MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Korg Monotron Euro Module DIY
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page Previous  1, 2, 3, 4, 5, 6  Next [all]
Author Korg Monotron Euro Module DIY
pulplogic
Tim,

It sounds like we are heading down similar paths. Here is where I am at:

1. Powered off of 5V using 7805, the Monotron draws 29mA. I will probably draw off of the 5V Euro rail once packaged and not use the +12V.

2. Built active output circuits for the LFO VCO and VCF. The VCO and VCF are capacitivly coupled with gain control and the LFO is dc coupled with offset and gain control (see schematic).

3. Worked on tuning the VCO for 1V/Oct. I found with the range in the center and inserting a signal at the pitch switch, I need an R of 37.6K to get close. I am working on an intonation circuit with offset so a bipolar signal can be used.

4. Padded the VCF input for large signals.

5. Added jack for gate with 100k input Z.


Still to do:

1. Active circuits for Pitch and Cutoff with dc offset and padding for large bipolar signals.

2. LFO reset

3. Lots of testing

4. Packaging


Parts removed so far:

Ribbon, R11, R42, R58, Speaker


--John
urbanscallywag
hi John,

the images must have gotten resized and are hard to read. can you try to repost them?

thanks,

Guinness ftw!
Tim Stinchcombe
The snow has cleared sufficiently to enable me to drive to work today, so no excuse to stay home and play! Consequently progress has slowed, but I just finished most of the wiring I'll need to get it going, but I daren't power it up until tomorrow (too late here now, and besides, after I squashed all the wires down and screwed it up, I realised I left the ribbon unconnected...). But here are some very crude snaps to give an idea - it is a bit 'frankenstein', but hopefully it'll work OK:




Tim
Tim Stinchcombe
pulplogic wrote:
It sounds like we are heading down similar paths.
Yeah, there is only so much one can wring out of this thing!
Quote:
2. Built active output circuits for the LFO VCO and VCF. The VCO and VCF are capacitivly coupled with gain control and the LFO is dc coupled with offset and gain control (see schematic).
For my messing with it, I decided against going that far (besides it almost seems like adding that amount one could almost build a stand-alone VCO and LFO from scratch...). Incidentally, the input and feedback resistors for your second LFO amp seem a little low - was there a reason for that? (I mean the gain is presumably OK, but I would have gone 10x bigger, i.e. at least 10k c/w 100k pot)?
Quote:
3. Worked on tuning the VCO for 1V/Oct. I found with the range in the center and inserting a signal at the pitch switch, I need an R of 37.6K to get close.
thumbs up Yep, that's close to my calcs and the 39k I'm using.
Quote:
4. Padded the VCF input for large signals.
Not sure I'm getting what you mean by 'padded'? But yes, I guess the resistors may need some adjustment for larger signals.

Tim
Tim Stinchcombe
OK, here are some more pretty pictures. First up, a quick plot of the filter self-oscillation frequency against the cut-off pot wiper voltage - the straightness of the log line over the middle section shows that it is not too bad an exponential response:

(The scaling through the 68k resistor (R50) on the pot wiper is about 0.23V/octave for that middle section.)

Now for the noise bit. It successfully runs off the +5V regulator (simply a 78L05), and the DC-DC converter seems to be disabled OK (via its 'disable' line). The following traces show the VCO, filter and headphone outs (AC coupled to the scope, and with a little resonance dialled into the filter) when running on the battery:

and when using the 5V regulator, powered externally via +12V:

The amount of 'mush' on the output whilst running on the battery is very evident, and running off the regulator is obviously a vast improvement. The noise is all from the switching DC-DC converter, apparently (mostly) feeding into the filter.

Looking at the rails (again AC coupled for the +5V), we see a huge amount of noise, at the frequency of the switching, i.e. 1.2MHz, when running off the battery, and there is quite a lot of noise on the ground rail too (at least at the point I probed...; note also the 5V rail and ground signals are not synced in any way in the following plot):


Most of the noise appears at the filter, and looking at the scaling on the filter input, it is easy to see why - the approx 1V VCO swing is dropped down to a mere 3.3 millivolts at the filter input - with the noise on the rails swinging through many times this figure, is it any wonder the output is so bad?!

What I'd like to do now is see if I can record the difference between the two set-ups, i.e. running off battery vs. regulator, to see if I can capture any difference in noise. However, just casually switching back and forth whilst listening to it, I can't really appreciate that there is any difference - thus my question to you all out there is, can anyone suggest a simple set-up, which you know to be noisy, that hopefully I can re-create so as to hear the noise for myself??!

Tim
bluemeanie
thanks for all your info Tim..... but I'm a bit lost as to what you're asking for. Feeding the monotron audio into an mixing desk with the speakers turned up maybe???? I'm probably wrong but i dont remember the monotron being really noisy... I do remember the gate on/off from the ribbon going chunk chink tho..... i know i make no sense!
Tim Stinchcombe
bluemeanie wrote:
but I'm a bit lost as to what you're asking for.
I'm not so sure myself! I'm not a musician, so am not constantly recording stuff to the point where I could say something sounds 'noisy' and hence is not so usable and/or a major irritation.

Normally when I am listening to stuff on my modular I feed the signal through an A-150 VC switch (so it is easily mutable), then to an A-134 Panner (basically acting as an attenuator), and then out to a pair of Edirol MA-7A 'monitors' (little more than a pair of computer speakers...). Sure, the output from the Monotron is quite feeble (of the order of +/-100mV or so, as the plots show), and so I need the volume on the Edirols cranked up, and so there is a bit of hiss, but is this because the gain is so high, rather than the Monotron being 'noisy'? Certainly there is little difference switching between battery and external power operation, but could this be because my set-up just doesn't have much high-end response? Even running the Monotron directly into the Edirols doesn't make much difference either.

Since several people have mentioned how noisy they find it, I'd like to know some detail of their set-up and under what sort of conditions they notice the noise. That is 'I run the Monotron through kit x with settings of y and z, and with the Monotron doing p and q and I get this awful hiss/whatever...'. Can anyone elaborate?

Quote:
I'm probably wrong but i dont remember the monotron being really noisy...
The way I'm using it, I tend to agree (but see above!).
Quote:
I do remember the gate on/off from the ribbon going chunk chink tho.....
Yes, that I have noticed! And it is also there when feeding an external gate in too (though perhaps this is no surprise really). I suspect it is caused by the way the circuit is gating the filter open/shut - there is a small cap involved, maybe beefing it up might help, but whatever, more investigation is needed!

Tim
bluemeanie
if you could find a way to fix the noisy gate that would be really cool.

thanks again for all your effors....if this thing ends up in the euro....well... Drugs Rawk!

i checked this arvo the noise from the monotron:

from monotron headphone out to behringer mixer input gain at 12:00 (pretty normal for me)

the noise was evident at high speaker volume, but at normal operating level was bearly noticable. You'd really here it on a big rig ...haha

Compared to my Frostwave sonic alienator this things quiet!

also this guys rockin... check it

http://beatnic.jp/monotron/V2.html
Tim Stinchcombe
I have just annotated one of the photos with all the component references and placed it at my website:
http://www.timstinchcombe.co.uk/synth/monotron_ref_des.jpg

Small (i.e. unreadable!) version:


Tim
pulplogic
Quote:
For my messing with it, I decided against going that far (besides it almost seems like adding that amount one could almost build a stand-alone VCO and LFO from scratch...). Incidentally, the input and feedback resistors for your second LFO amp seem a little low - was there a reason for that? (I mean the gain is presumably OK, but I would have gone 10x bigger, i.e. at least 10k c/w 100k pot)?


I am trying to do all the mods using 10k pots, just like the Monotron. I can't stand the unsupported plastic shaft pots. I plan on replacing all the pots with pots with bushings. I realize it is a little extreme to buffer all of the i/o with op amps, but I think it will provide a nice signal level and impedance match with other modules.


Quote:
Not sure I'm getting what you mean by 'padded'? But yes, I guess the resistors may need some adjustment for larger signals.


I just shed some signal on the input to tolerate larger signals before clipping. It still distorts nicely, but not with really small signals.

Nice job on the reference designator map. I bet that took some time. Thanks!
Scott Willingham
Hi all!

Looks like I finally found the best forum for serious Monotron tweaks and mods. I've been playing with mine since October and lately making lots of measurements and analyzing the schematics.

I think I have a pretty successful mod for lowering the noise floor. Some have speculated that it comes from the output amplifier. This is easy to disprove: just turn down the volume and the noise goes away. A look at the schematic shows that the volume pot is a simple attenuator before the output amp. With the volume at minimum, the noise from previous circuits is blocked, but the output amplifier noise is still present at the output. Since the noise goes away at min volume, we can conclude that it comes mainly from prior circuits.

As Tim S. has commented, the signal levels within the core of the filter circuit are very small. (BTW, thanks Tim for your MS-10/20 filter analysis document. It's been very helpful.) These signal levels are intrinsic to the way this filter works. Consequently, the output amplifier of the filter has to re-amplify the signal with a gain of around 70x (37 dB). With further analysis it turns out that the opamp input noise (IC4C) dominates other sources of noise.

So the trick is to find a suitable opamp with better noise specs. I'm not an expert in discrete opamps, but I think I found a good replacement, the OP462. It runs on 5V with near rail-to-rail outputs and has an input referred voltage noise four times lower (12 dB) than the LM324. It does have a higher input current noise, which unfortunately reduces the advantage a few dB. It has some drawbacks that I consider acceptable (YMMV). First, it consumes more power, but it's still a low-power design. John (pulplogic) measured the total current from the 5V rail at 29mA; the OP462 should typically draw about 2mA total for all four amplifiers. Second, it is much more expensive; I paid $7.94 for part number OP462GSZ-ND from Digi-Key. Also, while the OP462 has a very low offset voltage, its input bias currents are much higher than the LM324. This, combined with the large feedback resistor, produces about 80-100 mV offset at the filter output. This slightly unbalances the diode clipping in high-res situations, but I don't think this is too important.

There is one more consideration. The original LM324 has about 1MHz unity gain bandwidth. In the high-gain circuit used in the Monotron filter, this puts a pole in the ampifier stage at about 14 kHz. This limits filter resonance at high frequencies. The Monotron's designer added components R74 and C22 in the resonance feedback circuit to compensate this. Now the OP462 has about 15MHz bandwidth and with the compensation circuit, the resonance becomes harsh and unruly at frequencies above about 2 kHz. The solution I have chosen is to remove R74; this compensation is no longer needed. (An alternative solution is to add about 35 pF in parallel with R60, but I have not tested that.)

In summary, the modification consists of: 1) replacing IC4 with an OP462, and 2) removing R74. I measured the noise spectrum of my Monotron at frequencies of 2 and 8 kHz, before and after the mod. At both frequencies, the resulting noise reduction is 9dB. This is very noticable and makes listening to the Monotron in headphones much more pleasent.

My next project is trying to understand and reduce the filter gating "pops".

-Scott-
Scott Willingham
[quote="Tim Stinchcombe"]I have just annotated one of the photos with all the component references and placed it at my website:[/quote]

Thanks! Very handy to have this reference.
ether
Just came across this:

Monotron++

http://beatnic.jp/monotron/V2.html
Tim Stinchcombe
Hi Scott, and welcome!

Scott Willingham wrote:
Hi all!

Looks like I finally found the best forum for serious Monotron tweaks and mods.
Well it is a forum, and we often get serious about things - whether we are any good, certainly I shall leave others to judge! (And judging by the hits on my website, there are a lot of forums out there talking about modding the Monotron, mainly standard fare of sticking a resistor or two onto some of the 'TP's.)

Quote:
I think I have a pretty successful mod for lowering the noise floor. Some have speculated that it comes from the output amplifier. This is easy to disprove: just turn down the volume and the noise goes away. A look at the schematic shows that the volume pot is a simple attenuator before the output amp. With the volume at minimum, the noise from previous circuits is blocked, but the output amplifier noise is still present at the output. Since the noise goes away at min volume, we can conclude that it comes mainly from prior circuits.
Perhaps my mistake is in not using headphones, as I have mentioned above (several times) my difficulty in perceiving the noise at all, but your simple argument is very convincing.

Quote:
(BTW, thanks Tim for your MS-10/20 filter analysis document. It's been very helpful.)
My pleasure, and it is clear I can learn much from your post in return!

Quote:
So the trick is to find a suitable opamp with better noise specs. ... First, ..., Second, it is much more expensive; I paid $7.94 for part number OP462GSZ-ND from Digi-Key.
And an unfortunate 'third' for most people will be the difficulty in replacing such a small SOIC surface-mount IC. I seem to recall that one of the early Japanese blogs/sites modding the Monotron was explaining how to swap out chips - I'll have to see if I can re-find it, and see if it was specifically the filter IC, i.e. did that person come to the same conclusion? (I suspect at the time I was busy doing other stuff so didn't pay much attention to it.)

Quote:
Also, while the OP462 has a very low offset voltage, its input bias currents are much higher than the LM324. This, combined with the large feedback resistor, produces about 80-100 mV offset at the filter output. This slightly unbalances the diode clipping in high-res situations, but I don't think this is too important.
A trawl through the Synth DIY archives might be in order here, as there are often discussions about everybody's favourite op amp, so an even better replacement might yet be found (?).

Quote:
There is one more consideration. The original LM324 has about 1MHz unity gain bandwidth. In the high-gain circuit used in the Monotron filter, this puts a pole in the ampifier stage at about 14 kHz. This limits filter resonance at high frequencies. The Monotron's designer added components R74 and C22 in the resonance feedback circuit to compensate this.
Now this is where my ears really pick-up! I find the interplay between all these various parameters and design considerations really fascinating, and such knowledge is probably best picked-up through experience, but in general the opportunities for undergoing such experiences can be thin on the ground, so I am definitely all ears now! I had wondered what the reason for that little cap+resistor combo was - presumably it inserts a zero at roughly the same place as the pole, and so negates the pole's effect?
Quote:
Now the OP462 has about 15MHz bandwidth and with the compensation circuit, the resonance becomes harsh and unruly at frequencies above about 2 kHz. The solution I have chosen is to remove R74; this compensation is no longer needed. (An alternative solution is to add about 35 pF in parallel with R60, but I have not tested that.)
If my rough calcs are right, the 15meg bandwidth now means the errant pole shifts right up to over 200kHz, which suggests ignoring isn't going to cause any hassle...?

Quote:
My next project is trying to understand and reduce the filter gating "pops".
I've only had a quick look at that a day or so ago - C16 is an obvious starting-point, but having run some simulations I wasn't convinced increasing its value would help, especially as my initial look-see with a scope gave me the impression that the DC level in the signal is changing. I was starting to think along the lines of C14 charging whilst the VCO is gated off, and perhaps this introduces a thump of a voltage spike as the VCO starts up (but that was as far as I got...).

I've also been playing with the simplest way of showing what tweaks I've made so far - a combination of editing the Korg schemo, overlaid with some circuity-bits from my simulation package seems do-able. Here is my first effort - how I have 'grafted on' a 78L05, so that I can switch between battery and external (linearly regulated) power:

Basically I cut off the battery springs and replaced the battery holder altogether, then made sure the DC-DC converter was 'disabled' (via the new two-pole switch I added) when running on the 78L05 (but even without the 'EN' pin shorted to ground, without power via 'VBAT' pin 6, it doesn't look like the converter is doing anything). The 10k resistor was really an after-thought, just in case the regulator needs to be kept happy with some load when not connected to the Monotron.

Tim
Tim Stinchcombe
Tim Stinchcombe wrote:
I seem to recall that one of the early Japanese blogs/sites modding the Monotron was explaining how to swap out chips - I'll have to see if I can re-find it, and see if it was specifically the filter IC, i.e. did that person come to the same conclusion?
Well I found it, but I don't know that I am any the wiser - the Google translate is quite impenetrable! It is the filter IC though. [Edit: he does both LM324s.] Perhaps one of our Japanese-speaking contributors can tell us what his justification is for using the replacement op amp, an LMC6494:

http://airvariable.asablo.jp/blog/2010/05/01/5059566

Google translate

He even has a Youtube video of it!:


Product page for the Nat Semi LMC6494. From a quick squint at the datasheet, bandwidth is about the same as the LM324, at 1.5MHz; the input bias current is miniscule at 150 femto amps; I'll need to study a bit more on how to interpret the noise specs.

From reading the translated page, I'm wondering whether he did it more 'on spec', i.e. simply using a 'higer spec'd part', rather than out of any consideration of knowing it would lower the noise...? It also looks like he had a go at replacing the LED with a blue one, but perhaps didn't succeed? And what is the point of that big cap in what I assume is C15, one of the 'not placed' components, and the metallised tape...?!

Tim
Scott Willingham
Tim Stinchcombe wrote:

Quote:
So the trick is to find a suitable opamp with better noise specs. ... First, ..., Second, it is much more expensive; I paid $7.94 for part number OP462GSZ-ND from Digi-Key.
And an unfortunate 'third' for most people will be the difficulty in replacing such a small SOIC surface-mount IC.
Yes, this will be difficult for many DIYers. I'm blessed to have access to the right equipment for handling such jobs. After more than a decade of dealing with SMT parts (I design RF ICs), the SOIC and 0603 parts actually seem big to me! Still, I encourage others to give it a try. It is actually easier to desolder an SOIC part than a through-hole DIP. The easiest approach with simple equipment is to clip all the leads and then desolder them one at a time. Then, the only special equipment needed is an intermediate-grade soldering pen (and a magnifier). I like the Hakko 936 for home use.

Quote:
A trawl through the Synth DIY archives might be in order here, as there are often discussions about everybody's favourite op amp, so an even better replacement might yet be found (?).
I tried some general searching, but most traditional synth circuits are not constrained to 5V single supplies.

Quote:
Quote:
The original LM324 has about 1MHz unity gain bandwidth. In the high-gain circuit used in the Monotron filter, this puts a pole in the ampifier stage at about 14 kHz. This limits filter resonance at high frequencies. The Monotron's designer added components R74 and C22 in the resonance feedback circuit to compensate this.
Now this is where my ears really pick-up! I find the interplay between all these various parameters and design considerations really fascinating, and such knowledge is probably best picked-up through experience, but in general the opportunities for undergoing such experiences can be thin on the ground, so I am definitely all ears now! I had wondered what the reason for that little cap+resistor combo was - presumably it inserts a zero at roughly the same place as the pole, and so negates the pole's effect?
Yup. I wondered about R74 and C22 until I replaced the opamp. It became immediately apparent that the "peak" knob was too touchy at higher frequencies. Then, thinking about the bandwidth change, the reason for the components became clear. I simulated the filter's feedback section to verify this. R74 and C22 add a zero-pole doublet to compensate the opamp's gain drop-off. The actual zero is below the opamp's pole, overcompensating a bit so the response peaks a couple dB more at 9 to 10 kHz, then rolls off above 20 kHz.

Quote:
If my rough calcs are right, the 15meg bandwidth now means the errant pole shifts right up to over 200kHz, which suggests ignoring isn't going to cause any hassle...?
That's the idea. But the feedback response is now a bit flatter than before (with R74 removed). There's also a possibility that the higher harmonics are different when the filter is driven into diode clipping. It's a fascinating circuit, which is partially why I'm playing with it so much.

Quote:
Quote:
My next project is trying to understand and reduce the filter gating "pops".
I've only had a quick look at that a day or so ago - C16 is an obvious starting-point, but having run some simulations I wasn't convinced increasing its value would help, especially as my initial look-see with a scope gave me the impression that the DC level in the signal is changing. I was starting to think along the lines of C14 charging whilst the VCO is gated off, and perhaps this introduces a thump of a voltage spike as the VCO starts up (but that was as far as I got...).
The CV processing circuit for the filter is quite tricky and a neat little design in itself. I've taken to simulating the whole subcircuit in LTspice. There's definitely a thumping contribution from the VCO gating and C14, but this is much less than the contribution from the filter gating. Especially since the VCO thumping affects mainly the attack and doesn't contribute a big pop at the release of the note. It is helpful to look at the waveforms by recording some notes into a DAW or audio editor. Plugging an unconnected stereo plug into the aux input disables the filter gate, so you can compare waveforms and sound with just the VCO gating.

Quote:
I've also been playing with the simplest way of showing what tweaks I've made so far - a combination of editing the Korg schemo, overlaid with some circuity-bits from my simulation package seems do-able.
This looks very good. I may try something similar if I "publish" more complex tweaks. The last few days, I've measured a lot of the signal levels and control ranges in the Monotron, so I've been thinking of doing an overlay of my notes on the Korg schematic.

-Scott-
Cat-A-Tonic
Tim Stinchcombe wrote:
Perhaps one of our Japanese-speaking contributors can tell us what his justification is for using the replacement op amp, an LMC6494:

http://airvariable.asablo.jp/blog/2010/05/01/5059566

Google translate

My wife helped to confirm & clarify what it said.

He wanted to reduce the `sssss`noise that was constantly on the output,
so he replaced the opamp in the headphone amp output stage with a higher spec one.
This was apparently successful.

The aluminum tape is to shield the case.
I guess it had some handling noise or microphonics before. seriously, i just don't get it

The blue LED didn`t work as a replacement for the red because the blue required more voltage, and several resistors would have to be replaced to make this work.
Tim Stinchcombe
Scott Willingham wrote:
The easiest approach with simple equipment is to clip all the leads and then desolder them one at a time.
That's my normal method too.

Quote:
Quote:
A trawl through the Synth DIY archives might be in order here, as there are often discussions about everybody's favourite op amp, so an even better replacement might yet be found (?).
I tried some general searching, but most traditional synth circuits are not constrained to 5V single supplies.
Ah yes, I was overlooking that very important fact!

Quote:
Quote:
If my rough calcs are right, the 15meg bandwidth now means the errant pole shifts right up to over 200kHz, which suggests ignoring isn't going to cause any hassle...?
That's the idea. But the feedback response is now a bit flatter than before (with R74 removed). There's also a possibility that the higher harmonics are different when the filter is driven into diode clipping. It's a fascinating circuit, which is partially why I'm playing with it so much.
I need to spend more time simulating it myself, so that I can see the relative effects of the component value choices and the op amps.

Quote:
The CV processing circuit for the filter is quite tricky and a neat little design in itself.
Even though I'm relatively familiar with the way the Korg35 works, I still had to admire the clever use of the cut-off voltage to shutdown the filter completely.
Quote:
There's definitely a thumping contribution from the VCO gating and C14, but this is much less than the contribution from the filter gating. Especially since the VCO thumping affects mainly the attack and doesn't contribute a big pop at the release of the note.... Plugging an unconnected stereo plug into the aux input disables the filter gate, so you can compare waveforms and sound with just the VCO gating.
I now have a switch in there for that, so I will try both ways tomorrow. Another thing that occurred to me that I was also wondering about is the impact of the filter CV voltage on the signal through the filter: the CV voltage (R54/R53) jumps by around 200mV or greater as the filter is gated open, and this pushes Q12 emitter up, and hence also the voltages at C20 and C21... If I get time tomorrow, I may look at that/think on it further (unfortunately drainpipe repairs will have to take priority - the downpipe got completely filled with ice as the snow melted, and split the thing open for about half its height!).

Tim
Scott Willingham
Tim Stinchcombe wrote:
Quote:
I seem to recall that one of the early Japanese blogs/sites modding the Monotron was explaining how to swap out chips - I'll have to see if I can re-find it, and see if it was specifically the filter IC, i.e. did that person come to the same conclusion?
Well I found it, but I don't know that I am any the wiser - the Google translate is quite impenetrable! It is the filter IC though. [Edit: he does both LM324s.] Perhaps one of our Japanese-speaking contributors can tell us what his justification is for using the replacement op amp, an LMC6494:

http://airvariable.asablo.jp/blog/2010/05/01/5059566

Google translate

He even has a Youtube video of it!:


Wow! His technique for desoldering the chip is ingenious. He wraps a substantial hunk of metal tape (with thermal mass) around all the pins so he can heat them all simultaneously with a high-powered iron. That is cool. The way I did mine was using a special tool that blows a concentrated stream of hot air, enough to rapidly heat a controlled area. I wasn't smart enough to mask off the other parts though -- some of my capacitors came dangerously close to blowing out of place!

I looked up the datasheet for the LMC6494. A nice part, but the noise spec is 37 nV/root-hertz -- similar to the original LM324 (reportedly 40 nV/rootHz, but not easy to find out).

Quote:
From reading the translated page, I'm wondering whether he did it more 'on spec', i.e. simply using a 'higer spec'd part', rather than out of any consideration of knowing it would lower the noise...?
That seems to be the case. Similarly, for adding the low-ESR cap for C15. Some tweaking of the decoupling caps might reduce the boost-converter switching noise (at 1.2 MHz as you show in previous posts), but I don't think that's really an issue. I don't see any artifacts from that noise when I record the Monotron into an audio bandlimited sound card.

-Scott-
pulplogic
I like all the interesting mods, I hope to try some of them. First, I want to see how well the monotron can perform by only adding peripheral circuitry and removing components.

The noise performance was improved by creating buffered outputs for the VCO, LFO and VCF. If you take the output at the headphone / speaker output the noise can be fairly objectionable once amplified. Also, the max signal level at the headphone output in rather small to interface with euro modules. The scope shot is the VCO fed through the VCF, blue is buffered out yellow is speaker out with max volume (note the scale).
Scott Willingham
pulplogic wrote:
The noise performance was improved by creating buffered outputs for the VCO, LFO and VCF. If you take the output at the headphone / speaker output the noise can be fairly objectionable once amplified. Also, the max signal level at the headphone output in rather small to interface with euro modules. The scope shot is the VCO fed through the VCF, blue is buffered out yellow is speaker out with max volume (note the scale).


Interesting measurement. I think it is neccessary, however, to look at the spectrum of the noise. Your buffer is likely bandlimited and therefore filters out noise above the audio spectrum. In that case, the waveform looks cleaner, but may be audibly the same. For example, most of the fuzz in the yellow trace could be from power supply switching at 1.2 MHz. If you record both the signals into your computer and run a spectrum analysis, you can compare the SNR within the audible spectrum.

Also, the headphone output level was probably chosen for ear safety, rather than optimum line level outputs. The level can be boosted by changing R82 and/or R79.

-Scott-
Tim Stinchcombe
Thanks for that - very helpful to have it clarified/confirmed!

Tim
Cat-A-Tonic wrote:
Tim Stinchcombe wrote:
Perhaps one of our Japanese-speaking contributors can tell us what his justification is for using the replacement op amp, an LMC6494:

http://airvariable.asablo.jp/blog/2010/05/01/5059566

Google translate

My wife helped to confirm & clarify what it said.

....
plasticanimal
Hi, I don't mean to move the topic in another direction, but you guys seem to know what your talking about. So I was wondering would it be possible to send the LFO to modulate the amp then disable the re-trigger of the LFO and disable the gate to the amp and filter to make a simple decay effect?
richard
oh my god! Tim has multipied! applause Welcome Scott!
Scott Willingham
plasticanimal wrote:
Hi, I don't mean to move the topic in another direction, but you guys seem to know what your talking about. So I was wondering would it be possible to send the LFO to modulate the amp then disable the re-trigger of the LFO and disable the gate to the amp and filter to make a simple decay effect?


I'm afraid this is not possible because the Monotron doesn't actually have a VCA. Its simplified, economical design relies on gating the VCO along with gating and enveloping and VCF for contouring the sound. This is part of the reason for the noise complaints. I don't have one, but it's likely that the MS-10/20 filters are just as noisy, but people don't notice it so much because the following VCA fades the noise when the sound is quiet.

It is possible to disable the re-triggering of the LFO so it acts more like a simple envelope generator. There's a mod on one of the Japanese sites that does it. I'd like to do that on my Monotron, but first I'm waiting for inspiration in how to cleanly mount some switches in the case. The Monotron's physical design is very tight and I'm much more handy with the electronic mods than with the mechanical mods.

-Scott-
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page Previous  1, 2, 3, 4, 5, 6  Next [all]
Page 3 of 6
Powered by phpBB © phpBB Group