Synth-DIY Members' Substitution Info

PART #Descriptive TextAttribution
bc109c Lots of them. It's a NPN low noise (not too low) audio/general purpose tranny. Try others from BC family (very common here in my country, don't know about yours), like BC209C, BC149C, BC549C or BC550C (the last, the best). In the american family, the 2N5818 is a good and easy choice. Edu Silva
ICM7555, TSC555, CMOS version of 555 timer. Glitchless Scott Gravenhorst
MPF-102 2N3819 FET N-channel Scott Gravenhorst
From: The Old Crow 
Subject: Re: CA3046's

On Sun, 30 Jan 2000, Chris MacDonald wrote:

> I just bought some CA3080E from Jameco actually.  What I got were RCA
> 100135, for which I couldn't seem to find a cross reference.  Jameco
> says it's an exact substitution, does this part number sound familiar to
> anyone?  They work as 3080's anyway...

  Got the exact same parts (100135) for 3080Es.  Seem to work fine.  Must
be some house-marked chips for an old customer.

The Old Crow/Chris MacDonald
In a message dated 5/10/00 10:54:06 AM:


Unlike almost all opf the other currently available opamps, the LM301 is a 
uncompensated amp.  So, if you're trying to replace one in an existing 
circuit (or trying to clone a circuit that uses one), it's probably important 
to not try to substitute another opamp type.

I think that National is still making LM301's, and I usually buy them from 
Jameco.  Most electronic part stores carry Jameco's "Jim-Pak" line of parts, 
and I think the LM301 is in that series.  If you're having trouble finding 
one, you could also use one of the NTE substitute parts.  They have an 
uncompensated LM301 substitute called NTE975, which is the 8-pin DIP package, 
and the NTE1171, which substitutes for the metal can package.  The NTE parts 
will be a bit more expensive than the originals (and sometimes not quite up 
to the same quality level, unfortunately).


Michael Bacich
Michael Bacich
TL074 AD713 Alex Dickey
>Only problem with this circuit is that the FETs specified, 
>BF245C, weren't easy to get really. I could get 2 at one 
>shop, and 4 at another, exhausting both shops' supplies. You 
>cannot substitute BF245A or B, as the A, B, and C denote the 
>pinch-off voltage (A: -0.4..-2.2V,  B: -1.6..-3.8V,  C: 
>-3.2..-7.5V) and drain saturation current. Of course you 
>could choose another FET and adjust the LFO and voltage 
>levels accordingly, but I have no FET equivalents guide.

I have got a FET equivalent guide!! It lists the BF244 or the 2N5397 as

Jeremy Brookes
>> I have got a FET equivalent guide!! It lists the BF244 or the 2N5397 as
>> substitutes.
>So, what FET's are good replacements for FET's like 2N4856, 2N4859
>and MPF-102?

Aaagghhh!! I knew this would happen. Okay for now:
2N4856 = BSV78 or 2N4859
2N4859 = BSV78 or 2N4856 
(tee hee that was easy)
MPF-102 = BFW61 or 2N5486


Jeremy Brookes
>Clairex's CLM6000 optoisolator, are they
available these days, or are there other some kind of devices?
NSL-32 is a good substitute available from digi-key.

John Simonton
> Also, for the ASM-1, can the LM394 be substituted for the MAT-02?

>From a pure electrical point will the LM394 be usable as far as I can see.
The metal cap version LM194H, LM394H, LM394BH and LM394CH will work rigth away
from a pin view... but the 8 pin DIP/DIL LM394N and LM394CN will be a bit more
problematic since it will not directly work with the surrounding components.
You will need to cut away the pin 4 and 5 and let the 100k resistor be mounted
over the chip since other wise they will collide physically.

I have not tested the LM394 and have none to try it out. I do use the MAT-02
with great success. My VCOs goes from below 0.2 Hz to around 110 kHz so there's
about 19-20 octaves...

Magnus Danielson
> Can someone please tell me what a typical gain value for a BC109
> transistor is?  I'm looking for a suitable American substitute.

The BC109 (without a B or C suffix) has a typical current gain of
200 (at Ic = 2 mA), and is typically used for its low noise specs.

Therefore, I'd recommend using, for example, a 2N5089 (National, 
Motorola) as substitue, if the gain parameter isn't critical.
(The 2N5089 has a gain of 400).

Joachim Verghese
>>    According to my contact at Newark Electronics, the 2N4856 is a
>> discontinued part.  However, there are replacement parts that can be
>> substituted for this, the NTE466 or ECG466.  Has anyone tried these?
> I've used NTE466 for my ASM-1:s and it works excellent.
> I even think Gene suggested these at some occassion...

I'll second this.
The NTE466 does work very well - and, because it's NTE you can usually find
it in local electronics shops.

The stinky thing is the price - over $2 a piece (both from Mouser and

The tab on the can goes towards the top edge of the PCB. The pins line up
with the holes....

- CList
Christopher List
>>> does anyone know if there exists a substitute for the AD633JN ?
>>> (four quadrant multiplier/divider)
>>> bert

Also there is the MC1496, a balanced modulator/demodulator that can be used
as a four quadrant multiplier. It was used in the ETI vca/ringmod modules in
the 4600 series. I five of these modules myself. Don't ask me for quality
comparisons though.

Ken Stone
> I can't find BC557 or BC559 transistors (and I found only one place that has
> BC547's...). I need a whole bale of these for my slowly-growing Penfold. Is
> there anyone who can suggest substitutions, or, barring that, point me at a
> reference to let me find my own substitutions? ("Teach a man to fish," and
> all that sort of thing... I'm sure I'll need such things in the future.)
Use BCX70j (NPN) and BCX71j(PNP) as replacement.I always use
them to replace the general 2N series in USA and 2S series of Japan
abd the BC54X BC55X series of Europa.
Gretings BJ in Sweden
> I have a problem building my Paia MIDI2CV8:
> I have only ordered the PCB without the other parts.
> Now I need 12 Transistors 2N4124 (on the schematic Q4...Q15) which I canīt
> find at the local suppliers. The only substitute that I found in my
> transistor list was the BC413 that I even wasnīt able to buy anywhere.
> Is there a possebility to replace them with others?
> Thanks for the Answer!
> Heiko

Use "BC547" npn instead, they should work most fine alltime.
this is a no critical point in the design for substitutes.


2N5089, BC550C are NPNs
2N5087, BC560C are PNPs

Joachim Verghese
>I was buying parts yesterday, and when I got home I realised that the boy gave
>me the wrong op-amps.  Before complaining, I decided to see what it was that
>he gave me, and I came upon this description:

>Now what I am wondering is if this chip can be used as a substitute to the
>LM13600, because I can buy this chip easily and it costs .79 each. (or .70 in
>quantities over 49).
>I haven't seen any mention of this chip on any of the web sites that have
>designs that use the LM13600 though.  Any info would be helpful!

From the Datasheet It looks as if it was not only a substitute but really the 
same as LM13700/LM13600. (There are even some drawings identical in both

I'd say it's a 2nd source for the 13700 (not the 13600).

Rene Schmitz

Juergen Haible
Yep. Use a TL072 and you'll never notice.

Why didn't Roland use them?

A) not an "Asian product"
B) cost 9 cents moe

Paul S.
Paul Schreiber
Gerry Dahl Wrote:
>What is a good Roland IC source? I need a BA662 VCA IC for my JX3P.
>Anyone have any leads on these?

Tony Allgood's TB3030 project calls for one of these chips, but it also has
provision for a substitute chip, the BA6110 which is easier to get.

To quote his schematics:
"The Original (TB303) Bassline used the Roland Part BA662. This is only
available from Roland and could be difficult to source. The BA6110 could be
fitted instead, but it is not pin compatible with the BA662. Because of
this, 2 spaces on the PCB have been probided."...

You can find it on page 2 of the TB3030 schematics (available on the URL
above). The part seems to have the identical function. Only the pinout is

You should be able to adapt the part. The simplest way I can think of is to
use the "Dead Bug" technique(aka the "Dark Side" of synth repair :).

For those not familiar, "Dead Bug" is where you Hot-Glue the chip to the PCB
upside-down (so the pins point away from the PCB). Then you use Wire-Wrap
wire to route the leads to the appropriate holes. For more info, Search for
"Dead Bug Prototyping" on Yahoo.

-Dan Gendreau
I have now worked out a simple mod to allow the BA6110 to be used in the TB3030 with a perfect emulation of the BA662. Many purchasers of the TB3030 semi kit have got BA662s anyway, but those of you with BA6110s may notice a slight increase in 'bite' with the little mod. I'll put all the details on the site soon, and update the TB3030 User Guide. Regards, Tony Allgood Penrith, Cumbria, England
Dan Gendreau
Tony Allgood
The CA3046 and LM3046 are identical. National Semiconductor no
longer makes the LM3046. So if you get one, you are just luck
is all. Many of the places (like Jameco and JDR) in the past
would send you either part when you ordered the LM3046 (or vice

If the NTE912 is a cross for the CA3046, you can use that as well
(I am not familiar  with the NTE line, so I can't tell you if it
is or not). 

The part you need to be careful of is the CA3086. It is the same
as the CA3046, except it is not tested to make sure that the
transistors meet the matching spec, so you would be taking a gamble
with this part.
Jim Patchell
If you decide to go other than Digikey, I have been getting the TI -
LT1013CPs from Newark within 2 days of order.  I have purchsed almost 100
recently.  The minimum is only $25.  That doesn't take long when buying
1013s.  This part is also made by Maxim.   The part number is  MXL1013 (CN8
suffix, or something very close to that).
Larry Hendry
Larry Hendry
Subject: Re: [sdiy] LT1013CP alternative??
Date: Sat, 21 Apr 2001 20:14:41 +0100

You could also use Burr-Brown's OPA2277.


Tony Allgood  Penrith, Cumbria, England
Tony Allgood
From: Scott Bernardi 
To: Czech Martin 
Subject: Re: AW: [sdiy] discrete SSM2018?

In chip design it is fairly easy to get good matching between components,
but much harder to control the absolute value of a component. For
example, IC resistor values are typically +/-20%, but with care you can
match two resistors to 0.1%.
Vos is caused by manufacturing mismatches.  High precision design
techniques can minimize these (when we did the OP-27 at PMI, we typically
had 0.25mV or less Vos even before trimming), but the LM13700 is _not_ a
precision design.  In general an OTA is difficult because the main bias
current varies.  The CA3280 is one of the higher precision OTAs.
Scott Bernardi