Roland SH-101 – Missing pulse waveform, PWM and VCO MOD not working

Hi everyone,

I’m currently working on a Roland SH-101 that has a few issues I’m trying to track down. The main problems are with the VCO MOD slider, PULSE WIDTH slider, and PWM selector (MAN/LFO/ENV). I’ve highlighted the controls in the attached photo.

The sawtooth and sub oscillator are working, but I’m having trouble getting any usable sound from the pulse/square waveform.

Here’s what I’ve found so far:

Symptoms:

  • VCO MOD: Moving the slider doesn’t seem to have any effect on the sound.

  • PULSE WIDTH: Moving the slider from 0 to 10 doesn’t seem to change anything.

  • PWM selector:

    • MAN: No audible pulse waveform.
    • LFO: I get a rhythmic beeping/pulsing sound rather than a sustained waveform.
    • ENV: Usually just a brief pop when pressing a key, then silence.

The sawtooth waveform works normally, and the sub oscillator continues producing sound even when the pulse disappears.

Measurements taken at IC13 (CEM3340):

  • Pin 16: +15V
  • Pin 3: -5V
  • Pin 5 (PWM input), MAN: +7.77V
  • Pin 5, LFO: fluctuates between +3.9V and +4.7V
  • Pin 5, ENV: approximately +2.5V

Pin 4 (pulse output):

  • MAN: +13.3V DC
  • LFO: approximately +11V DC
  • ENV: +6.68V DC

What caught my attention is that pin 5 stays at +7.77V in MAN mode regardless of where I position the PULSE WIDTH slider. I checked it at 0, 5, and 10, and the voltage didn’t change.

I’ve also checked the PULSE WIDTH slider with the power disconnected, and its resistance changes as I move it. R178 measures approximately 121kΩ, which is close to its rated 120kΩ. The PWM selector also appears to have continuity through the expected contacts in MAN mode.

From what I’ve been reading, +7.77V seems pretty high for the CEM3340’s PWM input, and I’m wondering whether that’s causing the pulse waveform to disappear entirely.

I’m still trying to figure out whether this is an issue with the CEM3340 itself or something in the circuitry feeding it. I’m also curious whether the VCO MOD issue could be related or if I’m dealing with two separate faults.

I have the original Roland service manual and have been using it to follow the circuit, although I’m still learning my way around troubleshooting analog synths.

Unfortunately, I don’t have an oscilloscope, so I’ve been working with a multimeter.

Has anyone encountered these symptoms on an SH-101 before? Are there any particular components or areas of the circuit you’d recommend checking next?

I’d appreciate any suggestions before I start replacing parts unnecessarily.

Thanks!

Does the LFO-Modulation on the VCF work?

LakeEffectFix, fanwander: don’t pull the CEM3340 yet. Scope the CEM3340 pulse output before the waveform switch. If pulse is present there, the waveform switch contacts or pulse mixer are open; if absent, measure the PWM CV at the CEM3340 PWM input while sweeping the PWM slider and LFO. A CV pinned at either rail gives 0/100% duty and kills the pulse completely.

VCO MOD shares the LFO/mod bus, so a dead LFO or shorted summing op-amp (TL072/TL074) explains PWM and VCO MOD failing together. Clean the sliders with DeoxIT, check for cracked solder at slider pins, and verify LFO output at its test point. CEM3340 is static-sensitive and obsolete; use a grounded wrist strap and avoid shotgunning it.

What DC voltage do you read at the PWM input with the PWM slider at minimum and maximum?