The Singing Bush Tips 'n' Tricks

[ggetzoff]

This oscillator is epic and battery powered.

https://www.akitika.com/1kHzOscillator.html#OscillatorManual

 

[JeyDee]

Tweaking around with the amplifier potentiometer settings, the test setup, with the eyes laser locked on to the FFT-screen to get a feeling of it all. 2nd vs 3nd and THD, 4 and 8 ohm. And so on. 🙂🙂🙂

The mid node tops out at 49,65V and the THD seems to drop some what with higher voltages.

And the bias adjustment tops out at the same time at 2,4A.

Do you have any ballpark resistor values ideas to enable a increase of the upper adjustment range, to say something like 3A bias and lets say the mid node around 55V?

Dont seem to be possible to fully get rid of the 50hz Focusrite bump whatever i do. But i am thinking, since the amplifier is dead quiet with the Ipad connected or with the input RCA shorted. That problem is no biggie…🙂

 

[JeyDee]

connect the T to RCA positive, and the S and R together to ground.

It feels like i have tested all thinkable and even the unthinkable variations. But. I havent tested that variation yet. Will test it tomorrow. Very much to tired for any more experimentations this evening…
🙂🙂🙂

 

[ggetzoff]

JeyDee,

Have you listened to the amplifier yet in your normal environment to just see how it sounds within Alex’s target operating Iq and Vgs?

 

[Zen Mod]

Do you have any ballpark resistor values ideas to enable a increase of the upper adjustment range, to say something like 3A bias

increase R3 in 20% increments; say 120R, say 150R

and lets say the mid node around 55V?

remove R10, change R12 to 4K7

 

[ggetzoff]

Okay, I have a heat related issue with one channel. Once the channel heats up my Gate to Drain voltage is dropping off to some 15v. (Started today after I put the lid on) Cool down for a couple of minutes and power up the SIT is seeing 37V again. As the board heats up, it just starts descending. P2 may be the culprit or maybe the LM317 needs a sink.

1. All green leds stay illuminated when this happens and Q1 has a good size HS.

I’m going to try check IC1 output to check that there’s ~ -12v available to the bias circuit.

Any ideas Alex?

 

[Zen Mod]

hardly that LM317 heatsink is an issue, dissipation is nuttin'

check stability of voltage at LM317 output
check voltage at SIT gate in cold, then in hot, when it rolls down drain voltage

 

[JeyDee]

increase R3 in 20% increments; say 120R, say 150R

remove R10, change R12 to 4K7

It will be some fun testing done today! 🙂

Darnit that my Amazon cheap metal film resistor ”experimentation box” arrives first on monday. But ill manage today to series/parallel some resitors, dead bug style, that i already have on hand.

 

[JeyDee]

JeyDee,

Have you listened to the amplifier yet in your normal environment to just see how it sounds within Alex’s target operating Iq and Vgs?

No I havent done some proper listening tests since i only have one channel up and running yet. But i already head bob to what i hear out of one of the ridiculously small KEF speakers at the build/test-bench.

🎺🙂🎸

 

[ggetzoff]

hardly that LM317 heatsink is an issue, dissipation is nuttin'

check stability of voltage at LM317 output
check voltage at SIT gate in cold, then in hot, when it rolls down drain voltage

LM317 output static with V drop at Drain

Gate V of some 3.4v static during drop.

SIT Gate and Bias circuit maintain Vs during the Drain V-drop. I put a massive heatsink on Q1 and with the lid on Drain is holding. I've checked everything and, including IRF Source.

Could the IRF overheating affect the Output Node potential?

Lid is on and has been running in the shop for some time now and Drain V is holding.

 

[Zen Mod]

whatever change is, you must see change in gate voltage of SIT

it can be either input JFet buffer changing voltage potential at output, or it can be SIT itself

so, recheck SIT gate voltage

in cold, then observe in time; there must be a kick when Drain voltage changes

input cascode (IRF) will change some with heat but even V or 2 should not change buffer output voltage

 

[JeyDee]

Thanks Zen, now i have 2,5A bias, 47V node voltage, this one channel pulls 242W from mains. That was done by upping the R3 on mu/upper board from 100R to 140R.

I am happy with that for now, until proper listening tests can be done, and further tweaks from that most important aspect. 🙂

So. Now it is time to build the other channel, and for that i will start, for fun and change, with the Lazy Singing Bush. And listen to what that can do. 🎺🙂🎸

 

[ggetzoff]

Well, after an hour with the lid on the Node V held steady. If it drops again I will again check the Gate V. I put the same size HS on the other channel and it's all cooking again. Yeah, I was never able to see anything change except the node potential drop. It's probably something other than the input cascode but maybe the huge sink on Q1 is helping keep the board cooler. 🤔

It certainly will get fully heat tested where I live. 😉😎

 

[ggetzoff]

Listening tests tell me that the SIT certainly has the some of the characteristics of a triode: Harmonics become wide open as they propagate into the listening space and then they slowly dim into the background. With this much power transients are sailed through easily. With my dipole loudspeaker the entire venue is drawn in the room. I'm simply thrilled with how it sounds. For my set up the buffer is perfect as I don't need a VAS, just the gain from the SIT and the puck.

 

[Zen Mod]

 

[JeyDee]

This is the Singing Bush Tips And Tricks thread. But here is some tricks with my new Lazy Singing Bush build.

M1 and Q1 gets spring steel fasteners and Keratherm thermal pads against the bottom aluminium plate, in order to take care of the higher voltages and added heat.

I plan to test this Lazy one with 2SK182ES, since the THF-51ES got to take the heat in the other channels regular Singing Bush.

It is gonna be a pretty unique first stereo test, with these different amp channels. Just for fun. 🙂🙂🙂🥊

 

[Zen Mod]

 

[ggetzoff]

Choky,

Did you answer Steve on the SB kit?

 

[Zen Mod]

sort of
struggling these days with work, time, some back pain

combine all of those, you get ZM at 10% od usual productivity and 310% of usual mess

 

[Plott]

Okay, then I won't send you a question today, get well soon!