NAD 3225PE stereo amplifier repaired and tested

[feedback.loop]

Someone tossed this amplifier out with loose covers and missing screws. Apparently, this is an unfinished repair attempt. The amplifier was repaired and tested, and many details have been explained in the process.



Service manual attached

[Smoky]

Fantastic!

NAD built wonderful sounding stereo equipment.

Thanks for sharing your testing procedures too 

[mzacharias]

Quality-wise, they are absolute junk.

[Smoky]

That amplifier is about 32 years old, and after his fine repair, it'll probably go another 20.

I like seeing old stuff be gifted with a new lease on life

[DimitriP]

I'd be interested to find out the results of a Frequency response sweep with the tone controls flat, full boost and full cut.  Hopefully you are interested too

[feedback.loop]

I'd be interested to find out the results of a Frequency response sweep with the tone controls flat, full boost and full cut.  Hopefully you are interested too

Here you go:

[DimitriP]

Here is the LTSPICE simulation of the tone control circuit of the 3225. They use the same arrangement in the 7130 . It doesn't look very good either

[feedback.loop]

I run Ubuntu, and I don't think that LTSpice for Linux exists. Perhaps you could share the result, not just a simulation script.
What is your point? If I understand correctly, you are not happy with tone controls in NAD 3225PE, right?
Frankly, I do not care too much about the tone controls. I almost never use them. I just keep them in the flat position (or even press the "tone defeat" button on my NAD C350)

[DimitriP]

Here is the simulation result showing a wierd midrage hump.

[feedback.loop]

There is no legend on your plot. What do those three lines represent? Are you saying there must be about 5 db hump at 1 KHz with controls in the middle?
I do not measure any hump. Look at my scope screen captures. I used white noise signal, and enabled persistence on FFT (purple). The response with tone controls in the middle looks quite flat. The horizontal scale there is not logarithmic, unfortunately, but that should not hide a substantial hump.

[feedback.loop]

I experimented quite a bit more: hooked up HP 33120A function generator and HP 34401A multimeter, both controlled through GPIB, and wrote a Python script to measure frequency response.
This is significantly more precise and flexible than oscilloscope, but slow, not real time. And the results can be plotted properly on the logarithmic scale (both X and Y). Perhaps I might do a video about this.

Below is a plot of the flat response. Notice that it is within 1 dB throughout the whole range 20 Hz - 20 KHz

[trobbins]

For those without the HP grunt, having a soundcard setup to do frequency spectrum plots has somewhat overtaken the gpib control/record loop - and software nowadays adds so much more for audio amp testing like checking noise-floor and distortion levels at stepped output power levels in case something is not quite right in an amp.

[DimitriP]

Below is a plot of the flat response. Notice that it is within 1 dB throughout the whole range 20 Hz - 20 KHz

Wow. Thanks for posting the plot.
This verifies the arrangement works in real life but "fails" when simulated. I'll gather my screen captures etc and probably post on a new thread.

[DimitriP]

For those without the HP grunt,

When it comes to "HP grunt", there are those that have it and those that wish they did

[trobbins]

   In some ways I enjoy just using a soundcard and software for pretty much all my bench-top audio testing - perhaps similar to writing some python script.  It's only at the extremes of testing I have to extract something from a shelf, and apart from just fixing my 3325A, and using a 3497A for some cross-comparison calibration, it was a venerable 3400A that allowed some Williamson testing out to 1MHz.

[feedback.loop]

  In some ways I enjoy just using a soundcard and software for pretty much all my bench-top audio testing - perhaps similar to writing some python script.  It's only at the extremes of testing I have to extract something from a shelf, and apart from just fixing my 3325A, and using a 3497A for some cross-comparison calibration, it was a venerable 3400A that allowed some Williamson testing out to 1MHz.

What software do you use? Did you build some sort of a buffer to scale the input signal and protect your sound card from over voltage? Is there some calibration in the software to have realistic measurements?

[trobbins]

I lucked on an EMU 0404 USB soundcard interface a few years back - that type of USB interface (similar to 2i2) has input level control for the first line of protection against high signal level, and provides a single ended input connection with 1 Megohm loading which then allows a 10:1 or 100:1 scope probe to be used.  I use a BNC-RCA and then RCA-6.5mm metal interconnector between probe and interface, and the cheap 100:1 ebay clone probe needed its compensation adjusted for flat response (it starts to droop above about 60kHz so a calibration file can then be used to provide flat response out to 90kHz).  That allows probing within ss and valve amps (although I don't try to probe certain B+ level circuitry) and any speaker line output.  If the amp has a good internal ground connection from signal input to speaker output then a loopback connection through the amp allows the full gamut of software functions for testing.  For the last few years I've used REW software - a standout platform given its public forum and continual improvement that the designer has been putting in for many years, and being based on speaker/audio testing it is more than just for audio amp testing.

I don't even try to use a calibrated software setup (apart from a flat spectrum calibration sometimes), due to pot settings on the input level and output level of the interface, and just use a DVM on the speaker output and/or amp input to provide signal reference levels (that is one advantage of having two cheap Aneng AN8009 to provide bench-top quick convenience).  So not an embedded approach to displaying referenced levels on a plot, although that could be done for a particular setup of pot settings (although that requires confirmation each time pot levels are changed, but in practise I typically end up keeping the pot settings at 'min' and using software set levels and scaling due to the very large SNR of the soundcard).

[feedback.loop]

I finally managed to make a video. Here you go:
Some ways of measuring frequency response were discussed, and some of them demonstrated.



Python script is attached.

[Shock]

Great video as usual, watching the second one now.
A while back I got a NAD 7240PE Receiver given to me, haven't had a look at it yet to see what the problem is. I assume it's the integrated version of the NAD 3240PE amp. NAD makes great looking amps, I've never owned one but the few I've listened to were pleasing. Even though some would nitpick at the layout and I suspect they haven't gone all out on component choice they look really easy to work on. Not sure on the differences in the 3325PE and 3340PE but I also found the following video interesting as well. Apparently these are Class G amps, had to look that one up.

[Asycid]

Hello all!

I've just found the repair video on youtube, and it gave me big hope to repair a 3225PE too.

I'm very beginner in electronics and all, but I just started to repair a NAD myself, but I've out of ideas what should I change in the Amp.

I very appreciate if someone here could help me!

Symptoms:
I've bought the amp secondhand. I received it, I've powered it on. For about 2-5 minutes it was really good (I've used it at about 2/10th of the max volume)! After some minutes the right channel became quiet and distorted.
So I disassembled it.

I've figured out that the pre-amp is working.

The symptom is that F502 and F501 fuses are always blowing up when I turn the power on.
I already changed the big 50V 6800uF caps to 63V 6800uF caps, and I've changed the right channels main transistors (Q416 and Q418), but no dice, it still blowing the aforementioned fuses.

The thing is I just noticed that C505 cap has been changed before (oh, and I blew it up accidentally, this is how I noticed  ), so now I'm sure the seller was not honest when he said it was never repaired.

Please help me repair this Amp! Tell me please where should I look and what should I check!

Again, I'm really an amateur in amps and electronics, but I'm willing to learn, and this NAD3225PE is really a good base for learning, because it is really simple desing.

Thank you in advance!