Upgrading 4558's

[SonsOfThunder]

Nacho,

Here is my stab at your issue. I would only do the following to a single channel and then listen to it before proceeding. If you have access to a scope, it would be a good idea to use it to check for oscillations. I rather doubt that osc will occur using a 4558 but I'm just trying to cover all the bases here.

I would remove C9/13/15/32/35/37. I would then remove or replace the following with maybe 10pF to 22pF each: C10/16/33/38.

If you aren't concerned with the EQ performance, then the only OAs you really need to replace are M5 and M6. I could not find a thing on M1, the first OA (PC1010) is it a little potted OA or something?

The other mod that #might# help would be to add in parallel each of the series electrolytics with a 1uF film cap.

HTH!

 

[nacho459]

thanks, I'm gonna put in an order for a bunch of op amps and caps. As far as the oscillation problem, is it from the feedback loops?

The 1010 amp is this guy

And no one knows what it is comparable to.

Thanks again, if anyone else has any ideas don't hesitate.

 

[SonsOfThunder]

[quote author="nacho459"]As far as the oscillation problem, is it from the feedback loops?[/quote]
Too much bandwidth and/or not enough power supply bypassing...

The 1010 amp is this guy... And no one knows what it is comparable to.

I remember this now from another thread. Is the pinout the same as the JH990, etc? It might not be that bad anyway.

HTH!
Charlie

 

[nacho459]

OK, so as far as oscillation, the loop would feed back too much and draw too much current, feedback like a mic in a PA system? I think I kinda understand what you are saying. I understand negative feedback like in a guitar amp where you can adjust the tone and whatnot. Is this the same kind of thing? But we're trying to make sure the feedback doesn't get too strong where it starts to build up? By lowering the value down to 10pF were limiting the bandwidth of the feedback loop, right? I think I'm starting to get this.

[quote author="SonsOfThunder"]The other mod that #might# help would be to add in parallel each of the series electrolytics with a 1uF film cap.[/quote]

I'm not quite sure what you mean by series electrolytics? Do you mean two electrolytics together in series?

 

[SonsOfThunder]

[quote author="nacho459"]OK, so as far as oscillation,[/quote]
Smaller cap widens the BW. The cap looks like a short at high freq and lowers the gain at those freqs and above. For instance, the first OA has C10=470pF, R18=4.7k using 1/(2*pi*RC) gives a cutoff of 72kHz. If we change to 22pF, we get ~1.5MHz, though the OA won't go that high at the set gain of 5.

I'm not quite sure what you mean by series electrolytics? Do you mean two electrolytics together in series?

Nope, add a film cap across C2/6/11/12/14/36/34/39 etc. See?

HTH!

 

[Bryson]

What model TOA is this?

 

[nacho459]

ok, I kinda understand the bandwidth thing, but why would I need the bandwidth to go all to way up to 1.5MHz?

And as far as the series caps, these are there to filter out the DC from the audio signal, right? Why wouldn't I just want to swap them out with a film cap instead of just paralleling them? I'm thinking of them like an orange drop or other poly cap in a tube amp that keeps the B+ out of the audio signal.

 

[nacho459]

[quote author="Bryson"]What model TOA is this?[/quote]

an RX-7

 

[SonsOfThunder]

[quote author="nacho459"]...why would I need the bandwidth to go to 1.5MHz?[/quote]The better question becomes, "Why do you need to limit it at all?" Those limits are cumulative and it also adds to the capacitance that the OA has to drive. "If you don't need it, why put it in?" Thing is that the manufacturer has no control over the environment you are using the product in, so they are going to plan for the absolute worst case, even if it hurts performance some, so that the unit will work.

the series caps...filter out the DC... Why wouldn't I swap them with a film cap instead.

You could... but its not so easy to find or fit a 22 or 33uF film cap in those locations. Most electrolytics don't have really good high freq response and so its just really quick and easy to tack on a 1uF film. You can certainly do otherwise, this was just a suggestion. As always, try it on one channel and then LISTEN for yourself and decide.

Peace!

 

[nacho459]

[quote author="SonsOfThunder"][quote author="nacho459"]...why would I need the bandwidth to go to 1.5MHz?[/quote]The better question becomes, "Why do you need to limit it at all?" Those limits are cumulative and it also adds to the capacitance that the OA has to drive. "If you don't need it, why put it in?" Thing is that the manufacturer has no control over the environment you are using the product in, so they are going to plan for the absolute worst case, even if it hurts performance some, so that the unit will work.[/quote]

Hmm, so what is the benefit of limiting the bandwidth, does it cut down on the power consumption or something?

 

[SonsOfThunder]

Sorry Nacho, I just saw this question.

The answer is to prevent oscillation and/or RF interference from getting thru. This thing is totally OVER-protected. Its a similar situation to my Yammy MC1604. I tried replacing some 4558s with LM833 with no success in increasing the BW, because I didn't take the time to remove all of these stupid caps. It has 150pF across the OA inputs and 220pF across 13k feedback resistors.

HTH!