Nice. Perhaps I will try those when I finally break down and order those double-die Alfets...
I would measure the voltage drop across that 10R on power-up, when cold, and then again after the temperature has stabilized, and at least a few minutes of play at normal volume.
Yeah, I have the Cm set for about 1.5v drop whigh has it running between 37V to 38V depending on the house voltage. When I just measure it at 200mv on the 10R it was 38V rails. The vas transistors are too hot to touch.
Edit; you type too fast. The measurement I took was right after a cold start.
Nice. Perhaps I will try those when I finally break down and order those double-die Alfets...
I have matched pairs of the 992/1845. I found FTA for one and PTA for the other resulted in matched pairs in the 400 hfe range at 35V at 10 mA. They are also end of life.
These front end transistors should allow +/- 55V rails.
LOL. That must be the only thing then!
Measure again after the temp stabilizes, maybe tomorrow after you play a track or two, and compare the two voltages. The VAS current may (probably will) be higher at final temp. If you measured 20mA cold, it could be 25-30 mA hot, and that is indeed pushing it.
Use the worst case (hot) current to select the value of the current injection resistor, but also allow for the expected rise to be smaller than what you see now. In other words, it is like derating in reverse.
In any event, we have some other options for controlling the VAS current if the resistor change we are discussing isn't enough. Give it a try and report back your findings, I think it will likely work out.
We don't want to throw things off kilter by taking one thing or another too far. Its something of a balance. We'll get it tuned perfectly. I'm sure between myself, PMI and others who have these up and running you'll have one of the sweetest little amps going.
Little pic of a mostly stuffed board with nicer components than my first build. Vishay- Roderstein input film, Wima film bypasses, CDE silver mica comp and filter, Nichicon electrolytics, and metal film resistors. Will it make a difference? Dunno. I might open my wallet a little more and even spring for some double die outputs too, though they seem to be stocked only in Europe
, added shipping.Attachments
I definitely want to be able to run at ~50V, which means ~53V at light load given line and load regs. I hope to find the time to do some testing by next weekend.
Perhaps some of us can combine an order. I have noticed that the charge is once, per UK stock order, at least at Newark.com....I might open my wallet a little more and even spring for some double die outputs too, though they seem to be stocked only in Europe
Yes, but thanks for pointing it out. I have only used the BC550/560 in this particular circuit so far, and those won't get me there. I have not taken the time to research every possible part, so I am happy to start with parts someone else has already tried.
KSA992PTA Fairchild Semiconductor | Mouser
These are an excellent match @30V @2mA for the KSC1845FTA.
No, the excellent match is KSA992FTA + KSC1845FTA & they are still available.
I have the old version 2SA992+2SC1845 only 2 pair match & I use it in my first PeeCeeBee
but I change it back to BC550C-560C because only use +/-35VDC.
So I can use it in higher voltage version latter
Now I'm finish mounting the double die
(aka double K1058-J162) at heat sinkstill missing some parts to finish.
The building must be slowly to ensure no error
Regards
No, the excellent match is KSA992FTA + KSC1845FTA & they are still available.
I have the old version 2SA992+2SC1845 only 2 pair match & I use it in my first PeeCeeBee
but I change it back to BC550C-560C because only use +/-35VDC.
So I can use it in higher voltage version latter
Now I'm finish mounting the double die
still missing some parts to finish.
The building must be slowly to ensure no error
Regards
incredible
No, the excellent match is KSA992FTA + KSC1845FTA & they are still available.
I have the old version 2SA992+2SC1845 only 2 pair match & I use it in my first PeeCeeBee
but I change it back to BC550C-560C because only use +/-35VDC.
So I can use it in higher voltage version latter
Now I'm finish mounting the double die
still missing some parts to finish.
The building must be slowly to ensure no error
Regards
I tested a batch of KSA992FBU (bulk) and the hfe was 550 to 570 and do not match any of my 1845. I have not tried the KSA992FTA. I will have to order some next time I need some parts from Mouser.
Thanks
Ok, I sat down this morning to swap out the 18K resistors. Pulled them and then realized I didn't have any 20K. Closest I had were 22K so I tried them in there. Not good. 10R resistors now measure 65mv drop and the the highs sound distorted. I guess I'll pop over to Radio Shack and see if they have any 20K.
I'll report back.
I'll report back.
Or you could build a closer value with a series combination standing up vertically on the board to get closer to the 20K mark (which simulated well at your voltages). If you have to err a little then go slightly under the 20k mark but over 18k. It's a fine line.
Otherwise, put the 18k in and put a small heat sink front and back of the VAS devices if you can to run them cooler. Just try not to rework the boards too much, I'm not sure how well they will stand to repeated desolder / resolder cycles.
Otherwise, put the 18k in and put a small heat sink front and back of the VAS devices if you can to run them cooler. Just try not to rework the boards too much, I'm not sure how well they will stand to repeated desolder / resolder cycles.
Sorry, I should have reported back sooner. I piggybacked 180K on one board and 220k on the other. After letting it play for a couple hours I am getting 150mv across the 10R on the 220k board and 178mv on the other. Sounds better than it did with just the 22K in there. I'm thinking about making bigger heatsinks for the VAS and go back to the 18K. Seemed like it sounded better with that. It's kind of hard to tell at this point. I am getting listening fatigue. I have been playing the Peeceebee along side the DX Super A that I am finishing. They both sound so good and I have been A/B-ing them all day.
Fun stuff!!!!!!!
Fun stuff!!!!!!!
Sounds like a plan.
As a side note, I just checked my own PeeCeeBee boards to make a comparison with yours. Here's what I have:
Supply is +/-31.5V and my build uses 15kΩ resistors.
Cold start 197mV over 10Ω VAS transistor emitter resistors, so 19.7mA VAS current.
Full temperature 224mV over 10Ω VAS transistor emitter resistors, so 22.4mA VAS current.
Subjectively the transistors are pretty warm, but without any discomfort when touched and my contact-less thermometer says the VAS reaches about 45C.
I had a look at the data sheet to see the temperature derating information and I don't think there is a problem with these devices even at your 10C higher temperatures. Go back to the 18K and just enjoy the PeeCeeBee.