This is basically what I did, which of course converts your specified shunt compensation into miller compensation. I went searching for any posts by Tekko regarding this and came up empty, though I do recall reading an experience like this by someone.
I gave it a little resistive loading workout, nothing too heavy and all else seemed well. I did try a quick square wave test and found evidence of significantly slower slewing than expected, noticeable rounding on the leading edges of the transitions. No issue with overshoot or ringing, just seems dare I say sluggish. By the time we are into the upper audio band the output isn't even square at all. Otherwise seemed to be stable with steady 100mA per output device bias and offset at 12mV.
That said I hooked up one channel, the only one completed so far, and had a little preview. The sound is quite pleasant with the short listen I gave it, though I won't be able to be more specific until I have a stereo pair and can do a proper comparison to something else. Almost dead silent idle, slight hiss with ear on tweeter, with no hum or buzz.
Second SYMEF almost ready, just have to prep a heat sink and get things mounted up to it. Maybe on to testing the second one tomorrow afternoon.
I did however spend a little time woking with the completed module, rechecking some values and components to ensure I haven't made any errors. Reason for the added scrutiny is the apparent 3.2V/uS slew rate, measured from 10%-90% of a 10V p-p square wave output with an 8 ohm load. Simulation of the 'as built' circuit suggests about a 13.5v/uS slew rate, so a fair difference from reality.
Other observations are what seems to be good bias stability at 100mA per device and also good behaviour during the heat soak at 1/3 full output. Offset is also quite stable and remains at about 12mV during testing. Bear in mind by bench supply is +/-30V and not the specified +/-40V but I expect similar behaviour when the dedicated power supply is built.
Has anyone else noted the relative slowness during testing? Harrison, is the limited slew part of the design or is there something wrong? I will continue on to listening testes when ready regardless (hey, we don't listen to square waves anyway) but I curious about the limited slew and wether it is unique to my build.
I did however spend a little time woking with the completed module, rechecking some values and components to ensure I haven't made any errors. Reason for the added scrutiny is the apparent 3.2V/uS slew rate, measured from 10%-90% of a 10V p-p square wave output with an 8 ohm load. Simulation of the 'as built' circuit suggests about a 13.5v/uS slew rate, so a fair difference from reality.
Other observations are what seems to be good bias stability at 100mA per device and also good behaviour during the heat soak at 1/3 full output. Offset is also quite stable and remains at about 12mV during testing. Bear in mind by bench supply is +/-30V and not the specified +/-40V but I expect similar behaviour when the dedicated power supply is built.
Has anyone else noted the relative slowness during testing? Harrison, is the limited slew part of the design or is there something wrong? I will continue on to listening testes when ready regardless (hey, we don't listen to square waves anyway) but I curious about the limited slew and wether it is unique to my build.
Thanks Jason for your detailed description of your build. The data will be most useful in the publishing of the secret formula/algorithm.
In todays world each of the measurements we make in amplifier design has a weight when it comes to their contribution to the overall sound as well as its closeness to the actual sound in real life.
Post some pics of the amplifier and will go through them.
According to Pass* on most music the highest slew rate is 1.5V/us with 0.5V/us was most common.
Do post some pics of the square waves.
kind regards,
Harrison.
*https://www.passdiy.com/project/amplifiers/the-pass-a-40-power-amplifier
In todays world each of the measurements we make in amplifier design has a weight when it comes to their contribution to the overall sound as well as its closeness to the actual sound in real life.
Post some pics of the amplifier and will go through them.
According to Pass* on most music the highest slew rate is 1.5V/us with 0.5V/us was most common.
Do post some pics of the square waves.
kind regards,
Harrison.
*https://www.passdiy.com/project/amplifiers/the-pass-a-40-power-amplifier
Last edited:
I'll get to posting some pictures a little later, but first a few listening impressions.
First let me say that the SYMEF is a decent sounding amplifier, but it definitely has a sort of signature to its sound. I'll do my best to give a description of how I felt it presented the music.
Treble is smooth and rounded sounding without any excess sheen or metallic character. The midrange, particularly vocals are quite pronounced with real emphasis in the presentation, very forward. The amplifier sounds very rich in the bass region, almost artificially so, such that I think some other details end up somewhat masked.
The overall sound stage is almost a little odd, with a decent three dimensional effect for most of the music but the vocals are almost projected a little too far. I like to hear Diana Krall but she sound a little too close and somewhat separated from the rest. This seemed to occur with quite a few vocalists, both male and female.
I played a fair selection of tunes on the SYMEF to get a more familiar with it, then switched to my current benchmark, the PeeCeeBee version of the VSSA. I don't want to be the fly in your ointment, but I quite strongly prefer the VSSA due to its very balanced and what I feel to be very truthful or faithful performance, without any particular emphasis on specific parts of a recording.
With that said, the SYMEF sounds better than some commercial amps that I have and do own, but honestly didn't leave my jaw agape like the VSSA did when it played the first time. I think due to its strong bass performance the SYMEF may play well with smaller bookshelf type speakers, and not so much the larger sort of three way I'm using (a set of Paradigm 11se Mk3) which don't need any help playing deeper bass. Anyway, I hope the input is useful, even if it wasn't all rainbows and unicorns.
First let me say that the SYMEF is a decent sounding amplifier, but it definitely has a sort of signature to its sound. I'll do my best to give a description of how I felt it presented the music.
Treble is smooth and rounded sounding without any excess sheen or metallic character. The midrange, particularly vocals are quite pronounced with real emphasis in the presentation, very forward. The amplifier sounds very rich in the bass region, almost artificially so, such that I think some other details end up somewhat masked.
The overall sound stage is almost a little odd, with a decent three dimensional effect for most of the music but the vocals are almost projected a little too far. I like to hear Diana Krall but she sound a little too close and somewhat separated from the rest. This seemed to occur with quite a few vocalists, both male and female.
I played a fair selection of tunes on the SYMEF to get a more familiar with it, then switched to my current benchmark, the PeeCeeBee version of the VSSA. I don't want to be the fly in your ointment, but I quite strongly prefer the VSSA due to its very balanced and what I feel to be very truthful or faithful performance, without any particular emphasis on specific parts of a recording.
With that said, the SYMEF sounds better than some commercial amps that I have and do own, but honestly didn't leave my jaw agape like the VSSA did when it played the first time. I think due to its strong bass performance the SYMEF may play well with smaller bookshelf type speakers, and not so much the larger sort of three way I'm using (a set of Paradigm 11se Mk3) which don't need any help playing deeper bass. Anyway, I hope the input is useful, even if it wasn't all rainbows and unicorns.
Hi Jason,
Thank you again for your very informative review. That is one very serious set of speakers you have there.
Music is recorded and mixed in a very interesting fashion. Near field monitors as well as headsets are used. The lead singer is recorded in isolation and is usually cooing into the mic. You might be surprised that what symef is doing is bringing out the recording as it really is.
Congratulations and thank you once again. Your feedback is highly appreciated and does much for the research. Please keep us updated. Everyone's perspective matters.
The TYPE2 PCB is quite a piece of art.
kind regards,
Harrison.
Thank you again for your very informative review. That is one very serious set of speakers you have there.
Music is recorded and mixed in a very interesting fashion. Near field monitors as well as headsets are used. The lead singer is recorded in isolation and is usually cooing into the mic. You might be surprised that what symef is doing is bringing out the recording as it really is.
Congratulations and thank you once again. Your feedback is highly appreciated and does much for the research. Please keep us updated. Everyone's perspective matters.
The TYPE2 PCB is quite a piece of art.
kind regards,
Harrison.
Evette, I'm not opposed to some 'burn in' but to be honest I've never been a big believer in it making a real diference. If a capacitor needs to reform it achieves most of that process pretty quick and the rest takes place over time, but wether it makes a difference I can't say. Outside of electrolytic capacitors I can't see what else would need to be conditioned to achieve the best performance.
That said, I'll hook them up and let them run on the bench for a week and revisit. Maybe I'll hear a slightly different character afterwards. Oh and just to be clear, I generally just use relatively inexpensive parts because I feel if a design relies on the characteristics of a given part to function properly then it is operating too close to some limit and should likely be redesigned to be less dependant on specific device characteristics.
Anyhow, I was not suggesting the amplifier is bad. Rather that the SYMEF along with many others, exhibit a sonic character unique to that particular design or implementation.
That said, I'll hook them up and let them run on the bench for a week and revisit. Maybe I'll hear a slightly different character afterwards. Oh and just to be clear, I generally just use relatively inexpensive parts because I feel if a design relies on the characteristics of a given part to function properly then it is operating too close to some limit and should likely be redesigned to be less dependant on specific device characteristics.
Anyhow, I was not suggesting the amplifier is bad. Rather that the SYMEF along with many others, exhibit a sonic character unique to that particular design or implementation.
The "SE Class-A" effect.
My DOZ has exactly the same characteristics.
Jason, have you heard an SE yet?
If not, then you don't need to. As SYMEF has brought that magic of SE in PP with AB bias.
I like this amp. Efficient.
Many won't. So what?
Good job Harrison.
The "SE Class-A" effect.
My DOZ has exactly the same characteristics.
Jason, have you heard an SE yet?
If not, then you don't need to. As SYMEF has brought that magic of SE in PP with AB bias.
I like this amp. Efficient.
Many won't. So what?
Good job Harrison.
Shaan, no sir I have never built a Single Ended Class A amplifier before. I always considered them too inefficient from the theoretical numbers point of view, though I suppose it doesn't really matter for the hobbyist.
There was a comment earlier about the methodology of recording and the possibility of the amplifier rendering the performance 'as recorded'. I think there is also another way of looking at it too; more of how the artist and engineers intended the music to sound after final mixing rather than a more 'raw' portrayal. The latter seems to be my preference. Otherwise we'd just put up a stereo mic in front of the band, record it and call it a day.
I'm sure many will enjoy this amplifier, it certainly isn't a bad one. I have no doubt it will come out to play again, I do like to mix things up from time to time. I will give it an extended work out to see if that changes anything.
Mr,Shaan, Youi are correct Sir,in your hearing a SE Amp as a matter of fact I,ve found all of Mr.Harrison amp to sound that way. Mr.Jason,,Every amp I,ve build from Mr.Harrison I,ve built with parts from what I,ve collected over time from tv,s to old tube radios Every part I,ve re-cyled and you know something Sir After a couple hundred hours of burn-in and of course re-biasing to find a sweet spot, I,ve always found the amp to sound better. Evette P.S. your boards look great.
Welcome to the club Shaan and thank you for your feedback.
kind regards,
Harrison.
kind regards,
Harrison.
The "SE Class-A" effect.
My DOZ has exactly the same characteristics.
Jason, have you heard an SE yet?
If not, then you don't need to. As SYMEF has brought that magic of SE in PP with AB bias.
I like this amp. Efficient.
Many won't. So what?
Good job Harrison.
Hi Evette,
Thank you again for your feedback. It also looks like both SYMEF and IDIFFQC still provide pleasure in the long haul.
kind regards,
Harrison.
Thank you again for your feedback. It also looks like both SYMEF and IDIFFQC still provide pleasure in the long haul.
kind regards,
Harrison.