Marshall MG30DFX op-amps

[Hackinblack]

a new addition to the Obidia Trumpworthy home for badly behaved amplifiers

a slightly ill MG30DFX which wouldn't switch from clean to grubby
fixed that, but i am slightly puzzled,as the amp is full of TL072 op-amps;whereas the schematic i have (dated 2002) is full of 4558 DD's
(the high gain version of the 4558)

is this a factory revision, or has someone 'upgraded' it? anyone have a more up-to-date schematic to compare?

[J M Fahey]

RC4558 are an older design, TL072 a more modern one, but both are pin compatible and in the non critical world of guitars, practically interchangeable.

LONG ago I preferred 4558 for mic preamps, they had slightly less noise (hiss)  and thge slightly lower input impedance was not important, while Fet input TL072 had higher input impedance, great for Humbuckers.

But now 4558 are less available and TL072 became the workhorse.

[Roly]

the amp is full of TL072 op-amps;whereas the schematic i have (dated 2002) is full of 4558 DD's

Can't help good luck.   :dbtu:   


I must assume that you don't know that the gain of an op-amp in-circuit is determined by two resistors and would normally be set to x100 times or much less.  This is called the closed-loop gain.

The open-loop gain is typically x100,000 give or take.

When negative feedback is applied to an amplifier it improves the performance of that amplifier in several ways, noise, distortion, bandwidth, and output impedance, but at the expense of reducing the stage gain.  The improvement is related to the reduction in gain from the open-loop value to the closed-loop value.



It would be reasonable to assume that the higher the open-loop gain and the lower the closed-loop gain, the greater the improvement, and while this is strictly true, for gain reductions greater than about ten times you quickly run into the Law of Diminishing Returns.

When there is a large difference between the open-loop and closed-loop gains, say x100,000 to x100 or a 1000-fold reduction, even large changes in the open-loop gain have very little effect.  If we assume that a standard 4558 has an open-loop gain of x100,000 and the high gain "DD" version has an open-loop gain of x200,000, by the time feedback is applied in a practical circuit reducing the close-loop gain to only x100 the difference becomes miniscule, too small to measure, so the open-loop gain is of almost no interest at all to the circuit designer - a major benefit of using modern op-amps in the first place.

What becomes of far more interest is the inherent noise performance and where the gain starts to roll off with frequency, the dominant pole of the op-amp.  In these matters the 4558-series are seriously old hat compared to more modern op-amps and these days you can even do a lot better than the TL07x-series, however if the bandwidth is already wide enough (above the limit of hearing, never mind that most guitar speakers cut off around 5kHz) and the noise already low enough (inaudible) there is no practical reason to try and improve them.

For a full theoretical consideration see;
http://en.wikipedia.org/wiki/Negative_feedback_amplifier

HTH

[g1]

  There is no such thing as a "higher gain" version of a 4558 that is still labelled as 4558.  They are all high gain.  A version with a different gain would not be a 4558 anymore.  There is an upgraded version, 4559, then 4560, maybe more.
  If you can find a DD for sale, and they offer a datasheet, it is the standard 4558 or 4558D datasheet.
The Banzai info. is incorrect.

[Hackinblack]

Thanks for the replies...Sooo it looks like a factory mod :tu:
i did notice that the early MG series where criticised for excessive noise;maybe marshall get very good price break on 4558's

i'll be interested to play this amp and see how much difference the change of op-amp makes;my little MG15DFX is a pretty noisy amp as stock,a good mod was adding an metal screen between the mains transformer and the PCB;this alone halved the hum;then on to improving the filtering and smoothing on the supply rails

i noticed one poor bit of design on the '30 Watter;the O/P IC shares a heatsink with the main voltage regulator supplying the DFX card,any failure of the LM7805 would fry the DFX card.
i may run it into a dummy load and see if the PSU is adversely affacted by the power amp's heat..if it does i will seperate the siamese twins surgically :duh

[J M Fahey]

No need to.  :cheesy:
Trust me that it doesn't matter.
Too lazy on this nice Summer Sunday afternoon to write any longer.

[Enzo]

It's not a mod, it is a supply thing.   They designed it for common generic op amps.  The designer probably didn;t bother to spec one, knowing it would be one of any number of types they use.  On any given day the production line will need to have XXX number of common dual op amps in stock for the day's output.   They are not about to hold up production for 4558s if something else like TL072s are available.    In the amounts a company like Marshall buys, I doubt that cost was much of a difference.  But even 2 cents does matter.  One penny on 100,000 ICs is still $1000.  For us out here in the cold world, just about any op amp will work.


I am not aware of any heat problems with any of this series of amps.

[Roly]

any failure of the LM7805 would fry the DFX card

It would, but it's not going to fail due to being overheated by the heatsink because these three pin regulators are also thermally limited internally and if they get too hot they stop passing current and fold their output voltage back.  They are also very robust and hard to kill.