Noob with a soldering iron

[Jungle-Jim]

To Roly, and Phatt.
Firstly, updating this, responding to Roly's previous posts first...
I have put 22pF ceramic caps in those three spots which had missing caps. It seems to have calmed the hiss down a lot - but not completely. I will email Award-Session shortly and see if I can get any info about why those caps were missing - because it clearly needed them - if I'm lucky I might be able to contact the designer Stewart Ward.

But - the hiss not completely gone, so I am interested in continuing to work on this amp, and use Phatt's recent posts.

Also - next time I have the amp on the work table I will also think about replacing those input 4 1N4148 diodes - I have some at home bought for another project. (Roly had previously suggested that could be one possible problem). I could test each one individually, but it seems a lot of work for components worth pennys. It'd be just quicker to replace the four and see how that goes, can't go wrong.


Hi Phatt
Thanks for your posts.

I'll go over what you're saying to clarify your points...

Firstly - you've made two separate suggestions - and I'm taking it that the second one would be the one to try first?

Going over the two posts:
In your first post, you suggest scaling down x10 the resistors around IC2b. You point out that a possible consequence may be an altering of tone - I have to say that I don't want to change the tone. With this amp I am after blues and jazz sounds - not heavy metal. I only ever have it on the edge of overdriving, and there's a sweetspot on Input B/Ch B where I can pull a nice soulful blues sound (Ch A is better if I want to cut through with a rocky lead break).

I don't really want to experiment with dropping various resistors and caps in - though I guess with those resistors being much lower, I could probably get away with leaving the old resistors in and tacking the lower ones across the underside of the board in parallel - that would be a way to test different values without doing too much component replacement (and obviously I'd properly replace the resistors if I liked the change).

Looking now at the second post...
It sounds like you're more convinced about this modification: you are suggesting bringing IC3b into the same cap/resistor combination as IC1a, IC1b, IC2b and IC3a - the 10k res becomes 1M, but the cap goes down from 1nF to 10pF (the others are all on 22pF, though Roly preferred 15pF on IC1a and IC1b).

Roly - have you got an opinion on either of these modifications? I'm more inclined to try the second one first.

And Phatt, you're saying that the mods you are talking about refer to Ch B, and that you didn't get time to look at Ch A. Well as it stands, though the amp still has some hiss, Ch B is better than A. Ch A is more hissy, particularly once the overdrive knob is over 3-4.

I need to get the amp on the work table this week, and try those mods to IC3b and report back. I will also order the parts needed for Roly's power regulation circuitry.

Thanks
J

[phatt]

Hi Jim, Ok you like the sound/tone and obviously you don't want to loose that.

Yes I'd try tweaking IC3b first just go up x10's so 1k becomes 10k and 10k becomes 100k.

**If the hiss goes way down then forget tweaking IC2b as you then know which part of the circuit is causing the hiss.**

The tricky part is the cap value, as I'm not good with maths I cheat and use simulations to get a close value and then back to the bench and test for real. 

Simulation reveals that with 10k and 100k around IC3b the cap value becomes 100pF and this gives identical output and tone response. Of course sims are not perfect so once you have a ball park number you poke a few different caps in the circuit until it sounds right to your ears.

Worth a mention here,
Be careful with the PCB as it's likely the only non replaceable part in the whole amplifier.
Yes you certainly can just parallel resistors for a lower value thereby not disturbing the pcb tracks which can be easy to damage.
When you do need to lift a C or R just lift one end and float it rendering it out of service until you establish the new value, one less chance of damage to board.

If you do the above and hiss is still an issue for Ch B Then maybe look at changing values around IC2b (the mixer section)

With IC2b you can leave the 22pF cap in place for now just lower those R values and see if the hiss goes down, worry about the cap later. my guess is that those large value series resistors are causing a fair amount of the hiss or at least needlessly adding to it.

Re IC3a (Channel A) that 22pF cap you note as missing will certainly help with hiss but I have a hunch there are deeper issues.

I've still got some metal work to finish my new Cruize Control project but I'll try to get some time this week to ponder the A channel.
Phil.

[Roly]

I could test each one individually, but it seems a lot of work for components worth pennys.

A reason for taking the long way around (and you only have to lift one end of each diode) is that you can potentially identify exactly what is causing your excessive hiss, and know for certain you have found it.

I didn't expect the roll-off caps to totally cure your hiss, some of it will still be in-band, but they will cut off bandwidth that your guitar doesn't need and is only amplifying hiss.

I have to say that I'm a bit surprised that you still have detectable hiss with LM833's at these low gain levels.  As Phil suggests, there may be something else we have overlooked.

Roly - have you got an opinion on either of these modifications?

No, this is Phil's dance, I'll sit this one out.   

[phatt]

 :lmao: Thanks Roly but you better keep an eye on me cause my Tango looks more like a Tangle these days.
Phil.

[Jungle-Jim]

Thanks a lot Phatt and Roly
I will put the amp on the bench and do those experiments this week. I suppose I am a bit reticent to try tests which involve a lot of "solder, connect speaker, power up, power down, disconnect speaker, soldering, power up power down" etc, because the speaker cable is so short, and it's difficult to test the amp without almost fully reassembling it. Plus this circuit board seems weaker than most - one or two solder/desolders on these pads they seem to collapse (and no the iron is not too hot). I guess I could extend the speaker wire - that'd make life easier.

Roly - with the diodes - yes it'd be nice to know which one, but I have an image of the current hiss levels from this fkn amp in my head, and if replacing the 4 diodes made a change, then I might not know which one it was, but at least I'd be able to tell the problem was in that area.

But thanks to both of you - this is getting somewhere. I still think it's a nice little amp, worth persisting with.
J

[phatt]

Hi Jim,
The amp works with or without those protection diodes so just lift one end of all four and see if it is the issue as Roly has loads of experience along with a nose for problem solving so I'd certainly check it out first. 8|
Oh dear I feel for you if tracks are de-laminating,, Yes I've run into that problem on older PCBs.
You may have to rig up a better way to hold/mount the PCB with a temporary bracket until you get it all sorted out. I've screwed blocks of wood to pcbs to hold them up and out for many days trying to sort out problems like this. worth going to the trouble as constant removal of circuits will usually lead to a bigger disaster.

Meantime;
Here is a screen shot of Rockette30 Chb (clean) First only modifying values around IC3b.
The two boxes down the bottom show both the freq response as well as the signal are identical in every way after the changes. The green traces (modified values) are so much the same it completely covers the yellow traces of the original values.

(IC2 on this simulation is IC3b on rockette 30 schematic)
The great part of this little sim prg is that you can reload several plots to see how the changes effect the outcome. You can't see the yellow trace but the labels are shown in the corner of each box to verify they are indeed present, (IC2Norm, written in yellow top left of each plot box)
Sadly you don't know how much noise will happen until you test it out for real which as mentioned I've already done so I can say with some certainty that this will at least help.

BTW While your little story has been running I've had a few hiss issues with my own circuit so it's been good fodder to make me rethink some of my circuit. :tu:
As of the last few days I've finally finished my circuit and indeed found the above mod solved one hiss problem the other was a layout problem which I won't go into here.

So I'm fairly confident that given mods to IC3b will work. If noise is still a problem then next logical
candidate is IC2b (mixer) but simulation shows that changing those values does change the response curve by a small amount but adding the *Missing cap* on IC1a brings it very close to the original response. (2nd screen shot mods 02)
This part *I have not tested for real* so I can't say if it solves ALL noise issues but should help a lot.

At this point IF the clean Chb is now noise free or at least greatly improved then it's a fair bet that ChA has similar problems, attack that one latter.

Note; Added or changed values are marked for you. :tu:
Phil.

[Roly]

because the speaker cable is so short

Client to service tech:
"Don't just stand their mumbling 'short circuit' - lengthen it!"

I guess I could extend the speaker wire - that'd make life easier.

Use your head to save your ... um ... posterior.


If I seem a bit picky it may be because I spent a lot of time doing industrial and mine safety equipment, then bio-med, ICU gear and the like, and when people's lives are hanging on the end of your work you need to be really sure because there is always a real possibility you may have to answer for what you did in the witness box.

An audience is less likely to scrag you if things go pear-shaped on stage, but having had that happen to me on a couple of occasions you may wish they did.  It's not "safety of life" but it's still pretty damn important if anything seriously interrupts the flow of the show.  You may never take this amp in front of an audience or studio mike, but I like to try and maintain high standards and best practice, if only out of habit.

[phatt]

Stop press,, I just tested the whole CHb for real on the bench and it's not making much hiss/fizz. :loco

The 10k/100k mod I mentioned at IC3b does help a little but somehow I now think something else is going on.
Bare with me ,,If I recall correctly Jim mentioned that other owners have No issues with hiss.

WotIf??? there is wrong banding on resistors?  I've had that happen to me once.
You measure a 10k resistor only to find it actually reads 100k.

So before you dig a deep hole that you can't climb out off,, how about some checking?
Most obvious candidates, 560k series R at IC2b.
if it's 56k instead of 560k then the hiss would be horrendous.

Same goes for both input stages, that 100k R that goes to ground through a 10nF cap if that is 10k the gain would go way up and hiss would certainly be evident.
R's are made in batches so if one 100k is wrong it's a fair bet all are wrong.

Love the dog Roly,,but can he dance?
Phil.   

[Jungle-Jim]

Thanks Phatt and Roly
Tomorrow night definitely I will get the amp on the bench, and get some readings and results and we'll see where it goes from there.

Thanks guys, this is brilliant, the amp is improving and I am learning an enormous amount along the way.
J

[Jungle-Jim]

Hi Phatt and Roly

I spent some time tonight on the Session amp...

Firstly - I disconnected one leg off each of the 4 input diodes - no change to the hiss - soldered them back.

Following that, I did nearly all the tests and changes as suggested by Phatt - and I will leave what I wrote below so you can see what I did, but now... after again being a bit flummoxed by the PCB and there being a few resistors which were different to spec, and me not being able to find components which should be there, I scratched my head, then did another significant tracing of the PCB, and discovered one more thing:

The schematic got another pair of op-amps swapped in the wrong IC. FFS. IC3a and IC2b - mislabelled each other. The guys who made this board must have had a bloody good laugh because this has certainly confused me. I thought I'd correctly identified all the op-amps the previous time I discovered the other pair had to be swapped. I was correct with them, but I guess I didn't dig far enough into the circuit to see the error with the second pair.

So I'll do another PDF of the schematic, and undo all the changes I've made to the board so far, and get back to you with the new revised schematic, and maybe Phatt, if you don't mind having a look at the revised schematic when I post it tomorrow, and applying your theories to the new layout. We might have to rethink this again.

Below is a record of what I did tonight, though I now know that some of it will be useless because apart from in IC1, I was reading some of the op-amps round the wrong way...

--------------------------
Now - looking at Phatt's suggestions...
Firstly the possibly incorrectly banded resistors...
1 Phatt: "Most obvious candidates, 560k series R at IC2b. if it's 56k instead of 560k then the hiss would be horrendous."

Me before discovering the IC swap: On IC2b - both the 560k and 270k R's on pin 6 of IC2 are 470k, the 1M is 1M. Is this significant that they used 470k for both those? Should I try swapping them?

Me, later after discovering the IC swap: F*ckin glad I didn't: I now know that IC2b and IC3a are swapped - these readings are consistent with what IC3a should be - so ignore what I wrote, other than it confirming the op-amp swap.

2 Phatt: "Same goes for both input stages, that 100k R that goes to ground through a 10nF cap if that is 10k the gain would go way up and hiss would certainly be evident."

Me: On IC1a - pin 2 has 100K and 1M - which is correct, pin 6, the same - no errors on either. (NB: These readings are fine, because IC1 is correct in the schematic).

So I went to IC3b (correctly renamed after the first schematic revision):

3 Phatt: "IC3b pin 6 has 2 R's and 1 Cap and I just tested this for real and it is a big noise issue.
Raising the value of both R's x 100 will deliver same gain and noise is down a lot, So 100k to ground and 1Meg between pins 6&7. You will need to change the 1nF Cap to say 10pF for the same freq response. Left at 1nF you will have no treble."

Me: IC3b - I substituted the 10k across pin 6&7 for a 1M, and the 1k to ground became 100k. The 1nF cap across 6&7 became a 22pF. There was no particular change in hiss. I can't test the amp at any volume right now because it's late and people are sleeping, but I can say the hiss characteristics are more or less identical.

Given I've changed the res's and caps on IC3b - and it didn't solve it - should I replace the original parts back in seeing as that wasn't the problem?

So - I'm not sure what's next - I was out of my depth ages ago, and just relying on you guys. This schematic being wrong *twice* has cost me a bit of time - though I guess it's been a lesson is having to trace a board and solve a puzzle.

Thanks
Jim

[phatt]

Oh yes trying to correlate actual pcb and schematic can be frustrating but you are doing well. :tu:

Can I suggest isolating the sections to narrow down the field, divide and conquer.
The mixer is the last stage before poweramp and it mixes 3 signals, those being the 2 channels and the reverb return path.
By lifting one end of those 3 resistors one at a time might give a clue as to which section is causing the most trouble.

The other point
By lifting one end of those mix resistors from ChA and Rev return (points D&R)
Leaves just the clean ChB in circuit and if hiss is still present then just focus on fixing that section first.

For ref, IC1A has a gain of about x10 as does IC3B while IC2B (Mix) has a gain of x2. Obviously the higher the gain the more chance of hiss but having had this circuit working with little hiss evident I can only assume a wrong value somewhere causing way to much gain in one or more of those early stages.

There is one other possible cause of hiss which I'm reluctant to mention simply because I'm still not sure as to why it happens but Roly might know more about this odd quirk which I stumbled upon while building my latest circuit.

Both Rockette inputs are the same and just maybe,, IF? those 100k going to ground via 10n caps are close to the input cap it could cause excess noise as that was my experience with my build.
It took me a week of hair tearing trying to find the source of excess hiss/fizz when I turned up the gain in my test circuit.

I then noted when touching that resistor/cap section on the neg leg there was a lot of interference. Sure enough it was right next to the input and just by moving that pathway to the opposite side of the test board it stopped the excess hash/hiss considerably.
But of course if that be an issue it then becomes a layout problem which is a lot harder to fix.
I was so glad I found this quirk Before I etched the pcb.

Without a clear picture of your PCB (both sides) it would be impossible to know if that is the cause.

Sorry i'm not much help here and one other thing bothers me which I've only just realized and that is Ch Switch (SW1). It would seem that when SW1 is set to ChA (ODrive on) signal actually would pass through both channels via that 1k and 4n7. Which does then explain those extra 2 diodes hanging off point C which are then engaged,, maybe to add more clip.

This would then allow all front panel knobs control over the sound.

So Q? Does that happen in use.

Meantime I'll have to sleep on it.
Phil.

[Roly]

I disconnected one leg off each of the 4 input diodes - no change to the hiss

Smart.  Now you have eliminated them as the source of your hiss you don't need to change them.

The schematic got another pair of op-amps swapped in the wrong IC. FFS.

"You never find just one roach in a kitchen".

Our assumptions always bring us undone - check everything.

should I replace the original parts back in seeing as that wasn't the problem?

It's not critical, but I would so I was dealing with a reasonably consistent situation.  The more things you change, the greater the chance of confusion (and it's bad enough here that the board doesn't match the circuit numbering).

With repairs I try to be a bit Kung-Foo - hit it once in exactly the right place.


Phil - stray coupling?  The CR order might make a difference if the C was connected to ground or the op-amp input.  You could try screening the cap by wrapping it in grounded alfoil.

{and no, he doesn't dance - and my printer doesn't play MP3's either.  You seen the dingo singing the ABC theme?  Cracks me up every time.  :lmao: }


{ed speelinge}

[Jungle-Jim]

Hi Phil, Roly, and all riveted onlookers....

Here's a new version of the schematic with corrected op-amp labelling. Note IC2 and IC3 are now completely swapped around.

I will post another thread shortly with a few photos of either side of the PCB - if there's anything to be gleaned from these - eg component placement etc.

I have restored the original cap and res's to IC3b. None of these changes are impacting the hiss either way.

But there's one thing I have to report, and I don't know if it's significant: it's to do with the guitar signal that comes out according to different switch and input settings... bear with me... (I can't say when this problem started, but some time prior to me attempting to fix this amp)

I ran these tests:

Guitar in input 1
Switch on ChA: - Vol A/OD responds ok (while adding hiss), Vol B does nothing (but doesn't add hiss).
Switch on ChB: - same behaviour as above - Vol A/OD works ok (with hiss), Vol B does nothing (but doesn't add hiss).
Switch in middle position: - Vol A/OD responds ok (while adding hiss), but Vol B, when wound in, introduces hiss but no guitar sound. (this is the questionable one I want to highlight... surely the second channel is supposed to be audible on this setting)

Guitar in input 2
Switch on ChA: - Vol A/OD responds ok (while adding hiss), Vol B does nothing (but doesn't add hiss).
Switch on ChB: - Vol A/OD does nothing but adds hiss, Vol B responds normally (while adding hiss).
Switch in middle position: - Vol A/OD responds ok (adding usual hiss), Vol B responds ok (adding usual hiss). (if both channels are responsive when the guitar is in input 2, switch in middle, why is that not the case when the guitar is in input 1, switch middle??)

So - a question mark there: admittedly this amp has a slightly confusing set up with the channels, but surely this is not doing what it's supposed to?

Phil - I will detach legs off those resisters to isolate those pre-amp sections later tonight if I get time.

Jim

[Roly]

First up, I'm surprised that SW1 has three positions - on the drawing it is shown as only having two.

Whatever, it selects the active channel by shunting (shorting) the unwanted channel, so in the middle position both channels should be active.

(this is the questionable one I want to highlight... surely the second channel is supposed to be audible on this setting)

I would have thought so from the circuit.

(if both channels are responsive when the guitar is in input 2, switch in middle, why is that not the case when the guitar is in input 1, switch middle??)

And that is a damn good question.

The primary reason would be the shorting contact on the lower input socket that grounds the join of the 1k and 470k when nothing is in the lower socket.

[Jungle-Jim]

Hi there

Can I suggest isolating the sections to narrow down the field, divide and conquer.
The mixer is the last stage before poweramp and it mixes 3 signals, those being the 2 channels and the reverb return path.
By lifting one end of those 3 resistors one at a time might give a clue as to which section is causing the most trouble.

The other point
By lifting one end of those mix resistors from ChA and Rev return (points D&R)
Leaves just the clean ChB in circuit and if hiss is still present then just focus on fixing that section first.

I have done this experiment - here's the findings...
There are the three tracks which come into IC3a - from the Clean, OD Channel and Reverb, each with a resistor. I desoldered one leg of these resistors to disconnect these in this sequence...

Just the Reverb circuit removed (so OD and Clean)....
No change - same hiss as usual.

Reverb and OD removed, so just Clean channel...
Quieter, but some some hiss still present.

Rev, OD and Clean all removed...
Quiet, same as if the Return jack was shorted, as you would expect.

Reverb only...
Not totally silent, but acceptable levels of hiss given that it doesn't get noisy till the knob is 6/10.

Reverb and OD channel only...
Significant noise from the OD channel, on the OD and Vol A knobs, once they are past 3/10.

Reverb and Clear only...
Quieter than 'Rev and OD only', but still there's enough hiss to be annoying past 4/10 on the Vol B knob. Not quiet enough to record with. Still more noise than my other little Yamaha JX20 amp at comparible volumes.

So as it stands I have left it with the Clean and Reverb channels still connected, the res on the OD channel still disconnected. The Reverb channel isn't a problem.

The thing is that the Clean channel is just on the wrong side of being acceptable for hiss - just a bit less might be ok, meanwhlie, the OD channel has a worser hiss problem again.

Where do I go from here?

Is there a clue in that switch setting doing something a bit unexpected? Maybe not. What about that other idea mentioned about shielding the caps at the input stages.
Thanks guys for looking at this.
Jim