response
Through contolled listening tests. I have done alot of testing with SS devices in critical areas, R and D for 47 years. SS devices have a nack of changing the sound.
One of the big topics being currently experimented on by some, and in fairly wide discussion is ridding SS devices used for muting purposes in CD players. As you know, when music is playing, the SS devices (may I use the generic terms 'transistors') are turned completely off. Yet, they still, stubbornly, affect the sound.
Through contolled listening tests. I have done alot of testing with SS devices in critical areas, R and D for 47 years. SS devices have a nack of changing the sound.
One of the big topics being currently experimented on by some, and in fairly wide discussion is ridding SS devices used for muting purposes in CD players. As you know, when music is playing, the SS devices (may I use the generic terms 'transistors') are turned completely off. Yet, they still, stubbornly, affect the sound.
Hi Steve,
To comment on an O.T. post ...
Muting transistors are not completely off. The capacitance changes as do the leakage currents. The 2SC2878 is an example of a transistor specially designed for muting purposes. There are a couple newer ones, but the number escapes me at the moment. Normal transistors exhibit even worse performance in muting positions.
So, yes. You are further ahead to use a relay contact to short the signal to ground for muting. The same goes for the older players that used transistors to change the EQ. Newer ones do that in the DSP chip.
Now back to the topic at hand.
-Chris
To comment on an O.T. post ...
Muting transistors are not completely off. The capacitance changes as do the leakage currents. The 2SC2878 is an example of a transistor specially designed for muting purposes. There are a couple newer ones, but the number escapes me at the moment. Normal transistors exhibit even worse performance in muting positions.
So, yes. You are further ahead to use a relay contact to short the signal to ground for muting. The same goes for the older players that used transistors to change the EQ. Newer ones do that in the DSP chip.
Now back to the topic at hand.
-Chris
response
"Muting transistors are not completely off. The capacitance changes as do the leakage currents. The 2SC2878 is an example of a transistor specially designed for muting purposes. There are a couple newer ones, but the number escapes me at the moment. Normal transistors exhibit even worse performance in muting positions."
Yes, there is a very slight leakage current. There is also virtually no DC present, maybe mv to uv dc so I would suspect the base to collector leakage to predominate. The Z of the transistors should be extremely high.
Any ideas of the value of Z?
Nice point about the capacitance Anatech.
"Muting transistors are not completely off. The capacitance changes as do the leakage currents. The 2SC2878 is an example of a transistor specially designed for muting purposes. There are a couple newer ones, but the number escapes me at the moment. Normal transistors exhibit even worse performance in muting positions."
Yes, there is a very slight leakage current. There is also virtually no DC present, maybe mv to uv dc so I would suspect the base to collector leakage to predominate. The Z of the transistors should be extremely high.
Any ideas of the value of Z?
Nice point about the capacitance Anatech.
Hi Steve,
Now, consider an AC voltage and it's peaks. This circuit node is normally a reasonably high impedance compared to a DC circuit. It is therefore very sensitive to any leakage currents. Sometime you can see the signal being clipped ever so slightly using an oscilloscope.
When testing these I used a full scale audio sine wave recorded on a test CD. This allowed me to catch these leaky devices. You can test them out of circuit and you may possibly catch them using a curve tracer. My leakage tests would not show them at all unless they were really bad. This would be less than 5 uA. I'm guessing now that the test voltage may not have been high enough. Anyway, substitution was really the only reliable test to prove the transistor was bad and not some other component. This was also possible.
-Chris
The base will be held at a high DC potential. Typically 8 ~ 12 VDC in reverse bias. This is well beyond the normal breakdown of 5 ~ 6 VDC typical of most transistors. That's why the 2SC2878 is special. It was developed to withstand very high reverse bias base to emitter.There is virtually no DC present, maybe mv to uv dc. The Z of the transistors should be extremely high.
Now, consider an AC voltage and it's peaks. This circuit node is normally a reasonably high impedance compared to a DC circuit. It is therefore very sensitive to any leakage currents. Sometime you can see the signal being clipped ever so slightly using an oscilloscope.
When testing these I used a full scale audio sine wave recorded on a test CD. This allowed me to catch these leaky devices. You can test them out of circuit and you may possibly catch them using a curve tracer. My leakage tests would not show them at all unless they were really bad. This would be less than 5 uA. I'm guessing now that the test voltage may not have been high enough. Anyway, substitution was really the only reliable test to prove the transistor was bad and not some other component. This was also possible.
No, I was never interested in the magnitude. It was variable with the peak applied voltage. The principle is probably similar to a varicap diode (with no leakage currents mixed in).Any ideas how high Anatech?
-Chris
Z
"Now, consider an AC voltage and it's peaks. This circuit node is normally a reasonably high impedance compared to a DC circuit. It is therefore very sensitive to any leakage currents. Sometime you can see the signal being clipped ever so slightly using an oscilloscope."
I would estimate that the original design was in the megohms, much higher than the output Z.
Any measurements of the capacitance of these devices? (I know it varies with voltage). Any very loose ball park value, or variability from max to min?
"Now, consider an AC voltage and it's peaks. This circuit node is normally a reasonably high impedance compared to a DC circuit. It is therefore very sensitive to any leakage currents. Sometime you can see the signal being clipped ever so slightly using an oscilloscope."
I would estimate that the original design was in the megohms, much higher than the output Z.
Any measurements of the capacitance of these devices? (I know it varies with voltage). Any very loose ball park value, or variability from max to min?
Hi Steve,
It could be measured with my HP 4263A with external bias used. I don't have any devices to test though.
I don't think the actual capacitance is the biggest problem. It's probably the variable leakage of the transistor. These parts are misused as far as I am concerned. They even had to develop a transistor that would survive the application. That should have told them they were barking up the wrong tree.
A 4066, or better still a 4966 quad switch might be better in this application. The problem is that all semiconductor switches show higher distortion at lower impedances. The 4966 is a 15 volt device that was designed for audio use, the pin out is the same.
-Chris
It could be measured with my HP 4263A with external bias used. I don't have any devices to test though.
I don't think the actual capacitance is the biggest problem. It's probably the variable leakage of the transistor. These parts are misused as far as I am concerned. They even had to develop a transistor that would survive the application. That should have told them they were barking up the wrong tree.
A 4066, or better still a 4966 quad switch might be better in this application. The problem is that all semiconductor switches show higher distortion at lower impedances. The 4966 is a 15 volt device that was designed for audio use, the pin out is the same.
-Chris
Hi Steve,
What I think I'll do is to remove our discussion on muting to it's own, new thread. I will leave it in this forum although I am tempted to put it into "Solid State". Normally you see this in solid state preamps.
-Chris
Yes, for retro fitting a design, that's what you would do.A proposed solution that seem to work well from what I hear is to use a relay and clip out the devices altogether.
Well, for clipping this is the mechanism. The base circuit is supplied through 10 K resistors or higher often. This does lead to your next comment .....I think there is more to it than just impedance change or clipping.
Yes, when both are damaged, or one badly leaking, there is leakage. It sounds odd when it does occur. Never mind the distortion. This would be an odd occurrence I would think.Transistors are also notorious for 'bleedthru'.
What I think I'll do is to remove our discussion on muting to it's own, new thread. I will leave it in this forum although I am tempted to put it into "Solid State". Normally you see this in solid state preamps.
-Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Tubes / Valves
- Muting line level signals