Okay so I have here a NAD 304, in protect mode. One channel is good, the other not so good. Most of the time, there is 11 volts DC on the output, occasionally as much as 30 and sometimes only 100 mV.
After some basic testing, I decided to replace the 3.9V zener diodes as I wasn't reading the correct voltage drop over them. After that, the amplifier would click out of protect mode but the very strange thing is if I moved my hand over the pcb and close to the components, then it would click back into protect mode, so it may be picking up a tiny bit of EMI noise, just enough to upset things? When its not in protect, a quite hum can be heard from the speaker.
Voltages throughout the amp are all over the place and seem to change without anything else changing. I removed and did the diode test on each transistor individually and all transistors appeared to be fine, and I replaced the electrolytic caps in the bad side as well, but nothing seemed to make a difference. I also lifted a leg of each diode and did a voltage drop test on each and they all seem ok. Should I just replace all of the small transistors anyway? Or do I need to start checking resistors? This amplifier seems to have a very complex design compared to other amps, not sure what they were trying to achieve but it makes troubleshooting a tad more difficult.
Anyone who has any suggestions or ideas, your input is appreciated and I thank you for the time taken to help me with this amp.
After some basic testing, I decided to replace the 3.9V zener diodes as I wasn't reading the correct voltage drop over them. After that, the amplifier would click out of protect mode but the very strange thing is if I moved my hand over the pcb and close to the components, then it would click back into protect mode, so it may be picking up a tiny bit of EMI noise, just enough to upset things? When its not in protect, a quite hum can be heard from the speaker.
Voltages throughout the amp are all over the place and seem to change without anything else changing. I removed and did the diode test on each transistor individually and all transistors appeared to be fine, and I replaced the electrolytic caps in the bad side as well, but nothing seemed to make a difference. I also lifted a leg of each diode and did a voltage drop test on each and they all seem ok. Should I just replace all of the small transistors anyway? Or do I need to start checking resistors? This amplifier seems to have a very complex design compared to other amps, not sure what they were trying to achieve but it makes troubleshooting a tad more difficult.
Anyone who has any suggestions or ideas, your input is appreciated and I thank you for the time taken to help me with this amp.
Attachments
The Japanese (Hitachi) 2SB/2SD drivers have a peculiar failure mode of going intermitently open circuit base to emitter. Not NAD specific... I encountered this time and again in all type of gear and over similar devices.
The real answer is to try and be sure. I wouldn't use speakers for testing as a real fault could damage them.
The thing to check if you can is measuring across the base and emitter junctions on DC volts with it on. The reading should be in the 0.6 to 0.7v range for all the transistors. If one is faulty the reading will increase over that.
Obviously check for dry joints and so on too (on those transistors too). Make sure the solder "takes" cleanly to the leads. They have a habit of being oxidised.
The real answer is to try and be sure. I wouldn't use speakers for testing as a real fault could damage them.
The thing to check if you can is measuring across the base and emitter junctions on DC volts with it on. The reading should be in the 0.6 to 0.7v range for all the transistors. If one is faulty the reading will increase over that.
Obviously check for dry joints and so on too (on those transistors too). Make sure the solder "takes" cleanly to the leads. They have a habit of being oxidised.
I just got done debugging a Harman Kardon amp with kind of the same problem you described. Going in and out of protection mode and later just plain blowing fuses depending if the output transistor heatsink was touching the chassis or not and if my hand goes touched the heat sink or not. Finally, narrowed it down to a 2SA1306 driver transistor. Out of circuit, the base emitter junction would measure 0.85V with a Fluke DMM on the diode scale, while its counterparts in the other side and the other channels all read 0.55V. I think it was on its way out to complete failure, in circuit it would bias the power devices on full and blow the fuses. Replaced with 2SB649A/D669A and all is fine now.
A quick scan of a couple on line data sheet and I could not figure out the cut off frequency of the BD140 pair. The 2SB649 I suspect is much faster than them so I am not sure what would happen if you stick them in. At least you would have to tweak the Miller cap to ensure stability. 2SB649 type is quite common in receiver designs. I harvested mine from a non-functioning Insignia amp.
Look at Vceo and look at your rail voltages. BD139/140 are good to 40V rails and I think I see 50V rails here. They are reasonable drivers but not up to the task. The Hitachi/Renesas parts are now extinct so if you are limited to local retailers for your purchases, I think you will find only MJE340/350 but bear in mind that these will be stretched for current on 4 ohm loads (0.5A max).
Try Greg Erskine's spreadsheet for suitable parts listing and WES components in Sydney for more useful parts availability
Try Greg Erskine's spreadsheet for suitable parts listing and WES components in Sydney for more useful parts availability
Last edited:
Do you mean D305 and D307 or do you mean D309 and D311? If you mean D305 and D307, this is consistent with R333 being bad. So check that resistor. I have repaired a few 304s, and in every single one the problem was those 47k resistors. The symptoms were exactly as you describe, with the output voltage floating around uncontrolledly.
I'd check and replace the resistor in the other channel too. It might be on its way going open circuit. The circuit will work somewhat but with reduced bias current and increased distortion when the resistor increases in value, until that resistor goes completely open circuit at which point the output becomes uncontrolled.
Thermin mode
Capacitive effects by the hand can be caused but high frequency oscillation in the circuit. Have you scoped around the circuit ? Light sensitive effects by the hand can be caused by semiconductor photo sensitivity (like glass bodies).
Capacitive effects by the hand can be caused but high frequency oscillation in the circuit. Have you scoped around the circuit ? Light sensitive effects by the hand can be caused by semiconductor photo sensitivity (like glass bodies).
Well following your advice, I removed the R333 47K resistor, and sure enough, open circuit! Salvaged one from a junk pcb and hey, it works now! Set the bias and dc offset and it runs like a charm now, so will be buying some new resistors to replace them in both channels. Thank you to everyone who commented, you have saved me much time and unnecessary component replacement and I am thankful for that. Cheers! 
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.