This is greatly exaggerated and only applies in individual cases.
Although one cannot simply carry out a direct replacement in cases, where no Vbe multiplier and no driver stage is present in the front end (LTP-VAS) - OTOH - there are a lot of front-ends for MOSFET, where Vbe multiplier and driver transistors were provided in the same kind - than if one had planned to use BiTs from the beginning. In such cases nearly a direct replacement is possible - only other values for Cdom is necessary and a re-adjust of idle current so as additional 0,22-0,47R power resistors (emitter degeneration resistors) for the BjTs.
In the German magazines ELRAD - go to
https://musikelektronik.jimdofree.com/hobby-elektronik/bücher/elrad/
are to find two diy amplifier projects - one for MOSFET's (July 1990)
and one for BjT's, some months later.
The front-end and the driver stage from output buffer had from both diy projects the same basic topology.
Maybe one of the member can upload both schematic diagrams or know, where are already this diagrams here on diyaudio.
This circuit don't use a complementary front-end like on elektor's crescendo diy project and was close to this circuit under
http://www.stancurtis.com/Downloads/Moth 30 Stereo Amp copy.jpg
Last edited:
This minimal design aka Maplin is perfectly stable and sounds great. Prasi designed a PCB.
1000V Schottky diodes? I very much doubt it, nor would you want Schottky's in a high voltage situation unless heat-sinked - reverse leakage thermal run-away can be an issue. For PV cells only the diode DC performance is relevant, doesn't imply decent AC performance at all (probably the cheapest HV silicon diodes they can turn out in fact). Soft-recovery are typically the best-behaved rectifiers for low noise generation.
Read what you can and at least establish the topology of the circuit, which is the least you need to even start considering meaningful mods and upgrades. If you don't know what the markings in some components mean, share them and we'll help you out. If there are some you can't physically read because of the way they are installed, de-solder them if you can; if you can't or you're afraid to damage them, share what you have and chances are we can figure out what those components should be based on everything else, even easier if it turns out that it's just like one of the schematics with similar output transistors that have been posted above.
Hi thanks a lot for the advice Actually i cant find the Schottky word in the datasheet I guess the description was wrong
My main worry about diodes is the ability to pass high surge current when capacitance in the power supply is increased
That is what i would do almost always I know of many cases where a capacitance upgrade has been proposed even by the manufacturer as an improvement
My best solution would be big fast caps as close as possible to output devices Very very smart solution Put more uFs where they are needed more
I like this kind of diodes because the cathod is clearly marked I do not even know how to check a diode
I am about to buy a uF meter That would be very useful because i have a sack of unused caps to check Many years ago i bought a box of Siemens Sikorel
I never used them I had jut to have them What a waste They were wonderful in black with gold markings ... i must have them in the cellar somewhere
Hi thank you for the very valuable advice I guess that resistors can be measured directly on board without removing them ?
I think that a viable solution could be to take a picture of upper and lower side of the board, mirror one of the two and try to overlap them ?
The bjts should be quite common I will have the summer to do this
I will try to do and report
It can be possible
There is another thing that i do not like The output mosfets are screwed to the bottom plate like the mains transformer is Its vibrations are transmitted to the mosftes and this i do not like
I will have to put a rubber mat under the transformer and fix it with some plastic straps Just to keep it in place
I will not move the amp around for sure
Hi and thank you very much for this advice But then if
that i can also agree without problem
Then what's all this mythology about these Hitachi mosfets ? What do they have that components of the same type lack ?
Moreover there are great audio designers who use today's mosfets and still get great sound
I agree that replacement without changes in boundary components and bias values might not be feasible. That is also not discussed at all. Wish it were possible.
What is too nice is often inaccessible.
Hi thank you very much indeed I will try to take pics of the board The idea is
1) take pics of the upper and bottom side of the board
2) mirror the bottom side
3) overlap the two pics to show traces and parts
After that i will measure resistors and read the markings on parts I will take some time I have all the summer
If it only stopped to rain ....
There is nothing you can significantly improve there. It's a generic Hitachi schematic for their lateral mosfets, it always plays similar to other such amplifiers, no matter what you do to it. Those transistors cannot be replaced with cheap IRF vertical mosfets. So if you do something wrong, that's the end. I've seen J118/SK413 on ebay for cheap, I just doubt they're genuine. It would only mean adding DC protection for the speakers which doesn't exist so the device is a bit dangerous to use. But that can't be done easily either, because the output connectors are soldered directly to the PCB. So, make something better, and leave this as it is and sell it when you don't need it anymore.
https://www.ebay.com/itm/362255964346
https://www.ebay.com/itm/362255964346
Last edited:
They have a positive temperature coefficient (which makes them safe). They are designed and ready for heavy linear drive.
Furthermore they have low capacity.
You might ask why they (latfets) are not made anymore: because class d switching has taken over. For these amps verical fets are better.
Thank you very much and i have cleared up my mind Can i say that the project is quite good while the construction is poor ?
Because imho a DC protection circuit should be a must And that when well executed it will detract nothing to the sound
It is a pity that good sounding designs are poorly implemented
This amp should undergo a major rework from an expert
Anyway do you mean that vertical mosfets are worse for sound ?
because i understand that modern designs have to use them There is no other choice
Last edited:
I wouldn't know which ones sound better, but they can't be replaced without changing the schematics because they don't have the same thermal coefficient and some other parameters. That's why I said, drive it as it is, and watch something else.
Otherwise, those modern mosfets can be used in classic class A and AB amplifiers, but also in class D amplifiers. They are stronger and cheaper than those vintage Hitachi transistors. Many say they sound worse than Hitachi, but I have no experience with that. Once upon a time I built a number of Elektor Crescendo and Mini Crescendo amps, and they sounded much better than those Hitachi circuits. They were just unfinished, that is, they knew how to oscillate, especially the larger model.
Practically nobody is going to make transistors specifically for audio, because that's a very small part of the market. All are made for industry, for SMPS power supplies and various other applications.
Otherwise, those modern mosfets can be used in classic class A and AB amplifiers, but also in class D amplifiers. They are stronger and cheaper than those vintage Hitachi transistors. Many say they sound worse than Hitachi, but I have no experience with that. Once upon a time I built a number of Elektor Crescendo and Mini Crescendo amps, and they sounded much better than those Hitachi circuits. They were just unfinished, that is, they knew how to oscillate, especially the larger model.
Practically nobody is going to make transistors specifically for audio, because that's a very small part of the market. All are made for industry, for SMPS power supplies and various other applications.
Last edited:
Just because everyone does something a certain way, does not mean it's right. Like Taleb wrote, the majority is always wrong. (I like that quote because of the "always", which probably makes him wrong too, at least sometimes.)
There are online heatsink calculators, and if you use one of those, you're almost certain to find that by far the biggest bottleneck is the thermal pad. You can try to "tune" the pad by shopping for better materials, a thinner insulating layer or whatever, but the #1 thing you can do to increase heat conduction is increase the surface area. But the MOSFET is limited to only 20mm x 30mm? That's why I would consider a conductive metal-to-metal connection for the MOSFET, and then, if you need insulation, use a bigger insulator further downstream.
To go all out, you could try soldering a small copper bar, guesstimating say, 3mm thickness (the more the merrier), 20mm x 80-100mm, perhaps bent 90° as a heat spreader. Then you insulate that, and sandwich the bars between heatsinks that fan out on both sides.
you mean those from the power supply ? i would place some by pass there instead Like plastic small caps
@ginetto61
Seems, you´re set on modifying the amp, no matter what, but lack some electronics basic knowledge,
so why don´t start with the obvious?? Add som PSU electrolytics to (maybe) tighten the "mellow" sound??
If you look at the PCB, there´s room for an extra set. Just add two pieces similar to the two already there, and then listen
Seems, you´re set on modifying the amp, no matter what, but lack some electronics basic knowledge,
so why don´t start with the obvious?? Add som PSU electrolytics to (maybe) tighten the "mellow" sound??
If you look at the PCB, there´s room for an extra set. Just add two pieces similar to the two already there, and then listen
As far as I can see, the diodes are only 6A, I wouldn't add many more electrolytic capacitors. Maybe some larger plastic capacitors in parallel with the electrolytes, 10uF 100V or similar. Or some modern Low ESR electrolytes of 1000uF 63-100V. Pay attention to the polarity of course. It's just that nothing can be done there without disassembling the mosfets from the box.