I'm working on a 3020 that passed DBT and sounded good during a 15 minute listening test. The owner reported that F3 and F4 were blown when he acquired it. When I took a look, I found that one of the fuse holder tabs for F3 was nearly broken off, and F3 and F4 were 5A 250V fast blow instead of the 4A 250V slow blow specified.
I replaced the fuse holder and put in the correct fuses before proceeding to power up with the dim bulb tester (40W bulb). As I said, the bulb dimmed quickly. I took it off the DBT and checked DC offset. The right channel could be adjusted to 0V, but the left channel would only come down to ~60mV. I then moved to the brief listening session with my sacrificial speakers. No problem there.
Still curious about the cause of the blown fuses, I tried checking to see if any of the output transistors were shorted to the heat sink. Using a bare spot of metal on the heat sink, I tested for continuity with the transistor cases. All four transistors show as having the cases shorted to the heat sink. So I pull them all, replace the insulators, apply new thermal grease, and check the nylon bushings for the screws. Well, I'll skip to the ending: Even without the output transistors installed, I get continuity between the collector solder pads where the one mounting screw passes through and connects to the case of each transistor. In addition, I see that the heat sink is at chassis ground, due to the screw that holds it to the support bar running down the middle of the amp. So all four collector pads are shorted to the chassis.
The only other problem I've uncovered to this point is that one leg of R656, a 10 ohm 1W, was broken. I'm going to replace that tomorrow.
Any thoughts on what I'm seeing would be appreciated.
Thanks.
I replaced the fuse holder and put in the correct fuses before proceeding to power up with the dim bulb tester (40W bulb). As I said, the bulb dimmed quickly. I took it off the DBT and checked DC offset. The right channel could be adjusted to 0V, but the left channel would only come down to ~60mV. I then moved to the brief listening session with my sacrificial speakers. No problem there.
Still curious about the cause of the blown fuses, I tried checking to see if any of the output transistors were shorted to the heat sink. Using a bare spot of metal on the heat sink, I tested for continuity with the transistor cases. All four transistors show as having the cases shorted to the heat sink. So I pull them all, replace the insulators, apply new thermal grease, and check the nylon bushings for the screws. Well, I'll skip to the ending: Even without the output transistors installed, I get continuity between the collector solder pads where the one mounting screw passes through and connects to the case of each transistor. In addition, I see that the heat sink is at chassis ground, due to the screw that holds it to the support bar running down the middle of the amp. So all four collector pads are shorted to the chassis.
The only other problem I've uncovered to this point is that one leg of R656, a 10 ohm 1W, was broken. I'm going to replace that tomorrow.
Any thoughts on what I'm seeing would be appreciated.
Thanks.
Suggest you install one output transistor at a time, to ascertain where the transistor body is touching. For instance just one transistor might be doing this, rather than all of them... but finding that one will take a bit of time.
YouTube
Output collectors connect to the +/-30.5 V power rails. A "short" with outputs
removed suggests a failed component elsewhere that also connects to the +/-30.5V rails.
This supply is also used for the phono, and preamp, might also be a fault in the
power supply.
Is the measured resistance a short, ie, less than 1 ohm (MM probe leed resistance)
or unexpectedly low (a few ohms), this may point to the failed component.
Candidates (all very long shots) would be,
Phono C423
Power amp C639
psu caps C913
With a true short, expect the fuses to pop everytime. DBT does not stay bright.
Probably have a psu failure eg, Q901, have you verified psu voltages, eg at output
collectors, even on DBT expect xxVolts
removed suggests a failed component elsewhere that also connects to the +/-30.5V rails.
This supply is also used for the phono, and preamp, might also be a fault in the
power supply.
Is the measured resistance a short, ie, less than 1 ohm (MM probe leed resistance)
or unexpectedly low (a few ohms), this may point to the failed component.
Candidates (all very long shots) would be,
Phono C423
Power amp C639
psu caps C913
With a true short, expect the fuses to pop everytime. DBT does not stay bright.
Probably have a psu failure eg, Q901, have you verified psu voltages, eg at output
collectors, even on DBT expect xxVolts
I don't understand how the DBT dims normally if both + and - power supply rails in either channel are shorted to the heatsink and hence ground. The collectors of the output transistors are connected to the rails, so I would think about the nature of any shorts there, particularly if you are using a continuity test rather than a resistance test.
Continuity beepers respond at about 60 ohms, so perhaps it's the low resistance reading you typically read through the power supply rather than an amplifier short?
Continuity beepers respond at about 60 ohms, so perhaps it's the low resistance reading you typically read through the power supply rather than an amplifier short?
What exact model is this on hifiengine they show 3 different model variations and so 3 different service manuals?
NAD 3020 - Manual - Stereo Integrated Amplifier - HiFi Engine
NAD 3020 - Manual - Stereo Integrated Amplifier - HiFi Engine
Actually, when I reinstalled the output transistors and then redid my measurement, the resistance between collector and heat sink just kept rising. So bottom line is that using continuity to check for a short between the output transistors and heat sink can be misleading? Interesting.
Yes, I have nearly 60mV DC offset in the left channel.
Yes, I have nearly 60mV DC offset in the left channel.
60mV is a little on the high side, however have seen some specs where
100mV is "tolerated", yeah, I'd try and reduce it.
Assume that 60mV is at maximum adjustment.
NAD used some very ordinary trimmers, outside chance the trimmer is suspect.
Measure between trimmer outer legs (20k) and then between outer and wiper
repeat for other outer leg and wiper. One combo should be short while the other
should be max 20k. Replace trimmer if it doesn't measure 20k +/-x%
Diode check D603, looking at Vf, maybe drifted low. Long shot would be R615, 617
out of spec. Suspect trimmer of diode.
100mV is "tolerated", yeah, I'd try and reduce it.
Assume that 60mV is at maximum adjustment.
NAD used some very ordinary trimmers, outside chance the trimmer is suspect.
Measure between trimmer outer legs (20k) and then between outer and wiper
repeat for other outer leg and wiper. One combo should be short while the other
should be max 20k. Replace trimmer if it doesn't measure 20k +/-x%
Diode check D603, looking at Vf, maybe drifted low. Long shot would be R615, 617
out of spec. Suspect trimmer of diode.
"Continuity" is ONLY for a first-check of Trailer Wiring. Or other real-simple circuits where the "only possible answers" are Zero and Infinity.
Even at that, a continuity tester will NOT know the difference between a Good Wire (0.5r), a Bad Wire (10r), and a Good Lamp (3-12r). If the "Stop" lead on the connector beeps, is that a good lamp, a dead short, or a wire going bad, unable to light the lamp yet still "continuity" on the buzzer?
That's why there's numbers on that thing.
Simple lamp circuits cover a narrow range of Ohms. We in the Electronics racket can face Ohms from under 0.001 to over 1,000 Million. Aside from quick beep-test to know when something which should be connected really isn't, we normally look at the number (including any K or M!) and judge too-much or too-little.
Not knowing just what your box is or how much is disconnected, a 500 Ohm reading around the power stage is not prima-facie a "short", it may be biasing or reverse leakage. At say 50V it would pass 0.1 Amps, make 5 Watts heat, which may or may-not be a lot for that part but is unlikely to blow-up the entire amplifier.
Thanks mbz. I checked the resistance of the trimmers, and they're both fine. But after looking closer, I noticed that a tiny piece of the locking paint applied to the trimmers at the factory was interfering with full rotation. I can now get the DC offset in that channel down to -30mV, which, while not wonderful, is acceptable in my book
Last issue (I hope).
There's a channel imbalance (right lower output than left) that I've been trying to track down by tracing a 1kHz sine wave. It's in the amp section, and first appears at R632, on the side of the junction with C618 and C620. I pulled R632 (and R630 for good measure), but both tested fine. Should I be looking at those caps?
Also, for what it's worth, the right side has the C642 shown on the schematic installed, but the left does not have C641, nor is there a spot for it.
Left side of the amp schematic shown because it has the component values.
There's a channel imbalance (right lower output than left) that I've been trying to track down by tracing a 1kHz sine wave. It's in the amp section, and first appears at R632, on the side of the junction with C618 and C620. I pulled R632 (and R630 for good measure), but both tested fine. Should I be looking at those caps?
Also, for what it's worth, the right side has the C642 shown on the schematic installed, but the left does not have C641, nor is there a spot for it.
Left side of the amp schematic shown because it has the component values.
The attachment is too small and blurry to read. Let's just identify the model by the different design issues which were either A (original unpecified series), B, E, or i, including models that were also numbered in the 31xx series. The schematic snip you posted seems to be issue B. Is that schematic correct for the particular semis (e.g. Q607) specified?
Update: After confirming that the imbalance was occurring in the amp stage by feeding a sine wave directly to the "Normal In" jacks, there suddenly was no longer an issue -- both channels are now equal, both with the amp and pre separated and coupled, i.e. the jumpers installed. Perhaps it was oxidation at the Pre-Out/Main In jacks. I had treated them with Deoxit several days prior, seemingly to no effect at the time. I'm going to observe for a few more days to make sure it doesn't return.
On a different note, has anyone replace the source switches on a 3020? The Tuner push button switch only stays engaged about half the time it's pressed in. Looking at the service manual, there's only an NAD part number.
On a different note, has anyone replace the source switches on a 3020? The Tuner push button switch only stays engaged about half the time it's pressed in. Looking at the service manual, there's only an NAD part number.
The first apparent fault may return if the problem was due to a capacitor, as just suggested. From cold, old caps may require a considerable amount of time to reform and work normally. If not already done in recent times, I would routinely replace old electrolytics like that one - like for like.
Replacing a 35 year old, alternate action push switch bank is not going to be easy and NAD part numbers won't help you as their spares will have been cleared long ago. I think these were off-the-shelf items though, so replacement could be be possible from generic parts, by measuring and comparing pin spacings, overall length and distance from a given pin centre to the end of the shaft. More recent equivalents tend to be miniature versions so be aware of the scale of drawings and images. Otherwise, your only alternative is to source a scrapped 3020 or other similarly fitted amplifier etc.
It seems odd that that this switch would give problems if not just because of corrosion, worn contacts or accumulated grime. Personally, I would carefully examine the interlock trigger mechanism plus the switch and its soldered connections, repairing as necessary but not everyone is aware of their construction detail or has the patience to tinker with small mechanisms.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.