DIY F2 clone

Official Court Jester
Joined 2003
Paid Member
I've wondered about one thing, the output caps to the speakers are if I understand correctly just to remove DC and work in the same way as series capacitors in crossovers?

Because I'm planning 16 ohm speakers from 400 hz and upwards and if I understand correctly then I don't need as large capacitors as 15m uf but rather can get away with 1 m uf taking up less space =)......

you got it exactly

in fact - you can route direct DC connected cables to your xover , then from that node - connect separate caps for mid and for highs , in case that you intend using one amp for both ways
 
you got it exactly

in fact - you can route direct DC connected cables to your xover , then from that node - connect separate caps for mid and for highs , in case that you intend using one amp for both ways

Nice, I have enough heat sinks and chassi space for building 4x of them so why not? :D

But mostly it's that I'm realistic enough to admit that to do a really good current source optimized LR4 passive crossover is probably beyond me with perfect notches and everything, but to do the crossover digitally through DSP should be very easy and I like easy :D
 
Last edited:
Another question now: this time about adjusting P1.

Is there any difference in adjusting it with an 8 ohm load or a 16 ohm load (or even 4 ohm) or will the reduction be similar on all?

Why I wonder is because at the moment my test amp is adjusted with 8 ohm but I will use a 16 ohm nominal speaker and wonder if it's wort re-adjusting P1.

Thanks,
Olle
 
And another thing, is there some easy way of improving the output power into 16 ohm loads with the parallel resistors removed?

I just tested with a 15 ohm load and I got 0.65 W, and I would love to have a little bit more... I wish to drive a ~ 90 db efficient midrange with the F2J but I'm not sure if 0.65 W is enough.
 
And another thing, is there some easy way of improving the output power into 16 ohm loads with the parallel resistors removed?

I just tested with a 15 ohm load and I got 0.65 W, and I would love to have a little bit more... I wish to drive a ~ 90 db efficient midrange with the F2J but I'm not sure if 0.65 W is enough.

Strange... How do you measure/calculate max power ? you've got only 0.65W at clipping ?????
0.65W won't be enough, but you should have more than that even with the resistors in place
 
Strange... How do you measure/calculate max power ? you've got only 0.65W at clipping ?????
0.65W won't be enough, but you should have more than that even with the resistors in place

I hooked up an oscilloscope and pushed up voltage as high as possible until the waveform started clipping. So 0.65 W was the highest I got right before clipping. This was at 3.14 V at output so if my math is correct that is 3.14^2 / 15 = 0.66 W

And while I think I can get more power by putting back the parallel resistors that would destroy the whole point of having a current amplifier with high output impedance... which is why I'd rather solve it but keep the high output impedance =)
 
Last edited:
Would it be possible to say double rail voltage and then lower the current (which if I've understood correctly should increase performance in high ohm loads). And if so would this be easy to do or would it require modifying the entire circuit?

Everything is possible within reason... or rather ... within components specs.
With such a simple circuit, you can play with U and I by just playing with components values.
Just check Umax and Pmax of every components, and keep a safety margin.

I is set by R16//R17 in the current source... if it's not obvious to you (it wasn't to me), read the Zen Amp series articles to understand how, but in essence, increase R16//R17 to reduce I
 
Everything is possible within reason... or rather ... within components specs.
With such a simple circuit, you can play with U and I by just playing with components values.
Just check Umax and Pmax of every components, and keep a safety margin.

I is set by R16//R17 in the current source... if it's not obvious to you (it wasn't to me), read the Zen Amp series articles to understand how, but in essence, increase R16//R17 to reduce I

Nice, thanks for answering =)

I've read the Zen articles which is what gave me the idea since Nelson experiments with lots of different voltages and the F2 is if I remember correctly based on the Zen. I'm no electronics expert though so I thought I'd also ask to be sure ;)

So now another question if someone knows, is current linear in proportion to R16/R17 or is it some exponential whatnot such that I ideally need a pot to dial in amount of current?
 
Hello there again,

I'm considering adding an input transformer since I have a balanced source. If I've understood correctly the Jensen JT-11P-1 works well for this area.

So now the question: The Jensen is 10k:10k, and the F2J has 10k input impedance so I should just hook up the trafo before the F2J without putting any extra resistors on either side?

Thanks in advance,
Olle
 
Just powered up and measured my F2J with the input trafo and 48V rail voltages and halved current

Max clean output @ 8 ohm is ~ 4 Vrms so 2W if I've understood correctly.

Much more than 24V but still not a lot and hoped for more, I think it will be enough though since it will drive 300 and upwards with 94 db efficient drivers.