The picture actually shows 10 resistors in parallel. The schematic has a (20) marked on it. Hard to say what resistance is actually supposed to be there. Would 7.5k be reasonable there?
That was really som bad info, you managed to come up with 🤣
KT90, KT120, KT150 & KT170 are all Tungsol tubes.
Except for a few series of KT120 which had getter placed wrong, these tubes are all excellent
and btw........ NOT chinese 😉
That would require the resistors to have a pretty significant negative TCR. Usually TCR for even a crappy power resistor is on the order of a few hundred ppm/ºC, so they're unlikely to run away thermally.
What could have happened is that one or more of the resistors failed because they were running hotter than the rest. Once they failed to an open circuit the parallel combination of the remaining resistances had a higher resistance than the desired value so they limited the current and saved themselves that way.
From the picture they look like 1-2 W metal film types.
Maybe the designer couldn't get metal film resistors of the proper power rating. Or maybe it was cheaper to put multiple resistors in parallel rather than stocking one resistor of the needed power rating. There's nothing inherently wrong with putting resistors in parallel. But putting twenty (20) in parallel as indicated on the schematic seems pretty dumb to say the least.
On that note... The schematic says there should be 20 resistors in parallel, but the picture of the board only shows five. What gives?
Heh... Yeah. Gotta "love" that. Especially given that the tube characteristics will vary by at least ±10 %.
Also note that turning RV3 and RV4 will change the frequency response of the circuit.
Yep. The parametric sweep in the simulator said that 2.0024 uF was the optimum value, so by golly 2.0024 uF it must be. Never mind that capacitor tolerances are typically ±5 % for good ones.
Tom
the grid bias voltage for the 2 kt120 tubes is not specified/visible on the schematic, looks something the the bias circuitry is fishy.
What I don´t like is reliance on the pot wipers RV3/4 to keep the grids happy; should one go open even for a moment apart from a nasty crack, the wiper should be bypassed with a resistor to maintain bias. The circuit seems a risky way of going about esp with monster tubes.
Bench Baron
Bench Baron
The first amplifier designed by ChatGPT.
I think that should read "The first and clearly most accurate amplifier designed by ChatGPT. Valve amplifiers (often called tube amplifiers) were popular in the 1960s and are favoured by audiophiles for blah blah blah..." for several paragraphs.
kind regards
Marek
If you look closely, they’re piggybacked😁so 10, and another 10 of them on back side of the board
Thank you guys for your kind inputs.
We consulted with a retired technician who is a authorised repairer of this brand for many decades, his conclusion is the circuit and the tubes are fine, but there’re oxidation everywhere on the traces, dust all over the circuit board, causing high voltage arcing across those resistors’ legs. He has seen this all the time in his career😁
We consulted with a retired technician who is a authorised repairer of this brand for many decades, his conclusion is the circuit and the tubes are fine, but there’re oxidation everywhere on the traces, dust all over the circuit board, causing high voltage arcing across those resistors’ legs. He has seen this all the time in his career😁
Of course...... the chinese made copy´s.
The former mentioned KT´s were originally and first made by Tungsol.
Chinese copy´s doesn´t count 😉
Using two or more staggered in value capacitors is a common engineering method to reduce the effect of self inductance in larger wound capacitors like seen in the pictures.
As the data I attached shows the frequency of self resonance is affected by capacitor value.
Using spaced values of capacitors in parallel can be affective in reducing the effects of self inductance in the large film capacitors and provide reductions in possible phase shifts at high frequencies.
One could debate and disagree if this design is using the technique effectively however it seems the likely reason for the staggered values.
The schematic shows -200V rail, and a 420V B+ 'bootstrapped', but the schematic doesn't show the origin of the B+ rail (??). So as a minimum each resistor could experience over 600V. Even a quality 2W PRO-2 resistor has a 500V max rating, so if these weren't some very special 1W resistors then yes it sounds most plausible as the root cause.
What resistor solution are you going to use for a repair?
What resistor solution are you going to use for a repair?
The tech said this was caused by corroded traces, dusty board hence the arcing across the resistor under high voltage(this amplifier lives in a beach house for almost 10yrs can explain the corrosion). Solution will be scraping off the burnt marks on the board, seal with epoxy and then replace the components supplied by the factory so it’ll be the same resistor.
Forgot to mention, corroded traces need to be cleaned and re-flown with solder or even bypass with wire
Oh, dang! Yeah. I missed that. That's a manufacturing and service nightmare ... and not good for the reliability of the resistors either.
Tom
