Transistors and NE570 in modulation and delay.

[POTL]

Again, all welcome, I almost closed the questions on the technical aspects of the pedals, here's another 2:
1) The circuits below are transistors in the effects of BOSS BF-2 and MXR Flanger and Ibanez FL9.
I noticed that in almost all modulation schemes there are transistors between the LFO and the clock.
In the BOSS CE-2 chorus, I understood everything clearly on the electrosmash website, but for example there are no transistors in the EHX Small Clone chorus.
Tell me more about what these transistors are doing and how their combination is technically called.
I'm interested in understanding how each node works in effect and this is the last thing I do not know.

Ibanez FL9


MXR Flanger


BOSS BF-2


2) Compander in the delay effects of BOSS CE-2 and EHX DMM, I read the NE570 data sheet, but still do not understand how it works, I know it's a compressor and an enchaser, I understand what they do in the circuit, but I have questions
A) what is affected by the components' ratings in the piping of the chip
B) why the EHX circuits are so different from other circuits that they use more components and use more connectors than those of other manufacturers.
If you know this chip well, please tell us more.

[POTL]

bump

[ElectricDruid]

They're all current sink/source variations. Sink/Source are the same thing, really - only the direction changes.

Basically what affects the rate of the clock oscillator is the speed at which its timing capacitor can charge/discharge - e.g. the current it can draw/lose. If you can modulate that current, you can modulate the clock rate. So you need a method to get from LFO-output-voltage to charging-current. That's a current sink or source (depending on the direction of flow).

There are a huge number of different current sink/source circuits, and if you read up about them, you'll start to recognise a few from these chorus/flanger circuits. Each has advantages and disadvantages, and I can't say for sure why one manufacturer chose one over another. Suffice to say that they did, for whatever reason.

The simplest of all is probably the diode since the current through it depends on the voltage across it, at least over a limited range (so you'd want to stay in that range). This is how the EHX Small Clone works, and also the various synth "diode filters" that exist. It's a very basic form of voltage control, but has proved effective enough.

HTH,
Tom

[POTL]

Thank you
Now it is much clearer.
I will read about this design)
It remains only NE570)

[anotherjim]

Well now, does anybody understand the 570/571 datasheet?
I understand the block diagram, the rest is gobbledygook.

I think many have just copied the application circuit for 2:1 compressor and 1:2 expander to add companding noise reduction to noisy delay circuits. I have.

[greaser_au]

The NE570 is a dual compandor chip that was originally intended for the analogue telephony industry. Each half consists of a precision rectifier, a gain cell, and an opamp (741 grade).   The rectifier output controls the gain cell directly, with it's attack/devcay controlled by an external capacitor..

Depending on where the  rectifier and gain cell are connected, the device can perform either a compressor or expander function. the gain cell reduces it's impedance with higher control voltage.
With the rectifier sampling the opamp output signal, and the gain cell acting as the negative feedback of the opamp, it will reduce (compress) the dynamic range or 'voltage swing' of the  signal.
With the rectifier sampling the input signal, and the gain cell in series with the signal, it increases (expands) the dynamic range of the input signal 

In the DMM context, the low-number-pin half is configured as the compressor and the high-number half is the expander. I'd say it is intended to reduce the clock-switching noise in low-level input signals, and prevent distortion of high level signals, caused within the BBD, that is keeping the signal in the BBDs 'sweet spot'.   They look to be pretty much the template circuits from the datasheet with gain-setting resistor values (I found the MN3005 DMM over at FIS and there is a problem with one of TWO instances of pin 10... there is a hand drawn one that looks to be correct in this regard).

I hope this goes some way to answering the question - if not happy to EXPAND on it

For more useful and practical information,  search for Philips AN174  (NE570, 1, 2 application note - beware Figure 12,  it is wrong)

david

[GGBB]

Philips AN174  (NE570, 1, 2 application note - beware Figure 12,  it is wrong)

How is Figure 12 Fast Attack, Slow Release Hard Limiter wrong? Can you post or link to a corrected version? Thanks.

[greaser_au]

Philips AN174  (NE570, 1, 2 application note - beware Figure 12,  it is wrong)

How is Figure 12 Fast Attack, Slow Release Hard Limiter wrong? Can you post or link to a corrected version? Thanks.

Hi Gord,

From memory,  the reference voltages applied to the comparator inputs are the wrong way around. The transistor should turn on when the signal applied to the comparators is outside of the 0.8-2.8V passband.  With it as is,  the  comparators turn the transistor on with the signal in the passband. R12/13 (0.8V) should be on the top comparator and R14/15 (2.8V) on the bottom one.

Found this the hard way   

david

edit: fixed R12/R13 node voltage (thanks, Gord)

[GGBB]

Thanks David. Here is the figure 12 I have. Not certain about the math, but the lower comparator's R12/R13 looks to be about 0.79V to me. Just want to be sure we have the same drawing.

[greaser_au]

Thanks David. Here is the figure 12 I have. Not certain about the math, but the lower comparator's R12/R13 looks to be about 0.79V to me. Just want to be sure we have the same drawing.

That's the one I was working from.  It's also missing   
1:  the   +15V label  from the node where R15, R16 and PNP emitter meet.
2: a joining dot on the node where the  comparator outputs,  R16 and R17 meet.

Sorry,  yes you are correct.  R12/13 node is at 0.8V and R14/R15 node is at 2.8V intended to allow a 2.0 Vpp swing biased on 1.8 V DC offset   (it's been a few years    I had to read the AN again)

Basically for the comparator's outputs to be pulled high (so the PNP transistor is biased off - and thus no gain reduction), the (+) inputs need to be positive of the  (-)  inputs. 
With the circuit as shown
a: the upper (-) input is at 2.8V  and the no-signal point of the upper (+) input is at 1.8V the upper output will be pulled low
b: the lower (+) input is at 0.8V and the no-signal point of the lower (-) input is at 1.8V the lower output will be pulled low

So if the 2 bias networks are swapped (or instead the comparator +/- inputs are swapped on both) the no-signal 1.8V will be a happy place for both comparators.

I built 4 channels of this, and it took me a little while to figure out why my 4 limit LEDs were on all the time especially when one channel had it's input grounded...

david

edit: added image link to post

[anotherjim]

I think the app notes may mention the internal opamp can be replaced with a better external one. Also I think the main difference between 570 and 571 was the performance of the internal amp?
Anyway, has anyone tried this substitution? Do you simply avoid connecting pins 7 or 10 to remove the internal amp?

Does that application circuit require a negative supply to the chip - since R4 on the inv input is grounded? ...or is the internal amp a more special single supply job?

[greaser_au]

I think the app notes may mention the internal opamp can be replaced with a better external one. Also I think the main difference between 570 and 571 was the performance of the internal amp?
Anyway, has anyone tried this substitution? Do you simply avoid connecting pins 7 or 10 to remove the internal amp?

Does that application circuit require a negative supply to the chip - since R4 on the inv input is grounded? ...or is the internal amp a more special single supply job?

Jim,

The 570 has a very slightly better gain cell performance, better reference stability tolerance and is good for a Vcc of 24V,  while the 571 is limited to 18V.  otherwise I'd suggest they are a pin-for-pin replacement.

As you say, the application note shows how to implement an external opamp in Figure 13 (note the caveat about input common mode voltage):
- the inverting inputs of the internal and external opamps are tied togetther
- the non-inverting input of the external opamp is biased to 1.8V (the internal 1.8V reference can be accessed via the THD trim pin)
- the internal opamp output is left unconnected.
I have done this with  1/4 TL084 (see my '4 channels' link in my last post    )

Because of the internal voltage reference and biasing,  the applications can be operated with a single supply (see the deluxe memory man schematic - but don't get confused by the 'reversed' power supply). I used a bipolar supply with my external opamp implementation successfully.

david