Harrison 4032 restoration?!

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
nrgrecording said:
Yes... I read it several times.

There is an info about paralleling vcas and that there is an 6dB Improvement with 4 paralled VCAs.
They used 10k resistors in series with the signal input to stop "thumbs" and offsets when paralleling the inputs without 10k resistors. It feels that the "31K6" resistor (in the console circuit) at the imput needs to be changed when using more paralleld vcas because of the changed (internal) input resistance. (like shown on page 4) in the 'design note 127'. But I don't know the input resistance of the 2180 vca?

In the THAT2002 the used 4 THAT2181 in parallel. Input series resistor is 1,4k.
So the 31,6k reistor needs to be replaced with a 30,2k resistor.  (31,6k (existing input resistor) - 1,4k (internal input resistor) = 30,2k (resulting new input resistor)
Instead of 10k, the 4 VCAs will need four 5,6k input series resistors.
The recommended circuit "20002N" has a 3,57k resistor at the input and 4k99k resistor at the feedback of the following opamp.

I would still like to know the input resistance of the 2180. It feels like I have to add different resistors in series to the input when using 1,2,3 or 4 vcas in parallel?

The dn127 pdf is more or less made as an info for replacing the dbx202 with a that2002, isn't it?

Frank.
Just guessing, but the 2180 seems to have no built in additional input resistance. (2180 datasheet pg.4 says '...Q2/D2 and Q4/D4 take the bipolar anitlog of V3, creating an output current which is a precise replica of the input current...'
For paralleling 2,3,4,... VCAs you'd have to share the signal current going to each VCA by an additional resistor to each pin1. This paralleled value 1/(n VCAs/5K6) would have to be substracted from your common 31K6 input resistor to archive the current in=current out condition when Ec+ and Ec- are held at 0V. You further might need to scale this control voltage down for the 2180s 6.1mV/dB gain law. The consoles previous VCA might have a 50mV/dB gain law that could exceed the 2180s limit of max.Ec +/-1V. 
 
some good links to improve vca performance in case you don't know them already.

paralleling vcas
differential drive

and keep in mind:    dbx 202 "black can" is 6mv/db      dbx 202 "gold can" is 50mv/db 

regarding your replacement schematic...  I'd remove the 10k resistor from pin3 to pin5 and the 3k9 resistor at pin5 should be 5k1 for the 218x series

http://www.thatcorp.com/datashts/dn137.pdf

 
Thanks guys for your suggestions!  :)

Ok, I used the 2002VCA circuit now:  http://www.nrgrecording.de/temporary/vca2.pdf

Still some questions:
1. For 2180 series VCAs, is the symmetry pin internally (in the 2180vca) connected? (if yes, I may have to make 4 solderpads or something like this to cut the connection between the 4 vca's) ?

2. I would like to add a buffer at the control voltage input. With gain = 0 its -6,1mV/dB. Is there any online calculator or simple way to calculate the resistor combination for E24, E48 or whatever resistor ranges? Gain factor is "8,2" from 50mV to 6,1mV i think.
One 8,45k (8,49k) and 3,3k feedback resistor maybe?

Thanks,
Frank.
 
nrgrecording said:
One 8,45k (8,49k) and 3,3k feedback resistor maybe?

Noise on the control input(s) is a bad idea performance-wise; I'd pick those impedances an order of magnitude lower (if not more).

JDB.
[the folks at THAT are pretty explicit about this, to the point of suggesting an AD797 for the control port buffer]
 
Don't you want to drive Ec+ instead of Ec- with an inverting buffer at the control port?
Maybe a low offset non-inverting buffer with a simple voltage divider in front for Ec-.
 
Harpo said:
Don't you want to drive Ec+ instead of Ec- with an inverting buffer at the control port?
Maybe a low offset non-inverting buffer with a simple voltage divider in front for Ec-.

Good point, Thanks Harpo.  :D
I added a voltage divider infront of the buffer.
For flexibility I added:
Bridges for changing the control port ratio from 50mV to 6mV
Bridges for using pretrimmed or trimmable vca's
Bridges for choosing between driving the positive control port or the negative control port

Here it is:
http://www.nrgrecording.de/temporary/vca3.pdf

Hope this makes sense now.

I don't know... which opamp is cool and has a low offset voltage? NE5532 has 0,5V, 4570 or TL072 has 0,3V.Any "cheaper than 1€ opamp is welcome)  ::)

Thanks,
Frank.
 
nrgrecording said:
I added a voltage divider infront of the buffer.
...if your 715R would connect to the non-inverting input. Can your source drive this 815R load (715R series + 100R shunt) ?

I don't know... which opamp is cool and has a low offset voltage? NE5532 has 0,5V, 4570 or TL072 has 0,3V.Any "cheaper than 1€ opamp is welcome)  ::)
TL07xCP < 10mV (typ. 3mV); NE5532 < 4mV (typ. 0.5mV); same offset specs as NE5534 but offset might be adjusted within limits/temperature and needs compensation cap.
TL051CP < 1.5mV (typ. 0.59mV) to meet your < 1€, @Reichelt.de 0.98€ for single op amp, 1.05€ for the dual TL052CP.
The TL051ACP is < 800uV (typ.0.35mV) if it matters. What is your supply voltage to feed them.
 
Thanks Harpo and jdbakker!

...if your 715R would connect to the non-inverting input.
Now it is.

Can your source drive this 815R load (715R series + 100R shunt) ?

Good question. The full circuit is here... http://www.nrgrecording.de/temporary/VCA_circuit.pdf

Control port is driven by a Harris 4605 quad opamp. Can't find any datasheet. I only know its more ore less a quad 741 - pin compatible with TL074.

Mabe its better to use a 600ohm shunt and a 4,4k series resistor?

http://www.nrgrecording.de/temporary/vca4.pdf

Frank.
 
nrgrecording said:

Hadn't spotted that earlier, but as drawn the CV op-amp will do nothing (more precisely: output is ~0V regardless of input). Either make the shunt resistor a feedback resistor (between - and out), or connect the resistor divider to your non-inverting input instead.

nrgrecording said:
Can your source drive this 815R load (715R series + 100R shunt) ?

The full circuit is here... http://www.nrgrecording.de/temporary/VCA_circuit.pdf

Control port is driven by a Harris 4605 quad opamp. Can't find any datasheet. I only know its more ore less a quad 741 - pin compatible with TL074.

R414 (in VCA_circuit.pdf) limits drive current. You need to take this resistor into account when calculating the voltage divider, of course. I don't remember the control port input resistance of a 202, you may want to match that.

JDB.
 
The 2180 vca datasheet says:

The 2180 Series VCAs are designed to be
operated with zero source impedance at pins 2 and
3, and an infinite source impedance at pin 4. (pin 4
should be left open.) To realize all the performance
designed into a 2180, keep the source impedance of
the control voltage driver well under 50 Ω.
This often suggests driving the control port
directly with an opamp. However, the closed-loop
output impedance of an opamp typically rises at high
frequencies because open loop gain falls off as
frequency increases. A typical opamp's output impedance
is therefore inductive at high frequencies.
Excessive inductance in the control port source
impedance can cause the VCA to oscillate internally.
In such cases, a 100 Ω resistor in series with a
1.5 nF capacitor from the control port to ground will
usually suffice to prevent the instability.
Noise Considerations

So what about removing the 1,43k resistor (shown in series with the vca circuit) and removing the buffer opamp completely. Then adding the (715R series, 100R to ground) and a 1,5nf capacitor in series with the 100R resistor to ground? Like this:

vca_input.jpg


(saves me one opamp)
 
nrgrecording said:
The 2180 vca datasheet says:

The 2180 Series VCAs are designed to be
operated with zero source impedance at pins 2 and
3, and an infinite source impedance at pin 4. (pin 4
should be left open.) To realize all the performance
designed into a 2180, keep the source impedance of
the control voltage driver well under 50 Ω.

This often suggests driving the control port
directly with an opamp.

Will the (Thevenin) equivalent source impedance of your voltage divider satisfy this requirement?

JDB.
 
jdbakker said:
Will the (Thevenin) equivalent source impedance of your voltage divider satisfy this requirement
(haven't calculated the Thevenin resistance... but as a rule of thumb I thought the opamp will "see" the last resistor to ground?)

Maybe if the resistor to ground is 51ohm?  ??? (plus using a 374R resistor in series then) [recalculated voltage divider]

VCA_CV1.jpg


I don't know... there is already an opamp which should drive the CV port hopefully. Using the circuit as shown above sounds easier - the circuit here looks saver?  :-\
VCA_CV2.jpg
 
nrgrecording said:
Maybe if the resistor to ground is 51ohm?  ??? (plus using a 374R resistor in series then) [recalculated voltage divider]

Even if you short out the 1n5 capacitor the equivalent resistance will only be just below 50R. Personally I interpret well under 50 Ω as "no more than 10Ω in the audio band".

If you want optimal performance I see no way to do it without an op-amp buffer.

nrgrecording said:

As drawn the op-amp output will be clipping to either the positive or negative rail, depending on the polarity of the op-amp offset voltage, regardless of input voltage.

Remove the connection between the output and the positive input.
Connect the negative input to the output, if you want a standard unity-gain buffer.
Keep the original 1.43K source resistor, lose the 374R series resistor (short it with JP3), replace the shunt resistor between op-amp + input and ground with a 200R one. That should get you close enough to the intended CV response, without too much extra noise. This is fine for a 5532; lower-noise op-amps might benefit from a lower-R voltage divider.

If you worry about offset, replace the 5532 with a 5534 and a trimmer.

JDB.
 
nrgrecording said:
Ok... like this?

VCA_CV3.jpg

No, that outputs ~0V, like your earlier schematics. The 1.43K resistor connects to the '+' input as well. It's a voltage divider followed by an unity-gain buffer.

EDIT: Your offset trim circuit also needs to match the one shown in the 5534 data sheet.

JDB.
[you may want to review op-amp circuit topologies. This is a good start (more here)]
 
nrgrecording said:
Control port is driven by a Harris 4605 quad opamp. Can't find any datasheet. I only know its more ore less a quad 741 - pin compatible with TL074.
From an old data comparison table (IC linear '80 by ECA)
HA1-4605-5 (all numbers worst case specs.)
Supply voltage: +/-15V (obviously wrong, probably +/-20V)
Max.diff.input voltage: 24V
Temp.range: 0-70°C
Offset voltage: 4mV
Offset tempco: 2uV/°C
Offset current: 0.12uA
Bias current: 0.4uA
Output voltage swing (Rl 2K): +/-12V
CMRR: 80dB
Slew rate: 4V/us
Open loop response without comp.:97dB
Unity gain bandwidth: 8MHz
Noise figures, Zin and Zout impedances not listed

Improved version seems to be the HA1-4602-2 with higher temp.range (-55..+125°C), lower offset voltage (3mV), little less bias current (0.325uA) and better CMRR (86dB).

Comparing to HA1-4741-5, the HA1-4605-5 has lower offet voltage (5mV), more offet current (0.1uA), higher max.out voltage (+/-10V), higher slew rate (1.6V/us) and higher bandwidth (3.5MHz). Typical values from datasheet look better though.

All these HA1-4xxx outperform a 741.

From your 'Last try...'
A TL051A with +/-18V supply looks a little too close to the edge.
Just to make it complete, your NE5534 drawing is missing the negative supply @pin 4 and the 6mV would be -6mV.
good luck
 
Back
Top