MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Testing OTAs (CA3280s specifically)
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page 1, 2  Next [all]
Author Testing OTAs (CA3280s specifically)
devinw1
So I've ordered some CA3280s from a few ebay suppliers with decent feedback, but I know it can be a minefield. Unfortunately, as many of you know, sourcing these OTAs is becoming more and more difficult. You can buy them from Rochester Electronics if you want to pay $23 a pop and $250 minimum order. Ouch.

Anyway, I'm wondering if anyone has a simple test circuit handy that I can make to verify if these chips are good or DOA before I use them? Any help is much appreciated!
cygmu
Here’s what I’ve done in the past. Don’t know if it’s foolproof.

From the datasheet:
https://www.intersil.com/content/dam/intersil/documents/ca32/ca3280-a. pdf
I use the Figure 6 circuit, replace the 68k resistor with a 33k and a pot, ground one input, put a signal in the other one and make sure it comes out the other end and varies in amplitude as I twist the pot.

From about 15 ICs I found one that didn’t work in this configuration and got a refund on it from the eBay seller.
devinw1
cygmu wrote:
Here’s what I’ve done in the past. Don’t know if it’s foolproof.

From the datasheet:
https://www.intersil.com/content/dam/intersil/documents/ca32/ca3280-a. pdf
I use the Figure 6 circuit, replace the 68k resistor with a 33k and a pot, ground one input, put a signal in the other one and make sure it comes out the other end and varies in amplitude as I twist the pot.

From about 15 ICs I found one that didn’t work in this configuration and got a refund on it from the eBay seller.


Awesome, that is super helpful!

What do you use for the differential signal? (i have a feeling you're going to tell me I need a function generator.. grin)
cygmu
I just grounded one input and sent a single-ended signal through the other, from an oscillator.
devinw1
Nice. Thanks man. I will do this!

I have 10 of these chips coming from 1 vendor, and 5 from another. Both Chinese. We shall see what I get!
loki
This may be of interest to you.
cygmu
That's an interesting circuit.

I guess you are testing for more than just "does this thing work at all?" which is the main question with an eBay 3280 these days. Are you testing the ability to null the offset?

I would also like to understand the role of the FETs in the current sources. I don't think I've seen that configuration before and am keen to learn!
devinw1
loki wrote:
This may be of interest to you.


Oh wow that's great! Thank you!

And cygmu, isn't that JFET just a buffer?
cygmu
devinw1 wrote:

And cygmu, isn't that JFET just a buffer?


I’m not sure. It looks more like a resistor to me. I wonder if it’s there to limit the current or something.
guest
i think it helps stiffen up those current sources. in particular for the fixed current source driving the integration capacitor on the oscillator. since Vgs doesnt change much as Id isnt changing, it holds Vce fixed on the current source output transistor. this greatly reduces Ic variations that occur when Vce changes, and therefore makes a better current source. with the variable current source driving the OTA, the variation in Vgs is much smaller than the variation in Vce would have been, so it still helps.
Reese P. Dubin
I always just swap them into one of my Fritz Threeler VCFs. If it works they are good.

Probably not so helpful
devinw1
Reese P. Dubin wrote:
I always just swap them into one of my Fritz Threeler VCFs. If it works they are good.

Probably not so helpful


Hey, i mean it works as a "go/no go" test if you happen to have one of those!
fuzzbass
devinw1 wrote:
Reese P. Dubin wrote:
I always just swap them into one of my Fritz Threeler VCFs. If it works they are good.

Probably not so helpful


Hey, i mean it works as a "go/no go" test if you happen to have one of those!


I got a bit hung up on this before. I realized I had a Fritz Threeler filter using two CA2380. I don't know if swapping them into that filter is a thorough test, but it would certainly ID fakes.
loki
The change in Ic as the collector - base voltage changes is caused by the Early effect. The current sources are cascoded with FETs to minimize this. There is a nice description of the Early effect on Wikipedia.

The behavior of the CA3080 as the bias current is swept can be quite varied. I spent 3 weeks in 1975 studying what happens. I had about 50 or 60 3080s. The best of them would give an almost straight line when the offset is nulled. Some would give a parabola and some a polynomial. This is caused by the mismatch in the current mirrors in the part. The PNP current mirrors are made with lateral PNPs and their characteristics suck. Download the datasheet and look at the internal schematic.

The early synthesizer manufacturers who used 3080s all knew that they had to test them. When I went to work for 360 Systems in 1977 Bob Easton told me about a batch of 10,000 parts that had circulated among all the manufacturers on the west coast and that the yield of usable parts was zero. These were parts that passed the factory tests and met the specs in the datasheet.
indigoid
loki wrote:
The early synthesizer manufacturers who used 3080s all knew that they had to test them. When I went to work for 360 Systems in 1977 Bob Easton told me about a batch of 10,000 parts that had circulated among all the manufacturers on the west coast and that the yield of usable parts was zero. These were parts that passed the factory tests and met the specs in the datasheet.


this variability should then also be present in LM13700, yes?
oldenjon
Just wait for the AS3280 to be released?
cygmu
loki wrote:
The change in Ic as the collector - base voltage changes is caused by the Early effect. The current sources are cascoded with FETs to minimize this. There is a nice description of the Early effect on Wikipedia.

The behavior of the CA3080 as the bias current is swept can be quite varied. I spent 3 weeks in 1975 studying what happens. I had about 50 or 60 3080s. The best of them would give an almost straight line when the offset is nulled. Some would give a parabola and some a polynomial. This is caused by the mismatch in the current mirrors in the part. The PNP current mirrors are made with lateral PNPs and their characteristics suck. Download the datasheet and look at the internal schematic.


Thanks -- that is really helpful and interesting.
devinw1
loki wrote:

The early synthesizer manufacturers who used 3080s all knew that they had to test them. When I went to work for 360 Systems in 1977 Bob Easton told me about a batch of 10,000 parts that had circulated among all the manufacturers on the west coast and that the yield of usable parts was zero. These were parts that passed the factory tests and met the specs in the datasheet.


What was the application you are referring to?

For example, the Odyssey uses the 3080 for the VCA. I've personally swapped in several different batches of TO can versions and DIP versions and they all worked fine as a VCA. I realize that's not a huge sample set, but I didn't know the 3080s were so spotty. Or do you mean for more demanding applications ?
devinw1
oldenjon wrote:
Just wait for the AS3280 to be released?


Didn't realize this was happening... Good news for OTA fans I suppose. Seems to be very limited info on the web about it, but it seems it is happening. Sweet!
devinw1
Update: I have breadboarded a nice little test setup which is as follows:



When used with a function generator you get something like this (yellow is output, blue is input). I have a pot to adjust the voltage on the op-amp circuit so you can vary the gain. (Triangle wave is sorta clipped because it's just at maxing out on 9V battery signal amplitude):



Using this test setup, I am as follows for success:

-Ebay seller "Consolodated Electronics" (Dayton, OH): Bought 5 chips, all work perfect
-Synthcube.com: Bought 2 chips, 1 works perfect, 1 is dead! Dead one just has a -DC voltage on the output
-Ebay seller "Joytrin Electronics" (China): Bought 4 chips, all fakes. Has a "power" pin shorted to ground. Emits magic smoke.
-Ebay seller "cole-0520" (China): Bought 4 chips, all fakes, Output is DC.


I am going to make a PCB with a ZIF socket and a switch to test both halves of the OTA and get it made soon. If anyone is interested in something like this, let me know. I hope this is helpful for some!
dougt
devinw1 wrote:

-Ebay seller "Joytrin Electronics" (China): Bought 4 chips, all fakes. Has a "power" pin shorted to ground. Emits magic smoke.
-Ebay seller "cole-0520" (China): Bought 4 chips, all fakes, Output is DC.


See the clue there?
widdly
I've got a bunch I found in a surplus electronics store in Singapore.

I test them in this circuit..

http://www.noniandjim.com/Jim/archives/ca3280_12.html

If someone desperately needs some, I don't mind trading for interesting PCB's. PM me.
devinw1
dougt wrote:
devinw1 wrote:

-Ebay seller "Joytrin Electronics" (China): Bought 4 chips, all fakes. Has a "power" pin shorted to ground. Emits magic smoke.
-Ebay seller "cole-0520" (China): Bought 4 chips, all fakes, Output is DC.


See the clue there?


Haha yeah they seem to have the monopoly on faking chips! I think as a general rule if an IC or other item is fairly desirable, and no longer in production, and stocks are dwindling (this of course being the case for our beloved CA3280s), be VERY wary of buying these items from China.
alfa
Just to note:

AS3280 ( CA3280) in production:
http://www.alfarzpp.lv/eng/sc/application.php
Datasheet:
http://www.alfarzpp.lv/eng/sc/AS3280.php
Evaluation samples demonstration:
https://www.youtube.com/watch?v=B6qTiKN1Rn4
Borogove
Hey, alfa, the link to the 3394 datasheet is broken.
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page 1, 2  Next [all]
Page 1 of 2
Powered by phpBB © phpBB Group