Zen Mod Napisano Oktobar 16, 2010 Autor Share #441 Napisano Oktobar 16, 2010 what's offset on output now - I see 20V ( yikes ! ) on right pic , but what's on left pic ?and - do you have variac ? Link to comment Podeli na ovim sajtovima More sharing options...
Tea-Bag Napisano Oktobar 16, 2010 Share #442 Napisano Oktobar 16, 2010 Yes, I have a variac. But I test up to full voltage for reference. Offset on left pic is ~10mv. I found the biggest influence of this was R9, then R12,R25.Meddling with the pots on the right pic, has no influence, stays about 20v.Not sure what's broke here, although 2SJ74 pairs are a little off, and now R100's a little off Ugs. But I think the problem with Q8 has me most worried. Link to comment Podeli na ovim sajtovima More sharing options...
Impuls Napisano Oktobar 16, 2010 Share #443 Napisano Oktobar 16, 2010 It looks like R8 BC556 is short wired between collector and emitter. Link to comment Podeli na ovim sajtovima More sharing options...
Tea-Bag Napisano Oktobar 16, 2010 Share #444 Napisano Oktobar 16, 2010 Maybe I should replace Q8 again after Q14 blew. Link to comment Podeli na ovim sajtovima More sharing options...
Tea-Bag Napisano Oktobar 16, 2010 Share #445 Napisano Oktobar 16, 2010 no luck replacing Q8 again. Link to comment Podeli na ovim sajtovima More sharing options...
Zen Mod Napisano Oktobar 16, 2010 Autor Share #446 Napisano Oktobar 16, 2010 did you fired it up according to post http://www.diyaudio.rs/topic/1765-semisouth-power-jfets-fw-j2/page__view__findpost__p__51840 ?anyway - we can test it easy ;for easy testing , it's best to have outout jfets connected to pcbs with short wiresfirst step - unsolder both output jfets from pcb , and let me know what's Ugs for lower and same for upperthen I'll calc and tell you on which values you need to set every trimpotthen we'll test just front end and few other pieces - without output jfets in circuit Link to comment Podeli na ovim sajtovima More sharing options...
Tea-Bag Napisano Oktobar 16, 2010 Share #447 Napisano Oktobar 16, 2010 More or less went by this process, had to get voltage across R3 up, and fiddle with R12 for Source voltage up as high as I wanted to go to in test phase with Channel 1.Channel 2 more or less never behaved or has responded to R12,R25 Pot.I will take the Fet's off bad channel and do measurements as you ask. Link to comment Podeli na ovim sajtovima More sharing options...
Zen Mod Napisano Oktobar 16, 2010 Autor Share #448 Napisano Oktobar 16, 2010 is that "proper " channel making sound ? Link to comment Podeli na ovim sajtovima More sharing options...
Tea-Bag Napisano Oktobar 16, 2010 Share #449 Napisano Oktobar 16, 2010 is that "proper " channel making sound ?Yes it is! I am still using test speaker, but sound is promising. Link to comment Podeli na ovim sajtovima More sharing options...
Zen Mod Napisano Oktobar 16, 2010 Autor Share #450 Napisano Oktobar 16, 2010 did you test output jfets in bad channel ? Link to comment Podeli na ovim sajtovima More sharing options...
Tea-Bag Napisano Oktobar 16, 2010 Share #451 Napisano Oktobar 16, 2010 Yes, before to measure vgs. In circuit vgs, they appear to be quite close to original measurements, of course now the pair on installed is not too well matched.Q13 ~1.3v Q14 ~1.4v vgs measured in circuit at startup. This is of course without Q8 apparently conducting.I will do as you say and remove them from PCB, re-test, and test PCB without R100 Jfets. Link to comment Podeli na ovim sajtovima More sharing options...
Zen Mod Napisano Oktobar 16, 2010 Autor Share #452 Napisano Oktobar 16, 2010 ok - now I'll type how to test rest of circ without outputs after that , I must go to sleep ..... early morning is Philip Island ......so - wait a little Link to comment Podeli na ovim sajtovima More sharing options...
Zen Mod Napisano Oktobar 16, 2010 Autor Share #453 Napisano Oktobar 16, 2010 ok everything having on mind that Ugs of outout is in range of 1V3-1V4 output jfets desoldered both inputs tied to gnd output also tied to gnd R12 to zero , same as R25 now - power up just to have all leds lit , and up a little to - say +/-15V to max +/-16V set R9 to have 9V4 across it now - if all input jfets are ok , and BC are ok you'll have around 23mV across each of R5 , R6 , R7 and R8 across R35 you must have 10mAx220R= 2V2 ; because you have lowish Ugs comparing to mine Jfet outputs - you can decrease R35 to 180R , so you'll have broader range for setting R12 later ; in that case you'll measure 1V8 across R35 now - if you still have 0V across R23 ( and you'll always have around 0V65 across it , doesn't matter which value of R23 is - it's best to replace both Q7 and Q8 if you have 0V65 across R23 ....... taking in account that you place 60,4R there ( in previous schm maybe there was 68R , but 60,4R certainly wasn't ) , current throughthat CCS will be Vbe/R23= 0V65/60,4=10,7mA , which will make around 4V2 across R26 ; optocoupler is dead now - without current going through 3W source resistors , so LED isn't lit . it's best to change that optocoupler - who knows what you does ........ OK - let me know what you measure that way ....... and I'll see that in some 8-9 hours buona notte Link to comment Podeli na ovim sajtovima More sharing options...
Tea-Bag Napisano Oktobar 17, 2010 Share #454 Napisano Oktobar 17, 2010 ok everything having on mind that Ugs of outout is in range of 1V3-1V4 output jfets desoldered both inputs tied to gnd output also tied to gnd R12 to zero , same as R25 now - power up just to have all leds lit , and up a little to - say +/-15V to max +/-16V set R9 to have 9V4 across it now - if all input jfets are ok , and BC are ok you'll have around 23mV across each of R5 , R6 , R7 and R8 across R35 you must have 10mAx220R= 2V2 ; because you have lowish Ugs comparing to mine Jfet outputs - you can decrease R35 to 180R , so you'll have broader range for setting R12 later ; in that case you'll measure 1V8 across R35 now - if you still have 0V across R23 ( and you'll always have around 0V65 across it , doesn't matter which value of R23 is - it's best to replace both Q7 and Q8 if you have 0V65 across R23 ....... taking in account that you place 60,4R there ( in previous schm maybe there was 68R , but 60,4R certainly wasn't ) , current throughthat CCS will be Vbe/R23= 0V65/60,4=10,7mA , which will make around 4V2 across R26 ; optocoupler is dead now - without current going through 3W source resistors , so LED isn't lit . it's best to change that optocoupler - who knows what you does ........ OK - let me know what you measure that way ....... and I'll see that in some 8-9 hours buona notte Okay,Outputs removed, input shorted, output shorted. DC to 15.5v Changed R35 to 180R, reads 1.8v Changed R23 to 47R, still reads 0V R3 to 9.4v R12 and R25 to 0R 4.7R's all read 23.5mv, all 2SJ74's show 10.4v R26 is 2.0v Replaced opto-coupler as well, and Q7. Thanks a bunch, and rest well. I am no rush here. One oddity I have noticed is that with R27 installed (33R), the gate to Drain shorts, with R100 installed or not. I don't forsee it causing this issue, but, I cannot repeat this on the working channel. Link to comment Podeli na ovim sajtovima More sharing options...
Zen Mod Napisano Oktobar 17, 2010 Autor Share #455 Napisano Oktobar 17, 2010 Okay, Outputs removed, input shorted, output shorted. DC to 15.5v Changed R35 to 180R, reads 1.8v Changed R23 to 47R, still reads 0V R3 to 9.4v R12 and R25 to 0R 4.7R's all read 23.5mv, all 2SJ74's show 10.4v ..... well - to have more appropriate voltage for input jfets , one day you'll need to replace all these leds with decent 20mA type , which will gives you few volts up - as in schematic I posted ..... R26 is 2.0v Replaced opto-coupler as well, and Q7. ........ ??what's voltage across R23 ? ......... One oddity I have noticed is that with R27 installed (33R), the gate to Drain shorts, with R100 installed or not. I don't forsee it causing this issue, but, I cannot repeat this on the working channel. you made somewhere short ; you must solve that ( without output Jfet installed ) before anything else Link to comment Podeli na ovim sajtovima More sharing options...
Tea-Bag Napisano Oktobar 17, 2010 Share #456 Napisano Oktobar 17, 2010 well - to have more appropriate voltage for input jfets , one day you'll need to replace all these leds with decent 20mA type , which will gives you few volts up - as in schematic I posted ?? what's voltage across R23 ? you made somewhere short ; you must solve that ( without output Jfet installed ) before anything else I solved the short by drilling out the connection at C3 and one at R27. So some contact was made by the close track there it appears.Of course, I disassembled that section of the board to see if some component was doing this. I replaced the LED with 20MA last week. So I will re-assemble the other parts on there, and report back. The board is made nice, no track has lifted with my de-soldering. Link to comment Podeli na ovim sajtovima More sharing options...
Zen Mod Napisano Oktobar 18, 2010 Autor Share #457 Napisano Oktobar 18, 2010 few pics ; I simply forgot to post them ; detail - I made thermal contact between 4 LTP Jfets ( placed in plain row) with piece of thermally conductive silicone rubber ( usually found in old CD ROMS , as conductor between chips and case ) . rubber thickness is around 2,5mm . rubber is cut in strip , width same as Jfet width ; that strip is placed on top of 4 Jfets , then secured with small U shaped Al tin, placed on top of little buggers . voltmeters - left one is across 0R5 source resistor , right one is output offset with full blast temp on heatsink ; offset is starting at ~ 200mV at cold , then gradually decreasing to zero ; that little thermal bridge is really helping in final thermal and offset equilibrium . I wasn't in mood to ....... well - accu's in my cam weren't in mood for further shooting CRO screen ..... , so you'll must believe me ..... that they're flawless , at least in my book . Link to comment Podeli na ovim sajtovima More sharing options...
Zvu Napisano Oktobar 18, 2010 Share #458 Napisano Oktobar 18, 2010 Gledam. Čekam. Strpljiv sam Link to comment Podeli na ovim sajtovima More sharing options...
Zen Mod Napisano Oktobar 18, 2010 Autor Share #459 Napisano Oktobar 18, 2010 sta cekas ? Link to comment Podeli na ovim sajtovima More sharing options...
Zvu Napisano Oktobar 18, 2010 Share #460 Napisano Oktobar 18, 2010 Da neko odavde napravi do kraja, iznese slušne utiske pa da možda krenem i ja u neku izgradnju... Mislim, ako dozvoljavate... ??? Link to comment Podeli na ovim sajtovima More sharing options...
Mikorist Napisano Oktobar 18, 2010 Share #461 Napisano Oktobar 18, 2010 sta cekas ?da Bacim Miniku kad ti ovo proradi Link to comment Podeli na ovim sajtovima More sharing options...
Mikorist Napisano Oktobar 18, 2010 Share #462 Napisano Oktobar 18, 2010 iznese slušne utiske slusni utisci za Jfetove u Ulazu i Izlazu...... Link to comment Podeli na ovim sajtovima More sharing options...
Preporučeni Komentari
Kreiraj nalog ili se prijavi da daš komentar
Potrebno je da budeš član DiyAudio.rs-a da bi ostavio komentar
Kreiraj nalog
Prijavite se za novi nalog na DiyAudio.rs zajednici. Jednostavno je!
Registruj novi nalogPrijavi se
Već imaš nalog? Prijavi se ovde
Prijavi se odmah