A lot depends on the other components too....
What is the transformer, what sort of diodes for the bridge, what sort & value of supply caps and last, what series impedance of your new 2mH choke (or what voltage drop is a "little"?)
Presumably, dropping 0.5 volt on the O/P power resistors, (which are 0.4 Ohms, yes?) means running the amp at about 1+ amp bias - a bit low. The Fets sound better when above 70*C (or when the heatsink is above 50*C) which is definitely not "pretty cool" at all.
What is the transformer, what sort of diodes for the bridge, what sort & value of supply caps and last, what series impedance of your new 2mH choke (or what voltage drop is a "little"?)
Presumably, dropping 0.5 volt on the O/P power resistors, (which are 0.4 Ohms, yes?) means running the amp at about 1+ amp bias - a bit low. The Fets sound better when above 70*C (or when the heatsink is above 50*C) which is definitely not "pretty cool" at all.
If anything I would have expected you to have more rail voltage with CLC than CRC.
In a CRC supply, the voltage drop is I_bias x R.
In a CLC supply, the L should theoretically have 0 resistance and hence no voltage drop.
If you end up with less voltage with CLC than CRC, then it can only imply that the DC resistance of your coil is even higher than the R you use in you CRC. If this is indeed the case, all it means is that you have now a cleaner supply than before, due to increased filtering both with C-L-C and C-R_coil-C.
In a balanced F5, such as the F5X, the amplifier draws constant current in balanced mode in the entire Class A range. So the dynamic impedance of the power supply becomes much less critical.
Patrick
In a CRC supply, the voltage drop is I_bias x R.
In a CLC supply, the L should theoretically have 0 resistance and hence no voltage drop.
If you end up with less voltage with CLC than CRC, then it can only imply that the DC resistance of your coil is even higher than the R you use in you CRC. If this is indeed the case, all it means is that you have now a cleaner supply than before, due to increased filtering both with C-L-C and C-R_coil-C.
In a balanced F5, such as the F5X, the amplifier draws constant current in balanced mode in the entire Class A range. So the dynamic impedance of the power supply becomes much less critical.
Patrick
If you have a good quality scope try comparing the input signal to the output signal via inverting one channel (assuming the amp does not invert) and putting the amps into "add", this will give you the difference signal on the trace. Given that this is not an ultra low distortion amp one ought to be able to both see the trace pretty well with the vertical amps cranked up, then you can adjust.
One may see it go from sinusoidal to some other shape, OR change in amplitude with bias and other adjustments being made. Not quite the same as getting a number or seeing the curve of distortion vs. power, but could be quite revealing and useful.
_-_-bear
One may see it go from sinusoidal to some other shape, OR change in amplitude with bias and other adjustments being made. Not quite the same as getting a number or seeing the curve of distortion vs. power, but could be quite revealing and useful.
_-_-bear
With your choke's dcR of 0.31 Ohms, the voltage drop should be in the vicinity of about 0.5Volts on your rails if your current is about 1.5A, not the 2.5V that you are getting - suggest you measure the voltage drop across the new choke itself to check this out.
As Patrick mentioned, the choke will sound a bit different (IMO) but I would crank the bias current back up to get some heat back into those output Fets.
A suggestion - change out the chinese block rectifier for one of those IXY Shottky ones (from Parts Conn?) - not cheap but BIG improvement - or maybe use 4 individual Shottky fast, soft recovery types.
bcs caps = Vishay (Phillips) BC caps? They tend to be a bit "fatter" in the bass generally (IMO again!) similar to the old Roe caps, F&Ts, etc - should settle down a bit with use.
All the best.
As Patrick mentioned, the choke will sound a bit different (IMO) but I would crank the bias current back up to get some heat back into those output Fets.
A suggestion - change out the chinese block rectifier for one of those IXY Shottky ones (from Parts Conn?) - not cheap but BIG improvement - or maybe use 4 individual Shottky fast, soft recovery types.
bcs caps = Vishay (Phillips) BC caps? They tend to be a bit "fatter" in the bass generally (IMO again!) similar to the old Roe caps, F&Ts, etc - should settle down a bit with use.
All the best.
Ok I just went and remeasured the rails. They are 22.5. I am alse getting a. 7v drop accross the inductor.
I have thought about swaping the rectifier but I wasnt sure it was going to make much of a difference for the money.
I assume that is 0.7V drop across the choke (it'd be rather hot with a 7 volt drop!) So, if the current thru the amp is still biased up at about 1+ amp, the total series Z for the choke is about 0.5 Ohms or thereabouts - would be better at the 'aforementioned' 0.31R.
However, let it settle down and see how it functions in practice before chasing up other bits, I'd suggest.
Now, about those d*** bridges - literally, every component in this amp has a direct effect on it's final sound, some more than others - the bridges are one of the "more" effect bits and definitely worth the extra $s - oh yeah, keep them as COOL as possible - "Cool is Good" for diodes and "Hot is Better" for power Fets!.
Also, use good quality internal wiring, not the 'often used' scraps from mains power leads - these small things are easy to do and they all add up to a better amp.
In the meantime, possibly try adding your previous 0.1 Ohm resistor between the bridge and the first cap and see if it takes some of the "hard edge" of the upper mids/treble particularly noticeable if you like jazz with brass instruments in it, and song mixes like Celine Dion's, Tori Amos, etc.
Be patient with the BC caps - they also take awhile to settle down and the mid clarity improves after about 40 hours or so and the bass comes together a lot more after about 100 - lot of guys say this's all rubbish, but IMO it's another of those small things ...
Good amp design, this one, and simple - rewards a bit of care and attention - P3 is also a good addition.
However, let it settle down and see how it functions in practice before chasing up other bits, I'd suggest.
Now, about those d*** bridges - literally, every component in this amp has a direct effect on it's final sound, some more than others - the bridges are one of the "more" effect bits and definitely worth the extra $s - oh yeah, keep them as COOL as possible - "Cool is Good" for diodes and "Hot is Better" for power Fets!.
Also, use good quality internal wiring, not the 'often used' scraps from mains power leads - these small things are easy to do and they all add up to a better amp.
In the meantime, possibly try adding your previous 0.1 Ohm resistor between the bridge and the first cap and see if it takes some of the "hard edge" of the upper mids/treble particularly noticeable if you like jazz with brass instruments in it, and song mixes like Celine Dion's, Tori Amos, etc.
Be patient with the BC caps - they also take awhile to settle down and the mid clarity improves after about 40 hours or so and the bass comes together a lot more after about 100 - lot of guys say this's all rubbish, but IMO it's another of those small things ...
Good amp design, this one, and simple - rewards a bit of care and attention - P3 is also a good addition.
I want to say that the amp still sounds very good. I just wanted to point out how switching from CRC to CLC changed the sound subjectively for anyone who is considering what type of PS they want. I love the bass increase, it sounds much more balanced and weighty. I do not like the smooth/lower dynamics sound as much but until I re-bias the amp I am not sure if this is solely due to the switch to clc.
yes it's .7V drop. They are 14g solen perfect lay inductors.
I do understand that in a simple amp everything counts. This is my first amp build, and I have learned a lot along the way. All my wire is 14g or larger and is all copper. My bridge rectifiers are mounted to the case to keep cool. Now that I know it is a worthwhile improvement, what is a decent rectifier to try?
Is this worth the money or are there better options?
http://www.partsconnexion.com/product4794.html
What about this guy?
VS-GBPC3508A Vishay Semiconductors Bridge Rectifiers
Or is there something better out there for a good price?
yes it's .7V drop. They are 14g solen perfect lay inductors.
I do understand that in a simple amp everything counts. This is my first amp build, and I have learned a lot along the way. All my wire is 14g or larger and is all copper. My bridge rectifiers are mounted to the case to keep cool. Now that I know it is a worthwhile improvement, what is a decent rectifier to try?
Is this worth the money or are there better options?
http://www.partsconnexion.com/product4794.html
What about this guy?
VS-GBPC3508A Vishay Semiconductors Bridge Rectifiers
Or is there something better out there for a good price?
OK please ignore my previous review. It turns out that something was wrong with one of the rectifiers I was using, which was why my rails went down instead of up. I replaced my rectifiers and now my rails are closer to 30V and the amp sounds better then it has ever sounded before. 30V seems like to big of a jump for me but after switching from CRC to CLC + replacing junk rectifiers it sounds reasonable. I was hoping I was going to be able to give a subjective compare and contrast review of CRC vs. CLC in an F5 amp but now I don't trust what I was hearing before.
Here is what I can say now that the amp is working properly, I hope
It sounds HUGE on music like Shpongle. It has great bass, detail and dynamics. I honestly don't know how I am going to top this amp on my next build. I am not sure how much of this is due to CLC but I will be using CLC on my next class A amplifier build.One thing I have learned, I will never buy chinese parts again.
30 volt rails from an 18v transformer ...?
I may have that wrong, i have to go check when I get a chance.
Just to leave an update. I had the CL60's both going to the same coil on my amplifier, this was causing it to act really weird. Before I fixed it, I would get different measurements every time I measured something. Now that it is fixed I realize that everything I thought about the amp before was when it was not performing correctly. I am going to just spend some time listening to the amp as is and get used to it again before I make any more changes.
Sorry for the newbie questions, but if the input jFET has non-unity gain, then won't any preturbations of the rails will have a similar result on Vgs regardless whether they are coupled or not?
Won't the output MOSFET Vgs variation decrease with say using a regulated rail voltage on the front end?
The Supply rail can bounce up and down all day.
If the CURRENT through the jFET drain load stays the same, then the Vgs of the output mosFETS stays the same.
The bit that does change with bouncing supply rails is the Output MosFET Vds.
The current passing the output device will change very slightly with supply rail bounce.
We are relying on NFB to minimise the gain changes due to supply rail noise/bounce.
If the CURRENT through the jFET drain load stays the same, then the Vgs of the output mosFETS stays the same.
The bit that does change with bouncing supply rails is the Output MosFET Vds.
The current passing the output device will change very slightly with supply rail bounce.
We are relying on NFB to minimise the gain changes due to supply rail noise/bounce.
Has anybody here raised their rails voltage to squeeze some extra power from the vanilla F5, as suggested by Nelson at the start of the F5 Turbo .pdf?
I am thinking of getting a new transformer, maybe 22V or 24V 300VAC and uprating the resistors on the board to 5W, biasing a little lower until the heatsinks are happy.
Id like to hear how that turned out, if you've done it. It seems to make more sense to me than building a preamp gain stage. I need more power, but F5 Turbo seems too much hassle if this is doable.
I am thinking of getting a new transformer, maybe 22V or 24V 300VAC and uprating the resistors on the board to 5W, biasing a little lower until the heatsinks are happy.
Id like to hear how that turned out, if you've done it. It seems to make more sense to me than building a preamp gain stage. I need more power, but F5 Turbo seems too much hassle if this is doable.
