Hi Ale, I suspect it's the lower JBF862. I replaced the upper DN2540 MOSFET as I had some, but that didn't fix it. I haven't got back to the amp since, I have too much on at the moment. But it looks like BF862 is obsolete now at Mouser and Digikey - what would you suggest as an alternative?
Hello Simon! for BF862, look closer at hand .... Leeds! 19K stock, today.
BF862,215 NXP, JFET Transistor, JFET, -20 V, 10 mA, 25 mA, -1.2 V, SOT-23, JFET | Farnell UK
They will be gone eventually - after all, they were designed for AM radio-receiver front-ends.
I'll try to help, Simon.
I like the Infineon & Wolfspeed SiC rectifiers, too. They allow a power-amp build with large, high quality capacitors.
I removed the LCLC supply from my power amp, and installed SiC rectifiers -Wolfspeed C4D02120A - and large capacitors ( 820µF 550V Kemet ALS60) x2 per channel. After these, a capacitor-multiplier (using large N-channel FET FQA9N80C), followed output capacitor LCR 30µF 630V FP-AU.
Superb sound. The cap-multiplier can be configured to gently ramp the supply voltage, too.
To protect the filament when it is part-warm, the best solution is to apply extra grid bias, until everything is stable, and Vf is at rated value. Apply enough bias that the anode current is zero when warmed up. Example: for 300B, apply -120V at switch-ON.
Hi Rod, I'm pleased you chimed in 
. Good news regarding BF862 - when I checked last night the Farnell website was down for maintenance so I looked at Mouser and Digikey. Leeds is indeed a whole lot closer!
Thanks for the great info on rectifiers and PS too - I need to have a think about it tonight when I can have a read up on a capacitor multipliers. The variable bias that needs to vary with conduction, and be out of circuit when not in use needs some thinking about too!
(BTW the shunt cascode boards were built a while ago but I haven't had the time to try them in GM70 yet)
. Good news regarding BF862 - when I checked last night the Farnell website was down for maintenance so I looked at Mouser and Digikey. Leeds is indeed a whole lot closer!Thanks for the great info on rectifiers and PS too - I need to have a think about it tonight when I can have a read up on a capacitor multipliers. The variable bias that needs to vary with conduction, and be out of circuit when not in use needs some thinking about too!
(BTW the shunt cascode boards were built a while ago but I haven't had the time to try them in GM70 yet
)
He does not suggest the capacitance multiplier to be less suited to class AB circuits, but indicates that it is very well suited when well executed for that application.
This thread is about discrete preamps which are class A almost by definition; capacitance multipliers work well here.
This thread is about discrete preamps which are class A almost by definition; capacitance multipliers work well here.
Last edited:
Hi Felipe,
The Schematic is not easy to read - perhaps I can make a better one. I am thinking of making a PCB version.
Well, I don't want to open a can of worms and I know that there are other threads covering the use of SMPS.
I'm using a step up converter (200-600V) SMPS feeding an LC network (including same 820uF Kemet ALS60) and then a final DC-Link 40uF film cap including some CM choke and ferrite beads to clean up.
Result is outstanding in terms of sound and noise level. I was planning to add a cap multiplier but it wasn't necessary. As I wanted to use this as a variable HT supply to experiment with different DHTs stages the cap multiplier wasn't the best approach, probably a properly called simulated inductor (gyrator) instead.
I'm experimenting on the same lines with the LT raw supply but unfortunately haven't got to it yet. it's definitely more challenging.
Ale
hi Ale - DC-DC conversion for B+ is good, and not controversial, if implemented properly.
DC-DC not the same as mains-to-dc converters, which inject plenty of common-mode noise current into your ground network, risking contamination of DACs, MC cartridge stages, etc.
The noise current arises because the converters use the secondary circuit as EMC grounding (there's a cap from mains-side to the output). Eliminating this noise current can be tough.
DC-DC conversion does not have this problem, as it still uses a 50Hz transformer - and so benefits from the isolation this gives.
Glad to hear you had good results with my recommended big ALS61 capacitor!
DC-DC not the same as mains-to-dc converters, which inject plenty of common-mode noise current into your ground network, risking contamination of DACs, MC cartridge stages, etc.
The noise current arises because the converters use the secondary circuit as EMC grounding (there's a cap from mains-side to the output). Eliminating this noise current can be tough.
DC-DC conversion does not have this problem, as it still uses a 50Hz transformer - and so benefits from the isolation this gives.
Glad to hear you had good results with my recommended big ALS61 capacitor!
I quite agree on panel speakers - used to have Maggies, stacked Quads and Caliper Sigs. All marvellous but too big for my listening room which needs to have other functions as well.
I used 10 or 10Y, so thoriated. The 26 was a ST, but I didn't feel like swapping in a globe. No doubt I should have. I think the 26 may need another solution - I liked two 156C in series but they hummed like crazy. Maybe a permalloy shield....