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.
I have an adcom gfa-5400 whose design I think is based on the A75. However, the front end is only half of the A75's complementary circuit. The output stage is biased AB rather than A. I have not heard the A75. What sonic difference between the two do you expect? Are there any obvious ways to improve the 5400 (e.g., increasing bias?)

Do the bigger adcom's (5500 and 5800) have full complementary front end like the A75?
 
Saying that an Adcom is based on the A-75 is like saying that a pickup truck is based on a Porsche. Well, yes, they both have internal combustion engines, four wheels, and upholstery inside, but neither really has that much to do with the other; neither design is derived from the other's.
The complementary output follower stage is the most common output stage around. The only question is how hard it's biased: class B, AB, or A. Front ends vary somewhat more, but the most common front end design will have either one or two differentials, followed by a second stage (often called the VAS), which leads to the output stage.
As far as trying to hot-rod a commercial amp, you're pretty much limited by the original circuit unless you intend to gut the thing and use the chassis to build another amplifier entirely.
Feasible ugrades include:
--Increased bias. It's a rare case where you can take this much further than the original circuit. The available heatsinking and power supply are pretty hard limits. But if you can increase bias, it's generally a good idea, sound-wise. Speaking of wise, it's always good to remember that you can destroy your amp by pushing the bias too far.
--Upgrade caps. If you see an electrolytic, you might be able to get a film cap in...or you might not. Film caps are much larger (and more expensive) but sound better. Ceramic disc caps are obvious candidates for replacement also. As are tantalum caps.
--Increase capacitance in the power supply. This also runs into limits, both in terms of physical size (unless you intend to outboard the caps) and inrush current (which can be dealt with using varistors or soft start circuits).
DO NOT start randomly replacing active devices without thinking it through thoroughly. DO NOT start randomly changing resistors. If you want to redesign the front end, then go at it with a detailed game plan--including the contingency of going back to the original circuit if your circuit doesn't work.
Always keep in mind the dictum: First, do no harm.

Grey
 
GRollins said:
Saying that an Adcom is based on the A-75 is like saying that a pickup truck is based on a Porsche. Well, yes, they both have internal combustion engines, four wheels, and upholstery inside, but neither really has that much to do with the other; neither design is derived from the other's.
The complementary output follower stage is the most common output stage around. The only question is how hard it's biased: class B, AB, or A. Front ends vary somewhat more, but the most common front end design will have either one or two differentials, followed by a second stage (often called the VAS), which leads to the output stage.
As far as trying to hot-rod a commercial amp, you're pretty much limited by the original circuit unless you intend to gut the thing and use the chassis to build another amplifier entirely.
Feasible ugrades include:
--Increased bias. It's a rare case where you can take this much further than the original circuit. The available heatsinking and power supply are pretty hard limits. But if you can increase bias, it's generally a good idea, sound-wise. Speaking of wise, it's always good to remember that you can destroy your amp by pushing the bias too far.
--Upgrade caps. If you see an electrolytic, you might be able to get a film cap in...or you might not. Film caps are much larger (and more expensive) but sound better. Ceramic disc caps are obvious candidates for replacement also. As are tantalum caps.
--Increase capacitance in the power supply. This also runs into limits, both in terms of physical size (unless you intend to outboard the caps) and inrush current (which can be dealt with using varistors or soft start circuits).
DO NOT start randomly replacing active devices without thinking it through thoroughly. DO NOT start randomly changing resistors. If you want to redesign the front end, then go at it with a detailed game plan--including the contingency of going back to the original circuit if your circuit doesn't work.
Always keep in mind the dictum: First, do no harm.

Grey

To my eyes the front end of 5400 is almost exactly one half of the A75's. (If the Porsche had two engines then the pickup truck used one of the same).

I think it is safe to say that there are substantial differences between the two amps (the 5400 is somewhat harsh in the mids and treble and the latter is also rolled off somewhat). Would one expect this much degradation from using one half of the circuit alone or something else is responsible?

As far as increasing bias I might be able to put in a different heatsink (or a fan, ouch).

Speaking of caps, I noticed that many power amps do not use film caps to by-pass supplies, only electrolytic. Is that because it is not necessary to reduce the supply impedance at high frequencies or because of cost?
 
Nelson Pass said:
At the time I was re-designing the Adcom products, so you
will note some similarities. Keep in mind that the A75 was also
Norm Thagard's design.

:cool:


Thanks for confirming the similarities between the amps.

I just increased the output stage bias of the 5400 from about 190 mA (I was surprised to see how low it was) to about 340 mA. Now it sounds much better overall. The mids improved significantly although the treble is still somewhat rolled off. I left the amp cover open and the heat sink is still not too hot to touch. Maybe I can increase the bias more.
 
Administrator
Joined 2004
Paid Member
Hi ctong,
The Adcom is a pretty good amp, but increasing the bias is not the way to reduce distortion. It's an AB amplifer with overall feedback and there is a sweet spot for bias current. If you can't leave your hand on it, it is too darn hot.

I plan to do a bias vs THD demonstration at Burning Amp, you should attend. As for your amplifier, follow the bias setting procedures in the service manual.

Nelson,
Your hand is visible in many designs. Some merely inspired by your designs.

Did you get a chance to work with Walt Jung at all? I understand he was also involved with Adcom at some point in time. I'm not sure if you fellas overlapped at all.

-Chris
 
The one and only
Joined 2001
Paid Member
Originally posted by anatech The Adcom is a pretty good amp, but increasing the bias is not the way to reduce distortion.

On the Mosfet amps, you can reduce measured distortion by
increasing the bias, but you can't increase it by very much due
to the heat sinking.


Originally posted by anatech Did you get a chance to work with Walt Jung at all? I understand he was also involved with Adcom at some point in time. I'm not sure if you fellas overlapped at all.

We did not. I believe Walt advised them on some op-amp
based preamp circuits.

:cool:
 
GRollins said:
There's more to it than the temperature of the heatsinks. If the rails collapse, you're going to generate more distortion and problems.

Grey

The supply should be able to handle more than 2A per channel (approximately 220W). I doubt that I can increase the idle dissipation to more than 30W per channel due to the size of the heat sink unless I use a fan.
 
Administrator
Joined 2004
Paid Member
Hi ctong,
I doubt that I can increase the idle dissipation to more than 30W per channel due to the size of the heat sink unless I use a fan.
Don't forget about the thermal fuse in the transformer.

Hi Nelson,
I forgot this was a mosfet. I'm working on a BJT version now. Duh!

I did get some of the mosfets from the new series parts kits. One day I may try a Mosfet front end.
I believe Walt advised them on some op-amp
Yes, that is correct. I was just wondering. My guess is that his project was over by the time you had an interest.

-Chris
 
anatech said:
Hi ctong,
The Adcom is a pretty good amp, but increasing the bias is not the way to reduce distortion. It's an AB amplifer with overall feedback and there is a sweet spot for bias current. If you can't leave your hand on it, it is too darn hot.

I plan to do a bias vs THD demonstration at Burning Amp, you should attend. As for your amplifier, follow the bias setting procedures in the service manual.

Nelson,
Your hand is visible in many designs. Some merely inspired by your designs.

Did you get a chance to work with Walt Jung at all? I understand he was also involved with Adcom at some point in time. I'm not sure if you fellas overlapped at all.

-Chris

What is Burning Amp and where is it going to be?
 
I tried to increase the bias further and ended up with 370mA when the heatsink temperature was about 50C. The sound is still not satisfactory. If I want even higher bias I will need a fan or add another bank of mosfets with another heatsink. I wonder how much improvement I will get.
 
anatech said:
Hi ctong,

Not much I don't think.

I don't believe you are going in the right direction to solve your problem. What are you trying to eliminate in the sound? The front end has a lot to do with how an amplifier sounds.

-Chris

The sound is somewhat harsh and the treble is slightly rolled off (or maybe not as transparent?) I am interested in finding out how much of the problem is due to the relatively low bias and maybe I can infer the performance of the front end.
 
anatech said:
Hi ctong,
My guess here would be capacitors and transistor matching. Any ceramics should be replaced by mica, polypropylene or teflon. That should do it. Return the bias to it's original level.

-Chris

I have not looked for all the ceramics. I did notice some polyester coupling caps for the input and feedback. I know they are not good in signal paths. But is a coupling cap considered in a signal path? There is not a significant voltage across it.
 
Administrator
Joined 2004
Paid Member
Hi ctong,
Any cap in the signal path is ... well, in the signal path. What you are talking about is the area where some capacitors behave rather badly. Believe me, they can misbehave simply due to the current through them. It's just less of an effect, that's all. You are looking for a small change, not a gross change that causes great distortion.

-Chris
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.