I've been going through my inventory of old audio gear to give me something to do during the pandammit. (intentional misspell). I found my old Electro-harmonix Memory Man analog delay pedal, did some web research and found out it has since been crowned the holy grail of analog bucket brigade delay pedals. It has great value on the used market in fact.
I found a calibration procedure on-line and set about performing it. Not too far in I'm checking the LFO, low frequency oscillator which cyclically varies the delay time to give you what EH calls Chorus or Vibrato dependent on a switch setting. The switch simply goes between two cap values that determine the frequency of a traiangle wave oscillator. The two frequencies are 1hz and 1/4hz.
I set my scope input up on AC since the T wave was a 1Vac signal riding on 8Vdc.
I got this waveform that was hardly similar to a sawtooth. See the attachment .
I replaced the TL072 op-amp and the timing caps and generally traced the circuit against the schematic. The circuit did match the schematic but it was still putting out the erroneous waveform.
I got on line and looked up "triangle wave generator using a dual op-amp". Sure enough there was an example of the exact circuit. The first op-amp makes a square wave that is input to the second which is alternately charging and discharging a cap with a constant current - voila. this gives you a very linear ramp up and down not the typical cap charged through an R with a constant V like I was seeing on my scope.
Baffled, I decided to simulate it in Tina, Texas Instrument's simulator. It worked perfectly in simulation - beautiful symmetrical triangle wave.
Further baffled, I decided to breadboard it as this would be the first time I've ever had something not work in reality that worked in simulation (usually the other way around).
Damn if the breadboard did the same thing! - no triangle...................until I switched the scope to DC! Duh....the input coupling capacitor (inside the scope?) was not a large enough value to not distort a 1hz waveform.
Don't fall prey to this....I wasted, a couple days chasing this red herring down. (well, I learned something anyway.)
The pic is an image of both channel one and channel two that are probing the same point. The only difference is channel two is set on AC rather than DC. Excuse the sideways image.
I found a calibration procedure on-line and set about performing it. Not too far in I'm checking the LFO, low frequency oscillator which cyclically varies the delay time to give you what EH calls Chorus or Vibrato dependent on a switch setting. The switch simply goes between two cap values that determine the frequency of a traiangle wave oscillator. The two frequencies are 1hz and 1/4hz.
I set my scope input up on AC since the T wave was a 1Vac signal riding on 8Vdc.
I got this waveform that was hardly similar to a sawtooth. See the attachment .
I replaced the TL072 op-amp and the timing caps and generally traced the circuit against the schematic. The circuit did match the schematic but it was still putting out the erroneous waveform.
I got on line and looked up "triangle wave generator using a dual op-amp". Sure enough there was an example of the exact circuit. The first op-amp makes a square wave that is input to the second which is alternately charging and discharging a cap with a constant current - voila. this gives you a very linear ramp up and down not the typical cap charged through an R with a constant V like I was seeing on my scope.
Baffled, I decided to simulate it in Tina, Texas Instrument's simulator. It worked perfectly in simulation - beautiful symmetrical triangle wave.
Further baffled, I decided to breadboard it as this would be the first time I've ever had something not work in reality that worked in simulation (usually the other way around).
Damn if the breadboard did the same thing! - no triangle...................until I switched the scope to DC! Duh....the input coupling capacitor (inside the scope?) was not a large enough value to not distort a 1hz waveform.
Don't fall prey to this....I wasted, a couple days chasing this red herring down. (well, I learned something anyway.)
The pic is an image of both channel one and channel two that are probing the same point. The only difference is channel two is set on AC rather than DC. Excuse the sideways image.
Attachments
![IMG_0086[1].jpg](/community/data/attachments/837/837322-2b0b438397226f1891029227ec8078a5.jpg?hash=KwtDg5cibx)
I design low end USB scopes and have found with LF square waves etc the signal can start sloping off.
Unless you use a bi polar electrolytic on the front end its hard to get high value caps above about 470nF. If you do they are usually huge.
Test equipment can be misleading at times.
In m yearly years (1981) in electronics I was asked to look at a faulty modem.
I looked on the scope and it had terrible 50Hz on the signal.
It seemed to be all through the circuit.
I couldnt work out where it was coming from.
I told my boss and he laughed when he looked at my setup.
I had placed signal generator on top of the scope and magnetic field from sig gen mains transformer was radiating through the scope and modulating the signal.
Strangely I still do that now on my home setup with no problems so I can only guess someone had ripped off the earth on the scope in 1981.
I did get the problem on one piece of equipment I was working but that turned out to be my desk lamp with a CFL lamp in it radiating.
But at a much higher frequency.
Unless you use a bi polar electrolytic on the front end its hard to get high value caps above about 470nF. If you do they are usually huge.
Test equipment can be misleading at times.
In m yearly years (1981) in electronics I was asked to look at a faulty modem.
I looked on the scope and it had terrible 50Hz on the signal.
It seemed to be all through the circuit.
I couldnt work out where it was coming from.
I told my boss and he laughed when he looked at my setup.
I had placed signal generator on top of the scope and magnetic field from sig gen mains transformer was radiating through the scope and modulating the signal.
Strangely I still do that now on my home setup with no problems so I can only guess someone had ripped off the earth on the scope in 1981.
I did get the problem on one piece of equipment I was working but that turned out to be my desk lamp with a CFL lamp in it radiating.
But at a much higher frequency.
Last edited:
Most digital multimeters also have similar limitations. I also was recently looking at vibrato LFO distortion wrt the vintage phase shift valve oscillator, and confirmed the bandwidth limits of the meters, scope and soundcard I had available. Only one lab meter went down to 1Hz (but no further) and required the AC filtering setting to be changed. The old-but-not-too-old digital Tek scope luckily had a one-cycle rms and pp measurement capability with DC coupling. The soundcard was limited to 2Hz by the REW software, but could be calibrated flat to 2Hz.
'sic' is the expression you wanted: Shows the spelling is deliberate and not a mistype.