Bipolar PI design

[Detectorist#1]

Hi Carl,

No reason for different linearity in regulation if the SENS POT P4 is connected in series with R53. In this case the value of R54 have to be increased to 51K. If the POT is connected in series with R53 without change of the value of R54, the gain of the stage will be regulated from 2 to 1.02

[Carl-NC]

Another thing I want to make clear is that if I change the location of the P4 as you say, we will have a problem with the Threshold adjustment, which is the P5.

I think there is confusion because the part numbers keep changing. He's talking about the gain resistors on U12d, which are now labeled R53 & R54. You can put the Sens pot in series with either one. If you put it in series with R54 (as you have it) then the sensitivity adjustment is linear but the bandwidth changes with setting. This is not usually a terrible thing because you normally need to reduce the BW at high gains and not so much at low gains. If you put P4 in series with R53 the the sensitivity adjustment is non-linear but the BW is constant. In the future, it helps to keep part numbers constant until the design is finalized, then rename them if you like.

4. Max signal to pin 3 of LM386 have to be limited as +0.4V. Voltage amplifying in this connection is 20 and max output voltage will be +8V.

You can overdrive the LM386 input, no problem.

[eduardo1979]

Hello @Detectorist#1.
Once again, my haste has led me to new mistakes.
Of course, the 4066 has to be powered from +5v to -5v, and the output of the LM386 has to be connected to the speaker through an electrolytic capacitor. Of course it does. Another thing I want to make clear is that if I change the location of the P4 as you say, we will have a problem with the Threshold adjustment, which is the P5. By modifying the value of the P4, the Threshold will be affected. In this way, as it is shown in the diagram, the Threshold will remain constant despite modifying the sensitivity of the detector with the P4. I perfectly understand what you mean about the filter, but it is the harsh reality.
Attached is the diagram with the relevant modifications.
I look forward to your comments.
Regards,
Eduardo.

[Detectorist#1]

Hi Eduardo,
Unfortunately, this schematic also is full of problems, but we will help:
1. 4066 have to be supplied from +5V and -5V volts. No problem to use CD4066B. This will allows to be managed analog voltages in range +5v to -5V. Voltage translators from CPU to 4066 with transistors are OK.
2.Pot R4 have to be connected in series with R55 and the value of R56 have to be increased to 47K. This will preserve cut-off frequency of this stage to be independent from SENS setting.
3.No problem for D17 to be removed - no problem for Q13 to work with minus 5V at Base electrode.
4. Max signal to pin 3 of LM386 have to be limited as +0.4V. Voltage amplifying in this connection is 20 and max output voltage will be +8V.
5. The load of LM386 - SP1 have to be connected via electrolytic capacitor for stopping of DC current without the signal.
6. The diodes D18 and D19 have to be replaced with resistor with 22-47ohms. It is not possible to use diodes as filters - the dynamic resistance of the diodes is very low when a current flow via them.

[eduardo1979]

Hi Carl.
Thanks for your quick reply. I am attaching the corrected diagram. Please let me know when you can if everything is correct.

[Carl-NC]

The 74HCT4066 is connected between +5V and GND as Carl recommended me.

No, not what I meant. The 4066 has to be powered from +5V/-5V. By making R42,R47 = 0 then one side of each switch is connected to the virtual ground inputs of the opamps (U10) and the switches will behave more consistently.

[eduardo1979]

Hello again.
As promised, I have updated the schematic today with all the modifications suggested by the participants in this project.
The 74HCT4066 is connected between +5V and GND as Carl recommended me. In the end I have chosen to use the LM386 in the audio amplifier, just like Carl in his original project. For the preamplifier, in the end I have opted for the two-stage amplification, because if you want to use only one stage, with a few small modifications, you would have it and in this way we can test the two variants. I have replaced the two LT1112s with a TL064, since it is a cheaper operational amplifier. I have also replaced the LMH6646 with a NE5532, because it is cheaper and Carl recommended it to me that way and his recommendation seems very coherent to me. I hope you can review the schematic and if anything else needs to be added or modified, it will be modified.
Best rehards,
Eduardo.

[pechkata]

Hi everyone on the forum...
Chip file uploaded...

[eduardo1979]

You are right. D4 and D17 are reversed. Rushing is simply not good. I was in a hurry to finish the design since I had very little free time and I made those two mistakes. The C40 would leave it there. The volume potentiometer is fine where it is. I have used this audio amplifier circuit on more occasions and it works very well. The volume is regulated linearly. The LM386 is a good option and I don't rule it out at all. This Saturday I will make all the corrections to the schematic and publish it.

[Carl-NC]

Hi Eduardo,

The PCOUT output of 4046 haves +/-5V output voltage levels. J113 will be open always. Plus 5V will be limited by open junction Gate-Source to maybe +0.7V and minus 5V will be limited from connected diode to -0.7V. This voltage is not enough to stop J113 to conducting. For proper operation you have to remove diode D4 and to add resistor 10K-47K between 4046-pin4 and the Gate of J113.

Also the schematic of the audio amplifier have to be seriously corrected. The diode D17 limits the voltage on the Base of Q14 to 0.7-0.8V. With 0.6V on Base-Emitter junction of Q14 - max voltage on R71 will be 0.1-0.2V. This give collector current of 0.5 -1.0 mA. This current not gives enough voltage on R70 (1mA x 470R = 0.47V) for controlling of Q15. Also it is not serious to use pot with value of 1K for load of 50 ohms. The regulation of the level of the sound will be all or nothing.

I agree with all this. D4 is backwards from what you want, but you also need a resistor between D4 and U14 or you're likely to burn up the VCO output. Next, if D17 is intended to clamp the cap-coupled (C40) signal then it is also backwards. But C40 isn't needed because the signal on Q13-D already has the correct voltage levels (+5V/0V) so you can omit C40 and D17. I would then move the volume pot to the base side of Q14 and delete R69. Personally, I find the LM386 to both sound better and have better drive efficiency compared to most discrete solutions, and it's dirt cheap.

[Carl-NC]

Hello, Eduardo, you can use the same sound scheme on the TDI, it is perfect.

The TDI audio uses an incredibly expensive AD654 VCO and is also a power hog. There are better ways to do this.

[algan]

Hello, Eduardo, you can use the same sound scheme on the TDI, it is perfect.

[Carl-NC]

Yes, the micro runs on +5V/0V and the JFET clocks need to be 0V/-5V.

[Detectorist#1]

Hi Eduardo,
Also the schematic of the audio amplifier have to be seriously corrected. The diode D17 limits the voltage on the Base of Q14 to 0.7-0.8V. With 0.6V on Base-Emitter junction of Q14 - max voltage on R71 will be 0.1-0.2V. This give collector current of 0.5 -1.0 mA. This current not gives enough voltage on R70 (1mA x 470R = 0.47V) for controlling of Q15. Also it is not serious to use pot with value of 1K for load of 50 ohms. The regulation of the level of the sound will be all or nothing.

[Detectorist#1]

Hi Eduardo,

The PCOUT output of 4046 haves +/-5V output voltage levels. J113 will be open always. Plus 5V will be limited by open junction Gate-Source to maybe +0.7V and minus 5V will be limited from connected diode to -0.7V. This voltage is not enough to stop J113 to conducting. For proper operation you have to remove diode D4 and to add resistor 10K-47K between 4046-pin4 and the Gate of J113.