Frugal DOA twins ...

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
These BD135/136, 137/138 and 139/140 are all "the same thing", binned for voltage and Hfe rating.

http://www.fairchildsemi.com/ds/BD/BD135.pdf

These are 12'5W total dissipation parts (if properly heatsinked).

The 1'25W collector dissipation number is (in my understanding) valid for the 25°C ambient temperature, and "naked" transistor. IOW, a transistor without any heatsink in a place that doesn't exceed 25C.

I don't recommend running these DOAs without a heatsink.

some ideas here:
http://parts.digikey.ca/1/3/to-126-heatsink

While this is entirely non-scientific, I'd recommend choosing the heatsink "with a thermometer". IOW, don't let the output pair run hotter than cca. 40°C at full output into your chosen real-life load (if possible).

These DOAs are somewhat "pure" designs, so they don't incorporate any current-limiting (except base-current-starving) nor temperature stabilizing mechanisms, so the output transistor pair must be just simply heatsinked.
 
On the heels of my frankenstein tube pre, this project is becoming awefully appealling.  I have one of those P_re_son-us MP-20's and I've fiddle faddled with changing the op amps...  Still boring, even for something supposedly very clean.  I'm starting to consider how I could remove the twin servo opamps and implement one of these puppies. 

The reason I would even consider it is that I'm limited on time for new projects so this would be quick and fun, and after the way my tube pre sounds I doubt I'll ever even look in the MP-20's general direction if I don't do something about it.
 
Finally got one of these put together with SMD components.  I used whatever I had around, so that means PZT3904/6 for the Q1-Q6 and some other blasted thing for the current drivers...  I wrote it down somewhere but whatever, it seemed appropriate.

Anyway I used 0805 200mW parts, 4 51 ohm resistors in parallel for the current biasing.  I was struggling to get the thing running, then at one point I did, each half of the supply was pulling about 11mA with a 3Vpp input, which I thought was ok...  Current transistors got hot, probably need considerations for heat dissipation but I knew that.  I wired it up as an inverting unity gain amplifier, but then...  It stopped working.  Tried the unity gain buffer...  No current draw, but no smoke....

EDIT

So here's a bad cell phone pic.  I used bus wire and solder paste, and then reflowed it with a hot plate...  The power transistors are BCX55, but I have some MJD31 and MJD44H11 I might try.

The issue I was having seemed to be in how I connected common wire.  I had the reference lead from the single ended output of the sig gen going to the common, as well as the output reference which was the scope probe alligator.  Then I had the common node from the power supplies connected there too, but nothing else.  Well that didn't work, but it should have...  I removed the power supply common and that's when I did get it working for a few minutes before I moved the connections again and it stopped...
 

Attachments

  • Photo0144.jpg
    Photo0144.jpg
    396.5 KB · Views: 79
Sorry for the delay, but I'm not much around here because of being short on time.

- The 11mA current seems o.k.
- The output transistors should be at minimum 1W types, I posted some dissipation graphs before, IIRC the power draw goes over max. power of the BCX55. MJD44H11 may be better candidates.
- The "ground lead" problem seems too misterious to comprehend to me, sorry I can't be of more help here.

I'm not used to SMD prototyping, so my best advice atm may well seem unappropriate - I'd "ratnest" the damn thing with cheap tru-hole components and BC139 type TO126 transistors..

The "no current draw" suggests that something in the circuit failed into open-circuit mode.
 
Thank you for the response.  The common lead thing appears to be just a dumb thing to try on my part.  If the common lead is pulled from the dual supply circuit, the dual supply just turns into a 0 V and +30 volt rails and then everything in the circuit becomes referenced to 0V since the input and output reference leads will regularly be ground instead of being pulled up to 15V like they should. 

The real heart of the problem is why the parts aren't drawing current when wired up as normal, Either a mistake in the connections or I blew up a part...  I'll test each portion of the circuit by itself and see.  This is a great project though, thanks for your help!
 
Hi everyone !!
I've  just completed a sketch of an adaptation for the DIP8 format. I added some jumper ferrite bars and wide traces to dissipate also the heat  ... I believe this could be a good thing, or? ... Could someone kindly tell me which pairs would be matched transistors?
 

Attachments

  • MelChiOR1337.png
    MelChiOR1337.png
    128.1 KB · Views: 67
IIRC match (within reason):

q1-q2
q3-q4
q5-q6
and q7-q8

this project has been abandoned since, so ... good luck
use big heatsink on output transistors
your boards look very good.
 
Back
Top