Yes, both channels same problem. The only difference is the other channel gives 2.40V out instead of 2V for the same 4V in.
triple check values of all resistors, especially at buffer input and after buffer
most likely there is some heavy load, eating the signal
I mean - if you are able to set Iq and output node (prior to coupling cap) to prescribed values, you're almost there
reminder - post #2, almost everything you need it to sing
https://www.diyaudio.com/community/threads/the-singing-bush-tips-n-tricks.357497/post-6279704
most likely there is some heavy load, eating the signal
I mean - if you are able to set Iq and output node (prior to coupling cap) to prescribed values, you're almost there
reminder - post #2, almost everything you need it to sing
https://www.diyaudio.com/community/threads/the-singing-bush-tips-n-tricks.357497/post-6279704
With an oscilloscope handy, the signal level at the board input, at the JFET gate, and at the SIT gate may be checked to try to track down where the signal is lost.
Also check oscilloscope probes for X1 and X10.
Also check oscilloscope probes for X1 and X10.
Making some progress. I traced the signal and managed to short one of the jfets, eliciting the magic smoke. The jfets were toasted so I replaced them with matched lsk170bl (8-11mA Idss) from the diyaudio store. I fried the mosfet, the optocoupler and one of the diodes so I ended up replacing all semiconductors except for the 2sk182 which was fine. Replaced the jfets with Lsk170 in the other channel to match. Biased again both channels, 38.5V and 2.8A. one channel gives 22V out into 4R for 2V in, the other one 14.2V out for 2V in. What can I do to adjust this imbalance? Should I simply stick a pot in the input of the higher gain channel, and dial the input down to match the other channel? Thanks
pics
you did something wrong, I can hardly believe that difference in SIT gain is that high
first compare signal you get in and out of input buffer, either with scope or DMM
GND as reference is fine in both cases, meaning - where black probe goes
we will rather introduce some NFB than kill good signal simply with resistive divider
you did something wrong, I can hardly believe that difference in SIT gain is that high
first compare signal you get in and out of input buffer, either with scope or DMM
GND as reference is fine in both cases, meaning - where black probe goes
we will rather introduce some NFB than kill good signal simply with resistive divider
An update to my Singing Bush saga. After setting the first one on fire (I inadvertently shorted one of the jfets as I was trying to understand why I had different output voltages across the two channels despite similar bias and with the same input), I got a second set of boards from mighty ZM and rebuilt both channels from scratch. And still got widely different gains. It correlated with the vfet, so I bought a curve tracer matched pair of 2sk182es from Watanabe. With these, both channels sing equally well: 38.5V at output node, 2.55A bias, 16.45dB gain in one channel and 16.37dB in the other. Thanks for all the help! Now... what should the next project be?
Last edited:
Hello, i am planning to start populating these fine boards in the coming days.🙂
Since i have elevated rail voltages in this build, around 84V, some resistor changes is probably to be expected for proper function of the amps.
…And later for purely sonic fine tuning purposes.
In wich resistor positions could it possibly be useful to add these small sockets for ease of switcharoo of the resistor values?
From what i have gathered thus far is that R3 and R1 on Mu/upper PCB could be useful positions for these mini sockets?
Since i have elevated rail voltages in this build, around 84V, some resistor changes is probably to be expected for proper function of the amps.
…And later for purely sonic fine tuning purposes.
In wich resistor positions could it possibly be useful to add these small sockets for ease of switcharoo of the resistor values?
From what i have gathered thus far is that R3 and R1 on Mu/upper PCB could be useful positions for these mini sockets?
Okey, i am also planning to configure use of the entire amplifier aluminium bottom plate as a heatsink for the Q1, in order to take proper care of the added cooling needs. 🙂
Getting closer to release the magic smoke. 🙂
Just Now! One test-channel built for ease of component adjustment/tuning/switching without repeted heat torture of the circuit boards:
Easy to desolder and tune resistor config. When the right combination is found the resistor- holders will be omitted.
The low voltage rectifier is moved from the amplifier board to the PSU enclosures. That will make some additional CRC filtration via the resistance in the DC transfer cable. Because why not? 🙂
But now to the real question. My plan is to run full 240V mains to the small bias PSU, but i will use the variac for the main PSU. Dial that down to normal 60V rails for initial test with all the stock component values. And when that is dialed in and working: Slowly start rising the voltage, adjusting and swapping resistors as needed to reach the full rail voltage.
It is a 1000VA variac, hopefully good enough for driving at least one test-channel witout overheating.
What do you think about this idea?
🙂
Just Now! One test-channel built for ease of component adjustment/tuning/switching without repeted heat torture of the circuit boards:
Easy to desolder and tune resistor config. When the right combination is found the resistor- holders will be omitted.
The low voltage rectifier is moved from the amplifier board to the PSU enclosures. That will make some additional CRC filtration via the resistance in the DC transfer cable. Because why not? 🙂
But now to the real question. My plan is to run full 240V mains to the small bias PSU, but i will use the variac for the main PSU. Dial that down to normal 60V rails for initial test with all the stock component values. And when that is dialed in and working: Slowly start rising the voltage, adjusting and swapping resistors as needed to reach the full rail voltage.
It is a 1000VA variac, hopefully good enough for driving at least one test-channel witout overheating.
What do you think about this idea?
🙂
Alex,
What wattage do I want for the 22R between the gate and the gain PCB? Getting very close to powering up the first channel.
What wattage do I want for the 22R between the gate and the gain PCB? Getting very close to powering up the first channel.
0207 size proper
don't forget to heatshrink wire, resistor and gate stub, all together
only that way resistor isn't just hanging desperately
don't forget to heatshrink wire, resistor and gate stub, all together
only that way resistor isn't just hanging desperately