If you guys are talking about adding resistors across the main filter caps for discharging, I think 6.8K 2W is more appropriate. 680R across 60Vdc will be dissipating 5.3W.
Yes, I talking about resistors serving as bypass for quick discharge capacitors. Found somewhere need to be from 6,8K to 10K 2W.
New Year - new schematics
Since last post, I have had fun restoring two MOSCODE 300s. Sounds great and has lots of bass slam. For those who know a little about these fun amps, you may remember that they are based on a cascaded 6DJ8 with a 6FQ7 as a white cathode follower driving the MOSFET outputs. The signal has to pass through 4 capacitors. So the idea struck me: why not re-design the cascade and white cathode follower to eliminate at least two capacitors that we did not need? - I still like to keep the two caps between the MOSFETs and the tube section. Sure, a direct coupled output stage could work, - I just do not trust that the tubes will not drift over time. Still, going from 4 caps to 2 seemed well worth it. This also means that one can splurge a bit on the quality of the caps (I went w/ Russian Teflon in glass).
The MOSCODE cascade VAS section yields a lot of amplification. Using to TubeCad and plugging in the 33K and 39K plate values, the compounded open loop amplification came out at ~ 200X. Wow! Also the plate resistors seemed excessively large which made the 6DJ8s operate mostly on the curved section of their gridlines. Great for max amplification - and more distortion... The MOSCODEs reflect some older design ideas popular in the 1980's - lots of amplification and lots of feedback. Great for damping factor and bass - not so good for an airy top end. So lower plate resistors was a design goal.
Speaking of airy top end; here the SE 300B amp w/ zero feedback is a better choice. No bass slam though. This hobby is always about giving something up to get something else. But could a compromise be reached? Could we get good sound-stage and bass slam with a hybrid amp? The earlier SRPP version of my Hafler conversion sounded great. And it had no feedback. Yet, the MOSCODE 300 had better bass...
Since I have a few Haflers sitting around, and I did not wanted to butcher stock MOSCODEs, it was time to build yet another conversion. This time the design goal was:
a) Bass slam - so feedback = yes
b) Clean/warm midrange - so 6SN7s w/ Allen Bradley carbon comp resistors
c) Smooth treble - so as few coupling caps as possible
After some trial and error (as usual), here is what version 6 looks like (see attached).
There is a couple of tricks here I have not come across elsewhere. As you can see from the schematic, the two first triodes look very straight forward: cascaded w/ direct coupling. Each delivering roughly 10X gain so this VAS has only 100X amplification - which in turn meant that we could get by with half the feedback of the MOSCODEs. The 22K plate resistors offer a better plate load to plate resistance ratio for less distortion. Here we have a 3X ratio whereas the MOSCODEs have 10X. Direct coupling also got rid of the first coupling cap. Then there was the coupling cap from the plate of the first triode to the grid of the second triode in the white cathode follower driver. This had to go. Inspired by Broskie and some of his posts on Aikido, it struck me that we really need the bottom triode in white cathode follower to act as an active resistor to the top triodes cathode follower. The white cathode follower normally take this signal from the plate of the top triode via a cap... but since we already had another triode to spare, why not a) lose the plate resistor on the cathode follower (ala Aikido) and b) drive the bottom triode directly with the same signal as the second stage in the cascade? Voila! Done. No more cap and efficient use of the 4 triode available. Simple is best.
I have never seen this configuration before, so I might have "invented" something new. Probably not...But if someone reading this post has seen this approach elsewhere, please post a link.
So how does it sound? Great (as all amp builders will say). Lots of bass slam and a bit more aggressive treble. Midrange - Especially lower midrange is there. It is still breaking in - so perhaps the treble will get sweeter.
If you chose to build this MOSCODE 300 inspired amp, set the BIAS for 300mA (like MOSCODE) and then set the DC offset to 10-20 mV on the output terminals. Also like MOSCODE, you can go AC on the heaters without problems. Just make sure the AC heater is tied to the -60 rail via a 100 ohm resistor so that heater to cathode voltage of the 6SN7s is not too great. This also keeps hum in check. Speaking of hum: use a star grounding scheme and terminate all ground leads between the power supply caps.
Happy 2011
PS: I have posed a .PDF version of the schematic. That should be easier to read than the low res graphic.
Hey Vilfort
How is the amp working? Amy more versions or was this the top of the line? Just got a factory wired unit again for next to nothing and plan to gut and go with tube front end.
Before modification
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
After modification
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
DH-200 DH 101 bad boys
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Thank you Fab
Here I will try in a few words as possible to describe the difference before and after the modification, but it's all subjective and depends on the listener. The first big step was the replacement of the input capacitors and wiring, I tried several different manufacturers including Mundorf MCap PP cap. Nichicon MUSE, and finally Black Gate NX, which I decided to leave. You probably know that the Black Gate capacitor requires a little more time to start working corectly about three days is needed to go better . What I have noticed is that it is much more transparent but is still a lot missing, I can freely say it’s a big step and maybe the biggest since just heard.
The next major step was the replacement capacitor C3, C4, C20, C21, I noticed that the entire middle and upper range is starting to work much better, definition of high and middle range is incredibly better I used WIMA FKP2 must say that every time I change parts I listened music two to three days and trying to observe whether or not there is a difference. Basically it was the music that I listen to constantly could really tell the difference, it is also the best way to detect changes in the device itself. After all, this was followed by the replacement of capacitors in the power supply, must say is very interesting, I need to say that the DC voltage on the speaker before replacing the capacitor C7, C10 was about 100 mV after replacing the caps DC voltage has dropped dramatically to 50mV in both channels. Replacing the capacitors in the power supply and PCB improve the lower range, the bass is much more precise and tight, can say that with the greatest of ease is low frequency, vocals are much cleaner feel the breadth and depth of the scene. Current mode is set to 275 mA and the temperature measured at the heatsink was about 40 C after 1h, R45 was changed to metal film resistor of 10R all wires in amplifire is silver plated VDH CS 18 . Need to say that the device has no hum, no noise is absolutely quite and stable.
Last edited:
Bonjour rhr9
Thank you, of course the coper is originaly mounted by factory and indeed it makes sense to eliminate interference from the transformer.
Last edited:
Is it all around the transformer or just on top ? I could upgrade my amp
with such a copper plate .
@ + Philippe.
( PS : I don' t know how to say "Bonjour" in the Serbian language ... )
I do not know how this happened to you but 25 A - 40A only if there has been a short circuit on the unit, but in this case the first burned fuses that limit the current to 5A (first fuses are themselves semiconductors) believe me. Larger diameter wire is placed for several reasons one of the reasons is that to remove the grounding loop problem by default always pursued short wire paths and little resistance. What I noticed before modification is that the wire that was GROUNDING BUS has become oxidized which is normal after 30 years of listening at temperatures within the unit. Of course I thought about this but not in the direction that you mention, but something else entirely contemplates that I would, because of the high temperatures could reach the cold solders not only groundig bus but on all the more relevant points where high current flows rectafier, (power switch most important contact) and I turned particular attention twisting wire around the contacts. The idea that you mentioned is very good but is currently looking for a specific copper son that would fit nicely into the concept of amplifiers.