EHX Small Clone filter understanding.

[POTL]

Hi all.
I can't understand for what purpose a PNP transistor and a diode were used before the delay line?
Other modulation devices use NPN transistors or operational amplifiers. I saw diodes only in circuits using sad512 or sad1025. What prevents Small Clone from working without a diode and from an NPN transistor?

[idy]

PNP is because 3007, unlike SADs, is positive ground. I think. The diode I can't say.

[Rob Strand]

Hi all.
I can't understand for what purpose a PNP transistor and a diode were used before the delay line?
Other modulation devices use NPN transistors or operational amplifiers. I saw diodes only in circuits using sad512 or sad1025. What prevents Small Clone from working without a diode and from an NPN transistor?

It's a funky way to bias the BBD input. 

The 12k resistor is the emitter resistor for the PNP transistor buffer.  The base of the transistor sits one diode drop and one Vbe drop *below* the BBD bias voltage then the bias pot feeding the base sets the bias, which will be set to a somewhat lower voltage than the voltage on the BBD input.   

Why do that instead of the normal scheme?   I think the idea was to limit the negative input swing via the limited swing of the PNP buffer instead of adding clipping diodes - if you don't limit the input swing it can sound bad when the BBD input is overloaded.

In this circuit the diode doesn't clip it level shifts the voltage.

[POTL]

PNP is because 3007, unlike SADs, is positive ground. I think. The diode I can't say.

Yes 3007 and positive ground. but there are millions of other choruses that have a 3007 and an NPN transistor on board. CE-2 is an example of this.

[POTL]

Rob, thanks for the detailed answer. I thought it had something to do with CD4047. Even the LFO differs from the standard BOSS or MXR solutions, at least with a non-standard voltage divider, in which instead of 4.5 volts there are 2 different dividers for each half of the op-amp. Or a solution with a separate voltage rail, where BOSS or MXR will use an RC filter, here they used a transistor. It is necessary to carry out simclation, apparently, the circuit will work fine with an NPN transistor without a diode, but with a diode it will no longer work. Do I understand correctly that this voltage bias solution works better than BOSS?

[Rob Strand]

It is necessary to carry out simclation, apparently, the circuit will work fine with an NPN transistor without a diode, but with a diode it will no longer work. Do I understand correctly that this voltage bias solution works better than BOSS?

The main difference between Boss and MXR circuits is the filter has full AC swing then they add deliberate (diode) clipping to limit the level into the BBD.  The EHX unit is limiting the swing of the amplifier using the supply & level shifting diode.   

You can get the NPN circuit to work in the EHX unit.  You can even use the NPN circuit with a diode, if you flip the diode, but the circuit no longer achieves what it does in the EHX circuit, no point doing it really.   You would need to connect the 12k resistor to ground instead of the BBD +V rail, which is what Boss does on the BF2. 

The Boss circuits do more than just use an NPN:
- the filter output is AC coupled to the BBD and the BBD bias is separate doesn't pass through the filter.
- the diode clipper is before the filter so that the harsh distortion from the clipper is removed by the filter.

The Boss BF2 circuit would be an ideal model to use an NPN.   The MXR Flanger has the same structure to the Boss but the implementation details are different and it uses a passive filter instead of an NPN.