Yamaha P4500 / P3200 Power Amp | Distortion

[Sloan]

I'm working on a Yamaha P4500 power amp here with distortion on 'Channel B' and I'm trying to track down the source. There is no DC on the speaker outs.

I've been working on this for a few hours and have isolated the problem to the power section of 'Channel B'. I did this by simply swapping out the power channels to see if the problem stayed on 'B' or not - it followed the actual section/module.


These components test good with a multimeter out of circuit;

• All outputs - C3856's & A1492's
• Other transistors alongside outputs - C4793, D8LD40, A1859A, and FMX22S

All diodes, zeners, transistors, and resistors test OK in circuit. Capacitors look good.


There are transistors on the board which I can not find any information on, would like to figure them out. c3790, c3468 etc..




Yamaha P4500 / P3200 Schematic
> http://schnake99.sc.funpic.de/schema...00%20P4500.pdf

[Sloan]

Here is the schematic as an attachment.

[doctor]

Here is the schematic as an attachment.

Datasheets of aforementioned transistors are attached.

2SC3468 may have two package options.

Good luck

[Enzo]

Don't focus on "bad part" too hard. A broken solder connection can be just as likely, if not more. Or an open resistor. Back when ther were current, I worked on a ton of Yamaha power amps, and I found a number of consistent solder failures, notable on the small, single-transistor heatsink parts in the middle of the boards.


And as Doctor points out, all thos transistor type numbers have an understood "2S" in front of the number.


You should apply a sine wave signal and scope the result. The nature of the distortion is often a clue. For example if half the waveform is missing that tells us where to look.

[Sloan]

Don't focus on "bad part" too hard. A broken solder connection can be just as likely, if not more. Or an open resistor. Back when ther were current, I worked on a ton of Yamaha power amps, and I found a number of consistent solder failures, notable on the small, single-transistor heatsink parts in the middle of the boards.


And as Doctor points out, all thos transistor type numbers have an understood "2S" in front of the number.


You should apply a sine wave signal and scope the result. The nature of the distortion is often a clue. For example if half the waveform is missing that tells us where to look.

Thanks for the tips; I've scanned over it, but will take a closer look for bad joints. As for diagnostic tools, all I have is a multimeter - no scope or signal generator (on my to do list!).

I've only been repairing stuff for about a year now, so I'm still very green, especially knowing what transistors I can substitute. Is there a non NTE transistor cross reference somewhere? All my searching has failed to turned up anything.

Thanks again. I will report back after digging deeper into this thing.

[Enzo]

The internet. As long as you realize that C3856 refers to a 2SC3856 transistor and a A1492 refers to a 2SA1492, then whenever you wonder what a part is, just enter teh part number into Google, and up will come a number of places you can download a data sheet for the part.

Places that sell parts usually have data sheet links on their parts pages. For example, I just checked MCM Electronics. They have the 2SA1492 listed. Right on the first page it says this:
2SA1492 - SANKEN TRANSISTOR MT100 -180V -15A 130W BCE

And that pretty much tells me what I need to know about it. Oh and it also says they have them in stock for $2.99 each.

I usually check Mouser first, but for Asian type numbers MCM will have most, and I like B&D Enterprises.


I don;t know, there probably is some sort of cross reference of suggested subs, but really, there are so many types... I can usually just get the actual type. Making subs is just a matter of experience.

A scope is a very powerful tool. Even just playing music through the amp, my scope will show me that half my waveform is missing, or that I am clipping severely or whatever,.

[Sloan]

Thanks for the reply. I usually start with a Mouser search then I'll try Allied and then usually google. I added MCM to my bookmarks, looks like they have some of that stuff I couldn't find.

[Jazz P Bass]

Find Chips.com
Queries all the major houses.

[Enzo]

I look at parts houses in two catagories. To me there are general electronic parts suppliers - Mouser, Allied, Digikey, Newark - who sell common parts. In the USA, they sell largely what I consider domestic parts. I think of them as where engineers might look for parts.

Then there are suppliers of what I consider consumer electronics parts. Parts for TVs, stereos, home theater, home security, or anything Best Buy sells. And that is where most of the Asian semiconductor types seem to come from. That is suppliers like MCM and B&D. These are places where the local TV repair shops go.

And there are smaller houses like Jameco, which to me is aimed more at the hobbyist or serious amateur. Maybe "project electronics" is their catagory.

DO yourself a favor, any place you might do business with, request their paper catalog. If you want to look for something specific, a 2SC1815 transistor or a 1N4744 zener, or a 100uf 35v capacitor, then online is great. But it is hard to browse the internet. If you leaf page by page through the MCM catalog you will find things you never knew existed. Parts, tools, supplies, etc. And if I want to replace some speaker relay in a power amp, it needs to be the same shape and footprint. I can flip through the relay pages in a catalog and see what I am looking for a lot faster than downloading a million potential relays online.

And what if you need those plastic things the circuit board snaps onto? DOn;t know what to call them, how to search? Flip through the pages of the paper catalog in the hardware section adn you will spot them. AHA!! they are called "standoffs."

[Sloan]

Checked all resistors for 'open' in circuit - OK


Checked all small transistors out of circuit:
Q103 - OK
Q104 - OK
Q105 - OK
Q106 - OK
Q107 - OK
Q108 - OK
Q109 - OK
Q110 - OK
Q111 - OK
Q112 - OK
Q133 - OK
Q134 - OK


I did not see any bad joints when looking closely. After checking all of this, I hooked it back up and played audio through it and just poked around. No changes; still badly distorted.

I'm wondering if the capacitors would cause distortion, but not really sure how I can check that other than just replacing them.

[Enzo]

We really need to know the nature of the distortion.

[Sloan]

I know it's hard to explain. It's not like what I would call clipping distortion where you can still tell what the source sound is supposed to sound like - the whole signal is distorted and almost unintelligible. I was running some NPR talk through it and it sounds alien.

[Enzo]

I was being indirect. What I meant was, we really need to put a scope on it to see the way the waveform is being distorted. Explaining how it sounds is not going to help much, but thanks. Clipping is different from one side of the waveform missing which is different from some DC problem somewher which leaves only the peaks remaining, and so on.

[Sloan]

I was being indirect. What I meant was, we really need to put a scope on it to see the way the waveform is being distorted. Explaining how it sounds is not going to help much, but thanks. Clipping is different from one side of the waveform missing which is different from some DC problem somewher which leaves only the peaks remaining, and so on.

That's what I figured.
I'm also a musician which unfortunately also means 'broke', so I haven't been able to pick up a scope yet.

I'm not seeing any 'test points' on the schematic, do you have any idea where I should check some voltages? Also, perhaps a way to verify the capacitors are ok?

It's got me stumped right now since everything 'seems' to be correct via multimeter. I can switch out both pcb/output/heatsink 'modules' and the distortion definitely follows this one. It's further frustrating because I can test components on both 'modules' and not find any differences.

Thanks a ton for helping out so far.

[Steve Conner]

Well, if you only have a multimeter, all you can do is compare the DC voltages between the faulty channel and the good one. The two channels are supposedly identical, so the voltages should be the same at every point, too. Any large difference is a clue that you can follow.

This seems to be a big, nasty Class-H amp with two sets of rails, possibly not the friendliest project for a newbie.