Patchwork Reloaded: Circuit Optimization and Board Layout.

[MJL21193]

I have the other 5 boards done. Etched, silkscreened, tinned and ready for parts.



Looking like an amp factory. Wow is this repetitive. The tedious part is measuring each of the resistors. Some are good reliable Yageo but the others are the cheapies from Sure on Ebay.

[GK]

Originally posted by MJL21193
I don't know if it's my imagination, but I think this latest version has a difference in the midrange/treble. Seems more detail is coming through. Could this be attributed to the lower Miller cap value? This is the first listen with it at 47pF.

Thats a typical effect of marginal HF instability.

[AndrewT]

Originally posted by G.Kleinschmidt



Thats a typical effect of marginal HF instability.

yes, peaking on fast transients is heard as extra treble/detail/spitting/ssssing

[MJL21193]

Originally posted by G.Kleinschmidt



Thats a typical effect of marginal HF instability.

And this would be due to the low value Miller cap, right? I'm still operating without a Zobel or output inductor - could this make a difference?

The thing is, I can't be absolutely certain that the extra (apparent) detail is there or not. Like I've said, the best of my hearing is behind me. Especially high frequency stuff.

Here's what I'm going to do: I'm going to build the other 5 with 100pF for Cdom and leave this one as is. I can then quickly switch between each to listen for a difference. When complete, I will be taking this out for some real instrument testing at Chris' (anatech). His ears are probably better than mine.

With everything else fit and figured on this amp, it's not a big deal to swap out the Miller cap on these, when I arrive at the best value.

[MJL21193]

As I've mentioned earlier, the intended purpose for this amp design was for use in a 6 channel amp to drive my active three-way speakers. To this end I have been assembling 6 amp modules and I have good progress there. It' tedious so I have switched my focus temporarily to the chassis.
Shown earlier were my homemade heatsinks. To these I added angles to mount the amp boards. The thermal contact between the angle and the rather uneven surface of the heatsink is less than ideal. To fix this, I did something I've been meaning to do for a while - get shielding gas (argon) to allow me to use my Lincoln MIG welder for aluminum. After about 10 minutes of heating the sink up with my propane torch, I was able to start. I think I'm in love! It was amazing how easy it was to run a continuous bead across the joint. So smooth and spatter free it didn't need any grinding.

Afterwards, I cleaned them up again and painted the inside surfaces black again (I sanded the paint off to weld). They look sharp:



I cut and bent a sheet of 18 gauge stainless steel for what is the bottom, back and the front behind the wood front. I was able to unbolt the threaded rod and feed it through holes drilled in the stainless - makes for a very solid case.

[jaycee]

Looks fantastic. Wish I could do that too

ATM i'm contempating fashioning something new into the guts of an old Teac A919DC I picked up from freecycle. There's a lot of nice parts in there, but a lot of compromises, and IMO the output stage isnt big enough for the voltage used (one pair of 2SC2565/2SA1095 on 71VDC rails!!). At least it's good for a pair of heatsinks

[Nico Ras]

Originally posted by MJL21193
makes for a very solid case.

Hi John, After one year, now we are getting somewhere!

Nico

[MJL21193]

Originally posted by jaycee
one pair of 2SC2565/2SA1095 on 71VDC rails!!). At least it's good for a pair of heatsinks

Thanks jaycee,
Those are nice output though
Most of the stuff I find and scrap don't have such nice transistors. I have a few heatsinks but I can never seem to use them - they don't fit with my plans so I wind up buying or, like I did here, making my own.
The best chassis' are the ex-consumer audio amps or receivers. Just bolt on a new face plate over the old one and a new back plate to cover all of the old holes. I have gone down that road before:





That was an Akai receiver.

Originally posted by Nico Ras


Hi John, After one year, now we are getting somewhere!

Nico

Hi Nico,
It's always one step forward and a half step back with me. I can make steady progress once I have the prize clearly in my sight.

I will most likely have this finished in a week or so. I'm hung up a bit on the front though. I was inspired by the pic you showed me of your headphone amp, and I want to do this one in wood. Coming up with a suitable look, one that matches my vision for this amp, is proving elusive.I do know that I want to keep it clean and simple - no LED's. For the power switch, I'm trying something different - a touch switch. Just a small aluminum or stainless "button" or surface on the front panel.

[kalmara]

Hello, John. Could you share how did u make the boards ? They seem like factory-made, I can do the tracks that good as well with the laser toner transfer method, but how in the world did you do the perfect tin coating?

p.s. Would you care to email me the final review of the borad and schematic files (pdf's or pcb program files), so that I can try the patchwork as well - I saw it a long time ago and wanted to check it out, but didn't like the board and I didn't made it - the "reloaded" board looks far more better, so I might hear it out.


Niks

[MJL21193]

Originally posted by kalmara
Could you share how did u make the boards ?

Hi,
Here's how I do it for single sided boards:

1 - Print the design and iron on. I have been leaving the iron on the paper for longer, just letting it sit there for 5 minutes at a time then move it around rubbing. I find this takes less effort and the pattern transfers are crisper.

2 - Etch with 2 parts peroxide, 1 part muriatic acid. When finished I rinse with warm water and use wet/dry fine sandpaper to remove most of the toner. I then clean the rest of the toner with lacquer thinner.

3 - Print and iron on the silkscreen. After the paper has been cleaned off I dry it thoroughly and spray a coat of lacquer to protect the silkscreen and to make it more visible. You could use clear urethane also, water based or solvent. I happen to have the lacquer and it dries very fast.

4 - Tinning. Go to the hardware store and visit the plumbing section. Buy a roll of lead-free plumbing solder and some plumbing paste flux. Also, a cheap 25 or 40 watt soldering iron. Coat the copper side of the board with the flux, using your finger to spread it thin. Take the tip off of the soldering iron (if it's small. If it has a big tip, keep it on there) and let it heat up. I have ground a flat part on the end on mine that I've tinned.
Melt the solder onto the iron and use that like a paint brush to tin the traces. Be careful not to heat up the traces too much or they might separate from the board. You can use your regular iron for the small traces. Continue until all of the board is tinned. Wipe the board off with a paper towel to get rid of the excess paste flux and clean with steel wool. It's ready for drilling.

5 - Drilling is done after the tinning, even if you have printed the trace pattern with the drill holes open, they will not cover with solder during the tinning as easily as if they were drilled. This is not obvious and I found it out through experimentation.

You end up with a nice looking board that is lead-free and will not oxidize like bare copper will. It has the added benefit of beefing up the trace thickness, for improved current flow.