That remote server pic (what's wrong with attachments?) only shows the follower output stage.
Where is the amplifier?
Where is the amplifier?
That's the problem . I suspect the other part of the circuit shows some important detail . The other circuit I supplied seems similar . To be honest it seems overkill . I think the gentleman asking the question wants an answer without showing all of the circuit .
Last edited:
Yes.
Almost every ClassAB push pull amplifier has a DC output offset adjustment.
Hi Andrew,
This is actually not quite true. Most solid state power amps do not have a DC offset adjustment. Typically, most have a capacitor (an electrolytic, unfortunately) in the shunt leg of the NFB network to reduce the closed loop gain to unity at DC, leaving only the input offset of the amplifier at the output. This is usually less than 10 mV in a well-designed amplifier. Better amplifiers use a DC servo for offset control, eliminating the need for the electrolytic. This will often get output offset below 5mV, depending on the input offset voltage of the op amp used in the servo.
However, in the current context of a no-feedback amplifier, you are correct. But if one is willing to allow for the "feedback" of a DC servo in a no-NFB amplifier, then a DC servo can be used here as well to eliminate the need for an offset adjustment.
Cheers,
Bob
Perhaps my english is not enought but I don't understaind what is the problem. I don't whant to post the preamp section, It's not necessary. As I have say the final stage is drove by the cathode of half ecc83. The protection circuit protect by dc and short circuit. There is no dc correction, CIN1 and CIN2 prevent this possibility.
Rewind:
I found only a n-mos short. I can't found another ACD101NDD. I found a substitute, the question is: do I have to select 11 nmos and 11 pmos with the same caratteristic to obtain an accettable dc (near zero) at the output? Or the lateral mosfet haven't this problem?
This mosfet isn't cheap so I want to be quite sure that I don't spent money wastefully. Thaks.
Rewind:
I found only a n-mos short. I can't found another ACD101NDD. I found a substitute, the question is: do I have to select 11 nmos and 11 pmos with the same caratteristic to obtain an accettable dc (near zero) at the output? Or the lateral mosfet haven't this problem?
This mosfet isn't cheap so I want to be quite sure that I don't spent money wastefully. Thaks.

Here is a small amplifier I was playing with . If wondering the VAS supply originally was to be higher ( +/- 25 V ) . Sounds better this way . I never finished this amp . It looks promising . It outperformed a boosted OPA 604 technically . Note it doesn't need an output cap .
Note how the FET's sit at very different voltages 1.471 and -1.962 form 0V reference . Although at 0.7 A and lower voltage the differences are typical . So exact matches are unlikely to be acheived . Exicon , BUZ and Hitachi devices are very similar .
Hopefully someone will make a recommendation to allow you a simple solution . A reliable reference voltage is required .
These are my notes . I never show them so only in a form I need .
I think someone will also say you need a high drive current for this . Being as optimistic as I can be 20 mA absolute minimum . 50 mA would be ideal .
AndyVicus:
From what you are discussing it looks like you are building a MASSIVE source follower output stage to a tube driver? If this is correct, you do need a fair amount of current to drive the output stage with this many MOSFET pairs. Remember, they do have input capacitance. HAFLER and others used circa 10mA to drive 3 MOSFETs. You have 11. So shoot for 10mA X (11/3) = 40mA or so. You can get that easily from a EL34 or 2A3 or similar. Forget 6DJ8 or 6sn7 (unless you parallel 4 of them 😱).
Now for your DC offset problem: add a 100K trimpot in series with the 330K (R84) resistor after the zener on the negative rail. To give yourself a range, replace the 330K (R84) resistor with say 270K. Before turning the amp on set the 100K trimpot so that the 270K + pot = 330K. Turn on amp. Set BIAS. Allow to reach working temp. Now trim the 100K to 0mV between output and ground. In real life 5mV or so is acceptable.
From what you are discussing it looks like you are building a MASSIVE source follower output stage to a tube driver? If this is correct, you do need a fair amount of current to drive the output stage with this many MOSFET pairs. Remember, they do have input capacitance. HAFLER and others used circa 10mA to drive 3 MOSFETs. You have 11. So shoot for 10mA X (11/3) = 40mA or so. You can get that easily from a EL34 or 2A3 or similar. Forget 6DJ8 or 6sn7 (unless you parallel 4 of them 😱).
Now for your DC offset problem: add a 100K trimpot in series with the 330K (R84) resistor after the zener on the negative rail. To give yourself a range, replace the 330K (R84) resistor with say 270K. Before turning the amp on set the 100K trimpot so that the 270K + pot = 330K. Turn on amp. Set BIAS. Allow to reach working temp. Now trim the 100K to 0mV between output and ground. In real life 5mV or so is acceptable.
![]()
Here is a small amplifier I was playing with . If wondering the VAS supply originally was to be higher ( +/- 25 V ) . Sounds better this way . I never finished this amp . It looks promising . It outperformed a boosted OPA 604 technically . Note it doesn't need an output cap .
Note how the FET's sit at very different voltages 1.471 and -1.962 form 0V reference . Although at 0.7 A and lower voltage the differences are typical . So exact matches are unlikely to be acheived . Exicon , BUZ and Hitachi devices are very similar .
Hopefully someone will make a recommendation to allow you a simple solution . A reliable reference voltage is required .
These are my notes . I never show them so only in a form I need .
I think someone will also say you need a high drive current for this . Being as optimistic as I can be 20 mA absolute minimum . 50 mA would be ideal .
This is one of the reasons, why I hate so called true complemetary output power stages. Vbe multiplier could be a simple solution. But this isn't also the royal way.
If you want to avoid such problems, go to my overview about output stages topologies without such push pull output stages:
http://www.diyaudio.com/forums/soli...better-audio-non-complements-audio-power.html
The mentioned topologies from the last post perhaps of interest for you.
Last edited:
This I had observe in the same kind. My favorite solution isn't a DC offset servo; rather a MKP/MKS capacitor from 4,7uF until 47uF (for small sizes WIMA is a good solution) for electrolytic replace.Hi Andrew,
This is actually not quite true. Most solid state power amps do not have a DC offset adjustment. Typically, most have a capacitor (an electrolytic, unfortunately) in the shunt leg of the NFB network to reduce the closed loop gain to unity at DC, leaving only the input offset of the amplifier at the output. This is usually less than 10 mV in a well-designed amplifier. Better amplifiers use a DC servo for offset control, eliminating the need for the electrolytic. This will often get output offset below 5mV, depending on the input offset voltage of the op amp used in the servo.
Cheers,
Bob
The high pass filter character is usefull in such cases, where is present a fat punchy and bloated bass transmission in the aera arround 40-80Hz.
After adjust the capacitance to the right value, the bass transmission is more clear and tight, particularly in cases, where a near-wall/near-corner loudspeaker installation has been realized.
If the high pass character of this capacitor is unwanted, there is the solution without the GND connection of the secundary windings. Instead this each power supply capacitor get in parallel a resistor (please note: the required voltage now must be the twice than before) and various modifications are to carry out at the input of LTP - I have posted the suited schematic anywhere here on diyaudio.
Regard this capacitor in the G-NFB network here on diyaudio there is a detailed discussion.
Who know the URL(s) ?
Last edited:
![]()
Taken from item No 34
Could u share full schematic of this output stage ?
which lateral MOSFETs are still in currently production and thus not hard to find (except Semelab Al-FETs under
Magnatec. ALFET Lateral MOSFETs ) ??
Thank you for advices.
Magnatec. ALFET Lateral MOSFETs ) ??
Thank you for advices.
Magnatec. ALFET Lateral MOSFETs
As well as the ALFETs they make the BUZ lateral MOSFETs.
Semelab, Magnatec, make Buz, Alfet, Execon, all on the the same production line in Leicestershire, UK: Semelab Ltd.
Execon is Profusion's brand name and they sell direct. http://www.exicon.info/cross-ref.php
As well as the ALFETs they make the BUZ lateral MOSFETs.
Semelab, Magnatec, make Buz, Alfet, Execon, all on the the same production line in Leicestershire, UK: Semelab Ltd.
Execon is Profusion's brand name and they sell direct. http://www.exicon.info/cross-ref.php
Last edited:
![]()
Taken from item No 34
Hi,
how do you calibrate the bias on this amplifier? the rest of the amplifier where is it?
many thanks
Sanken's MLE20 are in use in Yamaha's circlotron power amp model M-5000 - go to
https://elektrotanya.com/yamaha_m-5000_pwr-amplifier_sm.pdf/download.html
the P-Channel version is the MJD20.
Are this versions from currently production a good replacement for the vintage (obsolete) versions from Hitachi and Toshiba from post #1 ?
https://elektrotanya.com/yamaha_m-5000_pwr-amplifier_sm.pdf/download.html
the P-Channel version is the MJD20.
Are this versions from currently production a good replacement for the vintage (obsolete) versions from Hitachi and Toshiba from post #1 ?
Attachments
- Home
- Amplifiers
- Solid State
- Audio (Lateral) MOSFET's for Class AB - Overview, P-Spice URL's