All, I have recently done a full recap of an acquired (for free) NAD3020i. I have done all the minor bits and bobs as discussed on various sites. What I am confused with is that one of the transistors in the preamp powersupply 2SB649A is very hot to touch. Please see the attached photo. Is this the way it should be? All other transistors are just about warm after 10 minutes?
Now I am newbie to electronics but know the basic stuff, so please go easy on me. Thanks
Now I am newbie to electronics but know the basic stuff, so please go easy on me. Thanks
Attachments
The amplifier will have worked this way for decades and there is no rule that says all components should run at the same temperature. The manufacturer has fitted a standard heatsink that is adequate to keep the transistor safely operating within specification and that's all you need to know about, as long as the voltages associated with it on the schematic are correct and the transistor is still attached to its heatsink.
If it ain't broke, don't try to fix it.
If it ain't broke, don't try to fix it.
Having looked at a fuller story on another forum about your recap. saga, perhaps there is something to think about. The details of what you have already done and what mistakes were made are usually necessary to see how any real problem might come about. Can you measure the transistor case temperate with a thermocouple, IR sensor meter etc?
Ian, I don't have access to any such instruments to check the temperature. All I can say is that I can put my finger on this specific transistor for about 3 seconds without burning myself.
As far as the history behind it. I acquired this from a friend of mine for free and in mint condition fully working order. Though idling current on one channel was stuck at zero. After replacing all the electros with Elna silimic II and installing new Bourn trimmers I could not set the idling current more than zero on this one channel. While I was checking voltages at different locations to find the fault, I had a short circuit and most of transistors in amp for that channel were gone. I replace all of these with exact same ones and other things that I have replaced include
All resistors are replaced with exactly same amount. Only thing which just to came to my mind is the new rectifier KBP01M intead of RS102? But the specs are same so not sure if that could be reason? Had the amp running in for few hours and all other transistors are just warm bar the one in question.
Hope this explains the lot. Attached some pics for reference of what I have done so far.
As far as the history behind it. I acquired this from a friend of mine for free and in mint condition fully working order. Though idling current on one channel was stuck at zero. After replacing all the electros with Elna silimic II and installing new Bourn trimmers I could not set the idling current more than zero on this one channel. While I was checking voltages at different locations to find the fault, I had a short circuit and most of transistors in amp for that channel were gone. I replace all of these with exact same ones and other things that I have replaced include
- Rectifier diodes 1N5401 with ultrafast UF5401
- 4700uf 35V power supply caps with 6800uf 63V caps
- Rectifier RS102 for preamp with newer KBP01M (could this be the culprit?) Preamp main 330uf 63V power supply caps to 680uf 63V.
- Driver transistors 2N3055 and MJ2955 with MJ15015 and MJ15016 and installed a 4.7uf 1/2watt Base and .27uf 3watt Emitter resistor to stabilize the new transistors.
All resistors are replaced with exactly same amount. Only thing which just to came to my mind is the new rectifier KBP01M intead of RS102? But the specs are same so not sure if that could be reason? Had the amp running in for few hours and all other transistors are just warm bar the one in question.
Hope this explains the lot. Attached some pics for reference of what I have done so far.
Attachments
Standard rectifier bridges like KBP01 and RS102 all do the same job, as long as their current and voltage rating (Diode markings relate to Peak inverse volts, not RMS AC volts) is adequate. I can't see any issues with your substituting the same pattern 100V bridge with only a higher current rating, but why?
There is no benefit in using UF diodes in linear power supplies. Their application is for fast square wave switching as in switch mode supplies. They will work OK but perhaps with a greater amount of HF noise getting into the DC power rails.
For interest, the fancy power diodes to use in audio amplifier supplies are categorised as "soft recovery" UF diodes. Some are also known as HEXFRED rectifiers. There are not many suitable types freely available and most are in single or dual TO220 or SMD packages - not as axial leaded single diodes or in bridge packages.
Generally, there is no benefit in replacing semiconductors that are working fine. Unless you are targeting lower noise transistors to use at the preamp input stage, there is little to be gained and possibly much to lose upgrading more than capacitors. I'm no 3020 sound expert but I think a lot of your enthusiasm for changing parts might have been unwise.
BTW, the thermometer types are dirt cheap devices these days - Some low cost DMMs already have a thermocouple wire socket. The thermocouple itself is cheap and many guys have IR sensor guns as toys now. Ask around, as you only need one for 5 minutes.
There is no benefit in using UF diodes in linear power supplies. Their application is for fast square wave switching as in switch mode supplies. They will work OK but perhaps with a greater amount of HF noise getting into the DC power rails.
For interest, the fancy power diodes to use in audio amplifier supplies are categorised as "soft recovery" UF diodes. Some are also known as HEXFRED rectifiers. There are not many suitable types freely available and most are in single or dual TO220 or SMD packages - not as axial leaded single diodes or in bridge packages.
Generally, there is no benefit in replacing semiconductors that are working fine. Unless you are targeting lower noise transistors to use at the preamp input stage, there is little to be gained and possibly much to lose upgrading more than capacitors. I'm no 3020 sound expert but I think a lot of your enthusiasm for changing parts might have been unwise.
I guess you meant ohms or R instead of uF (microFarad) for the resistors there. It would appear from those changes that you had oscillation and bias problems when you changed the power transistors. Presumably those base stopper resistors tamed it.
BTW, the thermometer types are dirt cheap devices these days - Some low cost DMMs already have a thermocouple wire socket. The thermocouple itself is cheap and many guys have IR sensor guns as toys now. Ask around, as you only need one for 5 minutes.
Last edited:
Ian this started as straight forward recap for all electros, but then because of me not able to set idling current current on one channel (it basically remained zero no matter what) and then short circuit causing most of the transitors up in smoke I have to replace transitors. So didn't really changed these for no reason.
I read in other posts related to NAD3020 that changing the rectifier diodes to Ultra fast soft recovery made the most difference in sound quality after recapping of course and hence that change. Vishay specs states these UF5401 as "Soft Recovery Ultrafast Plastic Rectifier" and as a direct replacement for 1N5401. Changed the RS102 as it was only €1.50 or something and I need to build up my order to get free delivery. With mouser I have to order minimum of €65 to get free delivery otherwise there is €20 charge.
I know I went bit over the top with this one. I guess I was treating this as my learning project. I haven't touched electronics in 18 years. Last amp I build was STK459 back in 1996 and that time I knew nothing about audio amps at all. I am a civil engineer by trade
.
I love music and I absolutely love the way NAD3020i sounds. Next step will be investing in good speakers to with NAD (any recommendations?) I am learning a lot by reading different forums on various sites but this is the first time i had the courage to start my own thread looking for help and I am glad I did.
I will look into getting these IR heat sensor guns or whatever its called, no harm having one in my minimilistic diy kit. I was reading the other thread which you have replied to and can see it is a common issue with these 2 transistors in preamp powersupply.
I read in other posts related to NAD3020 that changing the rectifier diodes to Ultra fast soft recovery made the most difference in sound quality after recapping of course and hence that change. Vishay specs states these UF5401 as "Soft Recovery Ultrafast Plastic Rectifier" and as a direct replacement for 1N5401. Changed the RS102 as it was only €1.50 or something and I need to build up my order to get free delivery. With mouser I have to order minimum of €65 to get free delivery otherwise there is €20 charge.
I know I went bit over the top with this one. I guess I was treating this as my learning project. I haven't touched electronics in 18 years. Last amp I build was STK459 back in 1996 and that time I knew nothing about audio amps at all. I am a civil engineer by trade
.I love music and I absolutely love the way NAD3020i sounds. Next step will be investing in good speakers to with NAD (any recommendations?) I am learning a lot by reading different forums on various sites but this is the first time i had the courage to start my own thread looking for help and I am glad I did.
I will look into getting these IR heat sensor guns or whatever its called, no harm having one in my minimilistic diy kit. I was reading the other thread which you have replied to and can see it is a common issue with these 2 transistors in preamp powersupply.
Hi,
It's just possible that it's oscillating, now that you've changed all the Cs. If you haven't already discarded it, try replacing C526 with the old one. The ESR (equivalent series resistance) of this cap may be important with regard to stability (in the Nyquist sense).
If you don't have the old C526, you could try adding a resistor is series with the transistor's base - close as possible to the transistor. Value should be a few 10's of ohms. I'd only do this if you were confident you weren't going to mess up the circuit board and/or components.
Of course, if you've got a decent 'scope, you could look for oscillation on the -28V power line - could be at a fairly high frequency though.
Of course, it's possible that the elevated temperature is normal, as others have said...
It's just possible that it's oscillating, now that you've changed all the Cs. If you haven't already discarded it, try replacing C526 with the old one. The ESR (equivalent series resistance) of this cap may be important with regard to stability (in the Nyquist sense).
If you don't have the old C526, you could try adding a resistor is series with the transistor's base - close as possible to the transistor. Value should be a few 10's of ohms. I'd only do this if you were confident you weren't going to mess up the circuit board and/or components.
Of course, if you've got a decent 'scope, you could look for oscillation on the -28V power line - could be at a fairly high frequency though.
Of course, it's possible that the elevated temperature is normal, as others have said...
So today I have some time to experiment bit more, as I was not satisfied with the bass reproduction after the recap. So I took out the rectifier diodes UF5401 and replaced these back with original 1N5401's to see what happens? Guess what, the bass was back but not in a good way. The only way to put this to the words is that instrument separation was reduced and vocals sounded as if they are echoing/elongated. You might not notice this unless you are listing to some Nora Jones I could not believe how I was listening to this before and thought it was ok. I guess I didn't have anything to compare it with. Anyway back in goes UF5401 and everything is nice and sweet again. Instrument separation is at different level to the extent that I am hearing instruments which i didn't even knew existed in my favourite songs. when I took out 1N5401 I saved these in my box to be reused if ever needed somewhere else, after this experiment out they go in bin. Now I can confirm that this is possibly the easiest tweak with major sound improvement in NAD3020i. Oh by the way the UF5401 are vishay Ultrafast soft recovery diodes.
I have some more questions. Looking at Schematics of 3020i and 3225PE. I can confirm that the only difference is main powersupply envelop, rest is all same circuit with same value parts. The specification for NAD3225PE rates it at 25W per channel compared to 20W on 3020i. What I am wondering is, as the PCB is same for both amps I can easily upgrade powersupply of 3020i to 3225PE. Which technically will make my 3020i to 3225PE. The question is, is it worth it? or should I just leave it alone? Attached photos your reference
Attachments
I am currently upgrading a 7020e to a 7220PE (same basic circuit design) that shares the same PCB. So far the only hiccup I have had is the main transformer. In my case the 7220e transformer was missing the power rails for the PE power supply circuit. I picked up a junked 7225PE off "The Bay" which is about the only way I know to get a NAD transformer today. I haven't started the rebuild yet, but I'll chime back in to let you know if it worked or not.
One other thing we have in common, is the heat issue with the pre-amp PS transistors. Back when my unit was a couple years old, it cooked the PCB. In may case it was R512 (the base resistor to Q508) that failed. But I've always been leery of that transistor. This time around, I'll be beefing up the heat sink on Q507,508.
One other thing we have in common, is the heat issue with the pre-amp PS transistors. Back when my unit was a couple years old, it cooked the PCB. In may case it was R512 (the base resistor to Q508) that failed. But I've always been leery of that transistor. This time around, I'll be beefing up the heat sink on Q507,508.
Attachments
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.