Don't fear the 2SC1583

[armdnrdy]

As some of you may know...Ibanez was very fond of using the 2SC1583 (DIL NPN common emitter transistor) in many of their flanger designs.

The obsolete 2SC1583 is very hard to come by and fetches a high price when you do find them. I sourced a few for testing purposes to compare the performance to matched NPNs.

I'm currently working on an Ibanez FL9 build for a going away present for a friend's son. (He was admitted to the Berkley School of Music) and while rewriting the FL9 calibration procedure (the original is a bit sketchy) I took a look at the Boss BF-2 files that I have.

I noticed that the circuit is very similar (okay....who copied who) but one big difference is the lack of the 2SC1583 in the LFO/clock section.

Boss uses two 2SC945Ps in it's place. There is no mention of matching/selected on the schematic and I do not see paint marks on the trannys in any of the numerous images of the BF-2 board available on the net.

I am assuming that Ibanez used the 2S1583 not for matched VBE or thermal matching but for convenience of the DIL common emitter package.

Any thoughts?

[slacker]

I had a look for the schematic to see what they were using them for and found this thread http://www.diystompboxes.com/smfforum/index.php?topic=71359.0 general consensus seems to be you don't need to bother with matching. So yeah perhaps they used the part for convenience.

[armdnrdy]

Thanks for the reply Ian,

Before this build gets "man gift wrapped" (hurriedly thrown in a bag on the way out the door) I'll do a bit of comparison testing between the 2SC1583 and a couple of matched 2SC945Ps to see if there is any scoped or audible difference in this application.

[armdnrdy]

Update:

I've scoped a 2SC1583 and compared it to a matched set of 2SC945P and an unmatched set of 2SC945P.

Here are the results:

The 2SC1583 in circuit.


The matched (to the nearest millivolt) 2SC945P in circuit.


The unmatched (7 millivolt difference) 2SC945P in circuit.


As you can see by the images, the 2SC1583 and the matched 945Ps are almost identical. The unmatched 945Ps are a bit more unsymmetrical at the top of the wave.

I didn't connect the 945Ps thermally, I just connected the emitter of the pairs.


Matched 2SC945P


I built a universal BJT matcher based off of a design by Ray Wilson.
http://www.musicfromouterspace.com/analogsynth_new/TRANSISTORMATCHER/TRANSISTORMATCHER.html

Noticing similarities in the design of the NPN and PNP matcher circuits, I was able to incorporate a 3PDT toggle switch to choose between NPN and PNP and house both circuits in one enclosure.
I also designed the board to accept three different transistor pin outs.

BJT matcher build image.

[Paul Marossy]

BJT matcher build image.

Nice piece of test equipment. 

[armdnrdy]

I like me the proper test equipment! 

[digi2t]

I used one in my array Superfuzz build for the octave section. Worked great. Overkill, I know, but I had it left over from the Japanese Guitar Synth trials, so what the hey!.

[Govmnt_Lacky]

@Larry

When I built my FL9 clone way long ago... I believe I used some standard NPN high gains (5088s I believe) and just matched them for gain. Threw them in the circuit and they sounded fantastic!!

So yes... I believe the whole "matched for this or that" mistique is just that.... mistique 

[armdnrdy]

I'll be honest....I don't know if it was luck (the sixteen 945Ps I tested were from the same batch) or what, but I don't see an extreme difference in the 2SC1583, the matched, and the unmatched 945P.

I don't hear any kind of audible difference either.....it sounds like a flanger to me!

I'm not sure why Ibanez would use the dual NPN when they had a large supply of single transistors at hand.
It couldn't be conserving board space with a dual in line package. The two single transistors fit in the same physical space.

[slacker]

If you want to see if they used it for temperature  reasons try heating one of the transistors in your pair up and see what happens to the waveform. For this application I'd imagine just putting two roughly vbe matched transistors in close proximity is more than adequate. They probably don't even need to be close together you're unlikely to get much of a temperature difference in different parts of a sealed pedal.

[armdnrdy]

You might get an adequate temperature change in a sealed pedal if you are playing an outdoor gig in Arizona, in the month of July, at 1:00 pm.

I don't anticipate such an event.....but if it were ever to take place.....I'm sure I'll have a few professional road crew personnel on hand to open all of my stomp boxes and fan them!   

[slacker]

Yeah you could get big temperature swings, what I meant was in a pedal both transistors are likely to be about the same temperature, what ever that happens to be, so there's probably not much to be gained thermally bonding them together or using a dual transistor.

Like the tester by the way.

[armdnrdy]

Okay then how about the same scenario in the scorching Arizona sun....but one side of the pedal is touching my large beer cooler (that I usually keep on stage) and it is enough temperature difference from one transistor to another to throw the whole effect off....forcing me to become frustrated....I throw my guitar (a cheap one!) down on the ground in disgust......blame, it on the roadies and the manager for not keeping my transistors thermally balanced (after all....I told them so....many...many....times!!) and walk off stage disappointing the whole crowd including a 16 year old terminally ill girl that got up out of her hospital bed to see me play one last time!

Well.......I don't know.....I think I'll use the original 2SC1583.....after all.....I don't want to leave all that good beer on the stage!!