Simulation - Pass MF-12 MOSFET Quasi Comp (Original) Modification for HK Citation 12

[PB2]

Old design - yes I know!
Simulated (LTspice) Nelson's old MF-12 mod to the HK Citation 12. I used andyc's EKV MOSFET models and Bob Cordell's transistor models all included in the .zip file. Just unzip the file in a folder and install the latest LTspice if you don't already have it. See the attached .zip file. I will discuss some of the issues with it here if others are interested in trying the simulation. Might be a good learning tool for Spice. Some of the issues are: Cross conduction during heavy overload
Slight crossover distortion on one edge only - any guesses as to why?
Fairly serious loss of power due to the front end not having higher rail voltages,
clips on the low side at 33-34V peak which is 8-9 volts below the rail.

I had to substitute devices for the models that are available and known to be good. Outputs are IRFP244 the rest you can see in the schematic.

Seems the complementary version is better without making mods to this design, however the IRFP240 has a higher current rating than any of the other devices and it would be nice to take advantage of that and also avoid the odd behavior of the IR P-channel parts.

[PB2]

I reread the article and noticed that the output stage bias is supposed to be 100mA, it is set for about 30mA in the above simulation. Change R15 from 1.35K to 1.26K to bump the idle up to 100mA. This cleans up most of the crossover distortion.

[Ar4]

So, complementary version would be better? IRFP9140 and IRFP240 is my choice. After trying different schematics in simulator, I returned to Citation 12, as it is simple, good, and not expensive. Is there any PCB design available for complementary version? Or maybe p2p is way to go with such a low parts count?

[TJ]

sorry i know is a bit off topic , can someone please show me how to get this amp working with 2sk134 and 2sj49 ? thk

[PB2]

Originally Posted by Ar4

So, complementary version would be better? IRFP9140 and IRFP240 is my choice. After trying different schematics in simulator, I returned to Citation 12, as it is simple, good, and not expensive. Is there any PCB design available for complementary version? Or maybe p2p is way to go with such a low parts count?

I've only simulated these and have no experience working on one, or listening to them. I would say yes the fully comp version is better especially if you use a better P type device as you have chosen there - watch out for capacitance some have much more. I simulated the fully comp version here but with a rather weak P ch device, you can put a second one in parallel to see what it will do with a better device - or find models for what you're planning to use:
Simulation - Pass MF-12 MOSFET Fully Comp Modification for HK Citation 12

I don't see any problem with P2P wiring, breadboard etc. but you might search here or on ebay for boards if you really want them.

[PB2]

Originally Posted by TJ

sorry i know is a bit off topic , can someone please show me how to get this amp working with 2sk134 and 2sj49 ? thk

I'm not familiar with those parts, people have used many different parts in this design so you might start a thread with those part numbers in the title and mention the Pass Citation MOSFET.

[MikeBerry]

I see you were talking about the complimentary design. I have a modified version of the circuit that is being discussed. I am using an IRFP244 and IRFP9240. It sounds pretty darn good in my opinion.

I will atach a "schematic". I am a novice and did not have the original schematic. It is a cross betwenn a schematic and an actual diagram of the PCB going to the connectors on the chasis. If someone could help me get this into LTSpice I would really appreciate it. I have a hard time visualizing how this looks in an actual schematic form. It took me about ten hours just to get the drawing done!

Hi Pete!

[PB2]

Hi Mike,

Have you loaded LtSpice on your system and brought in the comp version that I did?
You can then print out the schematic and yours, and mark it up in red taking the values off of your drawing, remember that it will be very close in design so it should not be much work.
Save my original and save yours in a new name. Just get started, look at the RC network across the output, make it match in the real schematic and cross them off of the printout of yours, then move on to the next component - it is not that hard.
In LtSpice just double click on the .asc file to open it, right click on a component to change the value, it is very intuitive to use the cut, wire, move, or copy tools.
If you find it is taking you hours then I'll do it for you.
Are the 1n5305 and 1n5297 two diodes in parallel in your drawing?
You will not be able to run the simulation once you change it since many of the semiconductors are not in the default library but at least you will get a schematic.

[MikeBerry]

I was able to load LTspice and open your diagram and have it printed out. I will hopefully have some time tomorrow night and I will follow your suggestions. Hopefully I will end up with a good schematic. Yes, the 5305 and the 5297 are in parallel. Thanks!!

[PB2]

It should be a good feeling when you get a finished schematic.

I once traced out the schematic for the Harman Kardon Citation II years ago when I had a lot more time on my hands!