Push-Pull Class AB Amplifier - nisanth

Lab Notes 28 Jan 2023 I've been looking for an electronics components vendor (in the USA) who offers a decent assortment of common through-hole capacitors. The capacitor assortments I'm looking for have to be reasonably well identified, as shipped by the vendor (and not just unlabeled grab bag goods). The point is to be able to select from various types of capacitors, from the limited information found in construction articles about the easier electronics projects that you can make yourself. Summary: Looking for an assortment of capacitors for projects with a focus on beginners' knowledge (I've always relied on kits with all the parts selected for you; those are getting harder to find, as are through-hole components in general). - - - recent falstad postings for what I am working on (push pull amp). I need the capacitors and all I have are 100 uF 16 V polarized electrolytics (which seem to distort a triangle wave input quite a bit) latest iteration of the push-pull bjt amp is here: https://tinyurl.com/2h7uxga6 on falstad ref circuit discussion was here: https://electronics.stackexchange.com/questions/601011/simple-push-pull-bjt-biasing I'm considering connecting the 'grounds' of two 120 VAC bench equipment front panel jacks together, but don't know the rules for it. ;) Trying to devise a safe test for excess current flow across the two supposed ground references. One's a function generator, and the other is a Korad KD-3005D bench PSU. Eventually the Owon oscilloscope (a third 120 VAC powered system) will be interconnected to this same ground ref. I've already had 'good luck' with the Owon scope and the (Lab-Volt) function gen, but not the Korad PSU. All Earthed via 3-prong plugs, on same power strip (itself connected to a ZeroSurge). I do own an isolation xformer for 120 VAC as well. ;) but it's a Vis Iso-tap II and has no third prong. (They were meant for hot chassis TV sets for repair shops in the 1960's). Was thinking maybe a PTC I have on hand in lieu of a 'fuse'. I probably still own a 250 mA fuse (melt type, glass cartridge) as well. (Bench supply has current limiting; Lab-Volt also has a PSU module, but no current limiting, so I'd rather not use it for this, for now). - - - Very confused by findings. ;) The Tektronix scope primer advice seems valid and informative (check what you 'think' is ground with the center conductor of the scope probe) on the DUT. Gave a nice low value for shared Earth ground (Lab-Volt has a front panel banana jack in the green color, with a ground symbol - that one seems to test 'good' using Tektronix' method). But all the black colored jacks did not test similarly. ;) Korad Bench PSU (green jack) tested the same as the Lab-Volt green ground jack did. Someone else mentioned: 'How about differential measurements? Use both channels in differential mode (i.e. X-Y), which requires no gounding of your scope. Just both tips (floating relatively to the circuit)' The bench supply and the Lab Volt fcn gen grounds seem to correspond well with true Earth & chassis ground. The fcn gen output on its own is fine - treat the black jack as 'ground' which afaict bonds it to Earth as soon as the scope clip is attached there. It'd be the same as bridging the black jack with the green jack on the lab volt and referencing the scope probe grounding clip to that green jack, I think. And with a stacking banana plug, I could do it just that way to be sure. I was mostly concerned about current flowing on the wire bonding the black jack with the green jack on the Lab Volt. After all it doesn't come wired that way. ;) Similarly with the Korad bench supply - referencing the scope probe clip to its black jack (minus terminal) and the scope tip to the positive terminal produces an expected value (matches front panel setting on the PSU, on the scope display deflection). I don't think I understand the client/end-user side of a power supply very well at all, in terms of floating above an Earth reference. ;) The isolation transformer (tried just with the Lab Volt) made things worse, not better. ;) At least one measured non-ground node shows 60 Hz ;) iirc it was either (ungrounded, voltage output) terminal of the bench supply, and was of rather low magnitude (500 mv or less?) 500 millivolts from the + terminal of bench PSU to scope probe - with or without bonding chassis grounds together (no change) (and nothing at all connected to the negative terminal of the bench PSU). I'm just not getting a strong enough stop signal so I'm pressing on with this lab experiment. ;) Someone else mentioned: 'sounds like a floating ground'. EBC L to R PN2222A ref https://tinyurl.com/2h7uxga6 falstad class AB amp w/100 uF electrolytics Wish I had better capacitors and not sure at least the input cap can be removed. 32 mA predicted, 26 mA measured in the lab, two 120 ohm resistors and two diodes, all in series, forward biased. CHECK. Everything's telling me the scope ground follows the measurement. So, reference the scope (clip ground) to the Korad bench PSU if measuring its output (in volts). Reference the scope to the black jack of the function generator when measuring its frequency and amplitude. That'd be trivial except for the fact that the fcn gen output is already connected in-circuit, through the electrolytic, to the jcn of the two biasing diodes.

TEST Circuit Before 20:59 UTC Sat 28 Jan

The DQ - lo power Sun 29 Jan 00:48:51 - 1 / 4 Watt - Vcc 4.6 V

The DQ - Mic Pre-Amp E-aa- 400 mv signal MAX and 600 Ohms out

The DQ - Mic Pre-Amp E-bb- 400 mv signal MAX and 600 Ohms out

Class AB low power amplifier 60 mW into 32 Ohms
The DQ - lo power PA E-cc- 1.8 V signal MAX and 32 Ohms out

Class AB low power amplifier 60 mW into 8 Ohms
The DQ - lo power PA E-dd- 1.8 V signal MAX and 32 Ohms out

Class AB low power amplifier 60 mW into 8 Ohms
The DQ - lo power PA E-ee- 1.35 V signal MAX and 8 Ohms out

Vbe Multiplier Biased Class AB Amp
The DQ - lo power PA F-aa- first iteration - not very good yet

Vbe Multiplier Biased Class AB Amp
The DQ - lo power PA F-bb- first iteration - not very good yet

Vbe Multiplier Biased Class AB Amp
The DQ - lo power PA F-cc- Balance found? Seems closer now.

Thu 2 Feb 2023:

Vbe Mult Bipolar PSU
The DQ - lo power PA F-dd- still looking to Balance

Thu 2 Feb 2023:

Vbe Mult Bipolar PSU
The DQ - lo power PA F-ee- Unbalanced bipolar - ERROR TODO:fix

Thu 2 Feb 2023:

Vbe Mult Bipolar PSU
The DQ - lo power PA F-ff- Low power fairly symmetrical dissipaton

NEXT: lower the high dissipation resistors at top and bottom of Vbe multiplier sub-circuit and examine results of doing so.

Fri 3 Feb 2023:

Vbe Mult Bipolar PSU Fri 3 Feb 01:51:02 UTC 2023
GOOD 3 Feb 23:52z
The DQ - lo power PA F-gg- Low power more accurate

NEXT: lower the high dissipation resistors at top and bottom of Vbe multiplier sub-circuit and examine results of doing so.

Fri 3 Feb 2023:

Vbe Mult Bipolar PSU Fri Feb 3 06:55:47 UTC 2023
The DQ - lo power PA F-hh- A bit closer > 50 mW out
PA xistors running hot

Vbe Mult Bipolar PSU Fri 3 Feb 01:51:02 UTC 2023
The DQ - lo power PA F-ii- comments
more comments

Vbe Mult Bipolar PSU Fri 3 Feb 23:37:10 UTC 2023
The DQ - lo power PA F-jj- comments
more comments

BACKTRACK to here:

Vbe Mult Bipolar PSU Fri 3 Feb 01:51:02 UTC 2023
GOOD 3 Feb 23:55z
The DQ - lo power PA F-gg- Low power more accurate

did BACKTRACK to here:

Vbe Mult Bipolar PSU Sat 4 Feb 20:33:19 UTC 2023
GOOD 4 Feb 20:33z
The DQ - lo power PA F-hh- backtrack then forward from there

Hosting for this site is provided by

The SDF Public Access UNIX System

Valid HTML w3.org

Sat 4 Feb 20:33:19 UTC 2023in recent drawing

Sat 4 Feb 20:48:07 UTC 2023last HTML update