Hi All,

New to the forum and could use some advice. Repairing our Denon home AV receiver and tracked the problem down to a shorted junction in one of the Sanken 2SB1647 output transistors. In my search for a replacement, I notice that some places sell them in "matched" pairs and was wondering if it's necessary to replace both for this channel using a matched pair set or can I just replace the shorted one? Here's a link to the data sheet, thanks for any assistance!

http://www.semicon.sanken-ele.co.jp/sk_content/2sb1647_ds_en.pdf

 

I've successfully replaced an output transistor not only unmatched, but with a completely different part number. I was unable to detect a difference between left and right. But, then again, I am not an audiophile.

 

Thanks for the reply Joey, appreciate it! I'm not an audiophile either, and doubt I could detect any difference. My main concern was that it may cause an issue with shortened component life, but it doesn't sound like it's a cause for concern. My local supply store said they carry the NTE 2650, which I assume is equivalent, and using that would be similar to what you did. But just to be on the safe side, I think I might stick with the Sanken.

 

Are both transistors in the pair the same part number?
If so, I would replace both at the same time.

 

Yes, they both have B1647 printed and below that is "OO P" which is the same on all the matched pairs (there are 14 transistors total). When I saw the listing for the replacement matched pair, the second set of characters on each was the same (but different from these), which is what made me wonder if they should both be replaced together? I'd initially assumed it was a date code but am not sure now.

 

The data sheet is for a high power PNP DARLINGTON transistor. I doubt they are matched. But, A good question to ask is, are there big fat emitter resistors for each transistor?

 

Here's a pic of the resistors. The design is such that the heat sink with the 2SB1647 chips are directly under the PC board and you can see them tucked into the space between with their leads sticking up and soldered to the edge of the PC board.

 

After measuring the size and calculating the resistance, I believe the larger (aqua) resistors are a half watt .46 ohm and the smaller (brown) is a quarter watt 22 ohm (I'm a bit colorblind so any confirmation would be great!). So I'm not sure if that qualifies Les? I need to replace these for this channel as they read infinite but no signs of overheating.

 

I say go forth and replace.

Also I looked into the NTE replacements. They do not list a matched pair of PNP transistors, they do list a complimentary pair of PNP and NPN transistors. The word "matched" is not mentioned.

 

Thanks for the reply Les! It took me a bit to get the parts straightened out as the 1/2 watt resistors they sent me were the same size as the 1/4 watt and I thought it was a mistake on their part. After some checking, seems they're making them smaller but the same wattage.

So I swapped the resistors and the B1647 yesterday, plugged it in, and unfortunately it toasted the new 22 ohm resistor that's connected to the base of the B1647! Very disappointing. I thought both transistors in each channel were B1647's but after what happened yesterday, I separated the heat sink to take a closer look and it seems the 1647 is paired with a D2560 for each channel and the small one in the middle is a C3964. I measured the junction voltages and they all seem OK, but am tempted to replace all 3 and see what happens? Any input would be greatly appreciated!

 

Ed
If you can get your hands on a variable transformer (auto-transformer) your trouble shooting would become less exciting, less smoke filled, but more productive. Once you have made a change/replacement then slowly increase the line voltage while monitoring the temperature of the parts and output voltage.
Sketching up a schematic will help also.

 

Thanks Les. I was thinking the same thing about sketching up the diagram. I have to believe this is a pretty common audio circuit. I haven't bee a tech for a few decades, so this stuff is very rusty for me (as well as my tool selection limited). Appreciate all the input. This is a great receiver and I hate to just toss it...

 

I think I found an alternative to your auto transformer suggestion Les. Read a tip on a AV repair forum that says to use a 60 watt bulb in series in the power cord to the amp before turning it on to test. If the bulb lights up when the amp is powered on, it indicates an issue still exists and limits the current to prevent the exciting smoke show. If it glows dimly, indicates that all is well. Seem reasonable?

 

ED,

What model a Denon is it? I have a 1612 that keeps giving me a red flash power button each time I turn it on from a cold start. After turning it off and on a couple of times, it finally starts. I'd be interested to hear what model you have and how you diagnosed it.

 

Gopher,

My unit is the AVR 3311CI. I started by searching the blinking red light safety mode and was able to see a few repair videos that were quite detailed. Unfortunately the models they repaired weren't like mine but it was enough to get me started. Seems the most common issue is the +15V regulator. Give this a watch, it covers the Deono AVR-788 troubleshooting, but lots of good info you can use. I like what he says at the end about keeping these units out of the landfill, that's my goal on this one. Good luck and post up if you make any headway.

 

Gopher,

My unit is the AVR 3311CI. I started by searching the blinking red light safety mode and was able to see a few repair videos that were quite detailed. Unfortunately the models they repaired weren't like mine but it was enough to get me started. Seems the most common issue is the +15V regulator. Give this a watch, it covers the Deono AVR-788 troubleshooting, but lots of good info you can use. I like what he says at the end about keeping these units out of the landfill, that's my goal on this one. Good luck and post up if you make any headway.

Thanks. I'll wait until mine becomes a brick. It only flashes red if it sits 12 hours. Kind of hard to fix if not consistently bad.

By the way, the desoldering technique used at 10:30 point in the video is a terrible idea. Besides flicking solder all over yourself, he is potentially flicking it to the nether regions of the board and possibly causing some shorts. Some good solder wick or a solder sucker is a much better idea.

 

How reliable is testing the junctions on these output transistors with a DVM on the diode setting to see if the transistor is good overall?