I've built a F5 using the diyaudio 5U case (BIG heatsinks).
It's seen a couple different mods...went from Fairchild Mosfets to IRF's, rewired the internals, power cords, etc. All these things brought subtle changes.
So I decide to go from 1.3A (.6v) bias to 1.5A (.7v) bias.
WOW. That was a big change. Amp has more punch, drive and basically more "wow". Soundstage is bigger and more forward. By far a bigger difference than any of the above mentioned changes.
Bias seems to have a dramatic effect on my F5.
How far can I push it? 1.7A (.8v)? My heatsinks are 45C now measured with a digital meat thermometer. I am not crazy enough to touch the Mosfet with the metal meat thermometer so I don't know the Mosfets temp. I am using K-10 silpads.
Is 1.7A safe? Even if the heatsinks can take it that doesn't mean the source resistors can. Anybody experimenting with higher bias on the F5?
It's seen a couple different mods...went from Fairchild Mosfets to IRF's, rewired the internals, power cords, etc. All these things brought subtle changes.
So I decide to go from 1.3A (.6v) bias to 1.5A (.7v) bias.
WOW. That was a big change. Amp has more punch, drive and basically more "wow". Soundstage is bigger and more forward. By far a bigger difference than any of the above mentioned changes.
Bias seems to have a dramatic effect on my F5.
How far can I push it? 1.7A (.8v)? My heatsinks are 45C now measured with a digital meat thermometer. I am not crazy enough to touch the Mosfet with the metal meat thermometer so I don't know the Mosfets temp. I am using K-10 silpads.
Is 1.7A safe? Even if the heatsinks can take it that doesn't mean the source resistors can. Anybody experimenting with higher bias on the F5?
Technically, the junctions of the devices are rated at 150 deg C. If you are
running as high as 50 watts per device, then you will want to measure well less
than 100 deg C on the case (you can use a cheap Radio Shack IR thermometer).
Don't expect them to have a long life span at those figures, but I have seen
reasonable reliability at 25 deg less junction temp.
running as high as 50 watts per device, then you will want to measure well less
than 100 deg C on the case (you can use a cheap Radio Shack IR thermometer).
Don't expect them to have a long life span at those figures, but I have seen
reasonable reliability at 25 deg less junction temp.
I've played around with loading IRF240/IRF9240 hard some years ago.
You can easily run them at 50W, but as Nelson wrote, the device should be kept at a reasonable temperature.
As I recall, I had mine around 85C case temp, and the heat sink was like 70C or so.
That worked for approx. 1000hrs., till the amp was taken apart, to make room for a new amp.
What I did learn also in that process, is that 50W+ is possible as well, if the heat sinks are big enough, and you are able to mount the device directly on the heat sink, with just thermal compound, and no insulator.
In that very same process, I too found the limit, which caused some rather interesting explosion.
The reason I mention this, is that at those power levels, an IRFP240 simply goes BANG! when it gives up the ghost.
Take that into account when working on the amp (safety glasses), and make sure that the box is closed when the amp is in use, and the vent holes are small enough to keep shrapnel in.
When they blow under that sort of load, all that's left on the heat sink, is the metal tab.
Magura
You can easily run them at 50W, but as Nelson wrote, the device should be kept at a reasonable temperature.
As I recall, I had mine around 85C case temp, and the heat sink was like 70C or so.
That worked for approx. 1000hrs., till the amp was taken apart, to make room for a new amp.
What I did learn also in that process, is that 50W+ is possible as well, if the heat sinks are big enough, and you are able to mount the device directly on the heat sink, with just thermal compound, and no insulator.
In that very same process, I too found the limit, which caused some rather interesting explosion.
The reason I mention this, is that at those power levels, an IRFP240 simply goes BANG! when it gives up the ghost.
Take that into account when working on the amp (safety glasses), and make sure that the box is closed when the amp is in use, and the vent holes are small enough to keep shrapnel in.
When they blow under that sort of load, all that's left on the heat sink, is the metal tab.
Magura
In my experience, active cooling for stuff that needs to be in the listening room, is something invented by the devil in anger.
All of a sudden, things gets expensive and complicated real quick, if it has to be safe, silent, and reliable.
Especially if you run your fets on the edge.
I've seen a few do it, and none that worked in the long haul.
Magura
All of a sudden, things gets expensive and complicated real quick, if it has to be safe, silent, and reliable.
Especially if you run your fets on the edge.
I've seen a few do it, and none that worked in the long haul.
Magura
In my experience, active cooling for stuff that needs to be in the listening room, is something invented by the devil in anger.
All of a sudden, things gets expensive and complicated real quick, if it has to be safe, silent, and reliable.
Especially if you run your fets on the edge.
I've seen a few do it, and none that worked in the long haul.
Magura
Points taken, although pro amps do it all the time, using common muffin fans.
1.5 amps is a lot of bias current.
I found I was getting away with 10mA/transistor and not getting any cross over distortion. I believe Peavey have the same attitude towards bias current, the less the better.
Having a smaller bias current means there is more current to provide real signal power.
It also means the amp runs cool at moderate output levels.
However if you think it sounds better with 1.5amps then go for it.
I found I was getting away with 10mA/transistor and not getting any cross over distortion. I believe Peavey have the same attitude towards bias current, the less the better.
Having a smaller bias current means there is more current to provide real signal power.
It also means the amp runs cool at moderate output levels.
However if you think it sounds better with 1.5amps then go for it.
They also they have all sorts of safety circuitry, that we normally don't see in DIY amps.
If you ask me, they're hardly comparable to a DIY hifi amp.
For a DIY class A amp to work well and be reliable, there are hardly any alternatives to passive cooling, and plenty of it.
Keeping things at a reasonable temperature, makes the entire amp a lot more reliable, and increases the life expectancy dramatically.
If you take a capacitor for instance, you'll see that mean time to failure is divided by 2, for every 10C of increased working temperature.
At least running things hot, asks for some pretty hefty capacitors, if they're meant to survive more than a test run.
Magura
The Aleph amps are based on an active current source, thus not "pure" class A" (or constant bias as we usually call it around here), but yet another active bias Class A.
It's fairly simple to define, as with class A without active bias, the highest efficiency I know of, is just shy of 20%.
Once you're past that figure, you've either done something that requires immediate patent application, or you've left the "no active current source" department.
To get close to the 20% factor, the only way I know of is inductor loading.
As I recall, that gets big, heavy and tremendously expensive real quick.
Magura
where did you get this?
First Watt is a brand not a category.
none of the First Watt amps only stay in class A for the "first watt" and most of the first watt amps ar SE amps. not push-pull.
the F5 gives 25W class A at 8ohm. and around 40W class A/B at 4ohm (13W class A at 4ohm)
Last edited:
Yes, the F5 is 25w/Class A.
Regarding bias here's what I've found:
At 1A, this amp is lacks punch and life. It was about on par w/a T2050 tripath amp I replaced it with, with better highs. This was the burn-in period.
At 1.3A now we are cooking. The amp came alive. It was different.
I am at 1.6ish amps right now. It's better. Not subjectively better. It is better. Mosfets like current I believe Nelson has said. Heatsinks are 45C.
When I first received my 5U diyaudio case I was shocked. It's huge. Wife was mad, friends made fun of me, mocking the extra space inside. Now I am glad I got it. People spend hundreds (thousands?) on capacitors, resistors, wire, interconnects, etc. that aren't going to make as much difference as jacking up the bias. I'd rather have more heatsink so I could bias higher than a $700 power cable.
Regarding lifespan, yes, it's probably shorter. I can live with that. The DIY stuff I build never lasts more than a few years before I replace parts with something better or move on to something else. My caps are 105 degree computer grade and they are far from the sinks. I don't worry at all about it. What I do worry about is blowing a Mosfet and sending DC to the speakers. But that is the risk you take in DIY. I figure with the $$$ I have saved doing DIY I can take that risk.
I am going to get an IR temp guage and measure the Mosfet. If it's not to hot I am going to go up to 800mv (1.7A) bias. I am sure the heatsink is enough but I want to make sure the mosfet is thermally well-coupled to the sink first. I am using K-10 silpads which some say are not as good as mica.
As-is this is the finest amp I have heard. Ever. I haven't heard them all, or even a lot of them but I enjoy it a lot. And I built it. So if it breaks I can fix it.
Regarding bias here's what I've found:
At 1A, this amp is lacks punch and life. It was about on par w/a T2050 tripath amp I replaced it with, with better highs. This was the burn-in period.
At 1.3A now we are cooking. The amp came alive. It was different.
I am at 1.6ish amps right now. It's better. Not subjectively better. It is better. Mosfets like current I believe Nelson has said. Heatsinks are 45C.
When I first received my 5U diyaudio case I was shocked. It's huge. Wife was mad, friends made fun of me, mocking the extra space inside. Now I am glad I got it. People spend hundreds (thousands?) on capacitors, resistors, wire, interconnects, etc. that aren't going to make as much difference as jacking up the bias. I'd rather have more heatsink so I could bias higher than a $700 power cable.
Regarding lifespan, yes, it's probably shorter. I can live with that. The DIY stuff I build never lasts more than a few years before I replace parts with something better or move on to something else. My caps are 105 degree computer grade and they are far from the sinks. I don't worry at all about it. What I do worry about is blowing a Mosfet and sending DC to the speakers. But that is the risk you take in DIY. I figure with the $$$ I have saved doing DIY I can take that risk.
I am going to get an IR temp guage and measure the Mosfet. If it's not to hot I am going to go up to 800mv (1.7A) bias. I am sure the heatsink is enough but I want to make sure the mosfet is thermally well-coupled to the sink first. I am using K-10 silpads which some say are not as good as mica.
As-is this is the finest amp I have heard. Ever. I haven't heard them all, or even a lot of them but I enjoy it a lot. And I built it. So if it breaks I can fix it.
