Circuit link below. Sorry for the poor quality. This amp was not popular and this is the best I've been able to find.
Background: I have some basic understanding of SS components and circuits. But I'm really a newb. Trying to increase my knowledge and perhaps mod this amp.
Signal comes in and through IC A / 4558 and moderated by the gain control POT 1. It then flows down into the other half of of the 4558, IC B (IC B is switchable by panel or footswitch through Q1). From there, the signal runs past a pair of Red LED clipping diodes LD1 & LD2. But, they are unswitched/directly grounded.
Are they always clipping? Or are the Red LEDs chosen because they only conduct over a higher gain signal coming through?
Possible Mod: What I'm thinking is that if the diodes themselves had a switch to ground that I might be able to get additional clean headroom with the primary gain control. But, I'm also thinking that by running the signal hotter (no clipping at all) that perhaps it is going to overdrive the poweramp... and maybe there's not much clean headroom to be added because of that (?). Perhaps the clipping is a "required limiter" or sorts? Any danger to the other circuit components by switch-controlling the LED's?
Also, there's a capacitor C9 in parallel with the LEDs. Is it shown there simply for circuit drawing convenience in the schematic? Or would it need to be switched along with the LED's if I switch them?
EDIT: Schematic cleaned and waxed: Zoomable circuit diagram: https://i.imgur.com/WjfO8n1.jpg
(https://i.imgur.com/WjfO8n1.jpg)
Well, before deciding what they do and designing a switch circuit. If you wonder what they do, unsolder one end of each and lift it out of the circuit. Now you'll know.
An LED has a junction voltage. Seems to me they start at something over 1 volt, and like blue ones up to well over 2v maybe even 3v. I don't mess with them so I am not up to date on current LED types. But until you hit that voltage, they don't conduct, so they only will clip when your signal level gets up to the 1.2v or whatever it is in your case.
Just my opinion, but I suspect they are not there as clippers, they are there as limiters. I could be wrong.
Thanks for the insight.
I'm OK with lifting them to test the sound... I'm just looking for confirmation that that's an OK thing to do. If they are limiters to subdue the signal of the dual 4558 amps and I'm going to fry the 1875 power amp by letting too large a current in, then I clearly wouldn't want to do it.
Also, I'm not sure what to do with C9 - leave it or lift it with the LED's. I don't know enough to understand, but I know enough to ask first :-).
They limit the signal by not letting it get larger than the LED drop. They don't act as an attenuator. Like your car seatbelt limits you to not going through the windshield in a crash. The only time they have an effect is on peaks. They will lop off peaks when they get too tall.
In my opinion, and it is only my opinion based on my understanding.
If you ignore the LED's,R13 and C9 would be a low pass filter, the LED's will clip the signal when it reaches the voltage to make the LED's conduct.
Look at the LED's while using the amp, full gain and the switch on overdrive and if the LED's light up its working as a clipper.
Q1 changes the gain of the op-amp, it puts R5 in parallel with R6.
If you want more clean head-room, you could add a germanium diode in series with each LED.
That would slightly increase the clipping voltage.
You could try different led's or different diodes in series with the led's to experiment until you are happy with the sound.
I think removing the LED's might not be a good idea, too much signal might get through.
Which amp model do you have ?
It appears that there might be 2 models using that circuit, the watts available depends on which power transformer is used.
I doubt you will blow the power amp chip by removing the leds.
I tend to think these little chips blow because quite often the modern metal shred players use very high out put pu's and thrash these chips to death. Often adding pedals to make it even louder.
Add to this it's a small amp and more often than not they have tiny heat sinks which are not up to task, so the combination of above will fry the chip.
At sane bedroom levels which these small amp are designed for you won't have issues.
So As Enzo mentioned lift the 2 Leds (or one at a time) and you will get more clean head room but as it's a small power chip it may not be much more because the power section will distort.
you can if you wish add a pot in series with the leds and that will allow you to dial in how early the leds clip.
(Leave C9 bridged from R13 to ground)
You could try better speakers but that will get expensive and still may not improve the SPL by much.
If you want more clean head room then you will need a bigger amp. Sorry
-------
Regards the C9 Q;
The cap is there to wipe off excess hifreq crud.
Valve amps wipe of hi freq by design while SS amps you have to add some hi freq roll off.
A lot of budget small amps don't bother to tweak this issue and hence sound very harsh Especially when in distortion mode.
Again lift it and hear the difference.
You will certainly have heaps more treble.
If the Amp is already bright then Leave it in place and just lift the Leds.
If you want, you can try different values on C9 to fine tune the top end. (lower values = more treble. higher = less bright)
One thing that may happen when you remove leds and engage the distortion.
The signal going to the reverb may start to distort the reverb drive. so you might have to raise the value of R201.
Just a note on the schematic;
Unless I'm missing something??? IC1B is labeled wrong.
Pin 5 is the Positive input and is normally grounded in this configuration,, It looks like it's + & - input labels are flipped. I can't read the label Numbers,, they both read 5 to me.
Hope it helps
Phil.
Thanks all.
Makes sense. I had a home-built distortion pedal I made that had a pot just like that... controlling the amount of clipping. I'll probably try that at least as a test to find the balance between clipping and power amp distortion.
Thoughts on the proper value for the pot to reduce LED throughput to zero? 10K? 100K? 250K? I know I can just head to the big side, but if ya'll have a suggestion, I'll start there.
Thanks for the insight on C9. I thought it might be a filter with that resistor, but I'm weak in that regard... and I definitely don't understand that tone stack. Different than anything I've seen or in Duncan's Tone Stack Calculator. Also, I am familiar with the "harsh treble" issue for SS amps! All to common. This one works though - the amp has very nice tone / range.
You may only need a low value pot for the leds ~ even 1k would likely be enough but use the lowest value you have on hand.
Yes odd ball tone design but if it works then leave it.
Phil.
This might help a little bit:
I just don't think they are there as an effect. Just don't want to overdrive the reverb or whatever. The power amp can only put out what the power supply rails allow. Drive it too hard and it just clips. I do agree with insufficient heatsinking.
Lift them to see what difference it makes, if any. No point in designing a circuit modification if it doesn't do anything. Find out.
The clipping LEDs make the amp a little more bedroom friendly. When operated with the Master Volume turned down so as not to allow the power amp chip to clip, the LEDs give the effect of more preamp gain and allow the Master Volume to be turned up a little farther past the point where the volume jumps from nothing to too loud.
Quote from: phatt on April 06, 2022, 09:42:26 PM...
Just a note on the schematic;
Unless I'm missing something??? IC1B is labeled wrong.
Pin 5 is the Positive input and is normally grounded in this configuration,, It looks like it's + & - input labels are flipped. I can't read the label Numbers,, they both read 5 to me.
Phil.
Yeah, this is actually the schematic from the manufacturer too. LOL.
From what I see, you are correct. Pins 3 & 5 are positive inputs. Pins 2 & 6 head to ground.
You're saying they'd normally flip the positive the other way when using both halves of the 4558? I'll look when I'm in there in a few days to see how it's really wired.
A better quality schematic for the Epiphone Regent 20,50R is available from
https://gibson.jp/support/schematics-and-manuals
direct link -
http://images.gibson.com/Lifestyle/Support/Files/Schematics/Regent20,50R.jpg
The schematic is readable - IC1B 4558 is pin5 positive and pin 6 negative.
The circuit is drawn right but labeled wrong, just to clarify a pic may help;
Wronglabels on opamp.jpg
Quote from: SemiConductive on April 07, 2022, 12:06:30 PMYeah, this is actually the schematic from the manufacturer too. LOL.
From what I see, you are correct. Pins 3 & 5 are positive inputs. Pins 2 & 6 head to ground.
You're saying they'd normally flip the positive the other way when using both halves of the 4558? I'll look when I'm in there in a few days to see how it's really wired.
Pins 2&6 are Neg inputs,, just don't assume that because they are Negative they go to ground.
As a rule of thumb the Positive input always needs a ground reference, often a resistor.
If you notice the Pos on IC1A has a resistor to Ground/Common.
While IC1B the Pos input goes directly to Ground/Common.
These chips are nearly all differential inputs, IC1A is non inverting while IC1B is wired as an inverting stage. both need the Pos input to have a DC path to Common.
Phil.
Thanks. I'll ask you one question, as I'll need to study a bit to ask more intelligent ones. (I'm really light when we get beyond basics and 'following directions').
Why do we invert at IC1B? It it because "that's the way a 4558 rolls"? Or, is it to accommodate what the power amp wants to see and/or another inversion happening there?
Update on the hands-on side: Successful test!
I lifted the diodes from ground and put a toggle in there so I could do back-to-back testing.
The diodes were definitely clipping at fairly low signal levels... as soon as the gain pot was set to above minimal. By disconnecting them with the switch, I can get a lot more clean headroom, which is what my goal is for this amp... and as expected.
Turning on the "overdrive" switch in the panel and bringing IC1-B into play bumps the volume fairly cleanly. But with that on and the gain turned up high you still get into distortion territory, again as expected.
There is a little bit of a treble cut when you switch on IC1-B, even at low levels where it stays very clean. Very slight cut, just taking off the upper edge a little. Not unpleasant but a minor change. Perhaps the resistors and caps hanging above IC1-B are doing some filtering when it's on? Or maybe it's the opposite... there's filtering when the signal runs fully through the resistors/caps above with IC1-B shut off and turning it on eliminates that? I'm a little confused by the fact that the caps appear to be "in-line" (not going to ground) and I would think they'd filter bass, if anything, in that configuration.
I stand corrected.
Quote from: Enzo on April 10, 2022, 03:54:24 PMI stand corrected.
Well that's a rare event for you :D
Phil.
Quote from: SemiConductive on April 10, 2022, 02:47:27 PMUpdate on the hands-on side: Successful test!
I lifted the diodes from ground and put a toggle in there so I could do back-to-back testing.
The diodes were definitely clipping at fairly low signal levels... as soon as the gain pot was set to above minimal. By disconnecting them with the switch, I can get a lot more clean headroom, which is what my goal is for this amp... and as expected.
Turning on the "overdrive" switch in the panel and bringing IC1-B into play bumps the volume fairly cleanly. But with that on and the gain turned up high you still get into distortion territory, again as expected.
There is a little bit of a treble cut when you switch on IC1-B, even at low levels where it stays very clean. Very slight cut, just taking off the upper edge a little. Not unpleasant but a minor change. Perhaps the resistors and caps hanging above IC1-B are doing some filtering when it's on? Or maybe it's the opposite... there's filtering when the signal runs fully through the resistors/caps above with IC1-B shut off and turning it on eliminates that? I'm a little confused by the fact that the caps appear to be "in-line" (not going to ground) and I would think they'd filter bass, if anything, in that configuration.
Good to hear,
Yes at higher levels the power chip will distort,, as well the power rails might be sagging a little bit due to the bigger current draw.
You can add a small value pot in series with the diodes if you want more control.
Re R5,R6 and C6;
These are not in the signal path, R sets the gain of the stage and C sets the hiFreq roll off.
you could lower the value of C6 which may balance it out a bit better but if you want the same freq response then you would have to switch the cap value to match the resistor change.
Likely not worth messing around with it.
Just search with *Opamp basics* or something similar as there is a stack of info on the web.
Here is a site with a massive amount of info on audio electronics.
https://www.sound-au.com/index.html
Click on *Articles* and search those for info.
Some of the *Projects* also have a good explanation of how they work.
Rod has written this site for folks like you (and Me :-[ ) with limited knowledge.
Phil.
Trying to understand what you said about R5/6 & C6 not being in the signal path...
So, they are doing some filtering, but the "signal" passes through pin 5(6) of IC1-B? And R5/6 & C6 are not "signal path"?
What I'm a little confused about is how the filtering is operating. I understand how an R/C network filters off treble e.g. a "tone control" (capacitor is grounded, bleeds off high frequencies, resistor limits bleed to cap). But in that scenario, the cap is always bleeding treble to ground. In this case it looks like it's bleeding it around IC1-B (?).
I'm also familiar with using a small cap to filter bass... and that looks a lot like the arrangement of R6 & C6, but again, turning on IC1-B filters more treble, not more bass. So if I carry that logic through, is R6 & C6 filtering (more) bass around IC1-B? So it's not reducing treble, it's adding bass?
Or have I missed this entirely - and R5 is involved in the filtering too... and somehow the grounding through Q1 is pulling off treble? I thought the Q1 circuit was strictly handling the foot (overdrive) switch. Does it do something else?
Thanks,
I said,
Re R5,R6 and C6;
These are not in the signal path, R sets the gain of the stage and C sets the hiFreq roll off.
*This is the FEEDBACK path Not Signal path.*
You said:
" In this case it looks like it's bleeding it around IC1-B (?). "
YES Correct ;)
C6 basically dead shorts hi freq (set by the value of Cap)
So the stage does not amplify above a set freq.
Higher values will wipe off at a lower freq.
Valve Amps used to have a Presence control built into the power amp stage.
Simply by adjusting a pot which progressively added or subtracted a Cap in the FEEDBACK path, More cap less treble.
Nothing magic Just another place to alter tone.
------------------
If the tone knobs can't give you good results then it's a matter of tweaking what I call *Whole System Tone*
The feedback path is often used to do this.
Side note;
I did try to simulate your tone control setup but it does not seem to work well.
I can't read some of the values so I guessing.
If you can give me the right values I maybe able to work it out.
--------------
Something that is often missed ,,well at least I did when learning how this stuff works, LOL;
When you look at a schematic you are actually looking at 2 circuits intertwined on one drawing.
There is the AC path, (signal you hear) Which floats on the DC potential which comes from your DC supply. You have to set the DC points so as to pass the best signal.
As I mentioned before just Google stuff.
Also down load a sim program and just fiddle with circuits relating to your Q's.
Often sims come with a library of circuits to help you get started.
Meantime trawl through old book stores you may find some electronics books.
I believe you can download Art of Electronics in PDF.
I have that book and it is not too hard to grasp.
How fast you learn depends a lot on what you already know.
For me it was a lot of reading, and Bread board testing.
But once I used sims my understand went ahead very fast. 8)
When I build/design a circuit I simulate it while also breadboarding it. The advantage is you not only see a plot on the screen but then you can hear it live.
Sims are not perfect but seeing and hearing at the same time speeds up the build process big time.
with BBoard and Sims you are able to tweak stuff to taste as well as find all the problems before you commit to a PCB.
Phil.
Ah, light bulb moment... so instead of bleeding the high frequencies to ground (like a tone control would), C6 lets them bypass amplification, thereby having the same rough effect... they don't get amplified and the tone is darker (but preserving at least some of the highs in the output).
So, if I were to change C6 to a smaller value, EX. 250pf, there would be less treble loss when IC1-B is turned on.
I'll head in there in the next couple days and try to get values on the tone stack and as many other components as I can read. If you don't mind mapping it out, that would be great. I usually try to fit tone-stacks into Duncan's Tone Calculator, but this thing doesn't look anything like the half dozen models they have.
Thanks for the pointers. I'll start looking for books as well as the sim programs.
Some quick reading, and playing with the calculator will help to understand maybe ?
https://www.learningaboutelectronics.com/Articles/Low-pass-filter-calculator.php#answer4
There are several calculators on this site for easy learning, just have to look hard for them.
Another good place to learn:
http://sim.okawa-denshi.jp/en/opampkeisan.htm
http://sim.okawa-denshi.jp/en/Fkeisan.htm
The Schematic on https://www.schematicsunlimited.com/e/epiphone seems a little bit clearer.
Look for the "epiphone regent 2050r" file.
https://www.schematicsunlimited.com/e/epiphone/epiphone-regent-2050r-amplifier-schematic
This is a good place for finding schematics.
Thanks, that is a little better. The one I had was sent to me by Epiphone/Gibson support!
EDIT: Took the better diagram you referenced and spent some time in the graphics program to clean it up further. I haven't edited any component value references... that will be phase 2. But this should be easier to read:
https://i.imgur.com/WjfO8n1.jpg
Here is my edit of Epi-Regent schematic, there may well be things I've missed.
Adding a screen shot of the tone curves.
Note; looks like C11 is 220nF which seems to work in the Simulation.
Screen shows 3 plots;
Yellow trace; Treb 10, mid 1, Bass 10
Green trace; Treb 1, mid 10, Bass 1
Blue trace; Treb 5, mid 5, Bass 5
Could be improved if Mid dip was shifted closer to 400hZ
Phil.
Thanks, @phatt. I am studying that and will compare to the sound samples I'm getting. Seems to make sense.
Couple more questions: To shift that mid dip... to 400hz as you suggest... is that a one or two component swap, or a rearrangement of the whole tone stack? I've played with stacks a bit in Duncan's Tone Simulator, but this thing doesn't really fit in there.
Back on the capacitor C6, which is effectively filtering out treble: I've installed a switch and a pot for the diodes and played with that in conjunction with the overdrive (IC1-B) switching. It really is cutting a lot of treble. I think they were going for that dark boost tone. Misses the mark.
I know I can change the value of C6 to get more treble. Is it reasonable to just life C6 and see what it's like? Just "eliminate" the bypass? I can also try some lighter values but I'm wondering if it's safe to go to infinity.
G'day SemiConductive
I'm fairly new to audio circuits myself, but I have heaps of experience with industrial electronics.
The way I understand the active inverting low pass filter of IC1B is as follows:
The gain is set with R4 and R6(and R5), we don't want to mess with this, unless you feel there is too much gain with the pedal engaged. then increase R5 till you are happy with max gain. (Pot ?)
So R6 is 150k, or 46.78899k with R5 in the circuit.
My understanding is that you calculate this the same as a RC low pass filter.
Capacitance is 470pF
Look at this page, the same formula for normal RC Low Pass filter as an Active Inverting Low Pass Filter like yours. So we can use the top calculator for frequency drop off.
https://www.learningaboutelectronics.com/Articles/Low-pass-filter-calculator.php#answer1
With R=150k and C=470pF you have a drop off frequency of 2.26 kHz
With R=46.78899k and C=470pF you have a drop off frequency of 7.24kHz
If you want 7.2kHz with the R=150k you have to change the capacitor to C=146pF (150pF)
You can easily put in a switch to change the frequency roll off capacitance to set it were you like it, rotary switch you get heaps of otions, but a bit overboard
But maybe I misunderstood the whole thing and got it wrong, If I did can someone please give me a hard smack over the ears please, That way I will remember next time.
Michael
yep that's sounds close to how the maths work. you are obviously already better than me at the maths, :-[
Just remember that this only tells you what the rolloff point is at that point in the circuit.
By the time it gets to the speaker it will have altered, often by a large amount.
Much like the Duncan tone stack app as a good example, Yes a big help but only tells one part of the whole system tone shape.
Phil.
Quote from: SemiConductive on April 19, 2022, 06:20:38 PMThanks, @phatt. I am studying that and will compare to the sound samples I'm getting. Seems to make sense.
Couple more questions: To shift that mid dip... to 400hz as you suggest... is that a one or two component swap, or a rearrangement of the whole tone stack? I've played with stacks a bit in Duncan's Tone Simulator, but this thing doesn't really fit in there.
You can change R15 to 10k which will move the mid dip a bit lower but you will loose a little bass.
Quote from: SemiConductive on April 19, 2022, 06:20:38 PMBack on the capacitor C6, which is effectively filtering out treble: I've installed a switch and a pot for the diodes and played with that in conjunction with the overdrive (IC1-B) switching. It really is cutting a lot of treble. I think they were going for that dark boost tone. Misses the mark.
I know I can change the value of C6 to get more treble. Is it reasonable to just life C6 and see what it's like? Just "eliminate" the bypass? I can also try some lighter values but I'm wondering if it's safe to go to infinity.
Removing C6 will give more treble but it also might oscillate at high gain levels. Maybe try 100pF see what happens?
Phil.
Thanks. I will give those mods a shot.
By "oscillate", you mean feedback? Or just an uncontrollable oscillation while operating?
Feedback is one form of oscillation. There are others.