Fender FR-10/FR-12 noise reduction mod

[aflynt]

That's great to know!

PS, i (sort of) managed to reproduce the issue you mention, but on my amp i needed to crank the volume all the way up and set tone knobs to their maximum values; the chirping sound is then faint, but audible. My running theory is this is switching noise from the power supply, somehow getting amplified.

A q&d mitigation would be to change the power-in decoupling caps (100uF 16v) to 470uF and see if it improves things...

View attachment 13277

...followed by the decoupling caps for the opamps themselves.

I did generally have to have the treble control fairly high to trigger the weird noise issue. There would be ranges on one or more of the other controls that would make it much worse. This was all with the volume off. The common mode rejection ratio rating for these chips does seem to correlate, with the TL084 being the lowest, the OPA4228 being the highest and the OPA1604 somewhere in between. Thank you very much for the mitigation suggestion! I'm going to keep mine as is for now, though as I'm pretty happy with the hiss level and sound.

-Aaron

 

[lolainnb]

How is this video for learning? Is the method ok?

 

[Lysander]

I did generally have to have the treble control fairly high to trigger the weird noise issue. There would be ranges on one or more of the other controls that would make it much worse. This was all with the volume off.

Rats I've just realized that the reason OPA4228 has a much higher slew rate than OPA4227 is that it is NOT unity-gain stable - in fact, it's optimized for gains way above what the FR-12 preamp is set up for.

So yeah, this would explain instabilities when replacing the first TL084, which handles the treble, middle and bass controls. Some EQ settings could definitely make the thing oscillate.

Most of the gain seems to happen on the next TL084s though, so if your already bought OPA4228s, just keep the leftmost TL084 in and replace only the next two, as suggested by @aflynt. This should be enough to kill most hiss.

 

[aflynt]

Rats I've just realized that the reason OPA4228 has a much higher slew rate than OPA4227 is that it is NOT unity-gain stable - in fact, it's optimized for gains way above what the FR-12 preamp is set up for.

View attachment 13316

So yeah, this would explain instabilities when replacing the first TL084, which handles the treble, middle and bass controls. Some EQ settings could definitely make the thing oscillate.

Most of the gain seems to happen on the next TL084s though, so if your already bought OPA4228s, just keep the leftmost TL084 in and replace only the next two, as suggested by @aflynt. This should be enough to kill most hiss.

Interesting. That would likely explain it, yes. So what do the other 8 gain stages from the other quad op-amps do (besides generate hiss)? Why aren’t they affected? I may grab some OPA4887s and OPA1604s and adapters to test out.

-Aaron

 

[quasimono]

Another mod success here!

Howdy folks, I received my order from Mouser last weekend and did the preamp mod too.
I'm by no means a professional solder jockey, but do have years of experience in building fx pedals and a tube amp head.
So, with a little bit of knowledge, experience in soldering and patience, the mod was painless and relatively easy to do.

I used three OPA1604AIDR SMD ICs + adapter as suggested by @Lysander to replace the original TL084s.
(Preparing/soldering the adapters and the SMD ICs was the hardest part to be honest...)
For replacing the RC4558s I had found two older OPA2134s in my parts stash so I socketed and used these:

It's a bit tight with the adapters over the existing parts, but if you carefully bow the ceramic and film caps down a bit, the adapter pins fit in just snuggly:

Along with the IC replacements I also did some updates to the input stage as proposed by WKSmith on the fractal board, mainly changing four caps to better quality box film caps and changing two resistors to higher quality metal film with the following two value changes for the non-inverting amp stage: 1K for R2 and 6.2k for R13 to set optimal gain for the OPA2134 IC I used as a replacement.

The cap changes were:
- C12, C13 use .22uf
- Output coupling caps (C14, C2) use 1uf
I used Wima MKP 2 caps, which are great, and red, but also a tad bit large... ;-)
I had no problem fitting them on the board though, the spacing is large enough to accomodate the "not so little buggers"...

So, what's the result of a few hours of work?
The hiss definitely is a lot quieter than before, but it's still there and audible when you are standing close to the amp.
I have no weird phasing or oscillating issues with my replacements too.

One thing I noticed on my amp is that not the loudness, but the "sound" or "tone" of the hiss changes when using the mid and treble controls on the preamp: They both add more pronounced hiss in their respective frequency range, which further goes to show that the hiss definitely is a product of the preamp (design).

I had initially hoped for even more hiss reduction from reading @Lysander 's initial post, but take a few steps back and you can't really hear it anymore. And it's a total non-issue once you start playing anyways! This thing sounds really great!

Overall I'm happy with the result, so massive thank you m8 to @Lysander !
I will still follow WKSmith and his preamp update(s) and see what he will eventually come up with to maybe further better this already awesome product!

Cheers folks!

 

[Lysander]

Interesting. That would likely explain it, yes. So what do the other 8 gain stages from the other quad op-amps do (besides generate hiss)? Why aren’t they affected? I may grab some OPA4887s and OPA1604s and adapters to test out.

Haven't fully traced down that section, but looks to me mostly as gyrators/filters related to the high cut control. Why would you ever need EIGHT OPAMP STAGES for that is beyond me (and a question for Fender - there's even a few transistors in there!), but stability will be a result of the preamp design: with proper gain and/or feedback setup, even the very fast OPA4228 can be perfectly stable.

The first TL084 runs the 3-band EQ, which is all active first order (+/- 6dB). Meaning, the opamp(s) will be running close to unity on most EQ settings, and it looks like that section's design is not as forgiving for the IC.

 

[aflynt]

I’m wondering if these would be a better adapter socket for the OPA1604s?

https://www.mouser.com/ProductDetail/Chip-Quik/PA0003C?qs=t7xnP681wgVpGgny%2Bv3YdQ==

-Aaron

 

[quasimono]

I’m wondering if these would be a better adapter socket for the OPA1604s?

https://www.mouser.com/ProductDetail/Chip-Quik/PA0003C?qs=t7xnP681wgVpGgny%2Bv3YdQ==

-Aaron

Yes, definitely! I saw these too, but my stingy self just wasn't ready to shell out over 6€ a piece...

 

[Lysander]

I’m wondering if these would be a better adapter socket for the OPA1604s?

They're all a bit of a pain to work with anyway but, yeah, those would fit just fine.

I used ChipQuik adapters on my FR-12, at ~$3.20 US each. These are slightly wider than the DIP socket, and require the SOIC parts to be soldered on top, and the pins in the bottom. They also won't socket properly and require soldering but, other than that, a good choice IMHO.

Part numbers are PA0001 (DIP-8) and PA0004 (DIP-14) in Mouser.

 

[vtgearhead]

How is this video for learning? Is the method ok?

As someone who repairs and restores vintage computer hardware I had to avert my eyes from that heat gun operation. Well intended, but in my opinion it's just begging for disaster. Instead use good quality solder wick (with liquid flux if needed) to soak up the stubborn residual, then gently wiggle the device pins from the bottom with needlenose pliers to break loose any "whiskers".

If you plan to do a significant amount of through-hole rework, pickup a Hakko FR-301. It's the best vacuum tool I've ever used and will clear just about anything, including pins that connect to a ground plane.

 

[quasimono]

Well, it's not easy to stop once you start fiddling around...
Having socketed the two OPA2134 I used to replace the original RC4558 with now, I'm thinking about doing a bit of IC rolling with these two...
Question to the knowledgeable folks: Would the OPA1612 also be an equally good fit as the mentioned OPA1656 (as used by @Lysander )?
It's not a FET-input like the OPA1656, but a bipolar-input device, but other than those fancy words, I don't know what to make of this information TBH...
I might be able to obtain one of each (1612 + 1656), and given that, which IC would you place at what position in the preamp board?
Thank you!

 

[Lysander]

Would the OPA1612 also be an equally good fit as the mentioned OPA1656 Lysander?

tl;dr; if you're replacing the RC4558s in your FR-10/12, you should use a JFET opamp at least for the differential input section (leftmost chip on the board). Bipolar opamps much lower input impedance, and hence more input bias current, than JFET counterparts. The combo line-in/XLR input in these amps connect pretty much directly to the opamp, so you definitely want to keep its input impedance as high as possible.

So yeah, i'd definitely recommend OPA1656 (or any other quality FET opamp) over the OPA1612. The main benefits OPA1612 has over the former is being cheaper, and having a lower noise floor - but that won't make any practical difference in this application.

The second RC4558 is just a crossover filter to split the preamp's signal between the main and tweeter amplifiers, so opamp selection is not as critical; pretty much any pin-compatible part will do the job. On my amp i used OPA1656 only because i couldn't bother to buy two different new parts

 

[quasimono]

Thanks @Lysander , I just found a european vendor who sells OPA1656 already soldered on nice little DIP-8 adapters and bought two.
So there's that.

Now I only need to recalculate my R13 / R2 values and if I should change them to something else than what I currently have for the OPA2134 (R13 = 6.19k, R2 = 1k) to accomodate the OPA1656s...
To my understanding, the OPA2134 has a max supply voltage of +/- .7V 36V (34.6V) vs. +/- .5V 40V (39V) for the OPA1656.
Which would mean I could raise the gain setting a bit? (1 + R13/R2)

 

[Lysander]

Now I only need to recalculate my R13 / R2 values and if I should change them to something else than what I currently have for the OPA2134 (R13 = 6.19k, R2 = 1k) to accomodate the OPA1656s...

Don't bother. Both OPA* opamps will take rail-to-rail inputs, and OPA1656 is also rated for higher supply voltages than either OPA2134 and RC4558 (40V). It will work as a straight drop-in replacement, without any other part changes.

I didn't need any gain adjustments on my amp using that IC.

 

[lolainnb]

 

[lolainnb]

Seriously guys. You need to start recording video of your mods. I’d subscribe and support

 

[laxu]

Charts and graphs, people!

 

[Lysander]

Seriously guys. You need to start recording video of your mods. I’d subscribe and support

Never even crossed my mind, tbh. All drama aside, it all boils down to desoldering an IC, and soldering in a replacement.

I'll set up a GoPro or something next time i crack open my FR-12.

 

[quasimono]

Charts and graphs, people!

*installs PowerPoint..*

 

[macula56]

@Lysander and @aflynt So I read thru the whole thread. IC's are not my strong suit so a little clarification would be great. The OPA4228PA and 4227PA are both available in DIP packaging. Which would actually be best for this mod?