Transistor substitute in vintage receiver

[mgl99]

I have been researching a substitute transistor in the preamp circuit of a Marantz 2270 vintage receiver.  The mfg. specified transistor is 2SA493, which appears to be no longer available.  Matching the transistor characteristics from the data sheet, I have located the nearest match to be a 2SA841.  The ft of the 2SA493 specifies 100MHz and Cc of 10 pF.  The 2SA841 has a ft of 70 MHz and Cc of 9 pF.  All other data points of the 2SA481 match up well with the 2SA493. 

My question is whether this is a viable substitution?

Thanks

[DC1MC]

Try getting the 2SA1015 or BC327-25, they are around 12c a piece and you can sort them for the similar gain.

 Cheers,
 DC1MC

[mgl99]

Thank you for the reply.  I am still learning as this is an area that I am now venturing into.  I did look into the two transistors that you recommended, however I don't understand why these are better substitutions than others.  If you can explain the reasoning, it will certainly help me learn what is important and what isn't when finding a substitute transistor.  I understand transistor theory and function, but I don't know the impact of every data point and the values specified on the transistor data sheet.

Thank you for the guidance.

[DC1MC]

The transistors are not the "black-boxes" that simulations made them be, especially for audio stuff.
In replacing transistors you have first in mind the amplification factor hFe, that have to be similar as much as possible to the original ones to avoid re-biasing.
The next important factor it's the noise factor, here are holly wars fought in the audio groups, because frankly measuring in a correct and repeatable way the noise factor it's a difficult enterprise and having measured in the same biasing condition as the destination circuit is a double hassle so subjectivism is showing up.
The cutoff frequency is also an important factor because it matters at high amplification factors, as it affects the linearity of the frequency response or worse, if part of a magnetic/crystal phono preamp, it can totally distort the response curve, that was calibrated for a specific cutoff frequency and capacities, this why the audiofreaks prefer replacing with the same transistors as the original ones to preserve the response curve, no matter how overpriced they are, because they "sound better".
And finally, the first option that I've suggested is widley used in this model and seem to sound good if you match the hFe of replacements with the originals, and the second option I used myself.

 As usual, YMMV.
 
 Cheers,
 DC1MC

[mgl99]

Thank you again for helping me.  I am still a little confused.  In your reply, you made this statement: "In replacing transistors you have first in mind the amplification factor hFe, that have to be similar as much as possible to the original ones to avoid re-biasing."

With respect to that, the 2SA1015 transistors (all of which I looked at) all have hFE MIN values that are not the same as the 2SA493, which has an hFE MIN of 200.

The 2SA841 does have an hFE MIN value of 200, which is exactly the same as the transistor I am trying to find a substitute for (2SA493).

Why would the 2SA1015 then be a better choice then the 2SA841?

Thanks

Thanks

[DC1MC]

Have a look here, I have no personal experience with 2SA1015, this is one of few forums indicating that they may be a good replacement:

http://audiokarma.org/forums/index.php?threads/2sa493-transistor-replacement.485887/

Also, it seems that there are 2 classes of 2SA493, sorted according to the hFE, Y or GR, this is important, what models are yours ?

On the BC237 I can guarantee that they will work, I've used some years ago and they were perfect, I even had a device to pair them and some kind of very sensitive amplifier to listen to the nose with static biasing, good times

In the end use what you see fit for your device, original Maranz PCBs are tough, with a bit of care you can test 2-3 pairs without damage, or even put little sockets made of Augat pins, I do remember doing this for a very picky guy that have listened to something like 8 pairs, until he said OK, and then I've removed the pins and soldered them permanently.

 Good luck,
 DC1MC
 

[David Hess]

I did look into the two transistors that you recommended, however I don't understand why these are better substitutions than others.  If you can explain the reasoning, it will certainly help me learn what is important and what isn't when finding a substitute transistor.  I understand transistor theory and function, but I don't know the impact of every data point and the values specified on the transistor data sheet.

That depends on how the transistor is being applied so I would need to see the schematics or have a very detailed description of the circuit.  I took a quick look at the Marantz 2270 schematics and did not find any 2SA493s.  If it is used as part of a differential amplifier, then both transistors should be changed for best performance.

For small signal audio amplifiers in low voltage gain circuits, (1) there is not much variation between different transistors as long as they have the needed ratings and are roughly the same size.  A modern replacement for the 2SA493 is the BC560 but watch out for the pinout which is different.

2SA493: (2-5B package ECB Pinout)
Vceo 50V   Ic 50mA   Pc 200mW
hfe 120 to 400 at 6V and 2mA
Noise Specified
Ft 100MHz Unverified
Cc 10 pF Unverified

BC560: (TO-92 package CBE Pinout)
Vceo 45V   Ic 100mA   Pc 500mW
hfe 110 to 800 at 5V and 2mA
Noise Specified
Ft 150MHz typ at 5V and 10mA
Cob 6pF max at 10V

I did not find detailed specifications for the 2SA493 but here is what I am looking at.  The maximum collector current specification is not too much larger or smaller, both are specified for noise meaning they are intended for low noise applications, and the hfe specifications are both given at 2mA meaning they are intended to operate at that collector current.

Another one which would be good is the 2N5086/2N5087:

2N5086/2N5087: (TO-92 EBC Pinout)
Vceo 50V   Ic 100mA   Pc 625mW
hfe 150/250 min at 5V and 1mA (specified at 100uA, 1mA, and 10mA)
Noise Specified
Ft 40MHz min at 5V and 0.5mA
Cob 4pF max at 5V

I disagree with others here about the importance of transistor hfe.  It is not likely to matter in this case but I would need to see the schematic to be sure.

(1) The output capacitance and reverse transfer capacitance become important in high voltage gain circuits like the VAS stage in an audio amplifier.

[mgl99]

I checked the original 2SA493 transistors on the board and there is no G following the A493 so I have made the (possibly wrong) assumption that it is a 2SA493.

In this particular case, my solution (as of an hour ago), was to buy a replacement board off of ebay, guaranteed to be good.  There are 8 transistors on this board of two different types.  4 of them being the 2SA493 and 4 of them being 2SC1000.  Since I have other Marantz receivers to restore, I thought this would be a good bet for only $30 including shipping.

I'm sure I don't know near enough yet, but thanks to people like you, I will learn.

Thanks again

[mgl99]

The 2SA493 transistors are used on the p400 board which is the tone amplifier board (preamplifier).  The board serves both channels and feeds the main amp boards. The Marantz designations are H403, 404, 407, and 408.

[David Hess]

With your description I managed to find them in the schematic.  They are all operating as the output transistor of a composite pair with a collector current of about 1.5mA so nothing special is required.  Any 35V or higher and 50 to 200mA cheap transistor like a 2N3906 would work fine as well as any of the above suggestions.  Hfe is irrelevant and I doubt Ft and/or capacitance matters much.

My guess is that Marantz was grading 2SC493s for some premium product and the extras went here.  Or maybe they just got them inexpensively because they were buying so many so they were the cheapest option.  As Sigmond Freud said, sometime a transistor is just a transistor.

[mgl99]

The problem I'm trying to repair is that when in operation, I am getting spurious load static in the left channel.  The static is independent of the volume control setting.  The right channel is clear and consistent.  When I swap (criss cross) the preamp to amp jumpers the problem goes to the right channel.  This told me that the problem was probably in the preamp.  A guy that I talked to, suggested that it sounded like a leaking transistor on the preamp board.  This remains to be seen.  I ordered a transistor checker because my multimeter doesn't have a diode checking capability.  It's an Ideal multimeter and good quality.  I wanted this one because it checks capacitors.  I bought it years ago, but it serves me well.  When I get the transistor checker, hopefully I can narrow down if it is a transistor.  There's not a lot on that board.  I suppose it could be an electrolytic cap, but I haven't pulled the board down yet to check.  I believe it's a transistor until I find out otherwise.  If it is, I will have a replacement due to buying this replacement board.

[duak]

I encountered a leaking transistor in a power amp once.  I can't recall how I figured it out - I probably used a 'scope and saw that the transistor's collector voltage was fluctuating randomly at low (<10 Hz) frequencies.  I sort of recall removing it from the PCB and applying 20 VDC between collector and emitter thru a 10K resistor to limit current and measuring the current.  If it is above, say 10 uA, or it fluctuating, the transistor is leaking.  (It could also be a germanium transistor, but that's unlikely.)

Best o' luck,

[David Hess]

Transistors sometimes fail producing "popcorn" noise which is exactly what it sounds like.  Below is an example of what popcorn noise from a bad transistor looks like on an oscilloscope.  Capacitors can become "staticy" and I bet resistors can also.  Cracked solder connections can do this also.

A transistor gain tester and a forward voltage drop check is unlikely to reveal a noisy transistor but of course low gain might correlate with intermittent noise.  A transistor curve tracer will show if the transistor has become weird.  I had like 20 2N3565s in the same instrument built in the 1970s where a couple had failed but more than half were "weak" with a soft knee, poor saturation, and low Vce.  I ended up just replacing them all whether they tested bad or not.

In your case I might start by using a plastic tool to push and tap on various places on the board to see if the noise can be localized, change all of the transistors for 2N3904s or 2N3906s or your favorite small signal transistors, and then relow all of the solder connections on the board with flux.

[mzacharias]

Sometimes freeze-mist (or hot air for that matter) can help locate the defective transistor. Also, I have seen a number of transistors that fail this way seem to have a black oxide, or tarnish on the transistor lead. Possibly this penetrates the semiconductor material itself.

Examples include the 2SC458 and 2SC1345 sometimes found in old audio equipment.