Help me troubleshoot Timo Rozendal Logic

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Kent, Joe., luketeaford, lisa

Post Reply
User avatar
wendallsan
Wiggling with Experience
Posts: 380
Joined: Thu Jul 04, 2019 1:06 pm
Location: Olympia, WA, USA

Help me troubleshoot Timo Rozendal Logic

Post by wendallsan » Wed Nov 27, 2019 12:48 pm

Hi all,

I have half of a Timo Rozendal Logic module that I need help troubleshooting.

The top half of the module is working great, but the lower half has both input LEDs lit at all times, and the OR out is lit at all times. Providing gate/triggers to these inputs doesn't change this behavior in any way.

I managed to solder a surface mount IC on backwards, and in removing it, one of the solder pads on the PCB came with it. :deadbanana: :deadbanana: :deadbanana:

The IC in question is a CD4081, and the pad in question is pad 10.

Documentation for the project is here:

https://docs.google.com/document/d/17Gw ... vfwHk/edit

But there is no schematic. :sadbanana:

The datasheet for this IC is here:

http://www.ti.com/lit/ds/symlink/cd4081b.pdf

I looked hard at the PCB and it looked like both pads 10 and 11 weren't connected to anything, so I soldered on a new IC hoping that it would work. In the datasheet, pin 10 seems to be one of the logic outputs, so maybe it DOES go somewhere and I just can't follow the board well enough to know. I'm hoping to find out if there is a connection there that needs to be jumpered to somewhere else on the board, or if I should look elsewhere on my board to troubleshoot this issue.

So far I've confirmed the values of each resistor on the boards, visually checked solder joints, and confirmed that IC's, polarized caps, and transistors are properly oriented. Any assistance with troubleshooting things from here would be greatly appreciated!
Currently building a whole mess of modules by Nonlinear Circuits. Troubleshooting a Nonlinear Circuits Sauce of Unce.

User avatar
SphericalSound
Wiggling with Experience
Posts: 366
Joined: Thu Feb 01, 2018 9:23 am
Location: Spain

Post by SphericalSound » Wed Nov 27, 2019 6:29 pm

You deffo need to connect that pin10 to wherever it belongs. I would ask directly Timo, he will tell you where

Synthiq
Veteran Wiggler
Posts: 602
Joined: Mon Feb 06, 2017 3:10 pm
Location: California

Post by Synthiq » Wed Nov 27, 2019 7:57 pm

The 4081 has 4 2-input AND gates but there are only 2 AND outputs from the module. So it's quite possible two of the gates are unused and the outputs on pin 10 and 11 are not connected to anything, especially if pin 3 and 4 are used.

[Edit] After looking closer at the pcb in the build document, it does look like pin 8 and 9 as well as 12 and 13 are all connected to Vss (pin 7) so pin 10 and 11 should not be used.

User avatar
joem
Ultra Wiggler
Posts: 963
Joined: Thu Mar 05, 2015 11:07 pm
Location: NYC

Post by joem » Wed Nov 27, 2019 9:26 pm

It’s basic advice but I’d try reflowing things, especially the IC’s. I had some issues that didn’t make sense logically (that I could tell) when I built one of these, and it turned out I just needed to reflow.

Orgia Mode
Common Wiggler
Posts: 105
Joined: Mon Nov 04, 2019 5:24 pm

Post by Orgia Mode » Wed Nov 27, 2019 9:38 pm

Can we get a few photos of your board?

EDIT: Oh my god the trace work on these boards is giving me a headache....

User avatar
wendallsan
Wiggling with Experience
Posts: 380
Joined: Thu Jul 04, 2019 1:06 pm
Location: Olympia, WA, USA

Post by wendallsan » Thu Nov 28, 2019 12:08 pm

Hi all,

Thanks very much for your advice and help.
joem wrote:It’s basic advice but I’d try reflowing things, especially the IC’s. I had some issues that didn’t make sense logically (that I could tell) when I built one of these, and it turned out I just needed to reflow.
I've reflowed everything on both boards that I can reach (there are 2 transistors that are not accessible due to the board connections being soldered into place, but every other component was reflowed) and the behavior remains. Hopefully this means that my SMD soldering technique is not to blame-- this is my 1st SMD board. :eek:

It sounds like there is consensus that pad 10 'PROBABLY' isn't used in the circuit, as its a quad OR gate and there are only 2 OR's used in the module. Great to hear that my 1st attempt at tracing a circuit on the PCB may have actually been correct-- it's pretty easy when you're trying to trace a pad that is not connected to anything at all, I suppose.
Orgia Mode wrote:Can we get a few photos of your board?


Absolutely!

Image

Image

Image

Any further assistance is greatly appreciated :tu:
Currently building a whole mess of modules by Nonlinear Circuits. Troubleshooting a Nonlinear Circuits Sauce of Unce.

User avatar
woodster
Super Deluxe Wiggler
Posts: 1351
Joined: Mon Mar 05, 2012 9:36 am
Location: UK

Post by woodster » Thu Nov 28, 2019 12:35 pm

R41 and R34 look well suspect to me, as do some other places.
I think you need a flux pen to match your solder, and a quick redoing of suspect solder joints.
Are any solder points touching the pin headers connecting the two pcb's?

User avatar
wendallsan
Wiggling with Experience
Posts: 380
Joined: Thu Jul 04, 2019 1:06 pm
Location: Olympia, WA, USA

Post by wendallsan » Thu Nov 28, 2019 1:58 pm

Alright, I've re-retouched any joints identified by woodster (R41 and R34), adding fresh solder so the joint will flow.

Things are still behaving as I originally described, the lower half of the module is still showing both inputs as active at all times and the OR output as active at all times, regardless of what signal is provided.

I have some voltage readings from the IC's, in case that is helpful in tracking down the problem:

4081
1 +0.67
2 +0.69
3 0
4 +11.83
5 +6.74
6 +8.18
7 0
8 0
9 0
10 0
11 0
12 0
13 0
14 +11.84

4071
1 +0.68
2 +0.69
3 0
4 +11.82
5 +6.74
6 +8.18
7 0
8 0
9 0
10 0
11 0
12 0
13 0
14 11.84

TL74
1 -10.40
2 +1.52
3 0
4 +11.84
5 +0.01
6 +1.53
7 -10.40
8 -10.35
9 -0.01
10 0
11 -11.68
12 0
13 +1.52
14 +11.22

4070
1 +0.68
2 +0.69
3 0
4 +0.01
5 +6.74
6 +8.15
7 0
8 0
9 0
10 0
11 +0.01
12 +11.83
13 +11.82
14 +11.84
Currently building a whole mess of modules by Nonlinear Circuits. Troubleshooting a Nonlinear Circuits Sauce of Unce.

Synthiq
Veteran Wiggler
Posts: 602
Joined: Mon Feb 06, 2017 3:10 pm
Location: California

Post by Synthiq » Thu Nov 28, 2019 4:12 pm

It looks like the general structure of the circuit is the same as the CGS39, see below. The four opamp in TL074 are used as comparators for the four input signals and the outputs are then level shifted from +/-12V swing to a +12/0V swing with a 2:1 voltage divider driving the logic chips.

The four inverting inputs should all see a 1.5V voltage, but your measurement says -0.01V on pin 9 so please check this again as it is not the correct voltage. The output from pin 14 should be negative just as the other 3 outputs so check the polarity again.

Pin 5 and 6 on the 3 logic chips all see 6.74V and 8.18V (so it's not a surprise the LEDs for the inputs are on) but should be close to 0V just as pin 1 and 2. It is possible the resistors in the dividers are incorrect and either the pullup resistors to +V are too low or the resistors from the opamps are too high so check them if they can be identified on the pcb.

Image

Orgia Mode
Common Wiggler
Posts: 105
Joined: Mon Nov 04, 2019 5:24 pm

Post by Orgia Mode » Thu Nov 28, 2019 8:17 pm

I agree that a flux pen will do wonders for you. YOu can clean up those resistor dividers in case they are pulling int inputs high and I see a few places on the ICs that are suspicious:

Picture file

Orgia Mode
Common Wiggler
Posts: 105
Joined: Mon Nov 04, 2019 5:24 pm

Post by Orgia Mode » Thu Nov 28, 2019 8:25 pm

great pics btw

User avatar
wendallsan
Wiggling with Experience
Posts: 380
Joined: Thu Jul 04, 2019 1:06 pm
Location: Olympia, WA, USA

Post by wendallsan » Tue Dec 03, 2019 10:32 pm

Thanks very much for the response and my apologies in delay in following up on my end. I'm glad the provided photos were helpful.
Synthiq wrote: The four inverting inputs should all see a 1.5V voltage, but your measurement says -0.01V on pin 9 so please check this again as it is not the correct voltage. The output from pin 14 should be negative just as the other 3 outputs so check the polarity again.
I've double-checked the voltages you questioned and found that the TL074's pin 9 is indeed +1.51V, so my original reading was wrong and it is actually at the voltage that you expected to see. Pin 14 is definitely reading +11.21V, though, so something is probably not right there.
Pin 5 and 6 on the 3 logic chips all see 6.74V and 8.18V (so it's not a surprise the LEDs for the inputs are on) but should be close to 0V just as pin 1 and 2. It is possible the resistors in the dividers are incorrect and either the pullup resistors to +V are too low or the resistors from the opamps are too high so check them if they can be identified on the pcb.
I've tried to make sense of these instructions, but I'm not sure what resistors I'll need to check to troubleshoot this further. I've confirmed that pins 5 and 6 of the 4070 are indeed at +6.75V and +8.12V respectively, and I've tried tracing the circuit from these pins, but didn't arrive at any resistors-- probably because I'm tracing it wrong. Here's an image of the tracing to the best of my understanding of the board, with the traces in question in orange and yellow.

https://dansteeby.com/wp-content/uploads/logic-trace.jpg

Can anyone point out what I'm missing in my trace, or otherwise identify what resistors I should be checking to make sure they are working as they should. I've already visually confirmed all resistor values on the boards and retouched all solder points a couple of times, so I'm fairly confident about those two points.
Currently building a whole mess of modules by Nonlinear Circuits. Troubleshooting a Nonlinear Circuits Sauce of Unce.

Synthiq
Veteran Wiggler
Posts: 602
Joined: Mon Feb 06, 2017 3:10 pm
Location: California

Post by Synthiq » Wed Dec 04, 2019 1:48 pm

Below is my own attempt to trace the signals of interest.

Input in2A is yellow and goes to resistor R36 (10kohm?) and with R38 (100kohm?) to ground and arrives at pin 10 on the TL074.
The output from this amplifier (blue) drives a voltage divider consisting of R44 and R45 (both 100kohm?) with R45 connected to positive supply. The midpoint of the voltage divider drives pin 5 of the logic gates as well as the driver for a LED.

Input in2B is green and goes to resistor R37 and with R39 to ground and arrives at pin 12 on the TL074. The output from this amplifier (red) drives a voltage divider consisting of R47 and R46 with R46 connected to positive supply. The midpoint of the voltage divider drives pin 6 of the logic gates as well as the driver for a LED.

Note that each color contains more than a single net but relates to a single signal.

Image

User avatar
wendallsan
Wiggling with Experience
Posts: 380
Joined: Thu Jul 04, 2019 1:06 pm
Location: Olympia, WA, USA

Post by wendallsan » Fri Dec 06, 2019 3:53 pm

Synthiq wrote:Below is my own attempt to trace the signals of interest
Thanks very much for taking time to trace this! I've taken voltage measurements along each trace that you colored that I can reach, and I'm hoping that these numbers will help to identify the root of the problem.

Image

I've confirmed that R36 is 10K and R38, R44 and R45 are all 100K. Please let me know if any of these voltage values look fishy, or if I can provide any other details to move troubleshooting this forward from here.
Currently building a whole mess of modules by Nonlinear Circuits. Troubleshooting a Nonlinear Circuits Sauce of Unce.

User avatar
EATyourGUITAR
has no life
Posts: 5798
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Post by EATyourGUITAR » Fri Dec 06, 2019 5:43 pm

I am no help. I just wanted to say that PCB design looks like a Ciat Lonbarde art project. he turned the grid off and slammed through it one click at a time. I don't think it is a real problem. but damn. that is some slapped together shit.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
wendallsan
Wiggling with Experience
Posts: 380
Joined: Thu Jul 04, 2019 1:06 pm
Location: Olympia, WA, USA

Post by wendallsan » Fri Dec 06, 2019 7:41 pm

EATyourGUITAR wrote:I am no help. I just wanted to say that PCB design looks like a Ciat Lonbarde art project. he turned the grid off and slammed through it one click at a time. I don't think it is a real problem. but damn. that is some slapped together shit.


Haha leave it me to pick a strangly designed board to build and troubleshoot! I wouldn't know a well designed board from a bad one at this stage. :roll:
Currently building a whole mess of modules by Nonlinear Circuits. Troubleshooting a Nonlinear Circuits Sauce of Unce.

Synthiq
Veteran Wiggler
Posts: 602
Joined: Mon Feb 06, 2017 3:10 pm
Location: California

Post by Synthiq » Fri Dec 06, 2019 7:53 pm

The input networks (yellow and green) looks fine except the -11.70 on pin 12 of the TL074, but you might have measured the negative supply on pin 11 instead as your earlier measurement correctly saw -11.68 on pin 11 and 0 on pin 12. So you can forget about these two nets for now.

The output on TL074 pin 8 (blue) is correctly -10.37, but once you go through a via to the other side the voltage suddenly changes to 6.17V/6.75V. I suspect a bad via so the input to R44 is floating and not connected to the output of the opamp. The 6.75V you see can be explained if that node is pulled up to the positive supply by the 100kohm R45 but also pulled down by the base resistor R3 for the LED driver if this resistor is about 88kohm. Try measuring the resistance between pin 8 of the TL074 and R44 to see if you have 0 ohm or not. If not, solder a piece of wire between them to fix this problem.

The +11.23V on pin 14 (red) of the TL074 still doesn't make sense if the positive input is 0V and the negative input +1.5V so it might be a problem with that amplifier internally. But given this voltage, the rest make sense. If resistor R47 is driven by +11.23V and R46 driven by the positive supply voltage and the mid point loaded by another 88kohm base resistor (R3) you should get a voltage close to the 8.19V you actually see. This voltage is only 69% of the supply voltage so this is strictly outside spec for a high voltage which normally is 70% but will normally work anyway. But I would have been happier if the 100kohm resistors would be reduced to maybe 50kohm or less.

I also noted I made mistakes when tracing the signals to the CD4070 and forgot the mirroring of the top and bottom layers so the red trace should should go to pin 6 instead of 1 and the blue trace to pin 5 and not pin 2. I apologize for that.

User avatar
wendallsan
Wiggling with Experience
Posts: 380
Joined: Thu Jul 04, 2019 1:06 pm
Location: Olympia, WA, USA

Post by wendallsan » Mon Dec 09, 2019 3:47 pm

Many thanks again Synthiq!

Your input helped me to identify leads on the TL074 that were STILL cold joints, even after at least 2 other rounds of touching these up. More solder was piled on until the connection worked, and I'll know to be more liberal with my solder on these SMD IC's in the future.

8-) 8-) 8-)

After this repair the 2nd Logic circuits worked as expected EXCEPT that the AND out and LED were still not going high. A long and painful process allowed me to trace this out back to the proper AND out on the CD8081 chip, and after confirming that it was going high and low at the chip as it should have been, I was able to find the signal going dead at a resistor along the way to the output from there. Touching this up allowed the AND output to finally go high.

First circuit tracing achieved! Many help for the hand-holding! I feel like I have a slightly better understanding on how to do this again next time. :nana:
Currently building a whole mess of modules by Nonlinear Circuits. Troubleshooting a Nonlinear Circuits Sauce of Unce.

Synthiq
Veteran Wiggler
Posts: 602
Joined: Mon Feb 06, 2017 3:10 pm
Location: California

Post by Synthiq » Mon Dec 09, 2019 7:54 pm

I'm glad to hear you got it to work! From your pictures it doesn't looks like you have too little solder so you may not apply the soldering iron long enough for the heat to spread to both the component and the pcb. Most electronic components are sensitive to too much heat but I don't think I have actually destroyed a component by applying too much heat while soldering, only a pcb trace or two lifting from the board from too much heat a long time ago. Just add a small amount of lead to the tip to make a good thermal contact with the component and the pcb and then apply solder and continue heating until the solder flows onto both the component leg and the pcb. A magnifying lens makes it easier to see what's happening, at least for me.

User avatar
EATyourGUITAR
has no life
Posts: 5798
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Post by EATyourGUITAR » Mon Dec 09, 2019 9:30 pm

I think if you are having problems reheating a cold solder joint then you may not have the right soldering iron or the right temperature setting on the soldering iron. You may have corrosion on your iron tip. You may have oxidation on the solder joint. You can swap in a new tip or use tip refresh chemicals to fix the tip. You can use flux and or a tiny bit of additional solder to get the joint to reflow. However, I have a special 0.032in kester RA type solder for this. In your situation, since you already had too much solder the first time, I would just hit it with the solder braid to suck up %90 of the solder first. Then I would wait a second to cool. Add a tiny bit of flux. Then solder the joint like normal with more heat, better contact, more time. However, if you have a low wattage iron, by the time you get the trace, the lead, and the solder all up to the same temperature, you may be getting close to the trace lifting up. The pads lift when there is too much dwell time and the adhesive beneath the copper fails. Because there is a lot of variation in the quality of adhesives used in PCB manufacturing it is impossible to know how resistant a PCB is unless you have some knowledge of the specific factory that was used. I have cooked components but they are components with huge mass, very small mass, components with short leads, or vintage germanium semiconductors. It could happen on a silicon transistor with the legs super short and a cheap iron. More likely if you solder all 3 legs without letting them cool one at a time.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

Post Reply

Return to “Music Tech DIY”