just go over with a cutter in 45° or a little file, sometimes there are tiny shorts
- instead of soldering in the resistors vertical i bend the legs like that, it is easyer to solder in and easyer to measure something when u need to debug |
Yeah, the great mistake has been to minimize the space and try to solder exactly as the diagram so I don´t get lost in what is what and where it belongs. Newbie error. Effecively reflowing and measurament is quite twisted in such a tiny space.
The dip socket is a 16 dip socket cutted in half (the 8 dips should have arrived 2 weeks ago) but I do measure directly from the chip pins and is working properly
Not sure about the upside method, I guess you mean in the non copper side. I thought in the beginning to do that, but was afraid of the mirror effect. Next building I will try like this. Thanks for your support
@AlanP LOL, yep, it looks paleolithic at least! |
|
Back to top
|
|
|
MapacheRaper
Joined: Feb 15, 2018 Posts: 166 Location: Spain
|
Posted: Sat Jul 14, 2018 5:16 pm Post subject:
|
|
|
I have started a new veroboard. I have checked all components before soldering. I have swapped the LM358 with another one that is in another circuit and working.
I get the same error. Only one LED light, the other off and the pots do nothing.
I have rechecked all the tracks, possible bridges, multimeter all the things.
I declare this circuit to not work and to be a joke from Mr Woollastron. This or some kind of course.
Next!
Description: |
|
Filesize: |
136.16 KB |
Viewed: |
184 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Filesize: |
136.16 KB |
Viewed: |
173 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
|
|
Back to top
|
|
|
gabbagabi
Joined: Nov 29, 2008 Posts: 651 Location: Berlin by n8
Audio files: 23
|
Posted: Sat Jul 14, 2018 11:41 pm Post subject:
|
|
|
dumb question - do u have cut all four traces under the opamp or just the three of them? |
|
Back to top
|
|
|
MapacheRaper
Joined: Feb 15, 2018 Posts: 166 Location: Spain
|
Posted: Sun Jul 15, 2018 1:09 am Post subject:
|
|
|
Just three.pins 3 and 6 go connected to gnd |
|
Back to top
|
|
|
gabbagabi
Joined: Nov 29, 2008 Posts: 651 Location: Berlin by n8
Audio files: 23
|
Posted: Sun Jul 15, 2018 9:34 am Post subject:
|
|
|
hm,
even in the simulation i did quickly the circuit worked well,
did u have tryed the full ciruit on breadboard?
may u also try to connect R3 directly to the wiper of P1? so u could test if the combination of the trimmer and R4 causes the trouble |
|
Back to top
|
|
|
MapacheRaper
Joined: Feb 15, 2018 Posts: 166 Location: Spain
|
|
Back to top
|
|
|
|