Ok then, will try to lower the 1,47V side LED triplets.
Made the mV drop measurements across the 100R:
LEFT: 1,28V ccs Vdrop and 823mV drop across the 100R
RIGHT: 1,47V ccs Vdrop and 900mV drop across the 100R
Thanks again Salas
Regards
Scorpion
Yep, gotta buy a temp probe presto!!! Especially true in the 1Amp ballpark. As you so cleverly pin-point Salas, it is true-class A-kind-of -heat around that thing !!!
Regards
Scorpion
Salas,
I tried to compensate by lowering the LED triplets as you suggested but now, triplets don't glow anymore even tough every single LED has been tested individually and were fully functional once soldered.
Very tiny smoke was coming from the board within seconds. I turned everything off right away.
Hope I haven't done serious damage because of my stubbornness to get measurements as equally balanced as possible. I have a bad gut feeling I might unfortunately juste have !!!! :-(
If the LEDs are individually ok, wich culprit should I be looking for then that would prevent triplet from being "ON"?
And what about the frickin' smoke?
Thanks a lot
Regards
Scorpion
I tried to compensate by lowering the LED triplets as you suggested but now, triplets don't glow anymore even tough every single LED has been tested individually and were fully functional once soldered.
Very tiny smoke was coming from the board within seconds. I turned everything off right away.
Hope I haven't done serious damage because of my stubbornness to get measurements as equally balanced as possible. I have a bad gut feeling I might unfortunately juste have !!!! :-(
If the LEDs are individually ok, wich culprit should I be looking for then that would prevent triplet from being "ON"?
And what about the frickin' smoke?
Thanks a lot
Regards
Scorpion
Stunning
I've got to say what a beautiful build - overkill but aesthetically you can't beat it.
I've got to say what a beautiful build - overkill but aesthetically you can't beat it.
I think I put them correctly but I'll double check again tomorrow.
In any case, should I change Rset, JFET and MOSFET as a security measure. I wouldn't like any of them to misbehave after the surge.
Thanks.
Regards
Scorpion
Such things we normally measure and decide. Testing the leds on board with a battery for health and orientation for starters. But if you wanna do it as a security measure alright, especially if you don't have enough test gear and expertise. I would power up the section from a current limited DC bench supply and follow a typical debugging procedure for instance. Checking the semis for characteristic voltages etc.
Right on, not enough test gears and expertise. Do not have a current limiting DC bench on hands. Will retest the LEDs and change MOSFET and JFET (next to the 100R) just to be on the safe side.
Btw, wich one is the Rset again?
Thanks
Regards
Scorpion
Rset is your 1.25Ω. Basically when for some reason the voltage bias is off in the CCS the PMOS will conduct through the Rset full on. About 25W could have developed briefly on the Rset and probably from there was the smoke bcs the MOSFET even if broken was on a large sink to get that hot that fast.
Rset measures perfectly. Can't see any visual damage anywhere (tarnish, cracks or anything). Same with MOSFET and JFET.
Will recheck LEDS right away.
Tiny smoke lasted maybe 3-5 secondes.
Regards
Scorpion
I got mine working! Rather embarrassingly I'd mixed up the 220K resistors with two of the 220R ones. At the time of construction I was 'painting by numbers' and the white print on the prp resistors could have been clearer but after the hours spent debugging in future if there's any ambiguity about a part's value I'll test it.
Incredibly frustrating but I'm sort of glad of the experience; I learned a fair bit about the circuit, and debugging that I wouldn't have otherwise. One tip for this is to get a very cheap battery powered computer speaker, run a test tone through the circuit from a phone app signal generator or a computer, ground the speaker ground somewhere convenient and use the speaker source input to follow the tone through the circuit until it stops.
Next project is a gainclone - and some research about the 'least damaging' (if that's the right phrase) way of protecting the speakers from two direct coupled devices...
Incredibly frustrating but I'm sort of glad of the experience; I learned a fair bit about the circuit, and debugging that I wouldn't have otherwise. One tip for this is to get a very cheap battery powered computer speaker, run a test tone through the circuit from a phone app signal generator or a computer, ground the speaker ground somewhere convenient and use the speaker source input to follow the tone through the circuit until it stops.
Next project is a gainclone - and some research about the 'least damaging' (if that's the right phrase) way of protecting the speakers from two direct coupled devices...
Thanks for your help Salas and also for coming up with a great design - and thanks to the others who chipped in when I was struggling.
... that's what I'd meant to say on my original message but didn't.
To anyone else struggling, stick with it and take it as a learning opportunity, time spent learning (however haphazardly) is never wasted.
... that's what I'd meant to say on my original message but didn't.To anyone else struggling, stick with it and take it as a learning opportunity, time spent learning (however haphazardly) is never wasted.
Stax why not try one of these.
http://www.diyaudio.com/forums/group-buys/262319-my_ref-fremen-edition-gb-fifth-run.html
http://www.diyaudio.com/forums/group-buys/262319-my_ref-fremen-edition-gb-fifth-run.html
Thanks for your help Salas and also for coming up with a great design - and thanks to the others who chipped in when I was struggling.
To anyone else struggling, stick with it and take it as a learning opportunity, time spent learning (however haphazardly) is never wasted.
In the end you followed the signal in a novel manner and found the culprit. That was a clever way when not having a scope. Is the sonic result up to your satisfaction?
Thanks Marra, good shout. I'd have gone for that but I already have the parts for an Audiosector LM3875 dual mono.
It sounds great Salas but I'm not sure if that's because I can now use it to bring my modded (not by me) Quad 405-2 into action instead of using an integrated T-amp (too many variables!). Mind you I tried the Quad previously with a £150 passive preamp and the DCB1 knocks spots off that. I only fixed it yesterday so I've a lot of listening to do. Another point is that I have two prp 220R in the signal chain and two cheap ebay metal films that I bought just to patch it up and test it. Prp's in the UK are cheap enough at about 30p but the only place I can source them charges a minimum £5 postage - that seems to be the case with a lot of component suppliers here and I sort of understand that. I'll maybe get some when I order the toroids for the amp.
It sounds great Salas but I'm not sure if that's because I can now use it to bring my modded (not by me) Quad 405-2 into action instead of using an integrated T-amp (too many variables!). Mind you I tried the Quad previously with a £150 passive preamp and the DCB1 knocks spots off that. I only fixed it yesterday so I've a lot of listening to do. Another point is that I have two prp 220R in the signal chain and two cheap ebay metal films that I bought just to patch it up and test it. Prp's in the UK are cheap enough at about 30p but the only place I can source them charges a minimum £5 postage - that seems to be the case with a lot of component suppliers here and I sort of understand that. I'll maybe get some when I order the toroids for the amp.
Last edited: