Hmm... I skipped building an F2J to compare with but I think I'll do that after all. I've now pulled out a big CPU cooler I'll convert into a TO-3P compatible heatsink =)
Now I just need to get my crossover parts... probably better to compare when the speaker covers the whole intended range and not just a single driver.
Now I just need to get my crossover parts... probably better to compare when the speaker covers the whole intended range and not just a single driver.
Some distortion measurements on the speaker with F2J vs L'amp. Both have 200 ohm in parallel so the F2J should have about 150 ohm output impedance.
The midrange is a fairly cheap one without fancy shorting rings and such =) The WIP OB-speaker has much higher quality drivers but to measure on that one I'll have to get those crossover parts...
The speaker (SQ-50 C2) is ~ 8 ohms. Ignore the ragged tweeter response on the crossover measurements since it has 4 tweeters so this close miced there is mad comb filtering. When measured from a longer distance in a real room it is of course much much more smooth =)
I think it is somewhere around 1W @ 1 m.
When I measured I decided to do some tests with a super high quality midrange too, the TD6M that has a full copper sleeve. The TD6M has almost flat 10-13 ohm impedance over the whole 100-20000 hz range. Your average quality midrange is probably somewhere in between those two.
It's interesting to see how much lower the distortion is with the F2J, especially 2nd order distortion and a fair amount of 3rd also, even on the high quality midrange.
Sensitivity wise the TD6M is 5 dB more efficient in comparison, I have modded my UMIK such that I have more gain and lower threshold @ 100 dB so to avoid clipping I had to run that with lower power since I didn't want to use any EQ this time to reduce the peak @ 1.6 khz. If you add 5 dB to the TD6M measurements they are the same voltage used as on the other speaker though, pretty good sensitivity for an OB speaker =)
EDIT: wrong graphs on wrong amp! The TD6M measurements are on wrong amps! Fixed now though =)
The midrange is a fairly cheap one without fancy shorting rings and such =) The WIP OB-speaker has much higher quality drivers but to measure on that one I'll have to get those crossover parts...
The speaker (SQ-50 C2) is ~ 8 ohms. Ignore the ragged tweeter response on the crossover measurements since it has 4 tweeters so this close miced there is mad comb filtering. When measured from a longer distance in a real room it is of course much much more smooth =)
I think it is somewhere around 1W @ 1 m.
When I measured I decided to do some tests with a super high quality midrange too, the TD6M that has a full copper sleeve. The TD6M has almost flat 10-13 ohm impedance over the whole 100-20000 hz range. Your average quality midrange is probably somewhere in between those two.
It's interesting to see how much lower the distortion is with the F2J, especially 2nd order distortion and a fair amount of 3rd also, even on the high quality midrange.
Sensitivity wise the TD6M is 5 dB more efficient in comparison, I have modded my UMIK such that I have more gain and lower threshold @ 100 dB so to avoid clipping I had to run that with lower power since I didn't want to use any EQ this time to reduce the peak @ 1.6 khz. If you add 5 dB to the TD6M measurements they are the same voltage used as on the other speaker though, pretty good sensitivity for an OB speaker =)
EDIT: wrong graphs on wrong amp! The TD6M measurements are on wrong amps! Fixed now though =)
Attachments
Last edited:
The F2J is also 193V loaded. When measuring distortion into resistors the SIT seems to outpreform the F2J with lower % distortion so I didn't expect differences this big, if anything I thought the SIT might perform slightly better except on the less expensive midrange where the absence of copper in the motor would cause excessive distortion in the not fully current source L'amp.
But even on the TD6M the F2J outperforms the SIT in distortion massively. Would be nice if I had a PASS-SIT around to see if a modern SIT performs significantly better than those 40 years old ones but as they are made from unobtainium I don't think I will =)
Last edited:
One thing I noticed with the L'amp when playing around with the scope: while it with the 193V does output a somewhat clean 1 khz sine wave up to 10W above at 2-5W large amounts of high order harmonics appeared. I guess Michael is right that while you could output more than 1W with this amp you'd probably be much better off getting more efficient speakers such that you always stay at 1W and below.
My more informal listening test confirmed this, currently the speaker isn't that efficient at 90 dB since the more efficient one isn't done yet and at low levels both the 193V loaded F2J clone and the L'amp sounded good but when pushed more there was a point where the L'amp started to sound mushed together where the F2J clone still sounded clean.
My more informal listening test confirmed this, currently the speaker isn't that efficient at 90 dB since the more efficient one isn't done yet and at low levels both the 193V loaded F2J clone and the L'amp sounded good but when pushed more there was a point where the L'amp started to sound mushed together where the F2J clone still sounded clean.
Last edited:
Happy 2014 Vfet Year
Dear Michael i am glad and say thanks to bring first
interesting project with this transistors to diyers.
With Your other projects like "Iron amp " and " Luminaria " this be a good time
and great learn for biginner and advanced again.
That really cool
Thanks Mr Rothacher
Wish for all diyers on
keep his solder station run hotBest regards
Still gatering all the parts I found a nice 1200VA transformer.
It's a beast, and it has two 18v secondary...do you think I can buld the bulb version hooking in series the secondaries (36v) and then using it to feed both channels in parallel?
I think I have to use a "strong" bridge rectifier, any hints?
Ciao!
It's a beast, and it has two 18v secondary...do you think I can buld the bulb version hooking in series the secondaries (36v) and then using it to feed both channels in parallel?
I think I have to use a "strong" bridge rectifier, any hints?
Ciao!
So I'll try...
I was misleaded by Duncan PSU designer: two caps after one rectifier is an "illegal combination".
Thank you Zen Mod!!!
You can put any lower value you like but current ripple filtering will be lower. Alternatively you can use double filtering chain using CRCRC setup instead.
Here is my version:
http://www.diyaudio.com/forums/pass-labs/233057-sit-variations.html#post3433887
Here is my version:
http://www.diyaudio.com/forums/pass-labs/233057-sit-variations.html#post3433887
