Please forgive me if I am being impolite.
But :
1) Compared to the original with 5 (functional) transistors, is this not a bit complicated ?
2) Is the distortion spectrum still dominantly 2nd harmonics, as in the original measured by Stereophile ?
First Watt J2 power amplifier Measurements | Stereophile.com
3) Did I miss your comments in how to deal with the positive tempco of the IRFP when used as the driver ?
Real quick building and testing. Will never beat that.
Cheers,
Patrick
But :
1) Compared to the original with 5 (functional) transistors, is this not a bit complicated ?
2) Is the distortion spectrum still dominantly 2nd harmonics, as in the original measured by Stereophile ?
First Watt J2 power amplifier Measurements | Stereophile.com
3) Did I miss your comments in how to deal with the positive tempco of the IRFP when used as the driver ?
Real quick building and testing. Will never beat that.
Cheers,
Patrick
The 2nd and 3rd harmonics are about 86 dB below the the 1 kHz fundamental (1W into 8 Ohm, with Semi South in lower position). This is about 0.005% each, summing to about 0.007% THD. This is quite a bit less than the 0.035% THD(+N) in the Stereophile test (Figure 6). How come? Is this the result of all the Babelfishing? Pretty impressive...
I'm never taking as impolite correspondence , which as goal is having sharing of knowledge and helping .... and I'm always trying to learn something
so :
1. as I said in few intention notes in post #1 ,my goals are clear and I pretty much reached them all ;
basic premise is here , number of stages is pretty much the same ( someone is counting current mirrors as stage , follower as stage , while someone is not counting ; I don't care , as long things I'm doing are intentional , not in any way or form patches from necessity
2.yes, with main iteration - all IRFP ; it's easy shaping whichever iteration to have dominant 2nd , simply playing with values of MOX resistors in OS vertical ; even simply removing emiter follower ( shortie in , few resistors and two bjts out) will invoke 2nd ; more current through each LTP Jfet will invoke 2nd etc. .......... but , I'm not interested this time in that , I already made (previous) Babelfish J2 with all SS , having same THD spectra as original ( nature yes , lower in numbers) ; measurements are somewhere around , posted in last few months
repeating these here would just clutter already big number of posted graphs
3.I'm not sure about which IRFP you're asking - lower one ( modulated directly from input) or upper one (part of Mu Follower)?
anyway - upper one is taken care with opto-functional block , lower one is having hefty 235mR in source and rest is taken care with feedback , what else ?
I'm for long time pretty content with Papamps having range of few hundred mV of output offset ( from cold to hot ; that's price of simplicity) , and this one is pretty much benign and well mannered ....... didn't write down but I think it is somewhere around 100mV from cold to hot , while Iq is rock solid
see next few posts and compare somewhat comparable circuits ; saying - more comparable is Babelfish J2 full SS with FW J2 ; B. J2 with SS down is not exactly same beast
reasons for so low THD are multiple - low current through each LTP Jfet (same trick Papa used) so low ΔI as result of input modulation ...... then cascodes ...... then easy load for JFets (CM) .... then emiter follower in front of output gate
not to forget obviously interestingly "opposite'' transfer characteristics of SS down and IRFP up , resulting in THD minimizing
Last edited:
Babelfish J2 SS both 8R
schematic , except logically Semisouth SJEP120R100 in both output positions, is practically the same as previous one
parts values around opto thingie are optimized to take care of setting same Iq - simple fiddling with trimpot .... and done.
schematic , except logically Semisouth SJEP120R100 in both output positions, is practically the same as previous one
parts values around opto thingie are optimized to take care of setting same Iq - simple fiddling with trimpot .... and done.
Attachments
Babelfish J2 SS both 8R
graphs
not sweet-spotted , regarding difference 2nd vs. 3rd , but obviously same nature of parts used up and down
as I said - not interesting especially for me , simply because I have more precious parts (than Semisouths) laying around , and I still didn't utilized them
graphs
not sweet-spotted , regarding difference 2nd vs. 3rd , but obviously same nature of parts used up and down
as I said - not interesting especially for me , simply because I have more precious parts (than Semisouths) laying around , and I still didn't utilized them
Attachments
as I promised - Eagle generated parts list , BOM , whatever
if in doubt - just take a look at schematic........ and it'll be crystal clear
you already have 2+2 IRFP240 with 1+1 IRFP150
cool them properly , and they'll cruise happily with 50W of dissipation each , for years
cool them properly , and they'll cruise happily with 50W of dissipation each , for years
Just curious: what is the spectrum of the residual harmonic distortion of your soundcard (loopback)?
EMU1212
just see that weed on bottom , on graphs with 1W ....... and then decrease it somewhat
in any case irrelevant for actual measurements - it's good enough
as I wrote , didn't even connected Quantasylum QA400 , when I have this one handy
just see that weed on bottom , on graphs with 1W ....... and then decrease it somewhat
in any case irrelevant for actual measurements - it's good enough
as I wrote , didn't even connected Quantasylum QA400 , when I have this one handy
Very nice work Zen Mod! I'm wondering about 2SK209 for the N-channel jfets. Is the dissipation too low?
they're perfect , 2+2 ..... I did though about them but , sorry, couldn't squeeze them in too, due to pinout ........ in this I prefer 2SK2145BL
2SK2145 is a nice bonus with 2 devices in one. Need to pick some up before they're gone!
Thanks Zen Mod.
Thanks Zen Mod.
I enjoyed listening all- IRFP iteration more than 6 hrs today ....... typical (part of) long term test, listening while working 🙂
finishing my own SissySIT , so pleasure was ..... quadrupled
what can I say , without straining in FiFi vocabulary ....... let's say it as - didn't found sole thing to complain
details , dynamics .... and bass is super-layered ....
finishing my own SissySIT , so pleasure was ..... quadrupled
what can I say , without straining in FiFi vocabulary ....... let's say it as - didn't found sole thing to complain
details , dynamics .... and bass is super-layered ....
