| Author |
Topic: Diode clipping mod, Ge and Si in parallel? |
Georg Sørtun
From:
Mandal, Agder, Norway
|
Posted 19 Jul 2018 1:20 pm |
|
Keith, you are only wrong about the current – drawing too much current through the offset voltage divider. The divider only has to bias the diodes, so not much current will have to go through those resistors if high enough values are chosen for them.
The capacitors connected between "v4" - "v1" and "v4'" - "v1'" to common/GND, have the dual purpose of stabilising the voltage at each of these points, and connect the signal to common/GND when the diodes conduct.
I suggest the following component values as a starting point…
… and then tune them for whatever the soft-clipper is supposed to sound like in the circuit it is supposed to become part of. With values somewhat close to the ones in my drawing, there won't be any current to worry about.
- "OUT" must of course be buffered with an OP AMP stage.
- The resistors between "v+" and "v4", and "v-" and "v4'", can either be chosen for fixed voltages (+/-.5V) at "v4" and "v4'", or suitable values – 1K to 10K – be used combined with variable voltage being fed to "v+" and "v-". The rest of the divider will self-adjust. |
|
|
|
Georg Sørtun
From:
Mandal, Agder, Norway
|
Posted 19 Jul 2018 2:02 pm
|
|
And, yes, my circuit is as simple as it appears. It is just different, and very flexible in that it can be set up (by varying component values) to emulate regular hard-clippers as well as tube-overdrive, or "roar" in its own, unique, soft-clipping ways.
I think any confusion about how it works, may be caused by the basic circuit's flexibility, and its expandability.
With only four stages (as in my drawings) it adds distortion – as it is supposed to. But, add enough clip-stages and match the components well, and it will form an "almost HiFi" sounding limiter that won't "pump" the signal, which may be suitable for those who want to run "hot but clean" through SS-amps. |
|
|
|
Keith Hilton
From:
248 Laurel Road Ozark, Missouri 65721
|
Posted 19 Jul 2018 2:04 pm
|
|
Yes, Georg you are correct about the current. I assume the values you listed are for the supply you mentioned earlier. "It is for +/-12-18 Volt supply."
Georg here is something you can smile about. It has probably be 18 or more years since I burned up a electronic part. I was fooling around with resistor values, and had a Texas Instruments 2426 rail splitter hooked up. I was thinking about referencing voltage off of the rail splitter ground. I think my volt probes were touching the rail splitter ground and positive rail, or it could have been touching the positive and negative rails. The negative rail had no resistors on it. Anyway, I smelled something, and put my thumb on the rail splitter and it burnt my thumb. The 2426 rail splitter is only good for 20 milliamps. Next time I hook up, I will just hook to the rails. Georg when people do stupid things, like I did, sometimes they get their thumb burned!  |
|
|
|
Georg Sørtun
From:
Mandal, Agder, Norway
|
Posted 19 Jul 2018 5:07 pm
|
|
| Keith Hilton wrote: |
| Yes, Georg you are correct about the current. I assume the values you listed are for the supply you mentioned earlier. "It is for +/-12-18 Volt supply." |
Yes, but…
| Quote: |
| - The resistors between "v+" and "v4", and "v-" and "v4'", can either be chosen for fixed voltages (+/-.5V) at "v4" and "v4'", or suitable values – 1K to 10K – be used combined with variable voltage being fed to "v+" and "v-". The rest of the divider will self-adjust. |
… the stated values will work with a 9Volt battery supply too, with the right values inserted for the "v+" and "v-" resistors. Some form for stabiliser is adviced when used with batteries.
About getting burned: I touched a 21KV wire with bad insulation once – decades ago, and burned my right hand thumb and index-finger pretty badly. Took a few months for the wound to quit hurting and heal. I'm sure glad I followed the old rule of working with higher voltages with one hand in the pocket, or else I most likely wouldn't have been around to tell.
Other than that I've only got burns by accidentally touching soldering irons 
Shit happens  |
|
|
|
Keith Hilton
From:
248 Laurel Road Ozark, Missouri 65721
|
Posted 20 Jul 2018 8:21 pm
|
|
Georg, I assume on the right side of the schematic, the 1K6 resistors are there to represent that another stage could be added, is that correct?
The line coming down from the 1K6 positive rail touches nothing, is that correct?
I am confused as to why the 330K resistor on the negative rail is tied to the output of the op-amp. Can you explain why it is there?
Will the presence of the 330K resistor only on the negative side mean there--is less or more-- clipping on the negative swing, than on the positive swing?
I am also interested why you chose the value of.47uf. I am guessing it is allowing a certain frequency, which has an effect on tone? What would be the difference if you used a smaller capacitor, say a .1uf?
Also, could the 1uf capacitor on the op amp output be a polarized cap, with + toward the op amp, and the - toward the right or out? Also, if you used a 2.2uf cap, wouldn't that add more bass?
Isn't the value of the 100K resistor in some ways dependent on the gain of the op amp? Seems to me the higher the output of the op amp the more clipping would take place. Is that correct?
Sorry for the questions, but this is a really interesting circuit. |
|
|
|
Georg Sørtun
From:
Mandal, Agder, Norway
|
Posted 21 Jul 2018 2:14 am
|
|
The two 1K6 resistors are in series each other, forming a 3K2 resistor that connects "v1" to "v1'". All those 3K3 and the 1K6 resistors on the positive and negative side, form one continuous voltage divider to bias the various diodes.
The 330K resistor is connected where the two 1K6 resistors meet, to pull the DC voltage on the output (after the 1uF capacitor following the OP AMP) to the exact center voltage of the divider that bias the clipping-diodes, for symetrical clipping.
| Keith Hilton wrote: |
| I am also interested why you chose the value of.47uf. I am guessing it is allowing a certain frequency, which has an effect on tone? What would be the difference if you used a smaller capacitor, say a .1uf? |
Smaller capacitor values means clipping at higher frequencies, and less DC-stability.
Note that as more clipping stages kick in for stronger signal, more and more of these capacitors contribute in parallel to shorten the signal-peaks to common/GND. With .47uF capacitors there the total capacitance will be more like 2uF for high peaks.
| Keith Hilton wrote: |
| Also, could the 1uf capacitor on the op amp output be a polarized cap, with + toward the op amp, and the - toward the right or out? Also, if you used a 2.2uf cap, wouldn't that add more bass? |
More bass for larger capacitor, yes, but it better be balanced – by ear – against those .47uF capacitors in the clipping stages. What lower frequencies the clipping stages cannot clip, will more or less bypass them.
Polarized (electrolytic or otherwise) capacitors are not very good at coupling signal/DC in reverse – most will fail over time if repeatedly exposed to more than about 1Volt in reverse.So, unless you modify the circuit to accomodate polarized capacitor, or couple two polarized capacitors end-to-end-in-reverse to become one non-polarized capacitor, you will experience non-symetrical distortion in the capacitor itself, premature capacitor aging and failure, and asymetrical drift relative to the clipping stages.
| Keith Hilton wrote: |
| Isn't the value of the 100K resistor in some ways dependent on the gain of the op amp? Seems to me the higher the output of the op amp the more clipping would take place. Is that correct? |
Yes, and no … that is the "drop resistor" that allow the clipping stages to dampen signal-peaks without (over)loading the OP AMP stage. Its value will affect the clippers for frequency and DC stability, and participate in "shaping" the sound. Keep that resistor larger than about 10K (to protect the OP AMP and diodes), and experiment away.
| Keith Hilton wrote: |
| Sorry for the questions, but this is a really interesting circuit. |
That's OK 
But, as a circuit like this is about as flexible as ones imagination whether it is to be used to add distortion or as "clean" limiter, don't expect me to go in depth for all possible variations. It's a question of time, and interest, on my part.
While it is summer here in Norway, I much rather play with these interesting beings…
… than with electronic circuits  |
|
|
|
Paul Arntson
From:
Washington, USA
|
Posted 25 Jul 2018 9:54 am
|
|
Thank you for sharing this, Georg. And that is a lovely photo, also.
_________________
Excel D10 8&4, Supro 8, Regal resonator, Peavey Powerslide, homemade lap 12(a work in progress) |
|
|
|
Georg Sørtun
From:
Mandal, Agder, Norway
|
Posted 25 Jul 2018 12:20 pm
|
|
| Paul Arntson wrote: |
| Thank you for sharing this, Georg. And that is a lovely photo, also. |
Thanks |
|
|
|
