A75 and adcom gfa-5400

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Administrator
Joined 2004
Paid Member
Hi ctong,
At this point I would like to suggest something.

Return the Adcom to stock condition. Build the new one from scratch.

The Adcom has value, but does not satisfy you. A modified amplifier has greatly reduced value. So, build your own and be happy.

-Chris
 
Hi Chris,

Thanks for the suggestion. The Adcom has a great front end, perhaps only second to the UGS in the X and XA amps. I think I can get very good sound out of it. After I am done modifying the Adcom, the bottle neck in my system is likely to be the digital players. I will probably build a Pass D1 analog stage for my sacd player. At that point, I will be in a better position to build and appreciate a even better amp (X or AX).
 
Administrator
Joined 2004
Paid Member
Hi ctong,
I figured that this would allow you to recoup some money after your build. Once you have built the amplifier the way you want it, the Adcom will have served it's purpose as a yardstick. You won't need it anymore, not unless you want to install it in your bedroom.

Don't laugh! I did something similar and now enjoy great audio in bed. My wife complained initially, but after a week she would kill you if you tried to remove it. ;)

Anyway, this would maximize your investment in time and money.

-Chris
 
The front end of 5400 is not regulated. How bad is this? I know Pass amps have regulated front end.

Also, the output stage rails have polyester film bypass caps. I am going to replace them with polypropylene ones. Is 0.1 u enough? This is a value often seen in bypass caps. It has a impedance of 159 ohms at 10KHz, hardly a small value considering that the output impedance of the amp is on the order of 0.1 ohms. Any suggestions?
 
Administrator
Joined 2004
Paid Member
Hi ctong,
Regulating the voltage amp section can be very beneficial, depends on the design. You will have to start with a higher voltage to give the regulator some "headroom" to work with.

Those bypass caps are there for the very high frequency impedance. Use larger ones for general bypass use.

-Chris
 
anatech said:
Hi ctong,
Regulating the voltage amp section can be very beneficial, depends on the design. You will have to start with a higher voltage to give the regulator some "headroom" to work with.

Those bypass caps are there for the very high frequency impedance. Use larger ones for general bypass use.

-Chris

Chris,

I was just wondering whether anyone has any experience with the problems of unregulated front end rails.

Those bypass caps seem small even for high frequencies (159 ohms at 10KHz). I have seen larger ones (0.68u) being used. But even that seems too small. In an article about power supply Nelson talks about keeping power supply impedance to a fraction of an ohm at high frequencies. That would require a fairly large film cap. I wonder what is a value that is practical to use.
 
Administrator
Joined 2004
Paid Member
Hi ctong,
I was just wondering whether anyone has any experience with the problems of unregulated front end rails.
Most amplifiers are designed that way. Others use a lower voltage that is zener regulated. Less common are the units designed with a separate set of windings for the front end and regulate that.

Those bypass caps seem small even for high frequencies (159 ohms at 10KHz).
Sometimes that is more than enough. You can increase the cap size, but understand that we are more interested in the ultrasonic to low RF frequencies. Consider your audio bandwidth as high up (or higher) as 1 MHz. I think you may understand better now. Also, don't forget inrush currents on the larger values and you should bypass those with .... what you started with. :cannotbe:

-Chris
 
I just realized that the front end of the 5400 is more similar to Aleph 0 than A75. I have not heard an Aleph 0. How does it compare with the X amps? I have heard a X5 (assuming that it has the X series sound) and was very impressed.

I am running my 5400 with about 0.8A of bias (with the help of 4 computer fans at a very low speed). It sound very good but is slightly laid back for my taste. The output stage has three pair of Mosfets. Will adding additional output Mosfets make it less laid back (improving treble performance)?
 
I'm not sure what you mean when you say that the treble isn't good enough, but adding more MOSFETs to the output stage will generally decrease the high end due to added Gate capacitance; probably not what you want to do.
If you're using high feedback amps as a reference, thinking that's the way high frequencies are supposed to sound, then you're barking up the wrong tree entirely. You'll need to increase the feedback on the Adcom to get that sound. I'm not a big fan of that sort of upper midrange to treble glare, but some people are convinced that it's really neat. Increasing the feedback (assuming you want to go that route) is problematic, in that you'll lose gain and may run into stability problems. Essentially, you're looking at redesigning the front end...at which point the amp ceases to be an Adcom and becomes something else. That will be a pretty involved project unless you can figure out some way to retrofit an existing design into your amp.
Film caps in the power supply can work wonders. My main tube amps use almost 1000uF of film caps--and only film caps--in the second leg of a PI filter. Sounds very nice. It also cost a young fortune. If you want to get serious about adding film caps to your power supply, see what Solen has to offer, but at all times keep in mind the fact that film caps get really big, really fast. Cramming 10, 20 or 100uF of film caps into an existing amp can be a problem in the purely physical sense.

Grey
 
GRollins said:
I'm not sure what you mean when you say that the treble isn't good enough, but adding more MOSFETs to the output stage will generally decrease the high end due to added Gate capacitance; probably not what you want to do.
If you're using high feedback amps as a reference, thinking that's the way high frequencies are supposed to sound, then you're barking up the wrong tree entirely. You'll need to increase the feedback on the Adcom to get that sound. I'm not a big fan of that sort of upper midrange to treble glare, but some people are convinced that it's really neat. Increasing the feedback (assuming you want to go that route) is problematic, in that you'll lose gain and may run into stability problems. Essentially, you're looking at redesigning the front end...at which point the amp ceases to be an Adcom and becomes something else. That will be a pretty involved project unless you can figure out some way to retrofit an existing design into your amp.
Film caps in the power supply can work wonders. My main tube amps use almost 1000uF of film caps--and only film caps--in the second leg of a PI filter. Sounds very nice. It also cost a young fortune. If you want to get serious about adding film caps to your power supply, see what Solen has to offer, but at all times keep in mind the fact that film caps get really big, really fast. Cramming 10, 20 or 100uF of film caps into an existing amp can be a problem in the purely physical sense.

Grey

Grey,

Thanks for the reply.

I compared the sound with the headphones directly connected to the Aleph P and to that connected to the 5400 output (with the Aleph P still in the chain). Through the 5400 the treble (and perhaps the high mids) is slightly recessed and not as delicate. I did increase the feedback to reduce the gain from 29db to about 19 db, which is comparable to that of Aleph 0. Before I reduced the gain there was a lot of glare in the midrange and treble, perhaps the open loop gain is not high enough, resulting in too little feedback.

The reason I am thinking about adding more output Mosfets is that
Aleph 0 (with more output Mosfets) runs at a much higher bias (2A). Therefore, the output stage has a much higher transconductance, resulting in lower distortion. The 5400 runs at a much lower bias and the source resister is higher, resulting in a lower transconductance and higher distortion.

I did eliminate some polyester caps, making the amp DC coupled. I also replaced the other polyester caps with polypropylene ones, resulting in significant improvement in the midrange and treble. I increased the power supply caps from 48000u to 108000u, which tightened up the bass. But adding more Mosfets will be more involved and I would like to hear suggestion.

I wonder how large is your 1000u of film cap. Are they polypropylene?
 
A few days ago I used a variac to reduce the AC supply and the DC rail (at 40V) and increased the bias to 1.3A with great results. The treble is much more transparent. Now I am considering adding four to six more output devices per channel so I can further increase the bias (A75 is biased at 2A) and at the same time reduce the dissipation per device. The 5400 uses IRF9240 for the P channel devices but I am considering adding Fairchild FQA12P20. Would it be problematic to mix IRF9240 and FQA12P20 in the same channel?
 
The one and only
Joined 2001
Paid Member
I believe that Fairchild has acquired the Samsung Mosfet production line, so
they would be Fairchild parts. It is possible that they will have the character
of the Samsung versions of those parts, but they are likely as good as the
IR parts. Getting the US versions is hit-and-miss.

:cool:
 
I measured the Vgs of some IRFP240 and FQA12PP20 at 570mA. The figures for IRFP240 are around 4.20V but those for FQA12PP20 are around 5.12V. Given the difference between the IRFP240 and FQA12PP20, can they even be used as complimentary pairs in the same channel?

BTW, the ones currently in the 5400 are around 3.88V. So I guess need to replace these instead of adding more to them.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.