the Tail of a Long Tail (transistors)

[Analog_Fan]

Or you are holding the leads to the device with your fingers and are measuring the parallel resistance of the device and your body.

i lay the resistor down on the work area than hold both leads of the meter and press down the pins of the leads on the lead of the resistor.
no personal contact is made.

 

[abbey road d enfer]

But it seams to move exponential, it amplifies (the upper part) a lot at small control voltage but seems to slow down above 6/7 volt.

That's not what I would expect, and not what I see in my simulation.

i tried adjusting resistor values to have full amplification at 10V and 50% of that at 5V, but whatever i did, it wasn't possible. i did spend hours on it.
But a VCA is supposed to be exponential, i remembered.

No, this kind of VCA, which is basically a transconductance amp has a linear relationship between gain and control current.
Of course control voltages below 0.6V are excluded.

"IC = (Voltage-0.7)/8.2K "

didn't know that

Current drawn by Q4 is same as current injected into Q3.
Current into Q3 is control voltage - 0.6V

 

[Analog_Fan]

That's not what I would expect, and not what I see in my simulation.

No, this kind of VCA, which is basically a transconductance amp has a linear relationship between gain and control current.
Of course control voltages below 0.6V are excluded.

Current drawn by Q4 is same as current injected into Q3.
Current into Q3 is control voltage - 0.6V

thank you, i just got the book now ... I'm a tortoise reader.
: )
the second book of Jjm has twice the amount of pages and is about 30% cheaper, will get that too.

 

[ccaudle]

if the input is 10V the output is 10V.p.p.

If you write only of the current mirror you have drawn in the first post, and not of the entire circuit, then you really should speak of input voltage, but output current.
Since the base and collector of the left transistor are connected together, the base voltage is emitter voltage + the junction voltage. Often estimated as 0.7V, but the junction voltage is dependent on both current and temperature, so consider that only an estimate, not an exact value.
The rest of the left transistor circuit can be analyzed as a resistor voltage divider, dividing the control voltage between R28, R27, and R22, with the addition of a 0.7V voltage source in the middle of the string.

Once you have calculated the current through that string of resistors and the transistor junction, note that since the base of the second transistor is connected to the base of the first transistor, and the resistance in the emitter circuit is the same, the voltage across the base-emitter junction must be the same, and therefore the current through the emitter resistor of the second transistor must be the same as the current through the emitter resistor of the first transistor.
Since the emitter current and collector current are very close, different only by the current through the base, the collector current of the second transistor must be very close to the collector current of the first transistor as well.

The voltage at the collector of the second transistor is not well defined, usually the collector voltage is determined by the rest of the circuit connected to the collector.

There are a lot of details that make the current not match perfectly between the two transistors, which has resulted in circuit variations using additional transistors (bipolar or MOSFET) that reduce the inaccuracy. Note also that the assumption is that the two transistors match characteristics almost exactly, and stay at the same temperature. That is why paired devices in the same package are often used for current mirrors (as well as differential pairs).

This web page has a good discussion of current mirrors, although most circuit textbooks will cover it as well:
Current mirror page at Electronics wiki at Analog Devices

 

[ccaudle]

check out this interesting video from Fran

That is neat, Jim Early presents starting at around 14 minutes into the video. He was one of the early pioneers in understanding transistor behavior and developing new types of transistors.
Jim Early bio at Wikipedia

 

[Newmarket]

mine was 25€ years ago.
good enough to do most things, accept precision.
that price came fro the local mayor electrics parts supplier where electricians and so buy there stuff.
You have to pay extra for the leads with alligator clips, i think it's 25€ per lead and why its sums to 350€

But maybe i could check if the leads are bad.

You don't need 25 Euro leads to measure a resistor. I'm guessing they insulated for HV or something. A DMM lead isn't complicated like 'scope probes can be.
And you don't need a DMM that an electrician would use to qualify electrical installations etc. to check resistors.
Plus - thru hole parts will have a colour code to read or the value explicitly stated (eg W21 type power resistors).
And SMD resistors will have a code indicating the value.
I don't want to be critical but you do seem to be making a problem out of the simplest of things. And whilst apparently having loads of transistors that you don't understand ?

 

[Analog_Fan]

You don't need 25 Euro leads to measure a resistor.

no problems, just ease and work handsfree.
the same with my scope probe, grab something and see it move.
now i only have the lower left side, these pins on my old 25€ multi meter.

... and 3 digit after the dot instead of 2.

 

[neil.johnson]

But a VCA is supposed to be exponential, i remembered.

Project 213

Did you actually read the webpage you scraped that circuit from? That circuit is roughly the same as a discrete version of the LM13700 family of linear VCAs. Don't expect great audio quality, as the webpage says that project is more for education than hifi quality.

If you want an exponential VCA (you probably don't if you're just learning) then you would be looking at the SSI and THAT range of parts (e.g., SSI2164, THAT2180)

Neil

 

[Analog_Fan]

Don't expect great audio quality

that's the idea, have some proper character.

exponential sound somewhat better.
(just don't understand the Roland that portion of Circuit: http://ua726.co.uk/wp-content/uploads/2017/02/130VCA-schematic.pdf)

i got a few VCA's, using LM13700 or CA3080.
It's only processes sine, saw, triangle, pulse waves or mixed combos of them, not really (mp3) audio.


https://frequencycentral.co.uk/product/dual-xvca/
I also made a stripped down Less than half the width version of the above but still dual, but using the original parts Roland used, like a 2SK30A.

Now i want a Long tail in the collection.



this is the is the LM13700 clone, routed a couple of month ago and probably going to be send in the next couple of days.

 

[JohnRoberts]

OTAs operational transconductance amplifiers.

The gain varies linearly with bias current.

JR

 

[Analog_Fan]

OTAs operational transconductance amplifiers.

The gain varies linearly with bias current.

JR

https://mutable-instruments.net/modules/ripples/"Fully analog, based on the V2164 low noise/distortion quad VCA chip, and LM13700 OTA for that extra bit of non-linearity."


Q1 maybe the source of that.

 

[neil.johnson]

https://mutable-instruments.net/modules/ripples/"Fully analog, based on the V2164 low noise/distortion quad VCA chip, and LM13700 OTA for that extra bit of non-linearity."

View attachment 94028
Q1 maybe the source of that.

Ummm......no.

Q1 is not in the audio path. Pretty much everything else is.

Neil

 

[Analog_Fan]

i know, but the could be why the LM17300 reacts somewhat non linear.

 

[neil.johnson]

exponential sound somewhat better.
(just don't understand the Roland that portion of Circuit: http://ua726.co.uk/wp-content/uploads/2017/02/130VCA-schematic.pdf)

That Roland module uses the hybrid BA662 VCA.

I think you would find a lot more useful commentary in the DIY forum on ModWiggler, or maybe Electro Music Forum.

Neil

 

[Analog_Fan]

Ummm......no.

Q1 is not in the audio path. Pretty much everything else is.

Neil

i know, but the could be why the LM17300 reacts somewhat non linear.

 

[Analog_Fan]

That Roland module uses the hybrid BA662 VCA.

I think you would find a lot more useful commentary in the DIY forum on ModWiggler, or maybe Electro Music Forum.

Neil

yes, i got more of them the average person has of them and BA6110.

 

[user 37518]

i know, but the could be why the LM17300 reacts somewhat non linear.

The OTA does not use global negative feedback, thus it operates best in a small signal regime (<VT), that is why it is not very linear, its not because of Q1, its the whole thing... particularly the differential pair. The input diodes help reduce distortion.

 

[Analog_Fan]

The OTA does not use global negative feedback, thus it operates best with a small signal voltage, that is why it is not very linear, its not just because of Q1, its the whole thing... particularly the differential pair.

i always use 100K and 560R, that's what Roland used, but most use 1K and some cases 1K5 or 220R in Case of the Korg ms-20 filter.
Korg late MS20 filter

Do you know how to use the formulas of page 17 of the LM13700, this Gm value.
the 2 pole high pass filter i made is awesome, but but doesn't go as low as 150/200Hz, but rather somewhere in the middle of 1kHz/2kHz.
using the same value as the Korg, 1nF.
but the again, i don't know how to extract the current from the control circuit (Korg ms-20 filter),

 

[JohnRoberts]

The OTA does not use global negative feedback, thus it operates best in a small signal regime (<VT), that is why it is not very linear, its not because of Q1, its the whole thing... particularly the differential pair. The input diodes help reduce distortion.

+1 OTA are not a low distortion solution. Indeed the linearity weak link is the input LTP operating open loop and there is no magic solution besides keeping the audio voltage at the input very small. Perhaps extremely low noise voltage bipolar devices in the LTP might support lower input voltages with decent S/N.

For today's [TMI] about OTA, and my apologies to regular readers because I have told this story before. Back last century while working for Peavey, one of my special projects was to investigate tooling up a dedicated full custom IC to replace the heart of our millions of general purpose amp channels.. This was a pretty simple driven rail audio power amp circuit. The basic circuit was probably ball park $1.00 and change for the glue circuitry parts; one opamp, one OTA (for clip limiting), and some small discrete components. A full custom IC is hundreds of $k NRE (non recurring engineering cost). I was made the point man and met with the junior engineer the IC company sent to MS. I tried to explain to him the significance of the design specifications. The OTA was used in the popular Peavey DDT clip limiter. In this application it is important that the OTA have low input offset voltage to minimize control voltage feed through. While I tried to explain to the junior engineer how important the LTP device match is for OTAs, he told me not to worry, they could just add degeneration resistors in series with the emitters. Apparently he had zero clue about how OTAs actually work.. No surprise the project never happened. [/TMI]

JR

 

[user 37518]

+1 OTA are not a low distortion solution. Indeed the linearity weak link is the input LTP operating open loop and there is no magic solution besides keeping the audio voltage at the input very small. Perhaps extremely low noise voltage bipolar devices in the LTP might support lower input voltages with decent S/N.

For today's [TMI] about OTA, and my apologies to regular readers because I have told this story before. Back last century while working for Peavey, one of my special projects was to investigate tooling up a dedicated full custom IC to replace the heart of our millions of general purpose amp channels.. This was a pretty simple driven rail audio power amp circuit. The basic circuit was probably ball park $1.00 and change for the glue circuitry parts; one opamp, one OTA (for clip limiting), and some small discrete components. A full custom IC is hundreds of $k NRE (non recurring engineering cost). I was made the point man and met with the junior engineer the IC company sent to MS. I tried to explain to him the significance of the design specifications. The OTA was used in the popular Peavey DDT clip limiter. In this application it is important that the OTA have low input offset voltage to minimize control voltage feed through. While I tried to explain to the junior engineer how important the LTP device match is for OTAs, he told me not to worry, they could just add degeneration resistors in series with the emitters. Apparently he had zero clue about how OTAs actually work.. No surprise the project never happened. [/TMI]

JR

John, Peavey stories are always welcome, you can drop the TMI brackets