Fender Pro Junior hiss (question about potential source, seeking advice)

[soapfoot]

I'd speculate that the Pro Junior was conceived to be a somewhat-specialized tube overdrive generator.

At that, it excels... its palette seems to range from "lightly overdriven" to "raucous and crunchy," and it sounds pretty good within that relatively-narrow sonic window. It's definitely not trying to be all-things-to-all-people.

There are other amps in Fender's lineup that do pristine cleans very well. This one is a bit more specialized (hence the two knobs).

I absolutely love it for exactly what it does. It could stand to be a little bit quieter at idle, but people have probably said the same about me

 

[abbey road d enfer]

I think the Pro Junior was a Bruce Zinky design; Rivera had left Fender by that point

According to the schemo, the engineer was a J. S. Sams.
Zinky had been involved, I would think his name would be mentioned.
Blues Jr was by a D. Lewis.
Zinky is a famous advocate of "no global NFB".

 

[soapfoot]

According to the schemo, the engineer was a J. S. Sams.
Zinky had been involved, I would think his name would be mentioned.
Blues Jr was by a D. Lewis.
Zinky is a famous advocate of "no global NFB".

interesting—if incorrect (about Zinky designing the PJ) it’s probably another example of internet speculation that got repeated so many times it became accepted as “truth”

Although the first-version schematic I’ve seen for the Pro Junior didn’t have a name (that I could see). The earliest one I’ve found with a named engineer was Rev. C

 

[abbey road d enfer]

Probably another example of internet speculation that got repeated so many times it became accepted as “truth”

What? The "no global NFB" thing?

 

[soapfoot]

What? The "no global NFB" thing?

No, the notion (if incorrect) that Zinky designed the Pro Jr (post above edited for clarity)

 

[abbey road d enfer]

Ah, OK...

 

[ccaudle]

Zinky is a famous advocate of "no global NFB"

In general, or just for guitar amps? I would think that no global NFB would be a really good idea on a design where the intended use would drive one or more stages into sustained saturation.

 

[abbey road d enfer]

I think Zinky is a guitar amp designer foremost: at least he isn't famous for being involved in HiFi amps.
The idea is that in absence of global NFB, the progression of distortion is more gradual. It is known that distortion is part of a clean undistorted guitar sound.

 

[Valveamp 101]

I've got a mid 90's Pro Jr. that I replaced the filter caps with Sprague Atoms when I got it about 10 years ago. They barely fit but I had extras from another repair. I use this amp all the time for recording and have never had any noise or hiss issues, although I have read many an article and thread about it. Over the years I've worked on numerous Pro Jr's and the only hiss issue I've run into was from a tube. It would be interesting to dig into one that has issues like previously described.

 

[soapfoot]

“Too much hiss” is obviously subjective.

It’s clearly not objectionable to everyone

 

[murrayatuptown]

I am working on a later version Pro Jr... Oops, had to correct Freudian typo ProbJr...I think the PCB had a 2002 (c) on it.

His complaints, multiple types of noise including sound with obnoxious but low-level noises other than hiss at 0 volume setting, rapidly-failing output tubes (but he kept calling them 12AX7's...he's musician, not a tech...I assured him he could not be blowing up 12AX7's and had misidentified them.

I put 1 ohm WW resistors in series with the EL-84 cathodes to pin jacks on the rear panel to measure(derive) bias current, and a pot to make bias adjustable.

I found the amp to be the noisiest (hiss & amp gastric noises) amp I've encountered.

If you look at the earliest version (schematic- there seem to be 6 revs but 4(?) Series). The earliest had no cathode bypass capacitor on the first stage. 2nd schematic and beyond there is a 25 uF. You can also see the reference signal levels on the later schematics are higher after the bypassed first stage.

If you compare R6 on first and later rev. schematics, it was 56k and became 22k. Apparently they had to discard more gain the bypassed 1st stage produced.

I exchanged some emails with Ulrich Neumann, and he sent me some suggestions without having played or experimented on one.

He also commented on the strange gain management. He created a sim file of the preamp only and noted two dominant hiss sources...the 470k at the volume control, and the input 10k resistor, which is already relatively low for guitar amps. He suggested going as low as 1k. I asked about the RFI-protection attributed to that 'lambda network'. He has never had an RFI problem, so lowering it worked in LTspice...raise it if necessary.

He also remarked the tone control didn't do a lot...reminded him of a hi-fi-leaning Bode plot.

His suggestions:

Reduce R1 as discussed.

Change the 470k R5 and 22k R6 (56k on original) both to 100k to improve s/n. This requires an additional resistor before the PI to 'normalize' the increased signal level back to what was there (by design, desire or destiny-my comment). Lastly, change the 22 pF C2 to 3300 pF to give the tone control a tilt function.

The sim file duplicates the stock and proposed changes and the Bode plots of the 2nd preamp stage can be compared, via stepped pot values

I am accustomed to making one measurement at a time and calculating gain by hand...so I found the number of plot traces for stepped volume and tone controls a bit overwhelming. I personally could not make decisions based on the plots alone.

I haven't described where the fixed extra 'gain normalization' resistor is because I have to find the sim file to describe where it was. I put a smaller fixed value plus a trim pot on a piece of PCB.

At this point the amp was gutted on a bench in my garage and I was doing non-signal-related changes.

I began trying to isolate the multiple chassis grounds and began clip-leading them experimentally. Despite being non-ideal wiring I found quieter ground/return results and tried a 5751. The hum and noises got much worse! This was annoying and made me think there was more than just gain distribution to blame.

In the PJ Rev. evolution (based on photos I found on the internet) it seems Fender made 1-2 or more grounding changes. This one has a metal chassis standoff grounding the main PCB to the chassis roughly (physical location) between the input resistor and the front panel controls. Another version had a pigtail ground lead from somewhere (? based on a web photo) to chassis.

This specimen clearly has excessive grumbling & whining with volume at 0 (before I gutted it), so I think they didn't finish their homework.

I detoured to some other distractions like mounting 800V MKP 60+20+20 uF box caps on perfboard in a subchassis, changing the output transformer and adding a screen+ preamp B+ choke.

It all fits but there is minimal clearance from the upgrade speaker (he did it so I have to look to say what it is) magnet...success or failure TBD.

So I'm way out in the weeds on goals without even running a guitar thru it yet!

I am in exterminator mode now and decided to apply Merlin Blencowe's grounding techniques with shielded twisted-pair wiring...going to cut some ground traces when I am 110% confident where, install local 'ground'/return terminals, and determine how they will be star-grounded.

I found isolating the output jack from chassis also helped so there is something non-ideal in the grounding.

I uploaded some screen shot traces to a TDPRI website discussion on stock PJ simulation but there was little response and I didn't share the sim file.

If anyone would be interested in seeing it, I can share it.

A lot of unrelated personal things have happened it the last 2+ years so I haven't finished the project. The friend can't use the amp at home & is playing thru a PA at church so he's patiently waiting.

I'm ditching the isolated plastic input PCB jack and using an isolated Switchcraft type with STP wiring.

Murray

 

[abbey road d enfer]

Reduce R1 as discussed.

Won't change a thing.

Change the 470k R5 and 22k R6 (56k on original) both to 100k to improve s/n.

This would improve the input stage S/N, but the noise problem comes from teh fact taht the power amp section is very sensitive (16mV for full power)

This requires an additional resistor before the PI to 'normalize' the increased signal level back to what was there (by design, desire or destiny-my comment).

This is useless because the power stag egain is still too high.

I began trying to isolate the multiple chassis grounds and began clip-leading them experimentally. Despite being non-ideal wiring I found quieter ground/return results and tried a 5751. The hum and noises got much worse! This was annoying and made me think there was more than just gain distribution to blame.

There is some confusion here, because grounding affects only hum and buzz, not hiss.
If you want to make this amp into a low-noise clean amp, you need to change the gain of the output stage. The easiest way is to increase NFB, by decreasing R27. I would start with 47k, then try 22k. Too much NFB can result in oscillations, so you must be ready to pull the plug when you do it.

 

[rock soderstrom]

This would improve the input stage S/N, but the noise problem comes from teh fact taht the power amp section is very sensitive (16mV for full power)

This is useless because the power stag egain is still too high.

If you want to make this amp into a low-noise clean amp, you need to change the gain of the output stage. The easiest way is to increase NFB, by decreasing R27. I would start with 47k, then try 22k. Too much NFB can result in oscillations, so you must be ready to pull the plug when you do it

This is interesting because I have a similar problem. I don't quite understand how the power stage determines the amount of noise? I would have rather located that in the preamp. How is the noise generated? In my case there is no difference with or without NFB.


I don't want to hijack this thread but this amp has to much hiss for my taste. Maybe the have similiar issues?

 

[abbey road d enfer]

This is interesting because I have a similar problem. I don't quite understand how the power stage determines the amount of noise?

All stages generate noise. In that case, the stages at the front of the power amp generate noise.

In my case there is no difference with or without NFB.

The 22uF capacitor across R31 kills NFB. You may try disconnecting it, but the Master Volume will not act normally.
This amp has been poorly designed.
Including the Master Volume within an amp with global NFB is a beginner's mistake. When they found out the incompatibility, they decided to shunt NFB.

 

[murrayatuptown]

A R d e - not confused..I completely agree with you there are more issues later. The pot at zero...well, the volume is after the first stage, so too early to exonerate anything later in the amp. The trim gain 'normalizer' pot addition seemed like a demonstration of a 2nd volume or gain control, or just an attenuator (depending whether it stays inside...probably will)...

I don't remember my shorting grids to ground with a capacitor details... I don't think they told me the first two stages were to blame, but the tilt feature and suggestions so far seemed interesting & easy to do (since it was disassembled anyway). I did hear less noise in an external speaker as I adjusted the volume extra 'norm' pot, and that gave me hope.

I have read elsewhere the PI does have too much gain and the cathode acts like a second 'input'. It that needs justification, I'm not prepared...I read it on...you know...

I do want to explore making some differences in the PI, too.

My friend prefers a fairly clean sound & isn't a big fan of OD at this point.

His favorite amp sound is old Princeton Reverb...so there is some room for reducing gain along with any other cleanup and not have him be disappointed.

Thank you for pointing out how large a problem the subsequent stages contribute...not just something to consider. You're confirming it, which is helpful to me.

Murray

 

[murrayatuptown]

I definitely want to try more NFB (smaller R27).

 

[murrayatuptown]

C41 on the Laredo schematic sure looks like a weird typo...only allows (a little) ripple to pass instead of B+at least it's a 1 kV part )

 

[murrayatuptown]

Not the face I drew.

 

[murrayatuptown]

16 mV for full power...ugh. that might have been a bigger reason for short-lived EL84's than somewhat high bias.

This was a helpful visit for me.

I've done many electronics repairs (consumer & industrial) over the decades but making changes to valve amps has been on my list of things to pursue, and I depend heavily on books & advice.

Had I been (numerically) aware of the degree of sensitivity at the PA, I would have started there first...but he doesn't mind me experimenting.

Then maybe I'll continue finishing the incomplete amp projects I still have collecting dust since the mid-80's. I try to get rid of some but others appear & occupy the recovered space.

 

[abbey road d enfer]

A R d e - not confused..I completely agree with you there are more issues later. The pot at zero...well, the volume is after the first stage, so too early to exonerate anything later in the amp. The trim gain 'normalizer' pot addition seemed like a demonstration of a 2nd volume or gain control, or just an attenuator (depending whether it stays inside...probably will)...

It's clear that anything done to decrease gain between the grid of V1 to the output will decrease noise, as well as signal, so some compensation will be needed, like just crankng the volume pot up.
Indeed, a "master" volume installed brtween V1 and V2 would have about the same effect, and be capable of dialing in the amount of dirt.