PI-GB Design

[Detectorist#1]

Hi Carl,
The connecting of the power of the uP in GBPI_Sch.pdf is also wrong. This uP have to be connected between +5V and GND for right control of 74HC4316. Now days, I'm waiting for my copy of ITMD3!
The controlling of upper P-ch MOS transistors in the Tx circuit of bipolar TX solution maybe will be more simple without using of 4053 (also 4053 haves big max Ron and the max switching time have to be in the range of microseconds) . Maybe in this part of the schematic is possible to use the same schematic as for making of VCP with 2N3904 and 2N3906 in GBPI_Sch.pdf. I will test this idea.
The name of the signal on the input of the 78L05 in GBPI_Sch.pdf have to be VCP - not +VB as the +VB and the GND are the same signals.

[Carl-NC]

The uP in the PI_GB circuit has to be powered from 0 & -5V because the charge pump creates +5V, so this is correct. What may not be correct is the use of the 4316 switches, especially with the logic reference tied to ground. The book schematic uses 4066 switches which should work (HCT variety) so I think this is the error.

You are correct that using a 4053 for gate drivers is not ideal, but it does work pretty well. One thing I've tried to do throughout the book is to present alternative solutions to the same problem to get readers to think about different ways of doing things. In that regard, it worked! The book designs are not intended to be final high-performance circuitry you build and go hunting, but rather teaching tools and a starting point for better ideas.

If I were design a TX with SPDT gate drivers, I would probably use a better part like the PI5A3157 (which I used in a low-voltage PI circuit) or the ADG5436 (which I used in a high-voltage VLF circuit).

[Detectorist#1]

Hi Carl,
Yes, you are right, when 74HCT4066 is used, uP haves to be powered with GND and "-5V". But in this case is interesting how the switches will work with positive analog signals - Vcc of 74HCT4066 is connected to digital ground (not to +5V).

[Carl-NC]

No, the 4066 is still connected to VCC and VEE, but the logic swing is 0V (high) and -5V (low).

[Detectorist#1]

Hi Carl,
The HIGH and LOW levels for controlling of the switches of 74HCT4066 are defined in table 8 on page 8/21of the datasheet as >2V and <0.8V at Vcc (4.5V-5.5V). For me it is not clear from all datasheet is it allowable for 74HCT4066 to work with power voltage more than 5.5V an what will happen with levels of control voltages in this case. Misunderstanding is provoked by the table 4 on page 4/21 where absolute maximum power voltage for both HC and HCT variants is defined as +11V, but in the table 8 on page 8/21 no information for switching levels of HCT variant in case when power voltage of HCT variant is more than 5.5V.

[Carl-NC]

Both the 74HC4066 and 74HCT4066 have an absmax VCC-VEE of 11V. The HCT is characterized at 5V because it's intended for TTL-compatible circuits, which are 5V. But it will run at 10V.

If you pretend that VCC=10V and VEE=0V then for the HC version the logic swing needs to be 7V (V_IH). But the HCT has a VIH of 2V which works nicely even with a 3V micro.

[Detectorist#1]

Still I not see information in the datasheet for switching levels of the HCT variant when power voltage is more than 5.5V. Maybe you are right but there is no explicit information for this.

[Carl-NC]

No, you won't find that in the data sheet. But it works just fine.

[Detectorist#1]

Congratulations for the manufacturer! Thank you!

[Detectorist#1]

New information for me. The company TI also manufactures 74HCT4066 (the name is CD74HCT4066M96). In the data sheet from TI, absmax Vcc is 7V. So, the mystery continues. Have you really tested the 74HCT4066 from NXP with 10V supply?

[Carl-NC]

It's not unusual for different manufacturers to have different absmax limits so it's possible that some will run at 10V and some will not. The oddity is I am using a TI CD74HCT4066 and it runs fine at 10V. Here is what ITMD3 says on p315:

[Detectorist#1]

Hi Carl,
It's not unusual for some amateurish designs to use some active components in "strange" way. But we have rights for this - we are experimenters!

[pechkata]

Hello Carl...
Will there be a Hexfile for this scheme...?
Thank you...

[GeoMax]

Hi Carl,

can you upload the source code files of the "GB PI" and "improved PI" designs from your book? Thanks.

[Carl-NC]

Yes, let me double-check first that the code is working correctly. Probably post tomorrow.