I meant for each output transistor pair. 150mA for each transistor pair?
What voltage on each .22 resistor?
What voltage on each .22 resistor?
Don't you think that 150 mA is a bit hot for bias current? Bias for lowest distortion varies, but it usually runs 5 ~ 30 mA depending on output transistor matching. In the Symasym, my bias ran about 5 mA for the lowest distortion. Running high bias like this will normally increase distortion and power supply noise.
-Chris
-Chris
Hi Veysel,
That's reasonable, a little on the high side, but reasonable.
Try adjusting the bias using a 4 R resistive load and a 10 KHz low level signal. Anything up to about 1 watt should be fine. You may find that your current can be much, much lower.
-Chris
That's reasonable, a little on the high side, but reasonable.
Try adjusting the bias using a 4 R resistive load and a 10 KHz low level signal. Anything up to about 1 watt should be fine. You may find that your current can be much, much lower.
-Chris
About the oscillation you mention on your schematic when you power the amp with the speakers connected, which should be the natural thing to do.
On my sim I got a glitch @ 20KHz on LTP stage. Tried several things to cure it, and what seemed to tame it was increasing the resistor between drivers emitters to 200 ohms, and decreasing the output transistors base resistors from 4.7 to 2.2.
Why don't you try it and see what happens?
On my sim I got a glitch @ 20KHz on LTP stage. Tried several things to cure it, and what seemed to tame it was increasing the resistor between drivers emitters to 200 ohms, and decreasing the output transistors base resistors from 4.7 to 2.2.
Why don't you try it and see what happens?
Base resistors in the output circuits are going to make the stage more likely to oscillate.
-Chris
-Chris
I said halving existing resistors, not adding them.
The oscillation I see is on the first stage, not the output.
Can you check the first stage output, feeding 20KHz, if there is oscillation or not?
The oscillation I see is on the first stage, not the output.
Can you check the first stage output, feeding 20KHz, if there is oscillation or not?
Well, you may try. You have parts and tools that they did not have.
Unless you have ALL the original parts, and they are not fakes.
But the simulation does show an oscillation, even with the original parts models.
Do you really think they designed an amplifier with that turning on limitation you described?
You can solder 4.7 resistors in parallel with the others, halving their value, on the copper side. And you can add another 100 resistor in series with the driver's. Not much to risk.
Unless you have ALL the original parts, and they are not fakes.
But the simulation does show an oscillation, even with the original parts models.
Do you really think they designed an amplifier with that turning on limitation you described?
You can solder 4.7 resistors in parallel with the others, halving their value, on the copper side. And you can add another 100 resistor in series with the driver's. Not much to risk.
Hi carlmart,
I also completely agree with your latest suggestions in your post directly above this one.
Hi Veysel,
-Chris
I was agreeing with you. It was merely an observation.
I also completely agree with your latest suggestions in your post directly above this one.
Hi Veysel,
You sure can. We understand so much more than they did back then. You may have better tools too! My ancient HP gear wasn't out yet when they designed the C-02, so I have better tools than they did. You probably do for at least one instrument, but the better knowledge we have today trumps the conditions back then. Now, no way would I even begin to suggest we are smarter than they are. Our tools are much better, that's all.
-Chris
Agree with you, Chris. This is DIY, and if we can change something to make things different, and that different sounds better to out ears, why should you not do it? Or at least try.
This cloning thing, when it's based on past designs where most active parts are not available anymore, can be tricky.
The ideal thing would be to have the original design available, and make a comparative listening. In my case, where I'm cloning a Luxman 5M21, a friend of mine has one, so I will try to do that comparison.
As he is on another country, a direct comparison may not be possible, but I did listen to that amp a lot, so I think I know how it sounds.
This cloning thing, when it's based on past designs where most active parts are not available anymore, can be tricky.
The ideal thing would be to have the original design available, and make a comparative listening. In my case, where I'm cloning a Luxman 5M21, a friend of mine has one, so I will try to do that comparison.
As he is on another country, a direct comparison may not be possible, but I did listen to that amp a lot, so I think I know how it sounds.
Hi carlmart,
I repaired (rebuilt) my C-02 with current parts, so there shouldn't be any problems. I used the On-Semi MJL21195 and MJL21196 outputs and the drivers were also from On-Semi. I can't remember those numbers, sorry. I did rebuild both channels due to the high temperature the original parts were run at. One channel had failed already.
The main thing is that the circuit remained stable with the current parts. You should be okay using the same ones for clones.
-Chris
I repaired (rebuilt) my C-02 with current parts, so there shouldn't be any problems. I used the On-Semi MJL21195 and MJL21196 outputs and the drivers were also from On-Semi. I can't remember those numbers, sorry. I did rebuild both channels due to the high temperature the original parts were run at. One channel had failed already.
The main thing is that the circuit remained stable with the current parts. You should be okay using the same ones for clones.
-Chris
Indeed. Luxman service manuals specifiy that the parts can be replaced with similar devices with the same specs.
Chris, those MJLs you mentioned are for power outputs. Did you use them on the C02?
Which were the parts that were hot and you replaced?
The C02 regulated supply looks very good.
Chris, those MJLs you mentioned are for power outputs. Did you use them on the C02?
Which were the parts that were hot and you replaced?
The C02 regulated supply looks very good.
Hi carlmart,
I used them in the M-02, sorry. I'll be playing in that C-02 as well. It is well engineered and mostly needs refreshing, probably PS capacitors. Then I will have a close look at the audio stages.
-Chris
I used them in the M-02, sorry. I'll be playing in that C-02 as well. It is well engineered and mostly needs refreshing, probably PS capacitors. Then I will have a close look at the audio stages.
-Chris
I can't remember any exact dates. You're talking about those front end modules - right? If so, they were earlier.
The M-02/C02 were earlier?
Because those DMLs had degeneration resistors inside, which I think was a very fine move from Luxman.
I tried adding some LTP degeneration on the M-02, but I didn't get it right. THD went up instead of down.
Because those DMLs had degeneration resistors inside, which I think was a very fine move from Luxman.
I tried adding some LTP degeneration on the M-02, but I didn't get it right. THD went up instead of down.
Hi carlmart,
No, the C-02 and M-02 came after.
If you add degeneration resistors, you linearise the diff pair a bit, but you also cut open loop gain. Sometimes that trade-off works, sometimes it doesn't. Using J-FETs (my choice) in the diff pair does the same thing, plus they are more immune to RF noise. They also have a higher breakdown than the E-B junction and will generally survive a DC offset situation where BJTs don't.
-Chris
No, the C-02 and M-02 came after.
If you add degeneration resistors, you linearise the diff pair a bit, but you also cut open loop gain. Sometimes that trade-off works, sometimes it doesn't. Using J-FETs (my choice) in the diff pair does the same thing, plus they are more immune to RF noise. They also have a higher breakdown than the E-B junction and will generally survive a DC offset situation where BJTs don't.
-Chris
Hi,
Some time has passed, and I have gone forward with my project to assemble an amp based on a Luxman classic design.
The models I have been playing with were quite similar in architecture: 5M21, L10 and M12. The main problem is that all active parts are obsolete and new types have to be found to replace them. Unfortunately LTSpice models for the old types are not available either, to know what happens when you use newer parts instead.
My main problem is with the Bode plot requirements, which I was not aware of, which seem to determine is the amp is table or not.
If someone is experienced with these plots, I can show the asc files for the M02 where the curves are created. They seem to show a potential instability for this amp, which would explain the oscillation when the amp is not powered in a certain way.
Some time has passed, and I have gone forward with my project to assemble an amp based on a Luxman classic design.
The models I have been playing with were quite similar in architecture: 5M21, L10 and M12. The main problem is that all active parts are obsolete and new types have to be found to replace them. Unfortunately LTSpice models for the old types are not available either, to know what happens when you use newer parts instead.
My main problem is with the Bode plot requirements, which I was not aware of, which seem to determine is the amp is table or not.
If someone is experienced with these plots, I can show the asc files for the M02 where the curves are created. They seem to show a potential instability for this amp, which would explain the oscillation when the amp is not powered in a certain way.