Bipolar PI design

[Detectorist#1]

Hi Carl, many of us, ITMD3 readers, are very grateful for your efforts in favor of DIY detectorists to be involved in this exciting hobby!

[Carl-NC]

Thank you. It's a 2-way street, I've learned a whole lot from everybody. ITMD has been a way to put it all together and share it back.

[boilcoil]

...I've recorded ...

Hello Eduardo,
I watched the video, really good design.
I noticed that all IC are mounted on sockets, which is a good choice for hobby electronics.
In the project, you use analog switches, type 4066, which, however, have one feature that may or may not be useful, depending on the specific application.
When a positive voltage is applied to pin nY (1Y, 2Y, 3Y, 4Y) relative to the corresponding pin nZ (1Z, 2Z, 3Z, 4Z) and the drop of voltage in the on state is greater than 0.4V, then current flows to pin nZ from the power supply of the chip (with reverse polarity, this effect is not observed).
You could try directly replacing 4066 with 4016. Of course, the 4016 has more resistance in the on state, but you put serial resistors of 1k in the circuits of the incoming signals, so it doesn't matter much.
IC 4016 is much simpler than 4066, but the aforementioned effect is not observed, i.e. it is completely symmetrical. Replacing 4066 with 4016 may or may not improve detector performance.

Please try this and share in the forum.

[Detectorist#1]

Hi boilcoil,
I just assembled one simple tester for 4016 and 4066 analog switches.
The schematic is attached.
You are right - these switches have problem when voltage between input and output of the switch in ON state is more than 0.4V-0.5V. Only in the datasheets of 4016 and 4066 from TI company have at first look strange information for MAX current via switch. Other companies are silent for this problem. For 4066 max current is 1.3mA and for 4016 max current is 0.5mA according datasheets at 10V supply voltage. The reason for those difference is different resistance in ON state. After your post, I understand why this strange parameter exists in the datasheets for 4016 and 4066 from TEXAS INSTRUMENTS company. Thank you for your post!

[eduardo1979]

Hello Boilcoil. I've tried with 74HCT4066, 74HC4066, CD4066, and CD4016. There's no difference in depth. The only difference I notice is that the Ground Balance drifts very slightly, and I have to readjust the potentiometer. The drift is minimal. I've tried with a Hot Rock.

[boilcoil]

Thank you, Eduardo for the tests done and for sharing them in the forum. Thank you again

[algan]

Hi Carl, if the power supply of the bipolar circuit is like in GB-PI and +12 volts will be GND, and the PIC processor is powered by -5 volts and GND, then it would be possible to use 4066 without translators, what would it work then and what would be the difference from the current one?

[Carl-NC]

Yes, that would work, with really no difference in performance. However, there is still the issue of getting enough logic swing on the 4066. If the 4066 is powered by +/- 5V then per the data sheet it needs a 7V logic swing. I have run these on 5V logic swing but it is not guaranteed to work with all chip producers.

[algan]

And this one is a bit more expensive MAX4616, but it should work without translators, maybe it's better than 74HC4066.

[Detectorist#1]

MAX4612-4616 works only at supply voltage 2.0-5.5V. You have to use MAX4610CPD+. This IC works OK with only 4V level of high voltage on the control inputs at 10V supply.

[Detectorist#1]

But this pleasure haves a price - appr. 7.50USD 1+

[Detectorist#1]

The best choice for analog switches in MD designs with +/-5V power of analog part and 0V/-5V supply voltage for control uP is MAX4066ACPD+. This IC needs only +4V high level of control voltage for switches for 12V supply voltage and the leakage current of open switches is below 100pA at 25oC. This current is 10 times lower than the typical leakage current (1nA) and 1000 times lower than max leakage current (100nA) at 25oC for popular 4066 switches. But the price is more than 6.60USD 1+

[pechkata]

What's wrong with the 74HCT4316 - from the schematic?

[Detectorist#1]

Hi pechkata,
Nothing wrong in this schematic. 74HCT4316 will work OK. The error in the controlling of 74HCT4316 was only in the schematic of PI-GB design in the first version of ITMD3 book.
In my post #387, I mean the best variant of 4066 (MAX4066ACPD) without need of level translators (when the micro is supplied from 0V and -5V) and with lowest leakage current for all 4066 type analog switches.

[eduardo1979]

I have completed testing the Bipolar PI design and posted new firmware. I changed the timing to use 1 timer and is now all interrupt-driven, there are no polling delays. This gives the delay adjustments more resolution and more consistency.

Two things I don't like about the design. One is that the demods are a bit glitchy in that they pass through some transients. I don't know if this is a particular problem with the 4316 switches or what. I have not seen this in other designs I've done. Second is that the audio is not all that great. The 555 VCO does not have as much frequency range as I recall from when I breadboarded the circuit. And the transient glitches from the demods gives the audio a little it of a raspy sound. With a little tinkering, both might be solvable but I don't have the time right now.