Hi All,
I need more amp for some new speakers (100W, 91db/2.82V) and I'm thinking of foregoing my typical affinity for tube amps to try something different. I'd like to hit 100W at 4 ohms just to be sure I have maximum headroom for peaks but this is negotiable. The F5 is a nice easy build that gets me 40W at 4 ohms but I'm worried I will regret not maxing out my headroom. A turbo V1 gets me to my the target wattage but has more unknowns for me to deal with. So, I'm a bit torn between the ease of building a standard F5 and the potential complications of building an F5T (note that turbo boards are out of stock as of this post). I've read through the F5 and F5T docs several times and burned though hundreds of pages of posts and it's all sloshing around in my head so please help me get a few things clear:
So, what you can probably ascertain from my questions is whether or not I can build a standard F5 today and keep a relatively short path to a turbo later, or if I can do something between a standard and turbo V1 now. For example, I could see myself hacking in a second set of output devices later without swapping out the boards, but adding cascodes not so much. Swapping out the transformer and filtering would be costly but easy, a new chassis not so much.
Sorry for the lengthy post but I've already spent several days running in circles and I find myself unable to pull the trigger on parts.
Brian
I need more amp for some new speakers (100W, 91db/2.82V) and I'm thinking of foregoing my typical affinity for tube amps to try something different. I'd like to hit 100W at 4 ohms just to be sure I have maximum headroom for peaks but this is negotiable. The F5 is a nice easy build that gets me 40W at 4 ohms but I'm worried I will regret not maxing out my headroom. A turbo V1 gets me to my the target wattage but has more unknowns for me to deal with. So, I'm a bit torn between the ease of building a standard F5 and the potential complications of building an F5T (note that turbo boards are out of stock as of this post). I've read through the F5 and F5T docs several times and burned though hundreds of pages of posts and it's all sloshing around in my head so please help me get a few things clear:
- The F5 doc specs distortion at 40W into 4 ohms as .2% but does it clip there or is there a little more power albeit at a slightly higher distortion level?
- Can the 4U deluxe chassis from the store handle the dissipation of F5T V1 if I ever want to reuse or modify?
- Can the Linear Systems jfets in the store handle the higher rail voltages of the turbo without cascoding or is that reserved for Toshiba parts only (with some established risk)?
- Early in the turbo document, after suggesting the higher rail voltages in a standard F5 circuit, Mr. Pass says, "Want to drive 4 ohms with this? You can do it, but I recommend that you consider the upgrade to the F5 Turbo V1". I'm wondering what are the consequences of running 4 ohms at higher voltages - particularly if you are not pushing the volume most of the time.
- Would parallel output devices or reduced source resistors increase power into 4 ohms at all? I'm thinking not, but would like confirmation.
So, what you can probably ascertain from my questions is whether or not I can build a standard F5 today and keep a relatively short path to a turbo later, or if I can do something between a standard and turbo V1 now. For example, I could see myself hacking in a second set of output devices later without swapping out the boards, but adding cascodes not so much. Swapping out the transformer and filtering would be costly but easy, a new chassis not so much.
Sorry for the lengthy post but I've already spent several days running in circles and I find myself unable to pull the trigger on parts.
Brian
Last edited:
All I can say is I'm running a standard F5 with a bit more bias voltage, 90dB efficiency 6 ohm nominal speakers, and I have plenty of power for my tastes. People are surprised that it's a 25W amp.
With your listening habits do you find yourself pushing it very hard? I'm curious if the onset of clipping is very obvious. If you are cranking it all the time and can't tell when it's hitting the limits then I guess I can relax and build a standard F5 with no worries.
Last edited:
You may have to cascode the front end to use higher power supply voltages.
The LS parts are only rated at 25V.
The LS parts are only rated at 25V.
100W sinewave into 4ohm is 5Arms, in other words a peak current of 7A, requiring 7A of bias. The Turbo V1 isn't even close to this, but this thread seems appropriate: F5 Turbo V1, modified for more Class A power into 4 ohms
[ 7A bias with 32V rails is nearly 500W of dissipation. ]
[ 7A bias with 32V rails is nearly 500W of dissipation. ]
I had it driving a pair of DQ-10s, very power-hungry speakers with a TU-8500 preamp at top volume, and it sounded wonderful. It takes a lot of preamp to make it clip.
With my regular 90dB speakers I can't remember wanting more power.
I’m not looking to hit 100W in class A. Correct me if I’m wrong but I don’t think any of the turbo variants promise that and I thought that even the standard transitioned to class B on peaks as well.
You can bias it hot to keep it in class A. You just need enough sink. Sinks on the JFETs as well.
Sure, but I’m more interested in being able to achieve more total output using as much of my initial F5 build as possible (or concluding that a standard F5 won’t meet my needs). If anything more class A bias seems to be antithetical to this.
100 W into a speaker with 91 dB sensitivity is a lot. A LOT. Too much for a living room?
What impedance level is your speaker? Do you have an impedance curve?
That said, my F5 is very much like a tube amp. If you want something different, you might want to look at something like the Honey Badger. They are typical solid-state amps (in a good way).
Troels has an impedance curve here:
SBA-16-MTM
I would never listen fully cranked at 100W, I just to make sure that less compressed music can play loud without offensive clipping on peaks. In truth I’m sure I’m over inking things... this is a very popular DIY amp. I’m going to build a standard F5 and go from there.
SBA-16-MTM
I would never listen fully cranked at 100W, I just to make sure that less compressed music can play loud without offensive clipping on peaks. In truth I’m sure I’m over inking things... this is a very popular DIY amp. I’m going to build a standard F5 and go from there.
Ok, that's a "4 Ohm" speaker. You'll want to tweak your amp towards high current output rather than high voltage. I'd say make the heatsinks large and crank up the bias current as much as possible.
The impedance curve is not perfectly flat, but no "sharp" peaks overall. I wouldn't worry.
The impedance curve is not perfectly flat, but no "sharp" peaks overall. I wouldn't worry.
Here’s the big thing: what volume levels do you listen at? Take out your phone, download a free spl meter and see.
Some people think 85db at the listening seat is really rocking the house.
Other people think it’s like 100db.
For others, they never go over 75db.
Unless you are looking for 90-95db levels I would think an F5 standard would get you there.
Even better would be a standard f5 (24v rails) with dual pair of outputs for more current into 4 ohms.
I understand you don’t care about class a “envelope” or operating range. But really that’s what these amps are all about. If you are going to do a low bias, high wattage F5 I would just build something class a/b.
Some people think 85db at the listening seat is really rocking the house.
Other people think it’s like 100db.
For others, they never go over 75db.
Unless you are looking for 90-95db levels I would think an F5 standard would get you there.
Even better would be a standard f5 (24v rails) with dual pair of outputs for more current into 4 ohms.
I understand you don’t care about class a “envelope” or operating range. But really that’s what these amps are all about. If you are going to do a low bias, high wattage F5 I would just build something class a/b.
That’s good advice about measuring listening levels. I could, and should, have done that. Much better frame of reference.
Can you quantify the increase in output power? I figured that because I’m working with the same rails as a F5 standard the amp won’t be asked to deliver more current. I asked about it because I’m curious but I don’t understand where the benefit is coming from.
Please! I didn’t say I didn’t care about class A envelope! If I didn’t I wouldn’t be here. What I said was that heating up the bias on a standard F5 wasn’t what I came to ask about, and it’s covered elsewhere. Dropping bias from 1.3 to 1 amp is exactly what Mr. Pass describes in the turbo document and doesn’t seem to be in conflict with the spirit of the design. Don’t misunderstand: I want a great low wattage amp first and foremost; I just wanted to probe a bit about where I can go after I build it if I need a bit more. Thank you
Can you quantify the increase in output power? I figured that because I’m working with the same rails as a F5 standard the amp won’t be asked to deliver more current. I asked about it because I’m curious but I don’t understand where the benefit is coming from.
Please! I didn’t say I didn’t care about class A envelope! If I didn’t I wouldn’t be here. What I said was that heating up the bias on a standard F5 wasn’t what I came to ask about, and it’s covered elsewhere. Dropping bias from 1.3 to 1 amp is exactly what Mr. Pass describes in the turbo document and doesn’t seem to be in conflict with the spirit of the design. Don’t misunderstand: I want a great low wattage amp first and foremost; I just wanted to probe a bit about where I can go after I build it if I need a bit more. Thank you
I was going to ask if you looked at the M2. But since the thread is about the F5 I'll keep quiet.
I think so. I’m pretty sure I scrubbed the whole “catalog” and ended up with the F5 as the best fit. It seems like Mr. Pass has something for everyone.
I've got a pair of Focal 936's which are about 92dB sensitive with an impedance low point of 2.8 ohm. I've been driving them with an F5 topology running 32V rails for several years now. I recently broke them into monoblocks and lowered the source resistance from .47R to .28R and marginally increased the bias over the stereo build to about 1.41 amp. Crunching the numbers, that gets me into the 90's dB while still staying inside the class A envelope at those speaker's impedance low point. That's actually quite a bit of mid-bass. Getting smacked in the face with 100 Hz at about 93dB will get your attention. I don't believe that ANYBODY listens at a sustained median volume of 93dB. That is getting uncomfortably loud and your ear is beginning to attenuate sound at that point so you're going to hear a little less detail. My F5 with those speakers has produced kettle drum strikes that have made me literally jump out of my chair. I don't have any use for amps more powerful than those.
If anyone NEEDS more current to drive an even tougher, less sensitive load, then it's probably best to just go for the F5T. I may do something like that in the future, but so far I've had no trouble at all with these amps driving those speakers reliably for years.
If anyone NEEDS more current to drive an even tougher, less sensitive load, then it's probably best to just go for the F5T. I may do something like that in the future, but so far I've had no trouble at all with these amps driving those speakers reliably for years.
Let me explain...sorry if you already know some of this but here we go:
These amplifiers are basically limited in wattage by the rail voltage (PSU voltage). Standard is 24V. Turbos start at 32V. 24V will get you 25W/8 ohm and 32V will get you 50W/100W.
More is better right? Well, not really. Because as we increase rail voltage the amplifier creates more heat. More than the heatsinks can handle. So we must turn down the bias. Which gives us less class A envelope ESPECIALLY into low ohm loads. The difference in sound between 1 amp bias and 2 amp bias into a 4 ohm load with this amplifier is DRAMATIC.
These amplifiers basically run all out hard until they run out of rail voltage. Consider that the power supply is the size of a 500W A/B amp. So, if you don't need the headroom (again, how loud do you listen?) you are better off with more bias and lower rails. Or, if headroom is important to you you need higher rails at the expense of bias. This won't be a better sounding amplifier...it will simply be an amplifier that lets you listen at the levels you want.
So, how can we have it both ways? Well, we need a large chassis. Maybe two. That way we can have rail voltage and bias. We need more mosfets, to spread the load. Monoblocks are one way. A 5U chassis and dual pairs of mosfets will get you 2.5A bias with 24V rails or 1.8A bias w/32V rails.
I have built the standard f5, an f5 with 24V rails and dual pairs of mosfets and an f5 with 32V rails and dual pairs of mosfets. Best sounding was 24V w/dual pairs and 2.5A bias. I am currently using the 32V rail version with dual pairs though. Because I like to listen at 90db+.
Honestly the difference between 25W and 50W is not as great as you would think though. If I listened at 85db or so I would be fine. But agin, this is room and speaker dependent.
If it was me I'd start with the biggest chassis you are comfortable with (most bias and most room for future experimentation) and try an f5. A lot has been written has the f5 is "tube-like". I don't think it is. It sounds great. But it doesn't sound like tubes...which is both good and bad Try it out first, see if you like it.