Good thought, Rick. I've made a slight change to the layout and the 2.2u/50v 10.2X8 LS7.5 will fit nicely.
I think Mouser has got the WIMA 10u/50v MKS2 size wrong by mixing the height and the width, as the datasheet states the height is 16 and the footprint is 11x7.2. If that is the case, the cap will fit in the now revised layout with no problems. Do you happen to have one of them and can verify? Thanks. Otherwise there is no way that I can make room for a 16x11 footprint for the input cap.
I think Mouser has got the WIMA 10u/50v MKS2 size wrong by mixing the height and the width, as the datasheet states the height is 16 and the footprint is 11x7.2. If that is the case, the cap will fit in the now revised layout with no problems. Do you happen to have one of them and can verify? Thanks. Otherwise there is no way that I can make room for a 16x11 footprint for the input cap.
Those are not so good for audio, polyester or for Panasonic it says it is plastic film nor specific wich type.
Better to choose polypropylene type like this for example: ECW-FD2W225K Panasonic | Mouser
BR Damir
That is a huge one, Damir, but I think one can bend the leads and somehow fit it in there. Our heat sink is only 19mm to the holes for that cap, I'd consider drilling a couple holes on the aluminum plate and zip-tie the cap on it for a good mechanical support.
That is a huge one, Damir, but I think one can bend the leads and somehow fit it in there. Our heat sink is only 19mm to the holes for that cap, I'd consider drilling a couple holes on the aluminum plate and zip-tie the cap on it for a good mechanical support.
OK there should be smaller type, just choose polypropylene, mybe some from here?
Film Capacitors | Mouser
Or keep it 1 uF, it will give -1 dB at 6.4 Hz if input resistor is 47 kohm or -1 dB at 15 Hz if input resistor is 22 kohm.
Last edited:
Chris, could you add those high frequency Zobels( R C) connected from power mosfer sources to the ground showed in Nattawa schematic here: http://www.diyaudio.com/forums/solid-state/243481-200w-mosfet-cfa-amp-24.html#post3698456. One of the my simulation shows that it can prevent high frequency oscilation on some difficult load.
Richard,
Are you not able to solder one up yourself? If not, since you are a solid contributor to this site, I can offer to do it for you. I am not going to do it for everyone, sorry. If so, I do expect that you supply the pcb & pay for the components. Providing the parts shows your commitment to me and I do not get stuck, if you decide to back out.
Assumption is that the group releases/shares the gerber data or does group pcb buy thing.
Looks to be 1/2 a days assembly work for me, also need to do a prelim test for functionality.
I am working on other amp designs, so I am getting the pieces together to do this sort of testing. Little bit more to it than just one PA pcb.
What I do for my projects is input the BOM into Mouser project manager. or the Digi-key version. Mouser prices are almost consistantly lower than Digi-key. I can share the project access code, so that others can order from it. Hopefully all the parts can be obtained from one vendor, as it just makes it easier and minimizes shipping costs.
Rick
Are you not able to solder one up yourself? If not, since you are a solid contributor to this site, I can offer to do it for you. I am not going to do it for everyone, sorry. If so, I do expect that you supply the pcb & pay for the components. Providing the parts shows your commitment to me and I do not get stuck, if you decide to back out.
Assumption is that the group releases/shares the gerber data or does group pcb buy thing.
Looks to be 1/2 a days assembly work for me, also need to do a prelim test for functionality.
I am working on other amp designs, so I am getting the pieces together to do this sort of testing. Little bit more to it than just one PA pcb.
What I do for my projects is input the BOM into Mouser project manager. or the Digi-key version. Mouser prices are almost consistantly lower than Digi-key. I can share the project access code, so that others can order from it. Hopefully all the parts can be obtained from one vendor, as it just makes it easier and minimizes shipping costs.
Rick
Sure...
Is there anything else that needs to be incorporated into the board?
I am going to do 3 types of emitter resistors and put the 1000uf on the front supply as a option.
How about some 1000uf or more on the OPS power connection points?
Any protections or such?
How about jumpers on compensation points? In or Out?
I will post a general layout before I do the detailing.
Chris, I thought that I told to put a jumper as a provision to choose between high and lower GNFB, look at Nattawa schematic, it's J22-1 and J22-2.
You can make possibility to use 1000 uF on the OPS too, I think that 220 uF is enough but if someone whants more one can choose in that case.
I thought about putting the big cap over the SMT parts but that would be against general practice, and, in my humble book, aesthetically less favored when other options are available. I seriously think it would be better leave the solution to individual DIYers to work out, such as, bending and shaping leads to make fit, or install a 2-pin pin-header and a pair of wires leading off the cap of choice zip-tied on the heat sink, and so on.
There might be a possibility that I give up the donor corner on the upper left for the F/B resister chain and use that area for the input cap instead. That area would be big enough for any insane sized input caps. The resistor chain can be on a separate PCB. But perhaps this would unlikely happen during prototyping phase.
I'm glad you've had an offer from member "rsavsa". However if you were referring to a more general concern of SMT being less friendly for DIY-ing, the design was made with professional SMT process in mind. I know it will go through the SMT assembly process where I work with no problems.
If there ever will be quantity demand, the board can possibly be ordered as SMT assembly in either 2x7 panelization or individually, though that will likely be off-shore sourced.
This way DIYers will need to cover thru-the-hole assembly only.
I don't have that option in my layout design. That resistor would require some sort of heat sinking, possibly sharing the main heat sink, and can make assembly complicated.
Resister chain on a daughter card seems to work perfectly in my design. It takes output sample exactly at the output inductor pin and returns the sample current right at the star GND without having to share any GND path with other part of the circuit. I don't see how I can do as good with other options.
I came up with an idea. The power resistor (isn't it TO-126, by the way?) can be mounted on the main heat sink with a screw driver going through the mounting hole of Q24. Q24 does not need a mounting hole as it is directly soldered onto the PCB, therefore, that hole is vacant.
I probably can put two holes in the resister chain daughter card for the leads of the power resister, so that one can choose either way. The assembly, however, can be a bit tricky, as the joints between the main board, the daughter card, and the power resistor are all rigid. One does not want to put a lot of stress to any of these when assembled. Perhaps some simple jig can be made for shaping the power resistor leads and when soldering it to the daughter card for a correct spacial positioning.
It is always a time issue with me. Multi-tasking like crazy. Trying to do a brain-dump for you guys where ever I can contribute something from experience.
I would be happy to pay for time and materials etc. Thx- Mooocho.
-Richard
I would be happy to pay for time and materials etc. Thx- Mooocho.
-Richard
Last edited:
