Just got a chance to build Mjohnson and Gtose mod cap multiplier as laid out by Prasi. Thanks to Prasi for a beautiful design and super easy to build. Parts count is minimal and I had everything in my bench drawers (best part of all). I am using ON Semi NJ/MJ's instead of Toshiba 2SC5200 and 2SA1943. I added a 100nF 100v filmcap bypass cap on the 1000uF (on bottom side). Still need to test once I find a suitable heatsink or suitable amp case bottom panel.
Attachments
Last edited:
All the best for testing.
I connected the Mjohnson Cap Mx to my 400VA 18v transformer fitted with dual square puck rectifiers, and then installed my old M2 amp into my 4U case. The Alpha 20 amp that was there has since been removed and sent to JPS64 in Germany as a gift. So this was a convenient Class A amp that I had available.
The cap Mx with no load was adjustable in voltage drop as low 1.3v and typically around 2v to 2.5v at the middle of the 1k pot. The 1.3v seems about right as that corresponds to 2x 0.6v drop across two BJTs.
The amp plays nicely with no hum - measured noise at speaker output was 400uV rms (within spec for M2). This PSU is a lot more compact and less expensive than a whole bank of four 22mF caps. I am using just two 10mF caps to smooth the rectifier bridge.
So all in all, the MarkJohnson Cap Mx is a success. It has less voltage drop than the Easy Peasy MOSFET cap (4v to 4.5v) so dissipation is about half as much and you get more voltage for higher peak power for your amp.
The heat from the two BJTs is minimal so bolting it to the bottom steel plate of my amp case works fine (it's the perforated part even). 2.5w is not much. You may need a heatsink if you plan on running more current - I am at 1.27amps. If you use it at 3amps - probably should be bolted at least to an aluminum floor plate. For my setup the bi BJTs are running at 50C and the steel plate is 41C.
Thanks to Mark Johnson, Gtose, and Prasi for this cool Cap Mx.
Here it is running with my M2 amp.
The cap Mx with no load was adjustable in voltage drop as low 1.3v and typically around 2v to 2.5v at the middle of the 1k pot. The 1.3v seems about right as that corresponds to 2x 0.6v drop across two BJTs.
The amp plays nicely with no hum - measured noise at speaker output was 400uV rms (within spec for M2). This PSU is a lot more compact and less expensive than a whole bank of four 22mF caps. I am using just two 10mF caps to smooth the rectifier bridge.
So all in all, the MarkJohnson Cap Mx is a success. It has less voltage drop than the Easy Peasy MOSFET cap (4v to 4.5v) so dissipation is about half as much and you get more voltage for higher peak power for your amp.
The heat from the two BJTs is minimal so bolting it to the bottom steel plate of my amp case works fine (it's the perforated part even). 2.5w is not much. You may need a heatsink if you plan on running more current - I am at 1.27amps. If you use it at 3amps - probably should be bolted at least to an aluminum floor plate. For my setup the bi BJTs are running at 50C and the steel plate is 41C.
Thanks to Mark Johnson, Gtose, and Prasi for this cool Cap Mx.
Here it is running with my M2 amp.
Attachments
Last edited:
Here is the sch I made for the layout being used for xrk's build.
Juma's Easy-Peasy Capacitance Multiplier
Post no. 512 gives the correct gerber files for anyone interested getting the PCB's manufactured.
regards
Prasi
Still in mono so far, and building up the second channel cap Mx while listening to the first one. I have to say it works quite well - no audible hum or anything coming through. With 10mF caps, the ripple at the smoothing caps is measuring 294mV rms with 1.27amp bias current. After the cap Mx, the ripple is measuing 1mV rms on my Fluke 101. The music sounds great. The M2 is an amp that I have particularly had trouble with hum thus far, and with the Cap Mx, it seems not to be an issue, at least in mono. I don't have dual trafos, so thinking of maybe running a complete separate set of rectifier bridges and second Cap Mx.
Here is how I have mine hooked up. Ideally, this is a monoblock and if you make two of these setups, you will not have any hum. I cannot say how it will sound when I connect two sets of bridges to the same secondaries.
I can say that, as wired above, there is no hum or ground loop noise. The normal hum in the M2 is not there.
I can say that, as wired above, there is no hum or ground loop noise. The normal hum in the M2 is not there.
The M2 has no on board rail caps if that's what you are asking. Totally dependent on getting a clean PSU. The Pass F6 PSU works pretty good and uses 8x33mF caps in a CRC-to-dual-RC. I made one P2P and was using that as my main 24v dual rail PSU for a while.
I installed the second stereo channel Cap Mx with its own set of rectifier bridges and a common ground point at the NTC to chassis GND/Earth GND. The right channel works fine, no hum with output noise at 400uV rms with source plugged into both channels. But when source is plugged into left channel, noise kicks up to 6mV rms. It's asymmetric - so strange as built the same and symmetry of placement is very close. It might be something in the amp itself. Even when the left channel is plugged in and RCA for right channel not connected there is noise in left.
I will have to debug this another day. It is promising that the right channel has no hum when the left channel is plugged in. In the past, that was automatic trouble the moment a second RCA jack was plugged in.
Now, I have noise in the amp itself regardless of left-right ground loops. Even in mono, so I would guess it is an amp issue.
Main heatsinks are running 40C, floorplate near cap Mx is at 44C, amp MOSFET bodies are at 56C to 60C, cap Mx BJT bodies are at 55C. Temps are all comfortable and manageable levels.
It occurrs to me that these dual rail cap Mx, if fitted with smaller TO-220 ouput BJTs could be suitable for smaller amps and other projects like headphone amps.
I installed the second stereo channel Cap Mx with its own set of rectifier bridges and a common ground point at the NTC to chassis GND/Earth GND. The right channel works fine, no hum with output noise at 400uV rms with source plugged into both channels. But when source is plugged into left channel, noise kicks up to 6mV rms. It's asymmetric - so strange as built the same and symmetry of placement is very close. It might be something in the amp itself. Even when the left channel is plugged in and RCA for right channel not connected there is noise in left.
I will have to debug this another day. It is promising that the right channel has no hum when the left channel is plugged in. In the past, that was automatic trouble the moment a second RCA jack was plugged in.
Now, I have noise in the amp itself regardless of left-right ground loops. Even in mono, so I would guess it is an amp issue.
Main heatsinks are running 40C, floorplate near cap Mx is at 44C, amp MOSFET bodies are at 56C to 60C, cap Mx BJT bodies are at 55C. Temps are all comfortable and manageable levels.
It occurrs to me that these dual rail cap Mx, if fitted with smaller TO-220 ouput BJTs could be suitable for smaller amps and other projects like headphone amps.
Last edited:
Perhaps add a second NTC to chassis ground - It might work. Or maybe just go dual mono with a second transformer - you could stand them on their sides to fit with one of those right-angle brackets.
In Mark's K-Multiplier, there's a couple of optional transistors for lower current - perhaps these would be feasible choices?
In Mark's K-Multiplier, there's a couple of optional transistors for lower current - perhaps these would be feasible choices?
I'll give that a try later. In looking at it, I think it may be the arrangement of the cap Mx location on left channel is next to power trafo primary wires. Maybe move rectifier pucks to front and cap Mx to back to keep mains AC away from cap Mx and amp signal trafo. This M2 is a finicky amp and if you read M2 thread you will see that it is not an easy maintenance amp, pages and pages discussing how to get rid of buzz and hum - very picky about case size (apparently 300mm deep is too small to allow enough room between power trafo and signal trafo) and needs shielding on trafo, mu metal, etc... to keep buzz at bay. This was just a test amp to put the cap Mx through its paces. I will go back to an Alpha as soon as I make another one. Alpha is dead quiet and not picky about positioning in case. A small unshielded signal trafo in a box stuffed with a mains transformer is a bad idea in general. If you look at quality signal transformers for hifi/audio like Cinemag, they come in a shielded capsule with flying leads coming out. The Edcor on the M2 is cost effective, but probably used for some other purpose with EMI interference not a consideration.
Last edited:
Hi Prasi,
the updated Gerbers+Excellon 2 don't work on jlbpcb Gerber viewer. If you upload it as it is, you get an error message, that points to missing drill file. If you copy the drill file to the output file, you get and error message that the board outline is missing and a lot of drill holes are displaced. My idea is that you have drill holes outside the board outline, that cause the trouble.
Btw, Gerberlogix's viewer shows all ok.
BR
Günni