Here's an updated schematic with measurements.
Red is measured over compoent and purple is relative to supply 0.
I've added values for Vbe T7, T8, T9, T10
I allso measured the output offset and voltages at the power transistor outputs relative to ground.
Thes measurements are not 100% exact as the card was run without heatsink so I had to be quick and the values were changing due to rising temperature in the power transistors. But hopefully, this is sufficient to give some indication.
Allso, all measurements have been done with out the DC servo connected, so this might influence/ explain some of the values?
Red is measured over compoent and purple is relative to supply 0.
I've added values for Vbe T7, T8, T9, T10
I allso measured the output offset and voltages at the power transistor outputs relative to ground.
Thes measurements are not 100% exact as the card was run without heatsink so I had to be quick and the values were changing due to rising temperature in the power transistors. But hopefully, this is sufficient to give some indication.
Allso, all measurements have been done with out the DC servo connected, so this might influence/ explain some of the values?
Last edited:
Fit temporary heatsinks before you damage some thing else.
No voltages around the input stage.
Give the 3significant figure Vdrop around the input stage.
The volts drop across R8 and R? do not add up. They should give Vbe of T7
Same error in the -ve side. I wonder if Vb is actually Vce?
You will get a better quality pic if you print the schematic and write on it.
The pdf you post is terrible.
No voltages around the input stage.
Give the 3significant figure Vdrop around the input stage.
The volts drop across R8 and R? do not add up. They should give Vbe of T7
Same error in the -ve side. I wonder if Vb is actually Vce?
You will get a better quality pic if you print the schematic and write on it.
The pdf you post is terrible.
See what I can doo abvout temporary heatsinks, but I' dont think I have anything suitable at the moment..
Will add Veb for input stage.
My scanner doesn't work with the new computer, so that schematic is unfortunately the only thing I can present... 🙁 (I do agree its not the best)
Well, the voltages across the resistors were all measured quite correctly, but I agree it doesn't seem to add up as you say.. could some other part of the circuit influence this??
Will add Veb for input stage.
My scanner doesn't work with the new computer, so that schematic is unfortunately the only thing I can present... 🙁 (I do agree its not the best)
Well, the voltages across the resistors were all measured quite correctly, but I agree it doesn't seem to add up as you say.. could some other part of the circuit influence this??
I have now added Vbe Measurements for T1-4 in the input stage.
Here I made an interresting observation
The Vbe voltage for T1 and T2 was stable, whilst the Voltage for T3 and T4 started to drop..
(This allso means that the Vbe measured for T3 and T4 are not 100% exact)
When I power on, the idle current allso rises (due to temperature increase in the output transistors without heatsink and thermal contact with T11), could there be a correlation to the Vbe drop over T3 and T4, and the stable Vbe of T1 and T2 be an indication of a problem??
Here I made an interresting observation
The Vbe voltage for T1 and T2 was stable, whilst the Voltage for T3 and T4 started to drop..
(This allso means that the Vbe measured for T3 and T4 are not 100% exact)
When I power on, the idle current allso rises (due to temperature increase in the output transistors without heatsink and thermal contact with T11), could there be a correlation to the Vbe drop over T3 and T4, and the stable Vbe of T1 and T2 be an indication of a problem??
Last edited:
Hi.
Rig up something to connect the outputs to the Vbe multiplier.
Input voltage at the bases of both sides of the LTP.
Tail voltage of the LTP where the emitter resistors meet.
Vdrop both NFB resistors.
Vdrop Rin
Vds & Vgs of the 2sk multiplier.
Rig up something to connect the outputs to the Vbe multiplier.
Input voltage at the bases of both sides of the LTP.
Tail voltage of the LTP where the emitter resistors meet.
Vdrop both NFB resistors.
Vdrop Rin
Vds & Vgs of the 2sk multiplier.
Andrew,
Sounds like you have a plan here.. all I need to do now is interpret it and understand it.. 😉
Lets see..
what do you mean by Veb multiplier? I'm not sure (haven't got a clue actually..) where in the circuit that is?
Base of T1 and T3 are connected on the PCB (input) and Base on T2 and T4 (feedback side) is allso connected. Do you mean that I should connect and apply the same voltage to the base of T1-4? and how much?
Both NFB resistors, that would be R22 and R23, I can't see any other resistors directly connected with the NFB path..
Which one is Rin? is it R3?
Vds and Vgs, that's drain-source and gate-source voltage of T11 (the thermal compensation transistor?, right?
Sounds like you have a plan here.. all I need to do now is interpret it and understand it.. 😉
Lets see..
what do you mean by Veb multiplier? I'm not sure (haven't got a clue actually..) where in the circuit that is?
Base of T1 and T3 are connected on the PCB (input) and Base on T2 and T4 (feedback side) is allso connected. Do you mean that I should connect and apply the same voltage to the base of T1-4? and how much?
Both NFB resistors, that would be R22 and R23, I can't see any other resistors directly connected with the NFB path..
Which one is Rin? is it R3?
Vds and Vgs, that's drain-source and gate-source voltage of T11 (the thermal compensation transistor?, right?
Last edited:
Just one thought...
Is there a chance that some of the electrolytic capacitors could have been damaged by the polarity reversal in a way that could explain the problem??
Say, if the capacitor C9 parallel to zener D4 is leaking or something??
Is there a chance that some of the electrolytic capacitors could have been damaged by the polarity reversal in a way that could explain the problem??
Say, if the capacitor C9 parallel to zener D4 is leaking or something??
Vbe multiplier is the BJT that is arranged as a constant voltage shunt regulator. It provides Vbias to the output stage.
Just 2 resistors and one small signal transistor for the most basic version.
The FET in the middle is your constant voltage shunt regulator, but it is based on a mosFET instead of a BJT. This makes it a Vgs multiplier.
It multiplies the Vgs of the mosFET by a factor determined by the two resistors.
That factored Vgs is the shunt regulated voltage.
If this Vgs multiplier monitors the temperature of the output stage, it can be arranged to provide some form of temperature compensation to the Vbias that it generates. It is not a temperature compensator, it is a Vbias generator.
What is the Vgs of your Vgs multiplier? It depends on the mosFET parameters. You can measure it or you can look up the datasheet, do both and see how they compare.
Just 2 resistors and one small signal transistor for the most basic version.
The FET in the middle is your constant voltage shunt regulator, but it is based on a mosFET instead of a BJT. This makes it a Vgs multiplier.
It multiplies the Vgs of the mosFET by a factor determined by the two resistors.
That factored Vgs is the shunt regulated voltage.
If this Vgs multiplier monitors the temperature of the output stage, it can be arranged to provide some form of temperature compensation to the Vbias that it generates. It is not a temperature compensator, it is a Vbias generator.
What is the Vgs of your Vgs multiplier? It depends on the mosFET parameters. You can measure it or you can look up the datasheet, do both and see how they compare.
Last edited:
Ah, then I understand what you mean by Veb multiplier (sort of).
But where exactly should I connect the amplifier output to the Vbe multiplier?
Will check the Vgs of the T 11 fet right away!
Well, that is if I can find a data-sheet for the 2SK537...
But where exactly should I connect the amplifier output to the Vbe multiplier?
Will check the Vgs of the T 11 fet right away!
Well, that is if I can find a data-sheet for the 2SK537...
Last edited:
Rin = R2 The resistor that sets the input impedance seen by the connecting cable.
Look at your schematic.
See the broken lines between T1 and the outputs. That is a Thermal connection. It provides thermal feedback and allows some Thermal Compensation to the Vbias that the Vgs multiplier generates.
Look at your schematic.
See the broken lines between T1 and the outputs. That is a Thermal connection. It provides thermal feedback and allows some Thermal Compensation to the Vbias that the Vgs multiplier generates.
Andrew,
Bit of a tough time finding a proper data sheet there, best i could find was a replacement in a toshiba catalog, and there it said Vgs 10v.
When I measure the the voltage between gate and Source, I get only 3,6V..
Can this tell us anything??
Bit of a tough time finding a proper data sheet there, best i could find was a replacement in a toshiba catalog, and there it said Vgs 10v.
When I measure the the voltage between gate and Source, I get only 3,6V..
Can this tell us anything??
Vgs =10V. I suspect you have found the maximum Vgs for which no damage will occur.
You need Vgs for a variety of Id currents, sometimes a graph some time a table.
Measured Vgs=3.6V tells us the schematic is wrong. The lowest Vbias you can get is 3.6V. The low Vgs output devices that are on the schematic will never need that much bias voltage for small Id values.
Do you realise that if you had put these voltages on a schematic 249 posts ago we could have saved 250 posts.
What a waste of everybodies time !
Yes, I thought 10v Vgs sounded a bit high, however, that is what the data-sheet i found for the equvelent replacement says.
Gate-source (Vgss) breakdown voltage is rated at 30 V, and gate threshold voltage is specified as 2V min and 4V max..
So, just so I get this right.. The Vgs value of 3,6V is much higher than it should be and is the cause for the high Idle-current in the output devices?
If so, why is the Vgs voltage so high...
I agree that I should have posted a schematic with voltages earlier on.. having said that, the problem that started this thread was a different one, and one that turned out much simpler to solve as a transistor was clearly defective (once located).. this time around I have not been able to find any defective transistor... 🙁
Gate-source (Vgss) breakdown voltage is rated at 30 V, and gate threshold voltage is specified as 2V min and 4V max..
So, just so I get this right.. The Vgs value of 3,6V is much higher than it should be and is the cause for the high Idle-current in the output devices?
If so, why is the Vgs voltage so high...
I agree that I should have posted a schematic with voltages earlier on.. having said that, the problem that started this thread was a different one, and one that turned out much simpler to solve as a transistor was clearly defective (once located).. this time around I have not been able to find any defective transistor... 🙁
Last edited:
Can you confirm the schematic values for the Vgs multiplier.
I see 130r, 270r, 1kVR and 2sk537.
What have you fitted?
Is the VR set to 1k at the moment?
What is Vbias? That is the same as Vds. It should also equal the 4.1V that you have measured.
I see 130r, 270r, 1kVR and 2sk537.
What have you fitted?
Is the VR set to 1k at the moment?
What is Vbias? That is the same as Vds. It should also equal the 4.1V that you have measured.
Yes Andrew, despite the sub-standard resolution of the schematic You got the right values.
The pot is set to full CCW, which gives 1k in series with R28 of 270 Ohms.
Voltage between drain and source for T11 was 4,01V
The pot is set to full CCW, which gives 1k in series with R28 of 270 Ohms.
Voltage between drain and source for T11 was 4,01V
I have allso measured all the reistors, including the output transistor Gate reisstors R32 and R33, and they are all fine. I allso checked for voltage drop across R32 and R33, and there were none, i.e. no currents flowing to or from the gate due to drain leakage. Assuming that the T11 I replaced is still OK, I can only see that the problem must be from somwhere upstream of the Vgs multiplier...?? Or could the pair of output transistors be responding differently than others I have used on other cards??
Last edited:
the output of the Vgs multiplier is Vgs * [1+{130r/(270r+1k)}] = 1.1*Vgs = 3.97V
That 4Volts is shared between the two output device Vgs and the Vdrop of both source resistors.
Let's simplify this and say the top & bottom halves are symmetrical.
2V = Vgs + Vrs = Vgs(300mA) + 0.22r*300mA = Vgs(300mA) + 0.66V
Vgs(300mA) = 2V-0.66V = 1.44V
Does this fit with the datasheet? Id=300mA and Vgs~1.4V
Replace the FET in the Vgs multiplier with a BJT.
Replace the upper 130r with 1k5.
Keep 270r and 1kVR.
The minimum Vbias when VR=1k is ~1.3V when Vgs =600mV
The maximum Vbias when VR=0r0 is ~3.9V
This is probably an adequate range of Vbias for the output mosFETs.
The PCB that works.
What is Vgs and Vbias?
What is Vrs and Vgs(output)?
I have a sneaky suspicion that the working PCB has the wrong output FETs in place, or very high Vgs versions. The tolerance on Vgs is very wide.
That 4Volts is shared between the two output device Vgs and the Vdrop of both source resistors.
Let's simplify this and say the top & bottom halves are symmetrical.
2V = Vgs + Vrs = Vgs(300mA) + 0.22r*300mA = Vgs(300mA) + 0.66V
Vgs(300mA) = 2V-0.66V = 1.44V
Does this fit with the datasheet? Id=300mA and Vgs~1.4V
Replace the FET in the Vgs multiplier with a BJT.
Replace the upper 130r with 1k5.
Keep 270r and 1kVR.
The minimum Vbias when VR=1k is ~1.3V when Vgs =600mV
The maximum Vbias when VR=0r0 is ~3.9V
This is probably an adequate range of Vbias for the output mosFETs.
The PCB that works.
What is Vgs and Vbias?
What is Vrs and Vgs(output)?
I have a sneaky suspicion that the working PCB has the wrong output FETs in place, or very high Vgs versions. The tolerance on Vgs is very wide.
Last edited:
Andrew,
You seem to be making a lot of sense here, at least more than I've been able to! 🙂
The best way to verify your theory is obviously to make measurements on one of the working cards.
I'll see if I can figure out a way to do it, with the card assembled in the chassis, it's a bit tricky to get to.
I share your sneaking suspicion allthough you have articullated it in much clearer terms than I've been able to.
The output transistors are definitively of the correct type (unless we are talking about counterfeits or components from a rejected batch), but they are not from the same source and could so very well be from different batches with different parameters.
Looking at the data-sheet for output transistor T12, 2Sk1530, a Vgs should correspond to 300-400 mA. Not so easy to see as the curve goes all the way to 12A, hence the resolution is not the best for the range we are looking at.
Annyway, I'll try and get those measurements off one of the working cards, that will certainly be interresting!
You seem to be making a lot of sense here, at least more than I've been able to! 🙂
The best way to verify your theory is obviously to make measurements on one of the working cards.
I'll see if I can figure out a way to do it, with the card assembled in the chassis, it's a bit tricky to get to.
I share your sneaking suspicion allthough you have articullated it in much clearer terms than I've been able to.
The output transistors are definitively of the correct type (unless we are talking about counterfeits or components from a rejected batch), but they are not from the same source and could so very well be from different batches with different parameters.
Looking at the data-sheet for output transistor T12, 2Sk1530, a Vgs should correspond to 300-400 mA. Not so easy to see as the curve goes all the way to 12A, hence the resolution is not the best for the range we are looking at.
Annyway, I'll try and get those measurements off one of the working cards, that will certainly be interresting!
Hmm.. no chance of doing that measurement today, I must get a couple of those long pincer type measurement probes in order to get safe access to the measurement points, see if I can get some tomorrow!
- Status
- Not open for further replies.