NAD 3020A Recap Woes

Earcraft

New Member
I dug out a 3020A (SN: A3243376) that I had packed away for about 15 years. An old boss of mine used to buy up defunct recording studios and this came out of one in NYC. Since it wasn't "Pro" gear and he had no idea what it was, he gave it to me. I decided to fire it up and it had the classic problems i.e Turn on thump (VRRRUUUMMPPP), and a major 60cycle hum in one of the channels. I have not turned it on since...

I set it aside till this winter so I could do a recap. The thing is in such good physical condition, i'm keeping the work to a simple restoration. I decided to replace all the electrolytics first to see where that got me. I used Elna SILMIC II in the audio circuits and a combination of Nichicon FG,FW,KA, and KL for power and other circuits. I replaced a few diodes and resistors that had corroded due to the brown glue. I finally replaced all the resistors in the “soft-clipping” circuits with metal film 1% equals to hopefully make it a little quieter.

After firing it up and adjusting bias and DC offset, I did a listening test. The audio is was completely distorted at low level and consistent across both channels. That's when I checked the fuses to find both 4A (F3,F4) blown. After checking cap polarity and a good inspection of the traces looking for shorts, I tried another pair and they went pop the way only slow-blows can.

After hijacking Leesonic’s post http://www.audiokarma.org/forums/showthread.php?t=460678&page=4 I got some advice to check out the output transistors and drivers which I did. While I cannot confirm they are working as spec’d, the tested out OK with my multi-tester using the diode setting. (I did not go as far as to pull the Bias transistors but plan to a little later today.) Also as suggested, I powered up with the drivers and output transistors out to check to make sure the rectifier was still good. It did not blow fuses, so I assume it is.

After a careful reassembly and making sure not to short the main output transistors to the heat sink, I again tested the unit. This time I built a quick dim-bulb-tester and used it. The 150W bulb glows dimly and pulsates at a slow rate (approx. 2 pulses a sec). I also noticed the the Output Transistors heat up fast and get really hot. I have a 7020e that I use and it is not nearly as hot after being on all day and driving music.

Since I am still suspicious about the Transistors, I started looking to buy replacements to have in case I want to pop them in to see what happens. The main 2N3055/MJ2955 are easy to find. The drivers are 2SD699A and Bias 2SD699A/2SB649. I have looked for a few days and cannot find direct replacements or a cross reference that I can understand. In earlier units I seen people using BD139 but to me, the data sheets don’t look all that similar to the 699A. Any AK’ers know of compatible replacements?

I’m a bit stumped, and I need some advice.
 

Attachments

  • Before 1.jpg
    Before 1.jpg
    91 KB · Views: 222
  • Glue mess1.jpg
    Glue mess1.jpg
    93.3 KB · Views: 187
  • Glue mess 2.jpg
    Glue mess 2.jpg
    50.5 KB · Views: 182
  • After 2.jpg
    After 2.jpg
    122.1 KB · Views: 226
Since I am still suspicious about the Transistors, I started looking to buy replacements to have in case I want to pop them in to see what happens. The main 2N3055/MJ2955 are easy to find. The drivers are 2SD699A and Bias 2SD699A/2SB649. I have looked for a few days and cannot find direct replacements or a cross reference that I can understand. In earlier units I seen people using BD139 but to me, the data sheets don’t look all that similar to the 699A. Any AK’ers know of compatible replacements?
Hi,

Not positive about replacements (never seen the 3020A) but as I just did the 3020e I thought I would show how these transistors are on it. Look at the pic. Its your pic but I marked what the factory used on the 3020e. I think they all have the same 2N3055/MJ2955 outputs - atleast my e and i -versions do and from what I have read also the earlier ones.

BTW. I don't remember where (maybe DIYaudio?) I got the data but I was planning to replace the 2SD699A/2SB649s with 2SC4793/2SA1837s. But as I didn't find enough of those I had to replace them with another replacement pair which was even further away from the originals. Cannot remember out of my mind what they were but anyways: the amp still sounds great to me.

BTW. aren't the 2SD699A/2SB649 pair drivers and the single ones mounted on the heat sink (in my case BD139) bias transistors?

EDIT: now checked also the 3020i specs and they are the same as 3020e except instead of bd139 they have used bd139-10s which only has minor difference between them. Maybe you could follow this path NAD took with the later ones?
 

Attachments

  • How-its-on-3020e.jpg
    How-its-on-3020e.jpg
    50.2 KB · Views: 133
Last edited:
Hi,
BTW. aren't the 2SD699A/2SB649 pair drivers and the single ones mounted on the heat sink (in my case BD139) bias transistors?

My understanding of amplifier circuits is defiantly a work in progress. I'm learning as I go. It would be easy to acquire BD139 which I think NAD started using for Bias transistors soon after my vintage. How well they will work and how they effect bias setup levels is what gives me pause.
 
Does the 3020A have trimpots for Bias adjustment?

Anyways, BD139 are not perfect, they are famous for causing thermal runoff which is said to be the main reason 3020 and its variations to self destruct sooner or later (1-30 years). And that is why bias resistors are recommended as an upgrade when rebuilding these amps.

But, if your amp uses only 2SD699A/2SB649 in both drivers and bias positions then you might try the ones I mentioned and forget the BD139. They are available in Mouser. Just double check the pinouts on these as they might be different to originals - or not. I had to install the replacements replacements I used other way around as the pin order wasn't identical to originals.

Mouser:
757-2SC4793
757-2SA1837

Ok, its late here in Europe so bye for now.
 
Thanks ToniCH... sweet dreams:boring:

My unit had fixed bias resistors originally. I swapped out them out with Bourns 20k trimmers. If you look close to my "After" picts you will see that I soldered temporary wires to the underside and brought them up to the topside (the red and black T) for easy Bias measurements.

I'll check out the recommended transistors.. Again, Thanks.
 
Since I am still suspicious about the Transistors, I started looking to buy replacements to have in case I want to pop them in to see what happens. The main 2N3055/MJ2955 are easy to find. The drivers are 2SD699A and Bias 2SD699A/2SB649. I have looked for a few days and cannot find direct replacements or a cross reference that I can understand. In earlier units I seen people using BD139 but to me, the data sheets don’t look all that similar to the 699A. Any AK’ers know of compatible replacements?

I used BD237 and BD238 to replace the 669 and 649 respectively, on the schematic they are Q611 and 612, and 613 and 614. Q607-610 I used BD139. Never had any problems with these replacements. Make sure you check the pins outs of your replacements compared to the originals. I know on the earlier amps, if you replace the 2N6551 and 2N6554 with BD237 and BD238, you have to mount them sideways and bend their legs in order to get them to fit (see picture below).

If you are having other problems, check and double check the circuit board for dry joints. There are a number of tiny jumpers that take the power across the board to the output transistors. Just one of these gone bad can throw you off big time.

Lee.

NAD3020second08.jpg
 
Remember that silicon washers cannot be re-used or even re-mounted. Replace them with new ones - or better yet - mica, otherwise thermal contact will be compromised.
 
Last edited:
Thanks for the transistor cross reference Lee. I spent hours looking and got nowhere (to many specs over my head to make a match). Yesterday I checked Q611-614 and those too checked out OK with a diode tester. After writing this, I’ll go further back in the circuit.

Even though all the transistors seem to have no shorts or opens, do you think it is worth replacing them with new? Especially the mains? If the answer is yes, we’ll need to start a new discussion on base blocking and emitter resistors.

I have inspected all the solder joints several times under magnification, but I know these are hard to identify so I will keep looking. There were a few jumpers that got "glued" and had some corrosion. If I can find them (I cleaned the board up pretty good) I guess I should replace. Are cut capacitor leads an OK replacement for jumpers or should I add real ones to my next Mouser order?

Baron, I did reuse the Sil-Pads from the first attempt. After testing and reinstalling, however, I used a meter to make sure the transistors were not shorting to the sink… they tested OK. I have several more pads on order as I’m sure I’ll need them throughout this troubleshooting process. I used the Tgard 5000 from Laird after seeing Lee’s work, but I did some digging this weekend and found Tgard K52 which uses Kapton film and looks like it has better insulating and thermal properties if you don’t over tighten the TO-3 screws. Any opinions?

Thanks for all the input… I’ll report back as time permits.
 
Sometimes it's easier to search on here to see if there is a replacement for a transistor, than try and find one yourself. Plenty of parts were used in other brands, and there are transistor lists all over this site that might give you a hit.

Don't forget to check the bridge rectifier for the power amp stage. I've seen one of these fail short circuited, I tested every transistor and they all checked out. As a last resort, I put new fuses in without any transistors on the board, and the fuses STILL blew. Sure enough, bad bridge rectifier. I think they are less than a dollar from Mouser, now I just change them out whenever I work on one of these amps.

Lee.
 
Well it’s Sunday again, so I am back to working on my 3020. I guess it’s obvious I’m a weekend Geek… too busy during the week.
Lee> Unfortunately, I missed the rectifier on the last order, but one is on the way. I did test the old one and it checks out, but I’ll replace it anyway next week.

My new Transmitters arrived and I have a hand full of BD237/238 and BD139. I am trying to avoid replacing the main output transistors if at all possible. So I plan to start with the Bias/Drivers and see if that resolves the problem. I still have questions about compatibility and how the different values will effect operation and Bias setup before I start replacing. Unlike the older 3020 (with the filter caps behind the headphone jack), my version uses 2SD669A for bias (Q609-610) and 2SD669A/649A for divers (I hope I have it in the right order now). The older 3020 uses BD139 for it’s Bias (Q609-610). I also noticed that my 7020e which is a few years newer uses the BD139 for bias (that one is next in line for a recap).

I need some help understanding with values are most important to match. To me, more of the BD139 values match the 3SD669A than the BD237 do. Neither the 237 or 139 come close to matching the “Collector to base” & “Collector to emitter” maximum voltage. (the spec sheets are to easy to find online, so I have not attached them)

2SD669A - C to B = 180V / C to E = 160V
BD237 - C to B = 100V / C to E = 80V
BD139 - C to B = 80V / C to E = 80V

I have a feeling that the “Maximum ratings” are like the “not to exceed” voltage on a capacitor, but that's obviously why I need some advice from AK’ers with a better understanding of transistors than I.

2 questions I would like advice on before I fire up the soldering iron are:
1 - Should I used the BD237 or BD139 for the bias transistors (see attached schematic section)? Any advantages or or disadvantages to either?
2 - Currently, the service manual recommends setting the Bias (Idle current) to 30mV - 60mV. Will that change?

Thanks in advance.
 

Attachments

  • 3020AMP.pdf
    143.1 KB · Views: 80
Last edited:
I just noticed that the schematic shows Q611-614 as 2N6553/6556. The parts list and what I pulled from the board are 2SD669A/649A.
 

Attachments

  • 3020AParts.pdf
    152.5 KB · Views: 60
Well it’s Sunday again, so I am back to working on my 3020. I guess it’s obvious I’m a weekend Geek… too busy during the week.
Lee> Unfortunately, I missed the rectifier on the last order, but one is on the way. I did test the old one and it checks out, but I’ll replace it anyway next week.

My new Transmitters arrived and I have a hand full of BD237/238 and BD139. I am trying to avoid replacing the main output transistors if at all possible. So I plan to start with the Bias/Drivers and see if that resolves the problem. I still have questions about compatibility and how the different values will effect operation and Bias setup before I start replacing. Unlike the older 3020 (with the filter caps behind the headphone jack), my version uses 2SD669A for bias (Q609-610) and 2SD669A/649A for divers (I hope I have it in the right order now). The older 3020 uses BD139 for it’s Bias (Q609-610). I also noticed that my 7020e which is a few years newer uses the BD139 for bias (that one is next in line for a recap).

I need some help understanding with values are most important to match. To me, more of the BD139 values match the 3SD669A than the BD237 do. Neither the 237 or 139 come close to matching the “Collector to base” & “Collector to emitter” maximum voltage. (the spec sheets are to easy to find online, so I have not attached them)

2SD669A - C to B = 180V / C to E = 160V
BD237 - C to B = 100V / C to E = 80V
BD139 - C to B = 80V / C to E = 80V

I have a feeling that the “Maximum ratings” are like the “not to exceed” voltage on a capacitor, but that's obviously why I need some advice from AK’ers with a better understanding of transistors than I.

2 questions I would like advice on before I fire up the soldering iron are:
1 - Should I used the BD237 or BD139 for the bias transistors (see attached schematic section)? Any advantages or or disadvantages to either?
2 - Currently, the service manual recommends setting the Bias (Idle current) to 30mV - 60mV. Will that change?

Thanks in advance.

Whats the status of the amp "right now"? Can you run it? Or does it blow a fuse when you try?
 
Currently I am blowing fuses... Actually I am lighting up my dim bulb tester now. Check the top of the thread for my trouble shooting so far.
 
I just noticed that the schematic shows Q611-614 as 2N6553/6556. The parts list and what I pulled from the board are 2SD669A/649A.
Just a quick observation: the parts list you posted above lists Q611-612 as 2SD669 and Q613-614 as 2SB649 - not 2SD649.

EDIT: I just had a chance to check it out: 2SB649 is PNP and 2SD649 is NPN. So, be careful out there! ;)

BTW. do you have a link to a 3020A service manual? I haven't found one.
 
Last edited:
EDIT: I just had a chance to check it out: 2SB649 is PNP and 2SD649 is NPN. So, be careful out there!

BTW. do you have a link to a 3020A service manual? I haven't found one.

Thanks for the warning ToniCH... It 2SB649's I pulled off the board. PM Me and I'll send you what I have... Very hard to find this version online.

I replaced the Drivers and Bias transistors using BD237's for The bias. When I fired it up, the DBT glowed bright (not dim as before). I quickly turned it off and hooked up my meter to the bias adjustment leads I had tacked on to the under side of the board. Turned it back on and measured 1.4V YIKES!!! Things started to smell hot but I was able to turn it off again before I let the smoke out. I cranked both bias trimmers way down and turned it back on and the bulb was dim until I adjust the bias down to roughly 38mV on both sides. Lights out... no smoke... fuses good.:beerchug:

I'll re-adjust tonight after a proper warm up and and hopefully give it a listen tomorrow after I pick up some 1/2A fuses. As much trouble as I've had, I want to protect my Polks just in case.

More to come...
 
After a Newbe blunder, and some work left to do, I have it running and it sounds so good. *I have it running through my favorite pair of Polk bookshelves an the sound is so dramatically better than the 7020e that's been powering them for years. Before I get to my blunder, I have to make a correction.

TonyHC called to my attention that the faceplate for the 3020A clearly has the “A” next to the model. Mine does NOT have an “A” so I have an original 3020 with an A-B style PCB. Those who have a true 3020A, please note, answers given here my not apply to your model. Tony and I have yet to fine a true 3020A service manual so it’s hard to say what the differences are.

My biggest struggle over the last 2 weeks has been with properly setting the bias (more on that below). But when I got it adjusted to where I felt it was right, I couldn’t wait to put sound through it. <Blunder Alert> So I hooked up my speakers with 1/2watt fuses in line for protection and the unit powered through the dim-bulb-tester. Turned it on and it sounded like angels coming out of my speakers. I continued to listen while I putzed on the computer. About 30 min later I start to smell something bad! R653 was glowing red and the main transistor heat sink was HOT! (I’m still apologizing to my wife for scaring the hell out of her at 1:am… she thought the house was on fire). I made 2 STUPID mistakes. 1) I didn’t reconnect the solder bridges to short the 1ohm resistors used to set bias (the Pros probably figured that out). 2) I did not screw the bias transistors back into the main heat sink. One good thing to come out of this is I now know the meaning of “Thermal Runaway”.

I have a few issues which I could use some advice on. *

1) I could not set the bias to 30mV with the 560 ohm resistor in parallel with R645-646 as stated in the manual. I have to believe this is a misprint or not meant for my exact model. *When I tried this, my dim bulb tester lit up and I could not get the bias below 115mV. *Without the 560ohms I can set bias to 30mV. *The resistors I pulled out of RX1-2 measured .457kohm and I was able to set the Bourns trimmers to that value. *I have the 2030B service manual and that procedure uses a 1ohm which on mine is already in place (R653-654). Right now everything is running cool and it sounds great. *Just not sure if it's right, or how to make sure. *I have attached the setup directions from the “late 3020 service manual”.

2) I still have some of the turn on Thumps, Vruumps, and Buzz that resolve in about 6 seconds. It's very minor compare to the first time I turned it on, but I hoped the recap would have fixed this. *The “Thump” has to be the on/off switch which is a well documented issue. I get quick Vrummp Vrummp one on each side a 1/2 second apart. A second after the Vrummps there is a Buzz that starts loud then fades out for about 3 seconds. When I pull the Pre Out to Normal In jumpers then power up the starts up is clean accept for the thump from the switch. *I'm guessing there is still work to be done on the pre-amp section. Any suggestions?

Sorry for the book…
 

Attachments

  • 3020A.gif
    3020A.gif
    135.2 KB · Views: 62
  • 3020Late.jpeg
    3020Late.jpeg
    26.6 KB · Views: 62
  • Late3020bias.pdf
    403.1 KB · Views: 52
There is no mention in your manual adjustment procedure about the solder bridge. Maybe it doesn't apply to your model?

When you are reducing bias does the resistance on your trimpot increase or decrease? Logically it should increase per your service manual #7 of the adjustment procedure. Maybe you need more resistance? May I suggest trying bigger ohm trimpots?

Again this might not apply to your amp but with my 3020e I could not get bias down to spec with the standard 100ohm trimpots. I installed 500ohm trimpots and I had no problems getting it down to 35mV. Don't know why (maybe some of my transistor subs caused this?) but now it works perfectly. Note that your trim pot values differ from mine so check what you have.

About the turn on thumps: Some threads on the forum suggested the J111 fets might be the cause. They are cheap so you should give them a try. Just check the pin order before installing.
 
ToniCH> Again this might not apply to your amp but with my 3020e I could not get bias down to spec with the standard 100ohm trimpots. I installed 500ohm trimpots and I had no problems getting it down to 35mV. Don't know why (maybe some of my transistor subs caused this?) but now it works perfectly. Note that your trim pot values differ from mine so check what you have.

You might be right about more resistance on the bias trimmer. I installed the Bourns 1k trimmers. Since I am so confused about how to measure bias on whatever version 3020 I have, I brought home an Oscilloscope from work to set bias. With pro amps, we crank down the bias until it starts square at the bottom of the sine, then bring it up slightly till it’s a perfect sine. I cannot get this amp to “under bias” with the 1K trimmers. However, I don’t know if consumer amps and live sound amps are biased the same way. Heat is always the enemy when dealing with >1000W amps…

I am SOOO! ready to put the cover back on this this amp this weekend and call this project done. But I have a have another question.

I received a new power switch that Leesonic and others have used with their rebuilds. This switch has [2] solder leads. The original switch has [10]… There are [4] wires that land at the switch. Two [2] from the transformer to the board and [2] straight to the board. On my board (P7, P8). Am I safe to parallel the main power and (P7, P8) cables together, or did I get the wrong switch?
 

Attachments

  • IMG_0306.jpg
    IMG_0306.jpg
    70.7 KB · Views: 95
  • IMG_1477.jpg
    IMG_1477.jpg
    37.3 KB · Views: 90
Last edited:
Back
Top Bottom