SMD Discrete Op Amp

Here is an SMD version of a 2520-style op amp. Placing the parts by hand is a bit difficult but soldering everything at once in one baking operation speeds up the process.

Bah... PCB still 1" x 1" :green:
Just jokin Good job, surelly need a lot of patience to handle those parts,
Nice Bonsai 2520 :thumb:

Nice job..... clean nails too..!!

Do you prefer the discrete input devices to a monolithic dual?

[quote author="keefaz"]Nice Bonsai 2520 :thumb:[/quote]

Indeed :thumb:

:green:

Well, we will see how patient I am. I need to build three more. It would be easier with a manufactured board because the solder mask prevents the solder from wicking under the part and causing shorts. I haven't had too much of a problem with that on these homemade boards but I do ring out all the connections that go under parts.

I am a programmer so my fingers hardly every get dirty although I did wash the solder paste off before taking this shot. Mmmm, solder paste . . . Maybe I could be a professional hand model.

I will try a monolythic input pair at some point. I was looking at the MAT02 as a possiblility. It requires some mod of the wiring though. I wanted to try the hand-matching process on these because that is the way the other DIYs are being done, I assume.

Very cool. I'm going to take a stab at a DIY DOA soon. This makes me think I should try SMD

That looks great. Is that your own circuit design? You may want to wash the entire assembly with alcohol. I think I can see some solder flux on the board. SMD is definitely the way to go. Except for the power devices, there is no need to use thru-hole components. Even the input transistors could be surface mount. Nice work. DW.

Even the input transistors could be surface mount.

Click to expand...

If there were as low noise parts out there as we have with TO-92...

Samuel

Wouldn't it be possible to have a batch of these boards made by a factory who also ads the smd parts, so that one would only have to solder the non-smd stuff? I imagine it might be not too much fun stuffing the smds and getting it soldered. Unless it's taken as a zen-pratice or so ;-)

But I admit I have no experience in soldering smd, maybe it is fun. Just feel like I'm getting less patient on such things over the years...

Michael

The idea of having them stuffed by an assembler is a good one. In fact, the board house I am using has an assembler nearby that they will deliver to. Need to develop some volume though because the setup charges are high.

The part that is tedious and requires a steady hand (i.e. no "Big Gulp" coffee before starting) is placing the parts. Once they are placed, the board goes into a small convection oven for about 10 minutes and its done.

I haven't found a good part yet for a surface mount power output device that matches the capability of the MJE243/MJE253 devices. Any ideas?

The circuit is the same as the one discussed in this thread:

http://www.groupdiy.com/index.php?topic=4796

It appears to need some tweaks though. I will post them after testing the "original" version. I just finished the actual preamp board layout last night so I am close.

One could surely tool these up and make these cost effectively in thosands at a time. Now selling thosands in a lifetime is another story.

JR

One caveat about SMD: evaluate your candidate resistors for excess noise. A while back I discovered huge amounts of flicker noise in thick film parts carrying d.c.---thin film, and MELFs that are little more than metal film axials with no leads attached, are o.k. generally.

As long as SMD doesn't get smaller than 0805, or maybe 0603, I find it's pretty easy to hand-solder resistors and SMD caps. Put a little blob on one pad only; slide the part over to it with tweezers and heat the blob. Solder that one side, then go back as solder the other side/pad.

Where I even prefer it is when values may need to be changed, as using two irons on each side, the (unglued) part comes off readily (adhering to the tip of one of the irons, typically---shake it off) without a tearup of a plated-through hole. Then clean the pads with a bit of wick and you're ready to install the new component.

Some progress on the preamp test bed has been made:

oooh looking nice. is that a 312 style pre or something of your own flavour? SMT makes the DOA look mysterious!

Its a basic 312 type circuit with switched gain, a defeatable DC servo, a simple DI circuit and a LED peak display (on a daughter board not shown that goes over the switches).

Do you self etched or are they manufactured??

looks like a 500 series??

Nice...how the DI switch works on your boards??

This design looks like it would be fun to build, is the artwork availible or PCB's for sale?

The boards are homemade right now. I am plating nickel on the card edge connectors in lieu of gold. It works and keeps the copper from tarnishing, but isn't as durable as gold would be. The format is a 500 module size and pinout.

The DI circuit is capacitivly coupled to a gain of one buffer and the DI switch selects the output of the buffer opamp or the output of the mike input transformer after the Zobel network to feed the positive input of the DOA.

I will try to draw up the circuit for the preamp and post it after I get the preliminary testing done. I will also try to make the artwork for the boards available on my web site.

The prototype has had some preliminary testing. The layout of the 2520 clone had a few problems and I inserted the output transistors backwards (duh-oh!) but the opamp is working well. It really cooks too. I was seeing clipping at about 30 Vpp. I am operating the amp at +-18 V and it seems to tolerate that well.





The total gain is a little lower than expected at +60db. I think it should be around +66db. I think this is due to the input transformer not being exactly the correct ratio. This transformer was obtained indirectly from a group buy and has some odd characteristics like the two input windings are not equal in DC resistance which the normal input tranny would be.