Nelson Pass said:I like the Panasonics. However if you are looking for the
ultimate (currently available) I think the Hovlands are getting
a lot of praise. I have to admit, however, that I take this from
people I trust, as opposed to personal experience.
Nelson,
I have a question regarding metalized polyester caps. Do they performe better than the regular polyester caps? It is hard and expensive to obtain polypropylene caps of large values. And they are physically large.
Try www.capacitors.com. There's an excellent paper there about capacitors. As a general rule, film and foil is better than metallized, and polyester is considered to be inferior to polypropylene, which is in turn inferior to polystyrene. Teflon is better still, but wasn't available when the paper was written.
The size and expense can't be helped. There's no shortcut.
Try WIMA polypropylene film and foil caps. They're difficult to get here in the US, but Mouser carries some of their values. They are more expensive than the polyester caps Nelson is describing, but far less than the boutique caps. Yes, they're large.
Grey
The size and expense can't be helped. There's no shortcut.
Try WIMA polypropylene film and foil caps. They're difficult to get here in the US, but Mouser carries some of their values. They are more expensive than the polyester caps Nelson is describing, but far less than the boutique caps. Yes, they're large.
Grey
Hi Grey,
The thing about larger caps is that the size can then become a source of trouble. Everything from heat to noise pickup (or radiation). If you are designing from scratch you can compensate for the size, but "upgrading" an existing unit can get you into trouble.
For upgrades, make sure the size of the capacitor is compatible with the space you have to play with. Do not ignore possible thermal issues.
-Chris
The thing about larger caps is that the size can then become a source of trouble. Everything from heat to noise pickup (or radiation). If you are designing from scratch you can compensate for the size, but "upgrading" an existing unit can get you into trouble.
For upgrades, make sure the size of the capacitor is compatible with the space you have to play with. Do not ignore possible thermal issues.
-Chris
Agreed. I noted that the WIMA polypropylene caps are large. Unfortunately, there's no easy solution to this. You can stick to the original footprint and lower the capacitance so it will fit (may work okay, may not, depending on where it is in the circuit), or you can squeeze something bigger into place. Other than that, you're pretty much stuck.
Grey
Grey
Hi ctong,
They are part of the protection. Leave them be.
There are local capacitors to decouple the amp circuits. I wouldn't worry too much about the fuses. Also, 5A has far less effect than say, 2A or less.
I fully expect someone to pop up and tell us how much open these sound with the fuses shorted out.
-Chris
They are part of the protection. Leave them be.
There are local capacitors to decouple the amp circuits. I wouldn't worry too much about the fuses. Also, 5A has far less effect than say, 2A or less.
I fully expect someone to pop up and tell us how much open these sound with the fuses shorted out.
-Chris
Hi ctong,
I don't understand the intent of your question at all.
If you use normal speakers with reasonable reactance without overdriving the amp, then I would doubt it. Normal speakers would be 4 ohm minimum rated impedance and high 80's or better efficiency.
If you use a silly speaker like a Kappa 9 or Ohm Walsh, then yes. And you deserve the results of a poor combination.
You must accept that circuit design is all about comprise. The one thing that should never be compromised are the safety or protection networks built into these devices.
I have noticed over many years of watching people and things that the least reliable (and poorly designed) amplifiers tend to be purchased by silly people chasing the myth of perfect sound. These items are almost always left on 24/7 and do not incorporate adequate protection circuits. It's like they are blinded by the nonsense spewed out from the manufacturer of said liability. The very best example of this would be the Bryston 4B.
So, what exactly are you asking about in this amp that would possibly degrade the sound and under what circumstances?
-Chris
I don't understand the intent of your question at all.
If you use normal speakers with reasonable reactance without overdriving the amp, then I would doubt it. Normal speakers would be 4 ohm minimum rated impedance and high 80's or better efficiency.
If you use a silly speaker like a Kappa 9 or Ohm Walsh, then yes. And you deserve the results of a poor combination.
You must accept that circuit design is all about comprise. The one thing that should never be compromised are the safety or protection networks built into these devices.
I have noticed over many years of watching people and things that the least reliable (and poorly designed) amplifiers tend to be purchased by silly people chasing the myth of perfect sound. These items are almost always left on 24/7 and do not incorporate adequate protection circuits. It's like they are blinded by the nonsense spewed out from the manufacturer of said liability. The very best example of this would be the Bryston 4B.
So, what exactly are you asking about in this amp that would possibly degrade the sound and under what circumstances?
-Chris
Chris,
What I meat was whether the protection circuit by its presence had some sort of effects on the signal.
I just shorted the input coupling caps and the the caps in the feedback loop. I readjusted the DC offset at the output, which starts at around 200mV cold and drops to around 10mV after 10-15 minutes of warm-up.
The bias is set to around 500mA and I am using a fan for each heatsink.
The sound improved quite a bit but the treble is not as smooth compared to that through my headphone. I wonder how much I would gain by further increasing the bias. What is the bias in X150 or X250?
What I meat was whether the protection circuit by its presence had some sort of effects on the signal.
I just shorted the input coupling caps and the the caps in the feedback loop. I readjusted the DC offset at the output, which starts at around 200mV cold and drops to around 10mV after 10-15 minutes of warm-up.
The bias is set to around 500mA and I am using a fan for each heatsink.
The sound improved quite a bit but the treble is not as smooth compared to that through my headphone. I wonder how much I would gain by further increasing the bias. What is the bias in X150 or X250?
Hi ctong,
You still have the transformer core temperature to worry about (and the thermal fuse), plus you are not gaining anything. Also, please please please, do not leave it running "24/7". I think that removing the input coupling caps was a bad idea. They were good quality parts.
-Chris
No, not under normal circumstances. That is what I was trying to get across.
Well, if you were going to do that, then add a DC servo along with a manual current injection. The manual adjustment takes care of DC offset conditions when the circuits are cold and the op amp servo will take care of the DC drift. May as well finesse this thing since you are improving it. No more turn on thump.
Put the bias back. The distortion actually increases beyond the optimum setting. What you are experiencing is psychological only. Certain circuit configurations need to run at high bias. Not that one.
You still have the transformer core temperature to worry about (and the thermal fuse), plus you are not gaining anything. Also, please please please, do not leave it running "24/7". I think that removing the input coupling caps was a bad idea. They were good quality parts.
-Chris
anatech said:
I agree that the protection circuit is probably not doing any harm.
The DC offset stabilizes after 10-15 minutes of warm-up so I am not concerned about it. I do not leave it on all the time.
The 5400 has a mosfet follower output stage. Nelson's measurements have shown that it will benefit from higher bias. My listening experience is also consistent. Now I am probably near the limit unless I add more mosfets and heatsink (and maybe another power supply). But I do not know how much improvement I will get for further increasing the bias (say to >1A).
Right now the amp is consuming no more than 150W idle. The rails to the output stage dropped for less than 0.1-0.2V so I am not concerned about the transformer, which remains cool.
I further increased the bias to 700mA and the amp sounds more like what I hear from my headphones. Much better. The effect of higher bias is stunning. I am running the fans at a higher speed and the noise is more than what I consider acceptable. So I think I am going to find some heatsinks and parallel more mosfets to get the bias up to at least 1A. I might have to get another power supply but I will see how this one is holding up.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.