Yes....Bigun may be right, this jump of bias looks to be a result of oscillation
Hook the scope and watch the output signal..... search of signal in the output without injecting signal in the input.
It may be a low level oscillation..... 1.5 volts peak to peak high frequency maybe.... do it after the bias jump, or increase too much..then watch the output.
It is a rarity, for me to mess with other amplifiers....i do not use to do that...because of Terry, that assembled two of my amplifiers and captured my empathy.... i never give opinions about any other amplifier then mine..this is an exception...the first or second in almost 10 years...but i am already getting ouf of here.
I hate to be nosy .. I dislike them, and so I do not want to be like them.
regards,
Carlos
Hook the scope and watch the output signal..... search of signal in the output without injecting signal in the input.
It may be a low level oscillation..... 1.5 volts peak to peak high frequency maybe.... do it after the bias jump, or increase too much..then watch the output.
It is a rarity, for me to mess with other amplifiers....i do not use to do that...because of Terry, that assembled two of my amplifiers and captured my empathy.... i never give opinions about any other amplifier then mine..this is an exception...the first or second in almost 10 years...but i am already getting ouf of here.
I hate to be nosy .. I dislike them, and so I do not want to be like them.
regards,
Carlos
Last edited:
Does anyone know what the voltage drop should be across R10 and R12? That would at least give me a place to look.
Thanks Carlos. I will hook up the scope and watch. I did try to play a signal through the amp. I think one side of is not turning on because I only get one half of the wave.
Blessings, Terry
Thanks Carlos. I will hook up the scope and watch. I did try to play a signal through the amp. I think one side of is not turning on because I only get one half of the wave.
Blessings, Terry
2SJ118/2SK413 or 2SJ 119/2SK414 have lower drain current but excellent data.
Yeah, I have 8v so obviously something is badly wrong. I tried adjusting the bias to 200ma but now I'm blowing the negative output fuse. Enough grief for one day. I'll try to get back to it tomorrow. I am currently trying to read through the thread on Greg's forum but I fear most of the useful info is in the owners thread.
Blessings, Terry
Blessings, Terry
the CCS resistors have two Vbe voltage drops across them. refer to my postpointing to voltages.
If the currents are very low then the Vbe is also very low, expect ~550mVbe giving ~1.1V across the resistors.
This is a two transistor CCS that has been modified to include an extra Vbe using the diode connected transistor D5.
If the currents are very low then the Vbe is also very low, expect ~550mVbe giving ~1.1V across the resistors.
This is a two transistor CCS that has been modified to include an extra Vbe using the diode connected transistor D5.
Hi Andrew,
This is probably very helpful to someone who knows what the LTP, CCS and Vbe are. To me you are speaking Greek. If you can tell me what part seems to be broken then I can replace it. It may be that I have bad match-ups. I did match hFE on the like transistors, but I have a few different "brands" of transistors and the hFE vary between them. If I knew how hFE affected opperation, perhaps I would be in better shape. For instance, the BD140's by SGS-THOMSON have a much higher hFE than those from STMicroelectronics. I don't know if that is my problem here. I have replaced all the small transistors at least once so far. As best I can tell, there are no shorts or bad joints and I am all but positive that the resistors are all correct.
Thanks, Terry
This is probably very helpful to someone who knows what the LTP, CCS and Vbe are. To me you are speaking Greek. If you can tell me what part seems to be broken then I can replace it. It may be that I have bad match-ups. I did match hFE on the like transistors, but I have a few different "brands" of transistors and the hFE vary between them. If I knew how hFE affected opperation, perhaps I would be in better shape. For instance, the BD140's by SGS-THOMSON have a much higher hFE than those from STMicroelectronics. I don't know if that is my problem here. I have replaced all the small transistors at least once so far. As best I can tell, there are no shorts or bad joints and I am all but positive that the resistors are all correct.
Thanks, Terry
Still4, you are no further forward than when you started 9years ago.
You have confused your self this time around by rushing to build 3 or 4 amplifiers at the same time. That rushing into multiple builds has incorporated avoidable mistakes.
Slow down. Concentrate on one build and LEARN what is happening.
It is no excuse to say it is "Greek" !
You have confused your self this time around by rushing to build 3 or 4 amplifiers at the same time. That rushing into multiple builds has incorporated avoidable mistakes.
Slow down. Concentrate on one build and LEARN what is happening.
It is no excuse to say it is "Greek" !
Member
Joined 2009
Paid Member
Hi Terry,
Let's look at those voltages you posted and see if we can narrow it down a bit more.
The voltage across R10 is too high - clue 1. It says we have too much current flowing through it 8V/1k = 8mA - where is it going ? Well not through the base of Q8 (this wouldn't be normal for Q8 anyhow) since it would then appear through R11. The voltage across R11 allows me to calculate that we have only 0.2mA there. So it's going through Q7 (as it should) and then provides current for the LTP that sits on the negative side of the circuit.
The current should then flow through R7 & R8 into Q3 & Q4. The voltage drop across R7 shows 14mV, so current = voltage / resistance = 14/22mA = 0.6mA which is about right. The voltage drop across R8 is 154mV, so the current is 7mA. Ah, now we know where all that current through R10 is going, it's going into the emitter of Q3. Q3 and Q4 measure the input signal (base of Q4) and the output signal (base Q3) and compare them - this is your feedback error amplifier. It's an LTP - Long Tail Pair. Invented by the Brit's during the war for radar it works very well. This design is a bit complex because it has two LTPs.
This excessive current flow into the LTP will unbalance it and produce a dc-offset at the output. The other LTP, the upper one will try to fight this so it will reduce the amount of dc-offset but can't completely null it out. This is what you are seeing.
But why is the current source flowing too much current in the first place. The finger points at the current source. This current source could be thought of as a simple feedback arrangement. Q8 monitors the current flow through R10 and throttles it back by modifying the voltage at Q7's base. So if Q8 isn't working properly we have a problem. Also, Q8 has D5 in it's emitter circuit which modifies how Q8 measures the current flow through R10. This is to allow the current to be further throttled back if the temperature of D5 increases. So if D5 isn't working properly we also have a problem. I'd look carefully at these parts, maybe replace D5 (a pitty as I notice you did a really neat job of wiring it up). It seems to me there is far too much voltage across D5 and I think this is where the issue lies. Try replacing D5, check it's not backwards or something, or that your version of the BD140 doesn't have a different pinout for some reason.
By the way, since there is a feedback going on in the current source it is possible for the current source to oscillate and in my design I included a small 1nF capacitor between the base and collector of Q8 to slow it down.
Let's look at those voltages you posted and see if we can narrow it down a bit more.
The voltage across R10 is too high - clue 1. It says we have too much current flowing through it 8V/1k = 8mA - where is it going ? Well not through the base of Q8 (this wouldn't be normal for Q8 anyhow) since it would then appear through R11. The voltage across R11 allows me to calculate that we have only 0.2mA there. So it's going through Q7 (as it should) and then provides current for the LTP that sits on the negative side of the circuit.
The current should then flow through R7 & R8 into Q3 & Q4. The voltage drop across R7 shows 14mV, so current = voltage / resistance = 14/22mA = 0.6mA which is about right. The voltage drop across R8 is 154mV, so the current is 7mA. Ah, now we know where all that current through R10 is going, it's going into the emitter of Q3. Q3 and Q4 measure the input signal (base of Q4) and the output signal (base Q3) and compare them - this is your feedback error amplifier. It's an LTP - Long Tail Pair. Invented by the Brit's during the war for radar it works very well. This design is a bit complex because it has two LTPs.
This excessive current flow into the LTP will unbalance it and produce a dc-offset at the output. The other LTP, the upper one will try to fight this so it will reduce the amount of dc-offset but can't completely null it out. This is what you are seeing.
But why is the current source flowing too much current in the first place. The finger points at the current source. This current source could be thought of as a simple feedback arrangement. Q8 monitors the current flow through R10 and throttles it back by modifying the voltage at Q7's base. So if Q8 isn't working properly we have a problem. Also, Q8 has D5 in it's emitter circuit which modifies how Q8 measures the current flow through R10. This is to allow the current to be further throttled back if the temperature of D5 increases. So if D5 isn't working properly we also have a problem. I'd look carefully at these parts, maybe replace D5 (a pitty as I notice you did a really neat job of wiring it up). It seems to me there is far too much voltage across D5 and I think this is where the issue lies. Try replacing D5, check it's not backwards or something, or that your version of the BD140 doesn't have a different pinout for some reason.
By the way, since there is a feedback going on in the current source it is possible for the current source to oscillate and in my design I included a small 1nF capacitor between the base and collector of Q8 to slow it down.
Last edited:
Hi Gareth,
Thanks so much for the detailed explanation. You nailed it right on the head. Had I mounted the BD140 on the pcb and then tied the flying wires to the backside I would have noticed that the wires cross. Unfortunately, I attached the wires to the top side and didn't notice it. It wasn't until I started checking per the schematic that I noticed that I had the emitter of D5 connected to the emitter of Q8. Good news is that I have one channel that is working perfectly. Bad news is that through all this I blew one of the IRFP9240's in the other channel. At least I have it figured out. Now I just have to get some more matched devices.
Very Cool!!!!!!!!
Thanks so much for the detailed explanation. You nailed it right on the head. Had I mounted the BD140 on the pcb and then tied the flying wires to the backside I would have noticed that the wires cross. Unfortunately, I attached the wires to the top side and didn't notice it. It wasn't until I started checking per the schematic that I noticed that I had the emitter of D5 connected to the emitter of Q8. Good news is that I have one channel that is working perfectly. Bad news is that through all this I blew one of the IRFP9240's in the other channel. At least I have it figured out. Now I just have to get some more matched devices.
Very Cool!!!!!!!!
Member
Joined 2009
Paid Member
I only played it for a couple minutes. I need to get it attached to a proper heatsink. I did notice that it is not very high gain. I will get it bolted up so I can push it a little. Always hard to tell how something sounds when you can only hear one side of the program. Pretty cool getting to the bottom of it. I have the replacement devices ordered. Probably be a couple of weeks before they arrive. I am tempted to remove one of the 240's so I can see how it sounds with one device per side. 