@@paulbishop1016 An "absolute maximum" rating is one that should never be exceeded to avoid the possibility of damage to the device. So the recommended operating conditions always have some margin to ensure you never exceed it. If the absolute maximum collector current were 10mA, you'd never see the hFE at 10mA specified on the datasheet. It's a pretty safe bet that 50mA per transistor is the absolute maximum collector current. Of course, the maximum usable working collector current may also be limited by the absolute maximum dissipation of 300mW per transistor, e.g. 20mA per device at 15V supply.
I've used the CA3046 in the past to make custom amplifiers. As well as some of its brethren like CA3096 (5 independent NPN transistors) and CA3054 (two full differential amplifiers). I particularly liked the 3054, it has almost everything one needs to make an instrumentation amplifier. And yes, the parts were originally developed by RCA (thus the "CA" in the part number).
The maximum collector current for the CA3046 is 50mA. It's right at the top of page 2 of the data sheet under the heading "Absolute Maximum Ratings", where you would expect to find it.
The classic 'Tandem Match' by John Grebenkemper (then KA3BLO - July 1987 QST) used the CA3146 (higher voltage version of 3046) as log/antilog amplifiers to compute and directly display SWR from voltages obtained from forward and reflected power detectors. I believe the circuit topology was called a 'Gilbert Cell'. I adapted the design for some homebrew RF power metering projects of my own. It worked well.
I have some CA3086s, which just have 3 NPN and 2 PNP transistors. These would be really nifty as I've been exploring diff amps lately. There is also the CA3096, which is configured for a diff amp, but in a slightly different way. Makes me think I need to find a comprehensive list of CA series chips.
OK, I'll take a hundred. Don't know for sure what I'll do with them but maybe they would come in handy in circuit to charge my fluxcapacitor or in a planetary force field detector.
Thanks I was looking for such transistor arrays (HFA3046, HFA3096, HFA3127, HFA3128)I am designing I am designing two stage cascading common emitter amplifier which are "DC Coupled" to find out resonant frequency of capacitors. I see many circuits in the industry , which are fairly high in frequency, and are AC coupled. The capacitor needed to couple these stages are specified, but such capacitors have resonant frequency less than the frequency of the circuit! One such circuit I see quit often is common emitter amplifier with emitter bias resistor shunted to ground with a large cap! The cap of such value has self resonant frequency of much less than operating frequency of the circuit!!!
Any amplifier has an operating frequency bandwidth which in turn has an influence on the capacitor values selected. Usually this is the lowest frequency the amplifier has to handle in as far as capacitor selection is concerned.
Use That300, That320 or That340 from the That cooperation and a tiny bit cheaper, the HFA parts are design for high frequencies. 8GHz for npn and 5.5 gHz for pnp.
I've always stuggled to grasp how the input diff amp in a solid state audio amplifier works, and this kinda helped, but not too much. You skimmed over the details too much for my pea brain to fill in the gaps. I'd like to understand the constant current source and current mirror better too. A deeper dive is requested.
If you're looking for maximum collector current, you really ought to be looking at the section "Absolute Maximum Ratings" (see 2:07 ) where you'll find that Ic max is 50mA and Vce max = 15V. It's really worth rehearsing the parameters you're going to talk about _before_ you start making the video.
Shouldn't the emitter of the emitter follower be connected to -V and not ground? Otherwise the signal will clip since it is oscillating around zero volts. Or am I missing something?
Ideally yes, to be fully on the conservative side, but not really in this example, as one of the inputs of the amplifier is referenced to ground and the collector of that transistor can only go down to about 0.6V anyway. Hooking up the emitter follower to ground, instead of -V in this case may lighten the dissipated power on the lil' thing. By the way, despite the Max IC is 50mA, the device Max Power is just 300mW, so it's quite unlikely that anyone would run those transistors at much more than 5mA each.
The output from either collector will have a quiescent bias point around (+V)/2, so there's no point in connecting an emitter follower to (-V), as it will clip on the positive excursion anyway as soon as the output approaches ±(+V)/2. This assumes that the supply rails are symmetrical. If the biasing isn't obvious, take an example where the supply is ±12V. Looking at the circuit at 6:01 the current mirror sinks (12V - 0.65V) / 18K = 0.63mA into pin 11. So ideally, 0.63mA is also sunk into pin 14. If the input voltage at pin 2 equals the voltage at pin 4 (i.e. 0V), then the current ideally splits equally between the differential pair, and there is 0.315mA through each. The quiescent voltage across each output resistor is then 0.315mA x 18K = 5.67V, and therefore each collector is biased to 12V - 5.67V = 6.33V above ground. The maximum symmetrical output swing is about ±5.6V before clipping and the output never reaches ground.
@@fabiotrevisan8922 That's pretty much the right way to think about the circuit, although the dissipation in the emitter follower would be independent of whether its emitter resistor is connected to ground or to (-V). The power dissipated in the emitter resistor would, of course, be different. Incidentally, a lot of the parameters and graphs given in the datasheet are specified at Ic = 1mA and Vce = 3V, so I'd aim for somewhere near that when designing in the absence of other constraints.