I'm rebuilding / modifying an Eico 2036 integrated amplifier. By the time I'll be finished, it will just be a power amp.
Eico (and other manufacturers) sometimes used PECs -- encapsulated circuits with multiple components. They have a reputation for drifting in value over time; it's usually considered best practice to replace them. So I did. I also wanted to increase the size of the coupling caps and match the grid resistors. I've attached pics of an original Sprague 29754 PEC and my replacement.
I used this stuff called Plasti-Dip to encapsulate the PCB. You can remove it if you have to, whereas epoxy is pretty permanent and unforgiving.
I measured the resistances between various points on a original PEC and the replacement. I've included the test results for just one PEC. Both OEM PECs were close in value to one another, as were the replacements, so I just included one data set:
OLD (Original) 29754 PEC
Test Points: Nominal Actual Variance
2-3 82K 83.1K 1.3%
1-4 330K 351.4K 6.5%
4-6 340K 357.6K 5.2%
4-5 82K 84.1K 2.6%
1-6 670K 709.1K 5.8%
NEW REPLACEMENT PEC
Test Points: Nominal Actual Variance
2-3 82K 81.5K 0.6%
1-4 330K 333.9K 1.2%
4-6 340K 343.6K 1.1%
4-5 82K 81.7K 0.4%
1-6 670K 677.0K 1.0%
Conclusions: The OEM PEC fared pretty well for being a 50 year old part -- better than my expectations. Note that the maximum deviation from the published (nominal) value is only 6.5%. Perhaps the critical measurement is between 1 & 6. It is the sum of two 330K grid resistors plus an additional 10K. Still par for the course.
I substituted 0.039 for 0.027 µF coupling caps. My ears are no longer good enough to hear if there is any discernible difference. But I know the specs will be tighter.
Eico (and other manufacturers) sometimes used PECs -- encapsulated circuits with multiple components. They have a reputation for drifting in value over time; it's usually considered best practice to replace them. So I did. I also wanted to increase the size of the coupling caps and match the grid resistors. I've attached pics of an original Sprague 29754 PEC and my replacement.
I used this stuff called Plasti-Dip to encapsulate the PCB. You can remove it if you have to, whereas epoxy is pretty permanent and unforgiving.
I measured the resistances between various points on a original PEC and the replacement. I've included the test results for just one PEC. Both OEM PECs were close in value to one another, as were the replacements, so I just included one data set:
OLD (Original) 29754 PEC
Test Points: Nominal Actual Variance
2-3 82K 83.1K 1.3%
1-4 330K 351.4K 6.5%
4-6 340K 357.6K 5.2%
4-5 82K 84.1K 2.6%
1-6 670K 709.1K 5.8%
NEW REPLACEMENT PEC
Test Points: Nominal Actual Variance
2-3 82K 81.5K 0.6%
1-4 330K 333.9K 1.2%
4-6 340K 343.6K 1.1%
4-5 82K 81.7K 0.4%
1-6 670K 677.0K 1.0%
Conclusions: The OEM PEC fared pretty well for being a 50 year old part -- better than my expectations. Note that the maximum deviation from the published (nominal) value is only 6.5%. Perhaps the critical measurement is between 1 & 6. It is the sum of two 330K grid resistors plus an additional 10K. Still par for the course.
I substituted 0.039 for 0.027 µF coupling caps. My ears are no longer good enough to hear if there is any discernible difference. But I know the specs will be tighter.
