MacOSMonkey
Well-known member
I can check C13 for you.
-
Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.
Supermac Spectrum/24 Series III and Series IV Issues
- Thread starter jmacz
- Start date
jmacz
Well-known member
Oh cool, thanks!
MacOSMonkey
Well-known member
To demonstrate properties of capacitors: Below are some example measurements on "new" (old) caps of different types showing their capacitance and resistance. If you have a capacitor and it is not showing very high (mOhm) or infinite resistance when measuring with an ohmmeter, then it is probably no good and has had internal breakdown/shorting (over-voltage damage, etc.). And, of course, the cap should have a readable value within its stated tolerance. I hope this info is helpful.
Mfg: Nichicon*
Type: Radial Electrolytic (Polarized)
Value: 330uF 35V 20%
Measures: 308uF (-6.67%)
Resistance: infinite/open
Mfg: Panasonic*
Type: SMT Tantalum (Polarized) - D Size
Value: 220uF 6.3V 20%
Measures: 243uF (+10.45%)
Resistance: infinite/open
Mfg: Kemet*
Type: SMT Tantalum (Polarized) - D Size
Value: 47uF 10V 20%
Measures: 47.6uF (+1.3%)
Resistance: infinite/open
Mfg: Kemet
Type: Radial Tantalum (Polarized)
Value: 10uF 35V 20%
Measures: 10.77uF (+7.7%)
Resistance: infinite/open
Mfg: Probably Kemet, but unknown
Type: SMT Ceramic (Non-Polarized) - 0805
Value: .1uf/100nF - probably 25V 5% / C0G
Measures: 101.7nF (+1.7%)
Resistance: infinite/open
Mfg: Kemet
Type: SMT Ceramic (Non-Polarized) - 0402
Value: 100pF 25V 5%
Measures: 100.1pF (...but there is stray capacitance and the meter has a variable 0 adjustment for values <200pF. It's probably within tolerance, but would need a higher resolution meter, known calibration source...or could check more parts, assuming they were all good)
Resistance: infinite/open
*Note that the higher value caps above may show transient mOhm resistance while charging. But, after that, they should read as open until discharged.
n.b. As above, all caps must be removed from a board/circuit in order to accurately measure them.
Mfg: Nichicon*
Type: Radial Electrolytic (Polarized)
Value: 330uF 35V 20%
Measures: 308uF (-6.67%)
Resistance: infinite/open
Mfg: Panasonic*
Type: SMT Tantalum (Polarized) - D Size
Value: 220uF 6.3V 20%
Measures: 243uF (+10.45%)
Resistance: infinite/open
Mfg: Kemet*
Type: SMT Tantalum (Polarized) - D Size
Value: 47uF 10V 20%
Measures: 47.6uF (+1.3%)
Resistance: infinite/open
Mfg: Kemet
Type: Radial Tantalum (Polarized)
Value: 10uF 35V 20%
Measures: 10.77uF (+7.7%)
Resistance: infinite/open
Mfg: Probably Kemet, but unknown
Type: SMT Ceramic (Non-Polarized) - 0805
Value: .1uf/100nF - probably 25V 5% / C0G
Measures: 101.7nF (+1.7%)
Resistance: infinite/open
Mfg: Kemet
Type: SMT Ceramic (Non-Polarized) - 0402
Value: 100pF 25V 5%
Measures: 100.1pF (...but there is stray capacitance and the meter has a variable 0 adjustment for values <200pF. It's probably within tolerance, but would need a higher resolution meter, known calibration source...or could check more parts, assuming they were all good)
Resistance: infinite/open
*Note that the higher value caps above may show transient mOhm resistance while charging. But, after that, they should read as open until discharged.
n.b. As above, all caps must be removed from a board/circuit in order to accurately measure them.
jmacz
Well-known member
Finally got the 24/IV working!!!
What an adventure... final set of fixes:
Card is working great in both my IIci and Quadra 700.
Thanks to @Phipli who hooked me up with a seller selling the RAMDAC, and @MacOSMonkey for all the help throughout.
What an adventure... final set of fixes:
- Replaced faulty RAMDAC (ADV473kp110es) with replacement (BT473kpj110 -- BT473kpj110ES2 did not work).
- Replaced blown F1 fuse with proper 2A rated fuse.
- Replaced faulty resistors R44, R52, and R19.
- Replaced faulty C13 capacitor -- couldn't get a reading so took a guess based on surrounding capacitors with the same ceramic exterior color (100nF).
- Replaced semi-misbehaving 47nF capacitors C4 and C7.
- Replaced semi-misbehaving 100nF capacitor C120.
- Replaced 47nF capacitor (although it was probably fine) C10.
- Replaced 1uF capacitor (although it was probably fine) C5.
- Replaced 100nF capacitor (although it was probably fine) C15.
- Replaced 10nF capacitor (although it was probably fine) C124.
Card is working great in both my IIci and Quadra 700.
Thanks to @Phipli who hooked me up with a seller selling the RAMDAC, and @MacOSMonkey for all the help throughout.
Phipli
Well-known member
Congratulations!Finally got the 24/IV working!!!
What an adventure... final set of fixes:
Clearly whoever had the card before did something bad -- whatever it was blew the F1 fuse, took out a few resistors, may have done something to 3 capacitors, and messed up the RAMDAC. And then they replaced the fuse with the one of the wrong spec, and eventually the card made it into the hands of a seller selling vintage hardware, who then sold it to me. The fuse replacement was preventative but didn't resolve my issue. The resistors/capacitors seem to help stabilize the image. But it was primarily the replacement of the RAMDAC that resolved the lack of 8bit video and blue tinted 24bit video (no red signal and weak green signal).
- Replaced faulty RAMDAC (ADV473kp110es) with replacement (BT473kpj110 -- BT473kpj110ES2 did not work).
- Replaced blown F1 fuse with proper 2A rated fuse.
- Replaced faulty resistors R44, R52, and R19.
- Replaced faulty C13 capacitor -- couldn't get a reading so took a guess based on surrounding capacitors with the same ceramic exterior color (100nF).
- Replaced semi-misbehaving 47nF capacitors C4 and C7.
- Replaced semi-misbehaving 100nF capacitor C120.
- Replaced 47nF capacitor (although it was probably fine) C10.
- Replaced 1uF capacitor (although it was probably fine) C5.
- Replaced 100nF capacitor (although it was probably fine) C15.
- Replaced 10nF capacitor (although it was probably fine) C124.
Card is working great in both my IIci and Quadra 700.
Thanks to @Phipli who hooked me up with a seller selling the RAMDAC, and @MacOSMonkey for all the help throughout.
Make me feel I need to try repairing mine. My PDQ is quite intimidating. Thats a lot of chips and not much clearance.
cheesestraws
Well-known member
Hooray! Well done.
MacOSMonkey
Well-known member
Great job! 
Really happy for you! Nice to have a fully working board, and it was worth saving!
Yeah, the -ES2 DAC substitution was probably a stretch, but glad you were able to find a comparable Bt473 part. Also, it is helpful for others that you confirmed that the AD/Bt parts are a direct swap (as expected). SuperMac used them interchangeably for multiple reasons -- availability, quality issues...and to angle for the best vendor pricing (or other vendor management reasons), of course.
I think your assumption about C13 is reasonable - .1uF is good enough, but I will still measure it for you for the sake of detail. Given that the board works fine, it's not going to matter and it's probably just a bypass cap.
re: Analog Devices vs. Brooktree
As I may have mentioned in another post, in the late '80s, Brooktree had an issue that caused shipping/RMA headaches for SuperMac. I think the problems related to the Bt478 and sudden palette-shifting on the Spectrum/8 Series III (exacerbated by temperature) that brought about the 2nd-sourcing to Analog Devices parts and use of the AD478. It was the only way for SuperMac to keep shipping at a critical moment in its history. So, if you happen to have a Spectrum/8 Series III board with a Bt478 on it and a v1.0 ROM and you warm the DAC (heat gun, but don't overdo it) while changing bit depths in/out of 8-bit and switching color/grayscale (maybe best tested in 19" 75Hz mode), if you notice that the colors suddenly go wonky (palette shown in the Monitors cdev), then you know why. If the board has a later ROM and a Bt part (post-resolution), it's probably OK. There was a failing run of parts, not all boards/parts showed the problem and not all early boards were returned. Bt eventually corrected the problem, but there was a stressful gap in manufacturing and shipping. If your Spectrum/8 Series III has an AD478 on it, then you know that it probably came from the period after initial production and detection of this issue. That being said, there were many Spectrum/8 Series III boards with Bt478s on them that were fine...and also from later in production after Bt addressed the issue.
Really happy for you! Nice to have a fully working board, and it was worth saving!
Yeah, the -ES2 DAC substitution was probably a stretch, but glad you were able to find a comparable Bt473 part. Also, it is helpful for others that you confirmed that the AD/Bt parts are a direct swap (as expected). SuperMac used them interchangeably for multiple reasons -- availability, quality issues...and to angle for the best vendor pricing (or other vendor management reasons), of course.
I think your assumption about C13 is reasonable - .1uF is good enough, but I will still measure it for you for the sake of detail. Given that the board works fine, it's not going to matter and it's probably just a bypass cap.
re: Analog Devices vs. Brooktree
As I may have mentioned in another post, in the late '80s, Brooktree had an issue that caused shipping/RMA headaches for SuperMac. I think the problems related to the Bt478 and sudden palette-shifting on the Spectrum/8 Series III (exacerbated by temperature) that brought about the 2nd-sourcing to Analog Devices parts and use of the AD478. It was the only way for SuperMac to keep shipping at a critical moment in its history. So, if you happen to have a Spectrum/8 Series III board with a Bt478 on it and a v1.0 ROM and you warm the DAC (heat gun, but don't overdo it) while changing bit depths in/out of 8-bit and switching color/grayscale (maybe best tested in 19" 75Hz mode), if you notice that the colors suddenly go wonky (palette shown in the Monitors cdev), then you know why. If the board has a later ROM and a Bt part (post-resolution), it's probably OK. There was a failing run of parts, not all boards/parts showed the problem and not all early boards were returned. Bt eventually corrected the problem, but there was a stressful gap in manufacturing and shipping. If your Spectrum/8 Series III has an AD478 on it, then you know that it probably came from the period after initial production and detection of this issue. That being said, there were many Spectrum/8 Series III boards with Bt478s on them that were fine...and also from later in production after Bt addressed the issue.
