How to substitute 2 CA3904 with 1 LM13700?

All frequent questions about ICs of all kinds, will be stored here.
Post Reply
User avatar
Duckman
Opamp Operator
Information
Posts: 1492
Joined: 20 May 2009, 01:45
Has thanked: 320 times
Been thanked: 134 times

Post by Duckman »

I read about and I think it's an interesting option, saving space and some money, but my knowledge is not enough to understand wich pin is equivalent between both IC's.

Anybody alredy do it or, at least, can help me??
Tanks in advance!!

Datasheets down here
Attachments
LM13700N.pdf
(635.22 KiB) Downloaded 209 times
CA 3094.pdf
(244.19 KiB) Downloaded 225 times

User avatar
frequencycentral
Cap Cooler
Information
Posts: 696
Joined: 30 Mar 2009, 09:37
Location: United Kingdom
Has thanked: 28 times
Been thanked: 61 times

Post by frequencycentral »

1 = 5
2 = n/c
3 = 2
4 = 3
5 = 1
6 = 4, 4
7 = 8
8 = 6
9 = 6
10 = 8
11 = 7, 7
12 = 1
13 = 3
14 = 2
15 = n/c
16 = 5

I think the Iabc input of the LM 13700 is two diode drops, whereas the CA3094 is only one. Your circuit may need some tweaking. :wink:

User avatar
Duckman
Opamp Operator
Information
Posts: 1492
Joined: 20 May 2009, 01:45
Has thanked: 320 times
Been thanked: 134 times

Post by Duckman »

Great! Thanks!

Now :oops: here we go with my noob questions:
Your circuit may need some tweaking.
:?: You think there is a better option than the LM13700?
I think the Iabc input of the LM 13700 is two diode drops, whereas the CA3094 is only one.
:?: How much affect two diodes drop vs. one diode drop? Higher drop than usual?
:?: Can I just put a single diode in front of pin 1 and pin 16 to fix it or we are talking about something a little more "elaborated"?

Hope not to boring you and big thanks for help me to learn :hug:

User avatar
uncleboko
Cap Cooler
Information
Posts: 509
Joined: 25 Mar 2008, 20:19
Has thanked: 58 times
Been thanked: 41 times

Post by uncleboko »

Not worth the hassle - only leads to divorce - I promise you :roll: , start again with 13600/13700.

User avatar
frequencycentral
Cap Cooler
Information
Posts: 696
Joined: 30 Mar 2009, 09:37
Location: United Kingdom
Has thanked: 28 times
Been thanked: 61 times

Post by frequencycentral »

As CA3094 are rare as camel's wings these days, LM13700 is your best option. (I have a stash of metal can CA3094 and CA3080, but I won't part with them sorry). As for the difference in the Iabc inputs, just be aware that you may need to tweak the circuit slightly around the Iabc inputs (pins 1 and 16) to get close to the same response that a CA3094 would have. I've never subbed like this myself, so I'd be interested to hear how it goes. What are you building? Oh...........breadboard it first!

User avatar
Duckman
Opamp Operator
Information
Posts: 1492
Joined: 20 May 2009, 01:45
Has thanked: 320 times
Been thanked: 134 times

Post by Duckman »

Ok. We are a little closer now
All start with this

https://www.freestompboxes.org/viewtopic ... microsynth

As you can imagine I even try to breadboard such a "Titanic" yet (I need a bigger breadboard first!!) but I'll give a try if I find an economic solution. Small Stone is in my list too... you see... a lot of 3094´s in my future!!
You still find 3094's and 3080's at Small Bear and here in Mexico you can find those 3094 metal can but $15 (we're talking dollars, not mexican pesos) per piece??? Obscene!!
I need something less expensive or my wife will bring me back to Italy (in a very painfull way, capicci?)
So, if you help me with the "tweaking affaire" I´ll be more than grateful with you. :hug:

User avatar
Bbcxs
Information
Posts: 4
Joined: 28 Jan 2010, 23:14

Post by Bbcxs »

hi, I'm thinking about doing the same thing and this is what bothers me:

shouldn't pin 3 of lm13700 be equal to pin3 of 3094 (positive voltage inputs) and pin4 of 13700 equal to pin 2 of 3094(negative inputs) :scratch:

or are they inverted for a reason?

thx

User avatar
Bbcxs
Information
Posts: 4
Joined: 28 Jan 2010, 23:14

Post by Bbcxs »

oh, sorry, this refers to eh microsynth schematics :wink:

User avatar
chicago_mike
Tube Twister
Information
Posts: 1754
Joined: 03 Jul 2008, 04:38
Has thanked: 22 times
Been thanked: 316 times

Post by chicago_mike »

CA3280!

I have a stash of these. Its a dual 3080 thats tweaked a little too. Awesome OTA! 8)
Skyline FX 2013

User avatar
Liquids
Breadboard Brother
Information
Posts: 143
Joined: 24 Apr 2009, 19:29
my favorite amplifier: Ampeg V4!!!!
Location: USA
Has thanked: 20 times
Been thanked: 27 times

Post by Liquids »

I've been looking this dilemma over a lot recently. With a little understanding of OTAs, really, only one of seven CA3094/OTAs in the microsynth is doing anything particular to that part. The rest are all acting like standard voltage-controlled (variable) resistors, more or less. Any other OTA can be implemented to do the same, and I would dare I say NO tweaking needs to be done in the majority of cases.

My understanding:
The CA3080 type OTA is basic/primitive single OTA (think 741 op amp) on a 8 pin package. It has not buffering, which comes in handy seeing as an OTA more or less acts as a variable (and sometimes high value) resistor, which can load following circuitry.

The CA3094 type OTAs are slightly less primitive 8 pin OTA which utilized extra pins for a basic transistor circuitry, most often used as an output buffer. The OTA portion CAN be used without the benefit of its 'buffer' or additional when preferred, but rarely is...but the option is there.

The CA3280 are 'the best of the OTAs' according to Don Tillman and spec, a dual chip with OTAs that meet better specs than the CA3094 and buffer, and possibly some additional capabilities. A lot of that is beyond me, and I suspect, beyond our uses (and the microsynths).

The LM13600 (LM13700) chips, to my understanding, are more or less a chip that houses 2 OTAs - which are fairly primitive - and similar to the 3080 OTA. The 3080 did not have output buffering capabilities, but the LM13600 adds simple darlington transistors onboard, or less one for each OTA, a slightly different (simplified) configuration than the added circuitry the 3094 has, but in most cases and for most of our purposes, functionally equivalent. Each OTA has an extra pin for a 'dual diode' drop functionality or whatever; if you ignore that pin and leave it unconnected, again, it's more or less the same as a CA3080/CA3094 OTA.

I say that at least from an audible perspective - So, after a bunch of simulation, that this weekend I spend time making the Microsynth filter a reality. I have a Microsynth I got used in a store trade for some gear I didn’t want. I’ve enjoyed it, but there are some things about it I've always wanted to tweak, the bypass scheme makes it useless. The internal gain boost/buffer is necessary for good tracking, but when ‘bypassed’ the gain boost, and hence the clean signal reaches self-distortion at that point, and too hot a signal for all gear following it. Likewise, it’s just too big to leave on the board. I'm not going to screw around with modding mine, and I will not even 'true bypass' it (since I plan to sell it and it’s unusable on my pedalboard), even though a 3PDT switch sits in there, stock, but with only one pole in use if I'm not mistaken! Moreso I've always wanted to build/imitate the triggered filter for a long time. I did it this weekend, and I did it with two LM13X00....

How I see it, the filter itself is 'nothing special' in terms of needing CA3094s. It's a standard OTA filter, it uses 3 OTA sections, and each section utilizes output buffering. Plenty of great filters use non-CA3094 OTAs in the synth world; the CA3094 has nothing ‘special’ for filters. Most if not all of Ray Wilsons filter and other noteworthy filter designs use 13X00 chips - they're not missing any filtering mojo.

The hard part here is the triggering circuitry which uses a CA3094 OTA to 'drive' the filter sweep. For this part, I realized I can simply used the OTA onboard the LM13X00 chip, added the two transistors & two resistors that you can see here, that function as pins 6 and 8 of the CA3094 circuit. Note that this is just a snippet from the CA3094 datasheet circuit itself:

http://forums.peavey.com/download/file. ... 4e2201fd24

If you connect the LM13X00 OTA to a discrete version of this two resistor and transistor circuit, properly as per the schematic, the filter really does work the same.

The schematic for the Microsynth is out there - there is a hand drawn one that is different, but I used the one that is more easily out there drawn by Fabian Hartley. There are one or two or three (critical?) resistor value typos in there, but otherwise it's fantastic, and the corrections for the errors are documented here on this very forum.

I mostly just wanted to say for the sake of other's trying it themselves, that the Microsynth it really can be done with 13X000! I realize the XO series redesign uses them as well, but that one seems to be also made to work with 9v as well, with at least some component value changes to follow suit. That would be cool too, but we don't have that schematic yet, and the schematic most widely available is with classic values, and bipolar power. Do note that your voltage (+/- 9v, 12v or 15v etc) will likely affect the sound of the filter, but it can be compensated as well. I simply used a ~9v power supply and a LT1054 to generate a ~9v rail which functioned, even if not ideal.

Anyhow, I did not build the octave, VCA, or Square wave portion of the circuit, but those do not utilize pins 8 or 6 of the CA3094, so really any OTA will do and sound just perfect, I suspect! I also can think of cooler circuits for generation and VCA, which I will likely house separately anyhow. A schematic of the circuit needed for the filter (only) is to follow in the relatively near future, when possible, for those of you who prefer to breadboard it with confirmed chip pin numbers rather than an explanatory write up… :)

User avatar
Liquids
Breadboard Brother
Information
Posts: 143
Joined: 24 Apr 2009, 19:29
my favorite amplifier: Ampeg V4!!!!
Location: USA
Has thanked: 20 times
Been thanked: 27 times

Post by Liquids »

Here's my schematic with the LM13X00 pinouts. If it looks wonky somewhere (undoubtedly it's got errors), just follow the standard schematic along with the 'corrections' noted by sergedeep on page one of this thread: viewtopic.php?f=1&t=304 It's meant to be the same circuit, but using the available chips and with some additional options you may or may not find useful.


Image

Have fun!

User avatar
Nocentelli
Tube Twister
Information
Posts: 2222
Joined: 09 Apr 2009, 07:06
Location: Leeds, UK
Has thanked: 1152 times
Been thanked: 954 times

Post by Nocentelli »

Awesome job! :hug:
modman wrote: Let's hope it's not a hit, because soldering up the same pedal everyday, is a sad life. It's that same ole devilish double bind again...

User avatar
uncleboko
Cap Cooler
Information
Posts: 509
Joined: 25 Mar 2008, 20:19
Has thanked: 58 times
Been thanked: 41 times

Post by uncleboko »

If you move the filter section down a little and the sweep generator to the left into the space, this excellent drawing prints very nicely in landscape.

User avatar
uncleboko
Cap Cooler
Information
Posts: 509
Joined: 25 Mar 2008, 20:19
Has thanked: 58 times
Been thanked: 41 times

Post by uncleboko »

Very odd - this doesn't show in the active topics??

Ah, it does now :roll:

User avatar
Liquids
Breadboard Brother
Information
Posts: 143
Joined: 24 Apr 2009, 19:29
my favorite amplifier: Ampeg V4!!!!
Location: USA
Has thanked: 20 times
Been thanked: 27 times

Post by Liquids »

uncleboko wrote:If you move the filter section down a little and the sweep generator to the left into the space, this excellent drawing prints very nicely in landscape.
I prettied it up for you (and fixed a few incorrect connections), but it doesn't seem to update the image here...hmm.

Here's a link...http://www.aronnelson.com/gallery/main. ... emId=45290

Post Reply