What is the purpose of this resistor in a preamp?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
user 37518 said:
Another dirt cheap opamp is the LM301, which  in my opinion is a very interesting part, I´ve seen Studer designs in which they replaced the entire input section of the 301 with discrete circuitry, it seems however that it is mostly frowned upon for modern work, the national semiconductor application handbook is full of 301/101s, it is my favorite opamp to use as a comparator thou, if im not using a lm311 /339

If by Studer you mean their tape recorders; I think you'll be lucky to find one in a studio these days.

I have a Philips EL3556A (mono) tape recorder which I serviced and it seems to work but I have no reels or tape for it. Fancy machine as it uses a tube. However, most professional applications would require a faster machine (this one goes 19/7,5ips)
 
user 37518 said:
I still consider the NE553x to be one of the best options out there both sound and pricewise,  surpased by the LME49720 or the LM4562 but not by much, the AD797 has really low noise but costs much more,  not to mention the prohibitive price on the OPA627. So yeah, like Douglas Self mentions in his book: you really need a very good reason not to use a 553x

Another dirt cheap opamp is the LM301, which  in my opinion is a very interesting part, I´ve seen Studer designs in which they replaced the entire input section of the 301 with discrete circuitry, it seems however that it is mostly frowned upon for modern work, the national semiconductor application handbook is full of 301/101s, it is my favorite opamp to use as a comparator thou, if im not using a lm311 /339
I used 301s back in the early 70's when is competitive technology. It facilitated a unique feedforward compensation (only worked with inverting topology) that delivered something like 10V/uSec (fast for the early 70s). 

They were mooted by early Bifet and 553x technology delivering similar slew rates without the requirement for trick compensation. (The feedforward cap bypassed around the slow input stage for high edge rates.)

JR
 
gridcurrent said:
I'll bite, as I used to sit under Deane's preaching.
inverting amplifiers were sinful,
Sins could be assigned to non-inverting op amps too.  ::)
but non-inverting amps especially those lacking emitter resistors in the input pair were more prone to stability issues.
common knowledge to IC designers. Degenerating the input LTP (long tail pair) with emitter degeneration resistors, reduces that input stage transconductance that can be manipulated for improved slew rate all else equal. The reason FET input op amps were naturally faster was because of the lower inherent transconductance of FETs vs bipolar devices.

There were old bipolar ICs with LTP emitter degeneration resistors integrated into the silicon, but in the classic "no free lunch" tradeoff there were downsides to these added resistors degrading DC performance, and adding noise (thermal noise of resistors effectively in series with input), etc.

Deane cleverly (he was a very clever guy) figured out he could substitute inductors for the LTP degeneration resistors to keep the good DC and LF noise performance, while enjoying improved slew rate and stability margin for HF performance.

Unfortunately we can not easily integrate inductors into op amp ICs.  :eek:  But by the mid-70s off the shelf op amps were more than fast enough for the relatively slow moving audio signals.

JR 
you can read Deane's notes in "Recording Engineer Producer", June 1978.
R1 has virtually no effect on the transformer,
neither is such series resistor found in time tested mic preamps of similar design (SSL).
 
JohnRoberts said:
Deane cleverly (he was a very clever guy) figured out he could substitute inductors for the LTP degeneration resistors to keep the good DC and LF noise performance, while enjoying improved slew rate and stability margin for HF performance.

Apparently he wasnt the first to use that trick http://www.proaudiodesignforum.com/forum/php/viewtopic.php?f=12&t=675

efinque said:
If by Studer you mean their tape recorders; I think you'll be lucky to find one in a studio these days.

I meant the consoles, specifically the 169 used this.
 
57sputnik said:
In answer to the OP, please refer to Jensen Application Note AN001. It requires a login to download.

Wow, that was your first post in 8 years!, I will check the application note you mentioned, its settled, it has to do with stabilizing the opamp then.
 
user 37518 said:
its settled, it has to do with stabilizing the opamp then.
Indeed, increasing the source impedance helps with stability, but actually the secondary impedance is high enough (even with the primary shorted) to make the additional 1k resistor unnecessary.
 
JohnRoberts said:
The LF 351 an early Bifet (similar to tl07x) so pretty much a different animal... Way lower input noise current but higher input noise voltage than bipolar input op amps (like 553x). So probably better for high impedance sources.

JR

Yes. Similar to TL07x as you say.  Nowhere near the Bipolar type NE553x characteristics but was available and relatively affordable.
 
Newmarket said:
Yes. Similar to TL07x as you say.  Nowhere near the Bipolar type NE553x characteristics but was available and relatively affordable.
Over the decades I used truckloads of both Bifet and 553x bipolars. They were both optimal for different tasks.

For audio they were fast enough (and quiet enough at modest gains). Modern uber op amps are better on paper, but not significant for many real world applications.

JR
 
The resistor is to keep the input impedance positive, ie not negative. Even the humble emitter follower can boast a negative input impedance, usually defeated by a few hundred ohms in series. In this case it is a non-invert opamp input; rigorous analysis will show negative input impedance which at some frequencies will combine with the 'nasty' transformer impedance characteristic to produce oscillation. Back when I was in time measurement we used an emitter follower to make a crystal oscillate.
 
JohnRoberts said:
Over the decades I used truckloads of both Bifet and 553x bipolars. They were both optimal for different tasks.

For audio they were fast enough (and quiet enough at modest gains). Modern uber op amps are better on paper, but not significant for many real world applications.

JR

Yes - when I first got a look at mixing desk circuits - DDA as it happens as involved in development work for automation etc on some of their desks - it was generally along the lines of 5534 where it matters (eg after the front end mic pre transistors) and TL07x elsewhere.
Both cost and current demand reasons ( 553x currents can add up in a large desk !) apart from circuit / impedance considerations.
I guess I'd be looking for something quieter now. For my DIY I have OPA2134.
In fact I recall I had some of the first OPA134 on my desk at Penny & Giles as we were in discussion with the TI guy about op amp options when they were released (derived from the more DC accurate OPA132).




 
Newmarket said:
I guess I'd be looking for something quieter now. For my DIY I have OPA2134.
In most cases (Zsource<50k), OPA2134 will not be quieter, with 8nV/sqrtHz compared to 3.5. The only justification I found for 2134 is for instrumentation, where the level of THD and the HF extension is an order of magnitude higher. Even for DI boxes, the difference with a pedestrian TL072 is marginal, and the price about 20dB more.
 
abbey road d enfer said:
, and the price about 20dB more.

I really love using electrical terms to describe everyday things. Back in the 70s when I was at Neve, my boss had a pretty loud voice. One day I brought in some sticky chewy toffees. I offered one to my boss and said " Here's a 20dB pad".

Cheers

Ian
 
ruffrecords said:
I really love using electrical terms to describe everyday things. Back in the 70s when I was at Neve, my boss had a pretty loud voice. One day I brought in some sticky chewy toffees. I offered one to my boss and said " Here's a 20dB pad".

Cheers

Ian
Many people do not realize how much life is related to exponentials/logs. Life itself is like a capacitor charging then discharging.
 
abbey road d enfer said:
Many people do not realize how much life is related to exponentials/logs. Life itself is like a capacitor charging then discharging.

I learned to hate programming 8 bit processors because life has more than 46dB of dynamic range.

Earthquakes and tornadoes are common examples of log/exponential metrics.

JR
 
abbey road d enfer said:
In most cases (Zsource<50k), OPA2134 will not be quieter, with 8nV/sqrtHz compared to 3.5. The only justification I found for 2134 is for instrumentation, where the level of THD and the HF extension is an order of magnitude higher. Even for DI boxes, the difference with a pedestrian TL072 is marginal, and the price about 20dB more.

Yeah - I think my post read ambiguously. I meant quieter wrt TL072 - not NE5534 (or more directly with NE5532 @ 5nV...)
For info - I used them (OPA134) in the front end of an ADC module using a Crystal Semiconductor device.
The equivalent Crystal circuit uses OPA627 but they were ruled out on the basis of cost as you might expect.
Measured performance of the circuit was a tiny (repeatable) bit (<1dB) below using OPA627 so with 3 opamps per channel = quite a saving in commercial terms.
IIRC discussion with TI technical advised that this was likely due to output Z remaining low at high frequencies thus better driving the ADC front end.
Other opamps tested were significantly worse in terms of measurement.
 

Latest posts

Back
Top