Balanced F5 question

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
EUVL

EUVL

Member
Joined 2003
xen-audio.com
Someone asked by PM :

"I saw your quote "You may also use the standard F5 in bridged mode, and reduce rail voltage to +/-18V, bias current up to 1.75A to optimize for 8 ohm load."

Could you do this with the IRF outputs and "X" the R1 and R2 ends to the other side - as you did for you XF5 - to get rid of harmonics? I wanted to try this, but was discouraged by many. They thought the IRFs had poor linearity below 20V. At 18V they don't look so bad. Would such a circuit be as worthwhile as the Toshiba version at 16V rails?"

a) You can get rid of the even harmonics also with grounded X (i.e. 2 box standard F5s fed with +ve & -ve signals) if the two (L&R) halves are properly matched. So in theory at least there is no need to float the X.

b) I have not used any IRF MOSFETs as gain devices since 2003. So although it will most likely function as proposed above, I cannot possibly make any comments to the results, especially in comparison with Toshiba MOSFETs.


Patrick
 
N

NYCOne

Member
Joined 2010
Thanks for the response, the question to Patrick was mine.

For those of us neophytes, could one bias up the stock 24V rail version to 1.75A and get the same effect? Initially I was told the stock version would be optimized for 16ohm loads when "xed" at 1.3amps. Does just biasing solve the optimization? The only obvious issue is the wattage per device - which goes to 42 watts (solved with parallel devices?)from 30w for the Toshibas and 35 from your suggestion here.

The only reason I return to this issue is it opens up balanaced F5s to a much larger audience if the scheme is viable.

EUVL said:
Someone asked by PM :

"I saw your quote "You may also use the standard F5 in bridged mode, and reduce rail voltage to +/-18V, bias current up to 1.75A to optimize for 8 ohm load."

Could you do this with the IRF outputs and "X" the R1 and R2 ends to the other side - as you did for you XF5 - to get rid of harmonics? I wanted to try this, but was discouraged by many. They thought the IRFs had poor linearity below 20V. At 18V they don't look so bad. Would such a circuit be as worthwhile as the Toshiba version at 16V rails?"

a) You can get rid of the even harmonics also with grounded X (i.e. 2 box standard F5s fed with +ve & -ve signals) if the two (L&R) halves are properly matched. So in theory at least there is no need to float the X.

b) I have not used any IRF MOSFETs as gain devices since 2003. So although it will most likely function as proposed above, I cannot possibly make any comments to the results, especially in comparison with Toshiba MOSFETs.


Patrick
Click to expand...
 
Last edited:
alazira

alazira

Member
Joined 2007
EUVL said:
> it is 4* 0.22 ohm resistors per amp, and not 2* 0.22 ohms and 2*0.18 ohm?

4x 0R22 is default.

The other values (deliberate detune from symmetry) is to trade 2nd harmonics against 3rd, and require a balanced IN/OUT THD analyser and a spectrum analyser.
The exact value is unknown & depends on P to N match.
You can always detune later on by soldering a 3R resistor in parallel, as NP & I did, for different reasons.

Patrick.
Click to expand...

Happy New Year Patrick. Thanks for your work on this project.

I was wondering if you could expand on your selection of 0R22 for the source resistors. Is this value specific for the 2SJ201/2SK1530, or because of the lower rail voltage or because it's a balanced circuit (in other words, is this value also appropriate for the single ended F5 using these mosfets?)

Also, is this formula from post 548 applicable for the SE F5?
Idss (2SJ74) = Idss SK170 * (1 + Rs.Yfs)
e.g. 7.5mA(1.1782)=8.837mA for 2SJ74
 
EUVL

EUVL

Member
Joined 2003
xen-audio.com
> Also, is this formula from post 548 applicable for the SE F5?
> Idss (2SJ74) = Idss SK170 * (1 + Rs.Yfs)
> e.g. 7.5mA(1.1782)=8.837mA for 2SJ74

This only applies with Toshiba MOSFETs (because of their true complementary nature).
It has already been explained in both threads in detail, and is even then optional.
During our test build phase, we shall experiment with both, support that with measurements and listening, before we report again in detail.

The F5X PCB will have all such options open.


Patrick

.
 
Last edited:
EUVL

EUVL

Member
Joined 2003
xen-audio.com
I was asked who I was referring to as the team.

They are the seven volunteers who are kindly helping me to bring you a complete set of PCBs, mechanical parts, building instructions, measurements and listening impressions of different configuration, which will allow you to build my version of the F5X as I have posted years ago at the F5 thread.


Regards,
Patrick
 
BobEllis

BobEllis

Member
Joined 2003
As the extrusion gets taller it loses efficiency compared to the same surface area on a wider platform. For example, you might expect that a Conrad MF35-150 to have half the thermal resistance of an MF35-75 since it has twice the area. But, from Conrad Heatsinks - Products you'll see that you only gain about 50% better performance with double the area. Also note that the MF30-100 has the same performance as the MF35-75, despite more surface area judging by its greater weight.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom