Quad Eight AM4 opamp.

This is a very simple circuit, I know. Still I got confused.

Does this looks right for you?



Aren´t + and - inputs wrong? Shouldn´t these be inverted?

Take a look at this schematic:
http://www.avensonaudio.com/tech/Quadeight/mm100schematicsm1.jpg

Thanks Brad!
Amperex carbon film resistors. So thats the secret!

Somebody take the BOM into the photoshop and earn your masters thesis.

OK, now I see something. So the second stage in the schem I´ve posted is a NPN, and in our schem it´s a PNP, that´s why inverting and non-inverting inputs are crossed... :idea:

:?: :?: :?: :?

Does this looks right for you?

Click to expand...

No. Inputs are inverted and Q3 has wrong polarity as you noted. In addition to this, R6 should go outside the feedback loop.

Samuel

Thanks!

So, is it what actually goes on inside a real AM4?



Here it´s in a API 312-like circuit.

Can anyone confirm if this is what´s inside a AM4 opamp?

Can anyone confirm if this is what´s inside a AM4 opamp?

Click to expand...

No--I don't have an AM4 at hand. The schematic linked above shows a slightly more complicated version. But what you have is surely something that will work!

Check the voltage drop across R5/R6 to measure the quiescent current of the output stage. If you get more than 20 mA it might be wise to change the diodes to 1N914B or similar. And while we're at it, add a pair of protection diodes to the input stage as seen in many schematics (e.g. the 990).

Having Q5 without any gain stabilisation looks somewhat dangerous as one opamp might be stable while another one is not due to beta-differences--but I guess it works in practice...

Samuel

mmm... I don´t know if that actually what you want, but here is a try:

2n2102 - R5 = 178.3mV

R5 - R6 = 158.5mV

R6 - 2N4036 = 138.6mV

I would expect much greater voltage drop??? It´s from circuit maker simulation.

when you talk about Q5 gain, should I select for lower HFe or higher HFe ones?

Anyway, having another look at my quad8 files, I found this exact same circuit in a quad eight preamp, that is also called 312 (like the API one). The feedback network is a little different, thought...

They also does indicate in4002 as the output diodes.

have a look:



Do you also think I should use this same feedback, with a 49.9k resistor instead of the 22k resistor like in the API circuit?

2n2102 - R5 = 178.3mV
R5 - R6 = 158.5mV
R6 - 2N4036 = 138.6mV

Click to expand...

So if I understand your voltages correctly you have 39.7 mV from emitter to emitter, right? that makes about 2 mA quiescent current (39.7 mV/(2*10 ohm)). That's rather on the low side for my taste and a good indicator that you'd better let the 1N4002 in as a 1N914B would get even lower current.

When you talk about Q5 gain, should I select for lower hfe or higher hfe ones?

Click to expand...

Hm, hard to say. IIRC the 2N4250 has a higher guaranteed minimum and maximum hfe than the 2N4403 which indicates that you should select for high hfe if you use the 2N4403. But before you start selecting simply build a few and see if they are stable at low gains. If you get some that are instable see if you can make them stable by replacing Q5 with a low-hfe one.

I suspect that you'll not get instable ones. My remark was more thought as a comment that I don't consider it good design practice to have a circuit whos stability is not well defined by passive components but by (very) variable semiconductor parameters.

Do you also think I should use this same feedback, with a 49.9k resistor instead of the 22k resistor like in the API circuit?

Click to expand...

I'd use the 22k as it gives lower noise. Shoudn't make much difference otherwise.

BTW note the zobel network at the output; it might be wise to include that, at least for the opamp that drives the output.

Samuel

Thaks a lot fr your comments, samuel. It´s really helpfull. I´ll include the zobel at the outputs.

HI,

I know this is an old topic, but what kind of mic tx was used with this opamp? Ratio? impedance?

thanks

Michael

have a 1:8 over here, Jensen JT-110K

so 64 times whatever mic is = sec Z

don't know the original, checking around...

looks like Reichenbach/Jensen were the transformer guys on the Quad 8 stuff,

there might be a tear-a-part on the output floating around as well,

Quad 8 name>

"Elaborating on this process a bit more, former Quad-Eight manufacturing engineer David M. Gordon told me "the process that was called 'Quad-Eight' was to print four strips of 8 mm film on one piece of specially perforated 35 mm film stock which would then be slit after developing. This enabled the lab to utilize the 35 mm processing equipment to develop consumer 8 mm film, hence 'Quad-Eight'."

Cinemag 3440A I think is the one, it's a 50K secondary.  I used some new ones along with the original Reichenbach units on some restorations, they seem identical.