Different resistor Question - Emitter resistor choice

C

cbrworm

Well-Known Member
I am working on reviving a middle aged amp (it's 22), I am trying to do this more as a chance to work on something newer than what I am used to. The owner of it has given up after a few short lived failures in the amplifier, but I want to finish it. It appears that the bad channel has had a lot of work done on it by a TV repair guy (no offense - it actually was a TV guy). The outputs have been replaced by NTE EGC36/37 replacements on the bad side. The negative driver label wiped off when I went over it with a paper towel with some solvent on it. I found some problems in the preamp for that same channel which I have addressed - I suspect they were contributing to the repeated left channel amp failure - as well as replacement parts that probably were not up to driving the 4 ohm speakers that were being used in the manner of the originals.

It is an Onkyo Integra integrated amplifier that seems to have been designed with a goal of pretty good sound quality - I would like to revive it without going crazy - and I am having a hard time finding a number of parts that match the original spec.

Anyway - onto the question of this post - in the output stage it has 2 pairs of outputs per channel (Sanken 2sa1186/2sc2837) being driven by +/- 57 volts. Each output has a .47 ohm 2 watt metal plate resistor as the emitter resistor. One of these has opened up corrosponding to it's output transistor (a non original part) which apparently shorted.

I am not having any luck finding this type of this resistor in the form factor of the original. If I replaced all the emittter resistors with identical metal film resistors of the .47ohm 2 watt variety would it likely be fine? Since this is directly in the audio path, I don't want to make a mistake here.

The original part is an MPC78 which is listed as a Metal Plate resistor which is a "low distortion factor product for acoustic products." It also says it can handle 10x it's rated power for 5 seconds and has a 10% tolerance.

It seems that if the ratings are the same it should function the same, but at the same time it seems that a metal film may not behave the same under load as a flat ribbon of metal that the original design called for. And the original is specified as a "low distortion. . ."

Seems like usually amps have cemented wire wound resistors, but I am a little worried about putting antennas/inductors in this spot unless it is ok.

Thanks!
 
Last edited:
Register to hide this ad
I wouldn't sweat trying to find the "low distortion factor designed for audio" resistors. Just get a 2~3 watt metal film or wirewound type. As long as those ECG/NTE aren't making the amp oscillate, the resistors will be fine.
 
First, GET RID of those NTE replacements and install the Sankens. Change any other replacement devices back to their originals.

Those output resistors are ubiquitous in Japanese audio gear. Buy something at Goodwill, or a garage sale, and they will be in there. Common values are .22, .33 and .47 ohms. Besides the manufacturer, you used to be able to get them from places like MAT Electronics, MCM, Dalbani, etc. Do not use a wirewound part there. That changes the impedance of the zobel. You could use a metal film or evaporated metal resistor there, but I would much rather see you use the original part.
 
PacificStereo said:
First, GET RID of those NTE replacements and install the Sankens. Change any other replacement devices back to their originals <snip>
Click to expand...

I have to agree - you can still get many of the originals - resorting to NTE should be absolutely a last resort.

Those resistors are a bit difficult - I tend to agree with Jon L regarding the resistors, although I would try to steer clear of wirewound types if possible. I have found T0126 case, 0.47ohm non-inductive resistors - intended to be surface mounted and rated at 30w - but these are hardly ideal.

OR, maybe someone else will chime in with a source for non-inductive 0.47ohm, 2w or 3w resistors of a more conventional design, with other characteristics (that you describe) to match the originals.
 
Last edited:
cbrworm said:
in the output stage it has 2 pairs of outputs per channel (Sanken 2sa1186/2sc2837) being driven by +/- 57 volts. Each output has a .47 ohm 2 watt metal plate resistor as the emitter resistor. One of these has opened up corrosponding to it's output transistor (a non original part) which apparently shorted.

I am not having any luck finding this type of this resistor in the form factor of the original. If I replaced all the emittter resistors with identical metal film resistors of the .47ohm 2 watt variety would it likely be fine? Since this is directly in the audio path, I don't want to make a mistake here.

The original part is an MPC78 which is listed as a Metal Plate resistor which is a "low distortion factor product for acoustic products." It also says it can handle 10x it's rated power for 5 seconds and has a 10% tolerance.
Click to expand...

By chance the original emitter resistors can handle 10X their rated power on short time intervals, because a 0.47 Ohm resistor will dissipate more than 2W if the amp is driven at high volume.

In my opinion, a (true) 2W resistor will fail. I would put 5W resistors here, particularly if you intend to use low impedance speakers.
 
Per the data sheet for the Panasonic part, it can tolerate a short term overload voltage of 1000V. I think it can tolerate the full output of the amp (56V) on peaks. If he is going to run the amp with music at full power (average) then the amp will be clipping on the peaks and the speakers or the output stage will be damaged before the resistors open up. I don't think he will be listening under those conditions.
http://industrial.panasonic.com/www-data/pdf/AOA0000/AOA0000CE18.pdf
 
dr*audio said:
Per the data sheet for the Panasonic part, it can tolerate a short term overload voltage of 1000V. http://industrial.panasonic.com/www-data/pdf/AOA0000/AOA0000CE18.pdf
Click to expand...

Warren,

Read notes 1, 2 and 3... You will see they can't sustain 1000V (not a 0.47 Ohm /2W)

And what is short term? 1% of the time? 0.1% of the time?

I've seen enough open emitter resistors in my life to know that a 0.5 Ohm /2W is just good enough for a 20W / channel amplifier, unless it is bypassed by a diode.
 
The voltage you get from the equation in the notes is so much lower than the ones in the chart above that it makes no sense, and I can't tell which to believe. However, I have seen 2W emitter resistors used many times in designs like this and have never seen one open unless the outputs are blown. Ok, let's look at the math. Lets say the rails are + / - 56V. Fine. Let's say the outputs can swing to within 5V of those rails, so we have 102V P-P. Let's assume a 90% efficiency of the output stage, so that's 91.8V P-P applied to an 8 ohm load. Convert to RMS: (91.8 / 2) / 1.414 = 32.46VRMS into an 8 Ohm load. Divide by 8 ohms to get the RMS current = 4.058 A. Now let's assume the amp has 4 transistors in the output stage, which most Integras do. So you divide that current by 4 to get the currrent through each emitter resistor. 1.014 A. P = IxIxR, so I get .48W through each emitter resister at full output.
 
I am curious if these emitter resistors are also being used as a fuse. I have ordered some 5 watt resistors that are the same MFG and package/style as original. I hope that the design of the amp doesn't rely on the resistance going up as the temperature of the resistor goes up - offering both some current limiting for the transistor and a fuse for shorts.

I do have some of the flameproof 2 watt resistors here, but after looking at the specs I suspect that they will cook pretty quickly under load.

This amplifier, should it work again, is used to drive a 4 ohm load full time.

I will post another thread about trying to find a replacement for the Toshiba 2SA1306 that isn't fake like I am 90% sure the ones I received are.
 
ecluser said:
By chance the original emitter resistors can handle 10X their rated power on short time intervals, because a 0.47 Ohm resistor will dissipate more than 2W if the amp is driven at high volume.

In my opinion, a (true) 2W resistor will fail. I would put 5W resistors here, particularly if you intend to use low impedance speakers.
Click to expand...


Don't forget that the duty cycle of an emitter resistor is only 50% in a complementary output stage (each output transistor is only on half the time). So a 2 watt .47ohm resistor can handle continuous operation at 60W RMS into 8 ohms for the amplifier, and be at it's rated specification. If there are 4 transistors in the output, then the power can double as each emitter resistor will be carrying one half the total load. So call it 120 watts into 8 ohms with 4 output transistors (two NPN, two PNP).

Of course, that depends on an honest power rating by the resistor manufacturer, and even then the lifetime of the part will be shortened somewhat. Still, the part will last a long time at it's rated power dissipation if the ventilation is adequate.

Power dissipation in the resistor will increase with a lower impedance load at a given output from the amp. As ecluser stated, if you intend to run it with 4 ohm loads, exceeding about 60 watts continuous into the load will run the 2 watt resistor at its rated power. Given the modest unloaded voltages of the power supply, exceeding this power output seems unlikely (but I have been wrong before).
:smoke:

Vance
 
Last edited:
This is the purely idealistic situation of a purely resistive load with idealy matched transistors. In real situations you should multiply the power rating by ~2 if you want to play safe.

In my opinion it is stupid to risk a set of transistors costing nearly 20$ to save 50¢ on emitter resistors.
 
The problem wasn't the money the proper resistors cost, but my inability to find exact replacements.

I was able to find very similar 2 watt metal plate resistors (the duals recommended by PacificStereo), replaced them all as a set (only using one side) and everything is working great.

It turned out that the original problem was the thermistor for the bias on that channel - it would occasionally open and cause bias current to spike - luckily I was wary and monitored it for a while to see it happen. After replacing the thermistor as well as everything from the drivers down, the amp has been working great.

There was quite a bit of damage.

Also a bunch of open resistors on the preamp side - I don't know what happened - it looked like maybe some current was connected to the chassis - possibly external, but it is fixed now.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom