Author |
Message |
Europanaut
Joined: Apr 04, 2014 Posts: 46 Location: Nanaimo, BC, Canada
|
Posted: Fri Jun 20, 2014 9:54 am Post subject:
Rev 1 oscillator build troubleshooting. |
|
|
Hi!
Well, I'm officially in the game now. The first oscillator is done. It's an old version, bought from a friend who never started his project.
It's running from +/-15 supply, and the recommended mods have been done.
U3 is Ray's matched 2N3904/tempco combination.
It's working, sort of. No sine, and pretty uneven volumes from the other outputs. No trims or front panel pots have any effect on the audio. I don't have a scope, so there may be subtle changes going on with the sine shape, but my ears are not detecting them. In any case, I need to figure out why the panel controls have no effect, then I'll look at the trim calibrations.
I've gone over the board in detail, and everything is oriented correctly. No backwards IC's, caps, or diodes. All front panel wiring is correct. Solder joints look good under 10x lighted magnifier. Pots have been bench tested. Grounds across the panel to all jacks are good. (Interestingly, in my initial check, I discovered that one of the CV IN 1/4" jacks was open across the tip and sleeve. Mechanically connected inside the jack. Never had THAT happen before. I'm not sure if that would have taken any other components out during power-up.)
Opinions? Troubleshooting has always been my weakness, so I'll try to keep up.
Thanks |
|
Back to top
|
|
|
Europanaut
Joined: Apr 04, 2014 Posts: 46 Location: Nanaimo, BC, Canada
|
Posted: Sat Jun 21, 2014 7:21 pm Post subject:
|
|
|
Crickets.....
I'm going to check input voltages on the IC's, and I'll report back. At least I'll know that the chips are getting power. Seems like a logical next step. |
|
Back to top
|
|
|
Europanaut
Joined: Apr 04, 2014 Posts: 46 Location: Nanaimo, BC, Canada
|
Posted: Sun Jun 22, 2014 11:17 pm Post subject:
|
|
|
Update:
All IC's are getting +/- 15v.
All IC's have been swapped for new components.
Output jack for TRI wasn't achieving consistent ground to the panel. Fixed.
Over all, no change. Square and Saw have output. Triangle has very weak output, and nothing at all from Sine.
And, as before, none of the front panel controls have any effect on the output. |
|
Back to top
|
|
|
electrotech
Joined: Apr 24, 2013 Posts: 38 Location: Ayrshire Scotland
|
Posted: Mon Jun 23, 2014 1:52 am Post subject:
|
|
|
Hi Europanaut,
Is your circuit the "VCO Old Design I" (in The Cave) on Ray's site ?
It's very difficult to suggest what might be wrong if I'm not sure which circuit you're referring to ! ( I'm not sure if it's acceptable to post the circuit here...)
Regarding your lack of a 'scope :- suitably attenuated, I believe (with the appropriate software) you can use the audio input on your computer to display waveforms. Maybe others can suggest an appropriate program to do this. ( I personally don't need to do this because I do have a 'scope )
Also, even a digital multimeter will tell you if any op-amp outputs are stuck close to the supply rails.
As a first step, I would double-check that all circuit grounds are present and correct.
Andy |
|
Back to top
|
|
|
Europanaut
Joined: Apr 04, 2014 Posts: 46 Location: Nanaimo, BC, Canada
|
Posted: Mon Jun 23, 2014 8:23 am Post subject:
|
|
|
Electrotech, thanks for the reply!
Yes, sorry about missing that important bit of info in my description. It's Rev 1, VCO Old Design, from the cave. I have three of them, actually.
By "circuit grounds", are you referring to points on the IC's? Or other particular points in the circuit? I've checked all ground points on the front panel, and they are all good, and are grounded to the pcb.
I'm trying to locate a scope at a reasonable price, so my current scopeless situation will soon change. I have a friend with a good scope, but I'd like to use as little of his time as possible, so I'm trying to identify and fix any dumb mistakes before taking the module to his shop.
Thanks again, the help is appreciated. |
|
Back to top
|
|
|
electrotech
Joined: Apr 24, 2013 Posts: 38 Location: Ayrshire Scotland
|
Posted: Tue Jun 24, 2014 4:02 am Post subject:
|
|
|
Hi,
That's a wise decision about getting a 'scope - it will make trouble-shooting so much easier !
Actually, setting up waveform symmetry and the triangle converter would be difficult without one.
Now I know what version of VCO you've got, I've had a look at the schematic and see that there aren't actually that many external ground wires to the panel potentiometers as I thought there might be. If you've already checked these that's fine.
If the oscillator is running but nothing is altering it's frequency, I would investigate the voltage at the left-hand side of R10 (R2P2) and make sure that it's changing as expected when you turn the coarse tuning control.
Are the jack sockets you're using the 'open' style type like the ones pictured on the MFOS web-site ?
f not, I'd check their contacts are wired correctly.
Also, I'm not sure what you meant by this :-
“Interestingly, in my initial check, I discovered that one of the CV IN 1/4" jacks was open across the tip and sleeve. Mechanically connected inside the jack”
Can you explain further ?
You need to get the ramp wave working first as the rest are derived from this. You can't set up the triangle until the ramp is correctly set up and likewise with the sine wave which is derived from the triangle.
One other thing, are all the IC's soldered to the board or are they in sockets ? - some times you can get bad sockets hence checking voltages on the actual IC pins themself.
I connect the negative lead from my multimeter to chassis ground, then go round the circuit to make sure that the expected DC voltages for the supply lines are correct on the appropriate IC pins. - take care when doing this as one slip with the prod can fry the chip !
With oscillator circuits though, any AC signals present can give funny results - hence why you really need a 'scope.
Hope this makes sense.
Andy |
|
Back to top
|
|
|
Europanaut
Joined: Apr 04, 2014 Posts: 46 Location: Nanaimo, BC, Canada
|
Posted: Tue Jun 24, 2014 8:49 am Post subject:
|
|
|
Thanks for digging into this, Andy.
To answer your questions......
I am getting a pretty clean ramp output.
I will check the voltage at R2P2, and report back.
Jack sockets are open style.
To clarify the CV IN jack issue. This jack was showing continuity from the tip to sleeve. A malfunctioning jack. It's been replaced.
The IC's are socketed. All voltage checks were done at the IC pins. |
|
Back to top
|
|
|
Europanaut
Joined: Apr 04, 2014 Posts: 46 Location: Nanaimo, BC, Canada
|
Posted: Tue Jun 24, 2014 8:40 pm Post subject:
|
|
|
Update:
The R2P2 is getting a range of 1.11 to -9.47 from the 100k Lin pot.
I bench-tested the pot before installation, and it's functioning correctly. |
|
Back to top
|
|
|
electrotech
Joined: Apr 24, 2013 Posts: 38 Location: Ayrshire Scotland
|
Posted: Wed Jun 25, 2014 10:22 am Post subject:
|
|
|
Thanks for reporting back. I put the input summing amp circuit into the simulator and those voltages are about right (with the mod for + & - 15V supply).
It is very strange if the frequency is not changing when you turn the coarse tune pot.
This suggests a problem with U1 (LF442) or the surrounding circuitry, but you say you've already changed the chip and checked the board over for constructional errors...
I think you'll have to get a 'scope onto it to figure out what's going on.
I don't know what else to suggest from this distance, sorry.
Andy |
|
Back to top
|
|
|
Europanaut
Joined: Apr 04, 2014 Posts: 46 Location: Nanaimo, BC, Canada
|
Posted: Wed Jun 25, 2014 12:29 pm Post subject:
|
|
|
Thanks Electrotech. All good. I appreciate the time that you've taken.
I've got an appointment with a scope later today. I'll be sure to keep updating this thread as the diagnostics progress.
(It drives me nuts when a troubleshooting thread is left hanging, just as it appears that a solution may have been found!) |
|
Back to top
|
|
|
electrotech
Joined: Apr 24, 2013 Posts: 38 Location: Ayrshire Scotland
|
Posted: Wed Jun 25, 2014 3:21 pm Post subject:
|
|
|
OK I'll keep trying
Now, please do not be offended ( because I don't know how much electronics experience you've got ) but with the symptoms you've described, it sounds as if there is either a dry joint somewhere (i.e. high resistance/open circuit) or a short-circuit somewhere, around U1 or the logging transistors U3. Voltage checks with reference to the circuit/schematic diagram should find this.
In fact, if you can measure the voltages in this area, write them down on the schematic at the appropriate points then scan/post a picture with your findings here, that will help diagnosis greatly.
Keep at it - you'll get it sorted eventually !
Andy |
|
Back to top
|
|
|
Europanaut
Joined: Apr 04, 2014 Posts: 46 Location: Nanaimo, BC, Canada
|
Posted: Sat Jun 28, 2014 6:19 pm Post subject:
|
|
|
Electrotech:
No offence taken! Treat me as a newbie. I'm not really new to this, but troubleshooting has always been my weak point, so I'm trying to learn as I go through this painful process.
I was at my tech's shop today, and have a bit more info on this misbehaving circuit. (Bear with me here, the tech is a very busy amp repair guy, and does his best to squeeze in a bit of time for my projects. As a result, I may not yet have all the info that you need, and certainly do not understand all of what I'm about to relate. )
He re-flowed some connections, and scoped what he had time for.
The High Freq Compensation circuit has been disabled for these diagnostics.
Here's what I've can provide from my hastily written notes:
Fine R3P2 is showing a range of .35 to 4.5.
Course R2P2 is showing a range of -9.5 to +1.1
U1 LF442 Pin 1 is showing a range of .122 to .338 when the course pot is turned.
Pin 7 of the LM394 (which is now two matched 2N3904's) is showing .338 at pin 7.
The 2N3904 closest to where R24 would normally be installed, is showing resistance between the base and collector. According to the tech, this is odd.
The junction of R28 to Pin 6 of the LF442 appears to be grounded, but should show somewhere around 8v?
The 1v/Oct Scale Trim appears to be non-functioning. UPDATE: I replaced it with a functioning pot. No change to the circuit behaviour. Also, I bench tested the removed "non functioning" pot, and it works fine.
One other thing that I may not have emphasized enough, is that the output from the ramp, triangle, and square are barely audible. Still nothing from the sine output. Also, I've tried swapping out three LF442 chips, with no change in circuit behaviour. |
|
Back to top
|
|
|
electrotech
Joined: Apr 24, 2013 Posts: 38 Location: Ayrshire Scotland
|
Posted: Sun Jun 29, 2014 6:08 am Post subject:
|
|
|
Thanks for this additional information and yes, I had assumed that the raw ramp waveform was at the correct level...
You need to get the ramp output working correctly before worrying about the others because, as I've already mentioned, they are derived from the ramp waveform and need it to have the correct amplitude and be centered properly or they won't work or have very distorted outputs.
Pin 6 on U1 would be at around +8V if it wasn't for the left hand transistor in the logging pair pulling current and bringing that point down to around ground level (this point is at high impedance). Also, the changes at that point are very small (less that 1mV) so are effectively not measurable. The output of the op-amp U1-B is measurable though and should swing negatively by a few volts when the voltage at the input summing point (U1-A pin 2) changes and this is what sets the frequency of the oscillator.
If you set the coarse & fine panel controls plus all other trimmers to 50% I'll tell you what voltages I got with the circuit simulation.
[ By the way, although simulations don't always give you the correct answers they're usually not far out and are certainly a good way of finding out how a circuit works. You really need to know how a circuit works before you can fault-find on it otherwise you won't know if the voltages/waveforms at any particular point are correct and so won't know which parts are causing the problem ! ]
The circuit had a ramp at U2-A (RAW) output going from 0V up to +1.5V.
A couple of simulated DC results :-
U3 emitters -0.5V
U1-A Pin 1 +61mV
Since you are not getting the correct anticipated voltage at the base of the left-hand transistor (on the diagram) I suspect the main problem is with one (or both) of the 2N3904's you're using in the logging converter. Please double check these are connected the right way round (!) - see attachment.
If they are 'suspect' in any way just change them. Transistors that have had the wrong voltages applied at any time or overheated whilst soldering can be permanently damaged or have altered characteristics even if they test out OK on a meter.
As a quick check you could temporarily fit a couple of new ones (they don't have to be matched for test purposes) and see if that makes any difference.
Good luck.
Andy
ps I've attached the schematic diagram so others know what we're talking about.
Hope Ray doesn't mind !
Description: |
|
Filesize: |
15.21 KB |
Viewed: |
11998 Time(s) |
|
Description: |
|
Download |
Filename: |
VCO20060319_WMODS_pg1_schem.pdf |
Filesize: |
33.23 KB |
Downloaded: |
235 Time(s) |
|
|
Back to top
|
|
|
Europanaut
Joined: Apr 04, 2014 Posts: 46 Location: Nanaimo, BC, Canada
|
Posted: Mon Jul 07, 2014 1:03 am Post subject:
|
|
|
A friend and I had a bit of time to look at the oscillator this evening. We made a few observations.
(Based on previous suggestions in this thread, we tried a couple of different 2N3904's, but didn't notice any change in behaviour.)
This bit is probably the most interesting.
When we check the sweep of the course pot at R2P2 BEFORE R10, it behaves as expected. When we check it AFTER R10, there is no sweep.
UNLESS....we physically remove U1 (the LF442). As long as that chip is not in the socket, we have a sweep from the course pot. As soon as we pop the IC back in - no sweep.
We tried a couple of LF442's, and a TL072. Same results with every chip. |
|
Back to top
|
|
|
m.o
Joined: Jul 05, 2014 Posts: 44 Location: Sweden
|
Posted: Mon Jul 07, 2014 1:54 am Post subject:
|
|
|
Europanaut wrote: | A friend and I had a bit of time to look at the oscillator this evening. We made a few observations.
(Based on previous suggestions in this thread, we tried a couple of different 2N3904's, but didn't notice any change in behaviour.)
This bit is probably the most interesting.
When we check the sweep of the course pot at R2P2 BEFORE R10, it behaves as expected. When we check it AFTER R10, there is no sweep.
UNLESS....we physically remove U1 (the LF442). As long as that chip is not in the socket, we have a sweep from the course pot. As soon as we pop the IC back in - no sweep.
We tried a couple of LF442's, and a TL072. Same results with every chip. |
Actually I believe that is the expected behavior.
In a circuit like this the op-amp will feed current back through R24 until both it's inputs are at the same potential (if it can), and with the + input at ground, the op-amp will make the - input also (virtual) ground.
As said before a scope is invaluable in debugging these things, so get one - a used analog 20MHz scope is quite ok.
I would take a look at the comparator and ramp-reset circuit, the s components around U2-B.
You should have a constant voltage at pin 6, this would be the "trigger value" for when the ramp is reset by emptying the C6 cap through opening Q1 (have you checked this transistor?).
Another thing that can go wrong with this circuit would be if the time/amplitude of the "reset pulse" going to Q1 is too short/not strong enough, this would cause low amplitude ramp that is offset to the + side.
Another tip, for debugging things is to start removing "extra" stuff from the equation.
For example if possible, remove LIN FM and the Sync-in (I think you could lift R34 without affecting the core circuit).
If you had a scope I'd suggest probing U2-A pin 1 and also lifting R12 to totally remove all the waveshaper circuits, but without scope it's not as simple but you could try lifting D2 and D3 to remove anything beyond the Saw output buffer... |
|
Back to top
|
|
|
Europanaut
Joined: Apr 04, 2014 Posts: 46 Location: Nanaimo, BC, Canada
|
Posted: Mon Jul 07, 2014 9:10 pm Post subject:
|
|
|
Thanks m.o.
Pin 6 of U2 is reading a steady 13.39.
A scope is the next thing on my buy-list. Unfortunately, it's summer, and my work is somewhat seasonal. This means that spare cash is a bit scarce for the next couple of months. Once Sept/Oct rolls around, it's all good again, but for now, I have to do the best I can with a multimeter and the assistance of you patient folk! |
|
Back to top
|
|
|
electrotech
Joined: Apr 24, 2013 Posts: 38 Location: Ayrshire Scotland
|
Posted: Thu Jul 10, 2014 1:03 am Post subject:
|
|
|
Pin 6 of U2-B should NOT be at 13.39V !
R47 & R46 form a potential divider between ground and +15V (in your case).
The voltage should be 475/(10,000+475) x 15 = 0.68V or thereabouts. The effect of R34 + R33 in parallel with R47 + R46 will be negligible.
I would check the value of these resistors with a DMM (you might have to de-solder one end to get an accurate reading)
If these measure OK make sure they are connected in circuit correctly i.e. good soldered joints with no damaged PCB tracks or shorts anywhere.
Andy |
|
Back to top
|
|
|
Europanaut
Joined: Apr 04, 2014 Posts: 46 Location: Nanaimo, BC, Canada
|
Posted: Sun Jul 13, 2014 11:34 pm Post subject:
|
|
|
Progress! Big time.
Electrotech, you were right on the money. R46 was pooched. I replaced it, and it all lit right up. All I had was 5% resistors, will that be ok at this point in the circuit?
Sine still wasn't working, but since I was getting audio, it was easy to locate a bad solder joint at R49.
The range was quite limited, so I experimented with different resistors at R65, and ended up using two 47k's in parallel.
So it's all back together, and sounds excellent. I'll calibrate it on my friend's scope this week.
As soon as my next parts order arrives, I can build up the next two oscillators! Pretty fun!
Thanks again Electrotech and m.o. I appreciate your time and expertise during this confusing little adventure!
Description: |
|
Filesize: |
3.51 MB |
Viewed: |
165 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
|
|
Back to top
|
|
|
electrotech
Joined: Apr 24, 2013 Posts: 38 Location: Ayrshire Scotland
|
Posted: Mon Jul 14, 2014 2:20 am Post subject:
|
|
|
I'm glad you finally got it sorted and even a 470 ohm resistor would be o.k. at that position so 5% is fine.
In future builds, I would check the value of every resistor with a multimeter before soldering them to the board, because some resistors are hard to read by eye - I do this when I'm building stuff.
To be honest, initially when you said that the front panel controls had no effect and that you were "getting a pretty clean ramp output" really sent us off down the wrong track and it was only when you measured the voltage on pin 6 of U2 (at the suggestion of m.o - cheers for that) that we got back on track.
All part of the learning experience but you'd have got it fixed much quicker using a 'scope !
Regards,
Andy |
|
Back to top
|
|
|
m.o
Joined: Jul 05, 2014 Posts: 44 Location: Sweden
|
Posted: Mon Jul 14, 2014 6:03 am Post subject:
|
|
|
Grats on fixing it!
It feels good when you find "it" doesn't it.
I was just looking at the schematic and following some of my own experiences (I have learned a lot by using LTSpice to simulate different synth circuits then building them from that).
I actually didn't realize until this thread that the MFOS vco's used that low value comparator / low amplitude raw saw wave (even though it clearly shows in the description) - in my vco builds I have always gone for something like 3-5 volts.
I wonder what the reasoning for it is? |
|
Back to top
|
|
|
Europanaut
Joined: Apr 04, 2014 Posts: 46 Location: Nanaimo, BC, Canada
|
Posted: Mon Jul 14, 2014 8:31 am Post subject:
|
|
|
Yes m.o. That moment when it (mostly) worked, was worth all the head scratching! Well, most of the head scratching. It did give me a pretty itchy head.
electrotech, this quote from your post needs to be repeated for emphasis.
"I would check the value of every resistor with a multimeter before soldering them to the board."
I was saying exactly the same thing to a friend last night. That simple step would have saved a bunch of time and hassle. (But, I did learn a lot about this circuit over the last couple of weeks, so it's not all bad!)
Fully agreed about the necessity of having a scope. As soon as my business kicks in again after the summer, I'll be on it. I was even considering ordering a DSO as a temporary fix, but I've read such mixed reviews.... |
|
Back to top
|
|
|
_Shorty
Joined: Jul 07, 2014 Posts: 2 Location: Parksville, BC
|
Posted: Mon Jul 14, 2014 12:11 pm Post subject:
|
|
|
I think for the calibration portion we'll do just fine with some oscilloscope software. |
|
Back to top
|
|
|
Europanaut
Joined: Apr 04, 2014 Posts: 46 Location: Nanaimo, BC, Canada
|
Posted: Fri Jul 18, 2014 1:17 am Post subject:
|
|
|
So, we now have oscilloscope software running on my PC.
Definitely some weird readings going on.
The scale and HFC trim pot doesn't seem to do much, and the Saw trim R5 seems to only adjust amplitude, rather than the shape.
The ramp has a curved leading edge, sort of a rounded hill. The sine is very pointed, and the square and triangle are also kind of odd looking, and none of the trims change the shapes, just the output volumes.
Thanks |
|
Back to top
|
|
|
Europanaut
Joined: Apr 04, 2014 Posts: 46 Location: Nanaimo, BC, Canada
|
Posted: Thu Sep 11, 2014 12:23 am Post subject:
|
|
|
So, against everyone's advice, I bought an oscilloscope. (Yeah, I know. You all were right, it's an amazing tool.)
Here are the readings from the four waveform outputs. Square and ramp look ok, but triangle and sine look pretty strange.
Any suggestions on where I should start my diagnosis?
Thanks. (Nice to be back in the workshop.)
Description: |
|
Filesize: |
3.13 MB |
Viewed: |
159 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Filesize: |
3.5 MB |
Viewed: |
146 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Filesize: |
3.37 MB |
Viewed: |
153 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Filesize: |
3.08 MB |
Viewed: |
148 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
|
|
Back to top
|
|
|
Europanaut
Joined: Apr 04, 2014 Posts: 46 Location: Nanaimo, BC, Canada
|
Posted: Sat Sep 13, 2014 10:50 am Post subject:
|
|
|
No more amazing ideas from the more experienced folks here? You guys sure helped me nail the problem last time!
Can anyone recommend a good source of info on troubleshooting circuits? I've found lots of good stuff about BUILDING, but no so much about the tactics and methodology for tracking down problems.
Thanks! |
|
Back to top
|
|
|
|