Temperature Change on IC Performace

Discussion in 'General Electronics Chat' started by johngalex, Jan 16, 2019.

  1. johngalex

    Thread Starter New Member

    Sep 7, 2013
    9
    0
    Hi All,

    I am refurbishing a musical synthesizer manufactured in 1979. I had tuning drift issues when the synthesizer was running for a couple of hours and got really warm. I believe I fixed the issue by replacing four old / original TL084 ICs in the S&H tuning circuit. Tuning is now much more stable when running warm.

    My questions:

    - Does temperate change an ICs performance?
    - Are general use ICs manufactured in 1979 more susceptible to temperature variances vs more modern version of the same IC?
    - Do ICs get more susceptible to temperature changes based upon their age / usage?

    I am trying to understand if these ICs were bad or if this was the expected behavior.

    John
     
  2. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    9,179
    2,164
    Yes. The datasheet for the IC should tell you which parameters are affected and by how much.
     
  3. MrChips

    Moderator

    Oct 2, 2009
    17,736
    5,540
    All electronic components are sensitive to temperature changes.

    In your case, if the synthesizer frequency is affected by temperature, then there is a tone oscillator circuit that is the culprit. The frequency of the oscillator is often dependent on two components, a resistor and a capacitor. It is most likely that the capacitor needs to be changed to one that is less dependent on temperature.

    If you can find the schematics for your synthesizer we will be able to advise which capacitor needs changing.
     
  4. Reloadron

    Distinguished Member

    Jan 15, 2015
    4,040
    1,832
    This is the data sheet for the TL084 JFET-Input Operational Amplifiers. The data sheet reflects several charts on the effects of ambient air temperature on various characteristics of the chip. Based on the data sheet I believe we can say yes temperature can very much the performance of this IC even when the IC is used within its working temperature range. Newer ICs in general are designed to be less susceptible to temperature but that really depends on the IC and the specifications for its intended application. Some ICs are more likely to be affected than others but always refer to the data sheet.Do ICs degrade over time? Again, it depends on the IC as some likely more than others. Overall the data sheet is your best friend.

    Ron
     
  5. crutschow

    Expert

    Mar 14, 2008
    20,813
    5,917
    I would say that, in general, they are likely to only change significantly if they are operated for long periods at near their maximum operating temperature.
     
  6. danadak

    Well-Known Member

    Mar 10, 2018
    2,222
    473
    Yes, specs in tables and graphs in datasheet disclose this.

    There are changes done to legacy parts, usually to adapt to a new process. But
    in general, unless datasheet discusses "this is the new best version of that raggedy
    part we used to make", you can assume its still the same old part. Only evidence
    would be if datasheet parameters are significantly different, or a process change,
    like from bipolar to CMOS, or smaller geometry process update.

    There is some complex long term drift associated with ICs in general, but the effects
    on non high precision analog designs minimal.


    Regards, Dana.
     
  7. johngalex

    Thread Starter New Member

    Sep 7, 2013
    9
    0
    Wow - thanks for all of the feedback. This is a great site to get help.

    I looked at a TL084 data sheet but the version I had did not have the graphs - these made it a lot clearer.

    Below is the schematic. The circuit is in the upper right hand corner. This is a S&H circuit to hold the fine tune offset voltages (FT1a, FT1b, etc.). that are derived from the microprocessor controlled (Z80) Autotune process. These voltages are then applied to the audio oscillators.

    I had a tuning issue on FT8a and FT8b when the unit got very warm. I swapped TL084 IC A71 with A70 and the problem moved so I knew i was a problem with A71 and decided to replace all of the TL084 ICs.

    My naive question is that I thought when an IC went bad, it just stopped functioning. This one seemed to work fine until it go to warm, but still in operating spec.


    [​IMG]
     
  8. johngalex

    Thread Starter New Member

    Sep 7, 2013
    9
    0
    Here is an easier to read version:


    OBX Schematic.jpg
     
  9. MrChips

    Moderator

    Oct 2, 2009
    17,736
    5,540
    Nothing in those drawings would cause a temperature dependent frequency shift.
    Can you post a pdf document of the rest of the circuitry?
     
  10. dl324

    AAC Fanatic!

    Mar 30, 2015
    7,914
    1,892
    Yes, but it's unlikely any consumer product would be affected because parts are generally rated for 0-80C. Industrial and military have wider ranges, but they're only used when necessary because they cost more.
     
  11. johngalex

    Thread Starter New Member

    Sep 7, 2013
    9
    0
  12. MrChips

    Moderator

    Oct 2, 2009
    17,736
    5,540
    Thanks for sharing the pdf.

    If you are experiencing tonal shifts, I would not be replacing chips aimlessly.
    The first step would be to observe and make notes of the symptoms. Listen to the sound carefully.
    Is the frequency shift occurring across all voices or is it only on one or specific voices?
    Play with the controls and see what effects these have.
    There are many other things that can give this temperature dependency.

    I would analyze the circuit and make a list of all the things that can cause this. We can help you with this.

    1) Power supplies
    2) Oscillators
    3) Benders
    4) Control pots

    One technique that can be helpful in isolating temperature dependencies is to use cold spray such as "Freeze Mist".
    You apply cold spray to one area at a time and observe the effects. When you localize the area you are often able to pin it down to one component.

    I will look over the schematics and see what I can find out for you.
     
  13. johngalex

    Thread Starter New Member

    Sep 7, 2013
    9
    0
    Thanks, that is how I am trouble shooting. The 4 TL084 S&H circuit controls the Fine Tune offsets. Voice 8 (FT8a & b controlled by A71) was slightly out of tune. I swapped A71 with A70 and then Voice 8 was in tune and Voice 6 (FT6a & b) was out of tune. Replacing A70 with a new TL084 fixed the issue.

    The power supply was rebuilt and I fully calibrated the power supply, benders, keyboard, etc. All voltages are within spec.

    The OBX has 8 voice cards with two oscillators, a VCF and a VCA on each. I have replaced all 11 trim pots on each board with Bourns muti-turns and replaced the electrolytic capacitors on each. The tuning is much more stable now.

    Great idea - thanks.

    The OBX is very stable in its tuning now. I just recently fixed the Autotune circuit, which makes it stay in tune even better as the unit moves from cold, to warm, to hot.

    I guess my real question is how did swapping the TL084 fix the issue. I thought an IC either worked or didn't work. This one seemed to go out of spec when hot while the other TL084s worked fine at the same temperature.
     
  14. dl324

    AAC Fanatic!

    Mar 30, 2015
    7,914
    1,892
    Your assumption that IC's either work or don't is invalid. Parts wear out and they don't have to fail completely. Some parts wear out before others, so expecting all opamps to either work or not work isn't realistic.

    Do the tires on your car either work or not work? Do you expect them to all fail catastrophically at the same time? That expectation just doesn't make any sense and neither does your expectation of semiconductors. All will degrade over time and they won't do it at the same rate.
     
  15. MrChips

    Moderator

    Oct 2, 2009
    17,736
    5,540
    Depends on which opamp you replaced.

    If the opamp is in a low/band/high-pass filter or envelope circuit it would not matter so much. If the opamp outputs a voltage that controls a VCO (voltage controlled oscillator) that would have an immediate impact. Let us know which opamp you are referring to.

    As the previous member said, expecting a go/no-go failure is far from the truth.

    Looks like you know what you are doing. Keep on doing what you are doing.
     
  16. johngalex

    Thread Starter New Member

    Sep 7, 2013
    9
    0
    This one directly controlled a VCO - so it had a huge impact.

    Thanks - this is what I thought and observed - but wanted someone with more experience to verify.

    Thanks again - I've only been doing this for about 6 months. After purchasing my OBX, I focused on learning on how to keep it functioning for another 35 years. I researched on different circuits as I read the OBX schematics. I purchased a Hakko soldering station and de-soldering gun. Researched on correct soldering techniques, soldering temperatures, the correct solder to use, flux, board cleaning,etc. I ripped apart some old electronics and practiced until I had it right before ever touching my OBX.

    I actually used my new found knowledge to fix an old 1983 Korg Poly 61 synth that had in my closet. It suffered from an alkaline battery leak which corroded many parts. I cleaned the main board, replaced a few chips, corroded capacitors, ribbon connectors, and even a board trace. It now works great!

    It's forums like this that helped me along the way...
     
  17. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    9,179
    2,164
    That's an impressive learning curve!
     
  18. johngalex

    Thread Starter New Member

    Sep 7, 2013
    9
    0
    Thanks - I am actually quite impressed with how far that I have progressed. I came in with some soldering skills and a basic understanding of electronics - resistors, capacitors, transistors, etc.

    When I was first reading the Oberheim OBX schematics - I was wondering what a mux / dmux and sample and hold circuit was. After researching, it is much clearer to me.

    From the input from this post - understanding that older ICs become "tired" - I have just troubleshot the next circuit in the tuning path. This consists of a pair of TL082's and CA3080's to control the voltage when a key is pressed. I methodically worked through each voice with an issue and replaced the pair and was able to measure the improvement. I ended up replacing all of the 16 ICs and now the OBX just sings...
     
Loading...