Hi,
I have been inactive for a good long time. So I decided to build something superbly easy.
I built myself an amp out of 74HCU04's. You probably know this is unbuffered. I picked up an application note on linear operation of 74C04's and saw few Japanese pages on building amp with these.
Anyways, going on to my question, I am having problems building dual PSU out of four 1.2V NiMH's. I don't want to incert a chip requiring big supply voltage since this thing has only +-2.4V.
I used the cheap method of using voltage divider to make a virtual ground. I tried reducing the resistors to increase the current going though the divider, but I wasn't able to reduce the offset below 0.1V (not loaded).
Can you guys help me?
Tomo
P.S. The amp schematics (basic one) is on this application note. Look at Figure 2.
http://www.fairchildsemi.com/an/AN/AN-88.pdf
I have been inactive for a good long time. So I decided to build something superbly easy.
I built myself an amp out of 74HCU04's. You probably know this is unbuffered. I picked up an application note on linear operation of 74C04's and saw few Japanese pages on building amp with these.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Anyways, going on to my question, I am having problems building dual PSU out of four 1.2V NiMH's. I don't want to incert a chip requiring big supply voltage since this thing has only +-2.4V.
I used the cheap method of using voltage divider to make a virtual ground. I tried reducing the resistors to increase the current going though the divider, but I wasn't able to reduce the offset below 0.1V (not loaded).
Can you guys help me?
Tomo
P.S. The amp schematics (basic one) is on this application note. Look at Figure 2.
http://www.fairchildsemi.com/an/AN/AN-88.pdf
Kudos for using a digital gate in linear mode, although I don't think the current driving capability of a '04 gate is that great. 
Just what do you mean by "offset" for the voltage divider? Output offset from the '04 gate? In that case, use a coupling capacitor.
Loading of the divider? In that case, use a buffer amplifier to provide a low-impedance source.
I would like to see measurements of how poor (or good) a circuit like this behaves. I can imagine PSSR being close to 1 or even less!
Also, what about stability issues. A circuit like this is dangerously close to becoming a full-feathered ring oscillator.
/M
Just what do you mean by "offset" for the voltage divider? Output offset from the '04 gate? In that case, use a coupling capacitor.
Loading of the divider? In that case, use a buffer amplifier to provide a low-impedance source.
I would like to see measurements of how poor (or good) a circuit like this behaves. I can imagine PSSR being close to 1 or even less!
Also, what about stability issues. A circuit like this is dangerously close to becoming a full-feathered ring oscillator.
/M
Nice! I have done something similar many years ago with a CD4007, which contains 3 inverters. I didn't know it was also possible with the 74 series.
The offset you mention, is that the imbalance between the + and - supply? That is not so bad, just a small decrease in max output. Otherwise, can you not use another 7404 as amp to provide the midpoint, just to stay in style?
Jan Didden
The offset you mention, is that the imbalance between the + and - supply? That is not so bad, just a small decrease in max output. Otherwise, can you not use another 7404 as amp to provide the midpoint, just to stay in style?
Jan Didden
Hi,
I thought so too. I have NEVER believed any logic chips which is sold like toilet paper rolls ... 74HCU04 costed me less than a toilet paper roll. It was supposedly an experiment. I don't have the equipments to make measurements. (My soundcard input is busted I think too) Nonetheless, it sounds fairly nicely. I posted mainly because I wanted the help of oh-so-well-equipped ones to test it out for me.
I believe this offset of 0.1V comes from the mismatch between NMOS and PMOS in the complementary pairs. I simply used 4.8V battery supply split into +-2.4V. This I understand is a bad thing. But I wanted it portable. You can see a stash of 8 resistors soldered togethered behind the 4 electrolytic caps. Without connecing to the amp, this will give me exact split. But once amp is connected, the supply voltages become 2.3V and 2.5V. (approximately) Then this leads to the output dc offset.
I thought about adding BUF634 to drive the ground to be the 1/2(Vcc), but the voltages seems rather too low for BUF634's lower operation limit of +-2.25V.
Adding coupling cap is definitely a solution, but I want to see if I can go without it. One way, I just read about yesterday is to fiddle with the gate voltages. I think I can do that.
As for PSSR, I don't think there is any. I mean I will probably get something better than basic common source amp, but it isn't much to begin with. But some good amps don't have any PSSR. Does it absolute necessary to have 70-100dB of PSSR?
I am not really sure about oscillations. I am sure there are some. But hey it costs a quarter! AD797 cost over 20 bucks a pop right?
It was a fun afternoon project. I will await you, Oh-So-Well-Equipped ones, to see to the poor-ness of the sound. I really would like to hear the result. (I am not being mean or anything, I simply don't have the equipments. I am very curious.)
Thanks in advance,
Tomo
I thought so too. I have NEVER believed any logic chips which is sold like toilet paper rolls ... 74HCU04 costed me less than a toilet paper roll. It was supposedly an experiment. I don't have the equipments to make measurements. (My soundcard input is busted I think too) Nonetheless, it sounds fairly nicely. I posted mainly because I wanted the help of oh-so-well-equipped ones to test it out for me.
I believe this offset of 0.1V comes from the mismatch between NMOS and PMOS in the complementary pairs. I simply used 4.8V battery supply split into +-2.4V. This I understand is a bad thing. But I wanted it portable. You can see a stash of 8 resistors soldered togethered behind the 4 electrolytic caps. Without connecing to the amp, this will give me exact split. But once amp is connected, the supply voltages become 2.3V and 2.5V. (approximately) Then this leads to the output dc offset.
I thought about adding BUF634 to drive the ground to be the 1/2(Vcc), but the voltages seems rather too low for BUF634's lower operation limit of +-2.25V.
Adding coupling cap is definitely a solution, but I want to see if I can go without it. One way, I just read about yesterday is to fiddle with the gate voltages. I think I can do that.
As for PSSR, I don't think there is any. I mean I will probably get something better than basic common source amp, but it isn't much to begin with. But some good amps don't have any PSSR. Does it absolute necessary to have 70-100dB of PSSR?
I am not really sure about oscillations. I am sure there are some. But hey it costs a quarter! AD797 cost over 20 bucks a pop right?
It was a fun afternoon project. I will await you, Oh-So-Well-Equipped ones, to see to the poor-ness of the sound. I really would like to hear the result. (I am not being mean or anything, I simply don't have the equipments. I am very curious.)
Thanks in advance,
Tomo
Tomo,
I think your experiment with the '04 is excellent! That is exactly the kind of stuff I want to see around here rather than pretty boxes and discussions on what resistor brand sound the best.
You've made me very curious and I can't wait to get to the lab to do some experiments!
To get rid of the DC offset you really need some sort of DC servo feedback loop. And about oscillations - every feedback amplifier regardless of cost will oscillate if not compensated properly. Its just a matter of measuring the loop response and compensate accordingly.
Stay tuned for an update...
/Magnus
I think your experiment with the '04 is excellent! That is exactly the kind of stuff I want to see around here rather than pretty boxes and discussions on what resistor brand sound the best.
You've made me very curious and I can't wait to get to the lab to do some experiments!
To get rid of the DC offset you really need some sort of DC servo feedback loop. And about oscillations - every feedback amplifier regardless of cost will oscillate if not compensated properly. Its just a matter of measuring the loop response and compensate accordingly.
Stay tuned for an update...
/Magnus
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.