Moog MG-1, Rogue, and Prodigy - Filter and cap mods

The Moog Forum Moog Music General Topics Forum
1

I’m working on a comprehensive guide on Moog MG-1 mods. I should be done in a couple weeks. Since I’m still working it all out, I thought this was interesting enough to bring up in it’s own thread.

While I have my MG-1 open I decided to see what I can do to make the synth sound better overall. You’ll see a lot of opinions regarding how the MG-1, Rogue and Prodigy sound thinner than the more desirable Moogs. Moog was trying to cut corners to make a more affordable synth. There are two odd choices with component selection that seem to make these synths sound different from the others.

Ladder Filter Cap Values
I spent some time studying every Moog schematic I could find. I found that the cap value choice in the filter section directly corresponds to how people perceive the depth of the synths. Here is a list of the capacitor values along with base resistors in multiple Moog synths:

My calculations were flawed in this. The different resistor values in the filter will also play a big role in the sweep. Calibration will also make a difference. I’m working to create a more accurate list of the different variations of the filter.

Source, Polymoog, Opus 3 - .01uF - 150 ohm
MG-1 - .027uF - 200 ohm
Prodigy, Rogue, Taurus II, Liberation - .027uF - 1k ohm
Micromoog, Multimoog - .033uF - 1k ohm
Minimoog, Sonic Six - .068uF - 150 ohm
Taurus - .1uF - 1k ohm



Low Value Caps Replaced With Resistors
There is something else that Moog did that I don’t quite understand. They used low value resistors in place of all ceramic capacitors. I’ve heard rumors of companies doing this as a cost cutting measure since resistors are cheaper, but can’t confirm. When you look at a Moog MG-1, Rogue and Prodigy pcb, you’ll notice there are quite a few bright green metal film resistors. All of these will be low value, under 800 ohms. These were put in place of most caps lower than .01uF. I had all the caps I needed on hand, so I replaced the resistors with the correct value caps. The synth became quieter, more stable and the overall quality of sound greatly improved. It seems more professional now.

So, for around $2 in parts, these “mods” brought my MG-1 into a completely different level. I hope this helps somebody out.



P.S. I haven’t measured or replaced the transistors in the ladder filter yet, but if Moog was cutting corners on the following parts, I’m pretty sure they were skipping out on transistor matching. I’ll post an update on this when I have it.

2

Thanks for posting that info, and really looking forward to the comprehensive mod guide!

Hoping to “degunk” and mod my MG-1 sometime this year ~
although I can’t say I’m looking forward to the “degunk” part of this!!

3

Good luck! Degunking doesn’t seem like very much fun. When I got mine, the previous owner got most of the gunk out, but used a cleaner that managed to melt together all of the keys. I had to use a thin hacksaw, dremel and xacto to cut them loose. Not a very nice process, so be careful as to what kind of cleaner you use.

4

Interesting thread. In regards to cost cutting in the design of the Realistic MG-1, here’s an excerpt from a message posted many years ago by the creator of the MG-1, Paul Schreiber (yes the same man that founded MOTM later on), about the requirements of the Tandy Corporation when it came to the cost of the new synth custom made for them by Moog:

“…I had to specify parts that Moog never had to use: cheapo pots. I’ll admit it:
CHEAPO. They were ALPS and I think we paid (back then) about 23 cents apiece.
That is because the RS [Radio-Shack] gross profit margin was an unheard of 63% (the
average of ALL the Forture 500 is like 8%)…”

You can read the whole interesting story here:

http://machines.hyperreal.org/manufacturers/Moog/MG-1/info/MG-1.history.txt

5

Wild. Thanks for posting. Makes me feel better about swapping out some parts back to Moog specs.

6

Might want to look at the modular Moog schematics of the 904 series filters. They used 4 pole rotary switches to change the filter response via different value caps in the ladder. The 904A Low Pass Filter had three different settings.
Setting 1: 1.2 uF caps - Frequency range: 1Hz to 5kHz
Setting 2: 0.3uF caps - Frequency range: 4 Hz to 20kHz
Setting 3: .075uF caps - Frequency range: 16Hz to 80kHz

A similar 4 pole switch could swap in different values for easier A/B comparisons of the cap changes. Working on a Minimoog filter now,with not only different values but also different types like the El-Menco and Mullard “tropical fish” types along with those latter box type caps.
My matched caps and 11 position 9 pole rotary switch.

Bob had it right on the early 904. Why have one value cap set when you can dial in different ones? More choices!

PS: So what was first, the MG-1 or the Rogue?. The Rogue was introduced at the summer NAMM 1981 with projected release by October 1981. While the MG-1 was listed in RS catalogs as “New for '82”. The Schriber article claims the Rogue was released after the MG-1 and a copy? Article states : “Tandy had 18 months exclusive. Moog then made the Rogue which is my design without the organ/ring mod, wheels back on.”

7

I saw your post on that and am really excited for an update. I’m not quite sure where a 4 pole switch would fit in the MG-1. It would definitely have to go into the back somewhere.

My plan was to purchase some tropical fish caps soon, but I’m curious to see if you thought there was any difference. I’ve swapped many different kinds of caps in guitar amps and pedals. I always hear a difference that is consistent with the type, but I’m not so sure I’m experienced enough to form an opinion on what cap sounds best for the synth. Would you mind sharing your opinion on this when you finish?

Also, I notice that you are matching the caps. I haven’t read anything about this being necessary. What is your reason for doing it? Will the small variance in value make a noticeable difference in the cutoff?

8

Something small like a mini 4-pole double throw switch might do the trick on a small Moog.
Here is a surplus UID brand mini 4 pole that would work.

The nice part about using a switch, either on the test bench or for final installation is the ability to quickly A/B the differences. The difference could be subtle and flipping back and forth under certain playing conditions help reveal any changes. The Moog 904A at first seems like not much difference in switching out those cap values, but white noise filtered shows a limited range in the higher frequency using those larger 1.2uF caps. Might be usefull and fun to be able to switch out the original Moog values with new ones on those ladder filters. The 904B High Pass filter only has the high/low range selection. Switching filter caps is a Bob Moog design that has it’s roots in the first Moog modules.

Not sure matching caps matters much, but if you can, why not. Perhaps more of a screening process to weed out any NOS caps that might be off a bit. In the Mini filter each side exits through a .22uF cap. Those might be more important to be close to each other in value. I plan on swithching those along with the four ladder caps, to see if there is a difference between all Mullard, El-Menco, or the plastic box type. And to try a few other values the Mini did not have. Might have to think about adding some smaller values.

My Mini project is going to take some time to get all together, but I am busy working on it. I had to scratch build a case to house the project and made some small 8U top cabinets while I was doing that.

I never did Tolex cabinets before, so there was a learning curve. They turned out OK for a first time project though. There were many design decisions on how to fabricate and mount the rails with 6-32 mounting screws. And had to power one of the small cabinets with a +/-15 supply for the non-Moog modules in the small three 8U cabinets. Those three 8U cabinets took a bit of time to fabricate, but those are done now and up and running.

The Minimoog part is more intense. I had to mount up the half of a Minimoog circuit boards I had. I have an original oscillator board and the power regulator board with the red, pink, white noise, modulation mixer with headphone amp. I looked at the Studio Electronics “Midi-Mini” mounting method. I was surprised they only used the AMP connectors to hold in the Moog boards. The AMP connectors are good for a third mounting point, so I had to make up some 1/8" aluminum plates to vertically secure the circuit boards. And mounted that all on an 1/8" aluminum base plate.




Since my Mini cabinet is a Moog P style, I had to fabricate the smaller height CP modules from full length Dotcom blanks. Along wth the Moog method of brass piano hinges for the bottom CP mounting. Here is the Mini CP power module.

Still working out the mounting detalis, but here is a view from the back.

And a view of the front of the cabinet.

I have a little way to go before I can post the differences I find in the Mini filter switching. I also plan to have pole selection, along with switching the feedback resistor and trim with each pole setting. My Mini filter panel will look much like the 2U 904A filter panel, with an additional small knob for selecting 1-4 poles. Have to get my +/- 10 volt supply up and running first, then the oscillators. But should be a fun project to have some +/-10 volt Moog modules to complement the existing system I have. :smiley:

9

The filter caps have little to do with Moog synths sounding different. Other people have treaded that road years ago.

The largest contributors are coupling between circuits, subtle overdrive in the VCAs, and - most importantly - the feedback architecture of the filter itself.

Feedback is what provides the resonance quality of the filter. Any EE designer worth their salt recognizes the value of feedback architecture on the sonority of a circuit. I played a model C minimoog prototype and its filter resonant quality was stale compared to the model D. Being a prototype built from vectorboards they had traced out the circuits and drawn the schematic. Instead of bipolar transistors for the recovery amplifier in the filter feedback loop, the model C had JFET transistors. They were trying to correct the problem of nonlinear resonance in the filter circuit, but the moogworks felt the sacrifice in sound quality was not worth it.

I drew out the filter circuits for various Moog models to compare them and found that the feedback architecture does indeed follow their evolution and sonority. Simpler is not always better (as was the case with the model C with JFETs), nor is more complex always the worse (Source and Memorymoog). The important measurement tool when designing a feedback architecture are THE EARS. During the Norlin years when Bob was out of the design loop, the synth design team neglected to use their ears.

Taurus pedals and modulars have a bigger sound because the AC coupling was optimized for low frequencies. Bob confirmed this in a private email exchange years ago. In fact Bob confirmed that the Taurus filter was virtually lifted from the modular. The Minimoog filter is also lifted from the modular. I designed the bass boost mod for the Micromoog but it will not make it a Minimoog.

Back in the old days, VCAs were not high fidelity. That was the best that technology had to offer at the time. They imparted their own subtle overdrive, and the overdrive varied with frequency and and input signal drive and control level. A very dynamic overdrive circuit - try and model that into a transfer function! But few people lend credit to this “overdrive” as a contributor to that “vintage sound”. The old Oberheim polysynths before the Xpander had a creamy edgy sound that got cleaner as the designs progressed through the SEM polysynths, OBX. OBSX, OBXa, OB8… and there was a later rev OB8 that reverted BACK to “dirty” VCAs because the new ones were too clean!!! Moog did value this enough to use “dirty” VCAs in the Memorymoog voicecards instead of the modern high fidelity CEM3360 VCAs.

And how does the filter produce its feedback architecture? With a - drum roll - VCA! That’s right, you will find nearly the same “grungy low fidelity” textbook differential amplifier circuit in both the VCF and VCA of the Minimoog. The most popular core component for the VCA was the CA3080 OTA. And it’s another one of those “grungy low fidelity” components! They didn’t exist when the Minimoog was being designed, but Moog did use the 3080 in the filter feedback of the Taurus and Source and Memorymoog. The 3080 made it possible to program the filter resonance though CV control. Why not the smaller models? Because they weren’t programmable and did not require the CV control of the others.

10

You guys are blowing me away here. The information in these posts are amazing. Thank you.

So, first I’d like to focus on understanding the feedback circuit on the MG-1. I’m still diving in here and am not an engineer, so I hope you don’t mind questions in laymans terms. The feedback circuit that I should be focusing on is in between the input and output of the ladder filter correct? The circuit that is controlled by the Emphasis pot and Regen Cal? In the Minimoog schematics, they use a four transistor (T1597) circuit in this section. The MG-1 uses LM431 and LF353 opamps. I’m wondering it would be beneficial for me to remove the exisiting feedback circuit from the MG-1 and patch in the Model D’s using stripboard.

11

In a word: I don’t think so. (wait, that’s four words ! Oh well… :laughing: )

Both filters might look similar but in fact they’re rather different. They use different supply voltages (+12V, -15V for MG-1 and +10V, -10V for Mini), different circuit configuration, different component values, etc…

You’d be better off building a Minimoog filter clone on a small board and bypassing the one inside the MG-1. But then again, you’d have to “adapt” the voltages, and you’d also have to use different value sliders for it, too, to match the pots on a Minimoog. And I know for a fact that finding a 50K reverse audio slider for the “emphasis” that will fit inside the slot of an MG-1 panel would be extremely difficult to find, if not impossible.

If you want the sound of a Minimoog D, just get a Minimoog D. Or the next best thing: a Moog The Source, which sounds very similar.
My opinion, for what it’s worth…

12

I get what you’re saying. I am planning on buying a Model D. And I don’t believe I’ll transform this synth into something as nice as a vintage Minimoog. I’ve contemplated building some of the DIY modules, but I seem to be set on this personal goal of getting the Mg-1 the best it can be.

Emphasis is A50K. Is it that big of a difference since I’m not worried about which way the pot turns? +/-10V is pretty easily doable.

13

I had a Realistic MG-1 for many years and, aside from adding an external potentiometer to control the PWM of the “Tone Source One” square waveform (effectively bringing the square wave trimpot adjustment outside), I was perfectly happy with the way it sounded.
As for your emphasis pot question, I wouldn’t know about that since I’m not an engineer but only an electronics hobbyist.

The Paul Schreiber designed Realistic/Moog MG-1 might not sound quite like a Minimoog D, but it doesn’t sound bad at all.

14

I really do love the synth. I guess that’s why I’m looking so deeply into it. Sometimes I forget about the big picture when focusing on the details.

15

LOL, kind of like looking for a radiator cap for your 1969 VW Beetle. (Air cooled!) To make a slider anti-log or reverse audio simply turn it over. Both the MG-1 and Rogue/Taurus II use standard 50K audio sliders and orient them so they are anti-log. Check the footprint on those sliders, the 50K emphasis are inverted. Learn someting new every day!

Like the idea of removing those OP-amps from the audio chain on the Prodigy. I have one and a complete spare board. Might be a fun project.
Here is an exerpt from a net article by Jim Scott who helped develop the Minimoog filter circuit.

Jim Scott, Moog Music engineer- 1969-1977.
“The original Mini had no integrated circuits whatsoever in it. None of us knew how to design
with them back in the bad old days and none of us had the time to mess around with anything unknown. All circuitry was implemented using discrete (individual) transistors. This resulted in a very wide-band audio chain with no feedback anywhere, which unlike IC op amp implementations when driven into distortion, did so softly like a vacuum tube amplifier, without clipping the waveforms. This allowed us to drive the circuitry rather hard, which we did to achieve a good output signal-to-noise ratio. As a result, each of the several stages of the sound chain ended up contributing a fortuitous gentle distortion which enlivened the sound. The ultrasonic bandwidth preserves the high frequency cross modulation components (far above the audible range) which are produced in this process, and which the distortion in later stages causes to reappear downshifted back into the audible range. This is probably a vital factor in the famous “fat” sound of the early Moog instruments.”

Jim Scott mentions this “ultrasonic bandwith”. Sounds you can’t hear that influence the lower frequency sounds you can hear. The discrete circuit Moog 904A specs read 1HZ to 80KHz.

16

I stand corrected. Thanks.

17

Good catch!

18

I’m getting pretty close to finishing up on testing all of the mods. Is there anything I’m missing on my list that anybody would suggest?

PERFORMANCE MODS
Replace wrong resistors back to capacitors according to schematic
Minimoog Filter mod - Increase filter coupling cap values, lower input impedance and change filter cap and resistor values to match Minimoog
Add continuous glide and re-trigger circuit to keyboard tracking (stripboard layout provided)
Replace 5837 Noise chip with analog noise generator (stripboard layout provided)
Tune Osc 1 & 2 one octave lower
Osc 2 Detune slide pot to full octave up and down
Increase filter modulation level
Increase audio output level
Change Ring Mod input from Osc 1 square only to post-switch/pre-mixer Osc 1


ADDITIONAL FUNCTIONALITY MODS
Add Sustain pot
Osc 1 and 2 PWM controls
Osc 2 Detune/pitch modulation pot (very cool flanger meets autowah effect with sync in!)
Poly pitch modulation pot
FM pot
VCA modulation pot
LFO speed switch (fast, normal, slow)
Osc 1 octave up switch
Ring Mod carrier from LFO (instead of Osc 2 Square) switch
Internal or External switchable Saw to Triangle to Sine convertor (stipboard layout provided)
Internal or External VCA ASDR (stripboard layout provided)
Separate headphone and master volume pots

BACK PANEL CONNECTIONS
CONTROL VOLTAGE:

VCA
VCF
Envelope
V-Trigger
Gate
LFO In/Out
Keyboard In/Out
Sync In

AUDIO:

Post Filter/Pre-VCA In/Out
Pre-Filter Audio In
Noise In/Out
VCO
Osc1 - Pre-mixer
Saw In/Out
Osc1 Post-Switch In/Out
Osc2 - Pre-mixer
Saw In/Out
Osc1 Post-Switch In/Out
S&H In
Poly Pre-mixer In/Out
Ring Mod Audio In
Ring Mod Carrier In

19

Back to the topic of filter cap mods, even though the MG-1 uses opamps in the following stage of the filter, I’ve found that the following modifications makes an improvement, IMO:

Swapping the filter caps for .068uF
Change the resistors in the ladder filter (except R77) to 150
Changing the coupling caps to .22uF
Changing the input impedance resistors from 1M to 47K

I simply copied the same specs of each from the Minimoog Model D. The filter now seems fatter and smoother. While I liked the stock MG-1 filter, these put it over the top for me and is especially noticeable during a filter sweep with the resonance up high.

20

Just the instruction manual(s) that show how to do all these mods! :wink:

All kidding aside, thanks for taking the time to do all this ~
Looking forward to taking your list along when I start my MG-1 overhaul!!