Time Machine (-) CV input & bypass modification – Revised 02/27/02
Notice: I have revised this document to remove one the modification for negative voltage protection of the External Mod CV. The modification designed to protect the Delay CV input actually protects both quite well. Thanks to Scott Juskiw for pointing this out to me.
This document contains all of the tested modifications I am currently aware of for the Blacet Time Machine module when constructed in the MOTM format. Certainly, I would love to hear from you if you have other useful additions or changes to this superb module.
The first modification involves adding a bypass switch to the Time Machine. What this
switch actually does is mute the "wet" part of the signal. Any dry signal in the mix is allowed to pass. This modification information is included in my assembly instructions if you care to use them. Or you can find the original documentation at Dave Bradley’s Hot Rod MOTM site: http://www.hotrodmotm.com/tm_mods.htm.
Here are some photos of my bypass switch modification too:
This one shows the resistor mounted to the switch:End of Page 1
Component leads left long to make it easy to attach wires:Wires attached to the PCB:
End of Page 2
Bypass switch theory: You can see from the schematic that the external “cancel” (or bypass as identified on the stooge panel) causes current in U13c by applying positive voltage through a 100K resistor to pin 9. The bypass switch modification does also.
The positive voltage is pulled directly from the power supply +15v rail (on my PCB, I used the untrimmed lead of PS1). The voltage is switched and then applied through a resistor to pin 9 of U13c (on my PCB, I used the untrimmed lead of R42). The output of U13c in turn applies negative voltage to pin 3 of the SSM2164 causing its output to go to zero.
The other modifications is the result of a problem that can occur if you apply negative control voltage to either the Delay CV input or the External Mod CV. The negative voltage can cause the clock to lock-up and your Time Machine will not function until you turn it off and back on allowing the clock to reset. We are protecting two inputs with only one modification.
Theory of the Delay CV modification: This modification protects the Delay CV input. My first attempt at this solution did just that, shunted negative voltages to ground at the input with a diode and resistor. However, I found that had a negative side effect. Since the (-) input of U1d is normally held negative to set the initial range of the HF VCO, the negative shunt on the input was a path to ground for current that otherwise would have been summed into U1d. The effect was limiting the upper frequency of the HF VCO to about 250 KHz. John specifies calibration to 400KHz. Therefore, that idea was not as good as originally hoped. What I found by measurement is that the voltage on the output of U1d at pin 14 is about 1 volt positive at 400KHz and goes toward zero and significantly negative as the delay time is increased. I found that negative input voltages push this voltage over 2 volts positive (it is an inverting summing amp) causing the HF VCO to approach a frequency of 1MHz and then the clock locks up. So, the solution is to limit the output of U1d at 2 volts positive and not effect its range below 1 volt positive. The solution does just that. In fact, the voltage is limited to only 1.4 volts with a full 5 volt negative voltage. I could not apply enough negative voltage to get the clock to lock up. And, protection at this location fortunately protects both “Delay CV” and “External Mod CV” inputs.
Construction of the CV input protection modification: This assumes you have already built your Time Machine and you are retrofitting this modification after the fact. If you are just building your Time Machine, just follow the direction of my assembly instructions. You need one 1K 1% resistor, and two 1N4148 or equivalent signal diodes. I also recommend a little heat shrink.
De-solder and lift one end of R11 (45.3K ohm) so that it is in a standing position. You are lifting the end closest to the edge of the PCB. R11 is now standing in the hole that is closest to Q1. At the opposite hole (toward the edge of the PCB), insert your extra 1K ohm 1% resistor also standing on end. Now, tie together by twisting the two tops of these two resistors. They now have a “tee pee” style appearance on the PCB. Do not solder the tops together yet. Do solder to PCB.
End of page 3
Take your two remaining 1N4148 signal diodes and connect them end to end about ½ inch apart (removing a significant portion of their leads), Connect so that one cathode connects to one anode, or so the cathode stripes are both facing the same direction. I stuck about 1 ½ inches of heat shrink over this so that I end up with a component inside (composed of two diodes in series) and I have a nice long lead sticking out of each end. The heat shrink is not required, but makes the modification bullet proof in my opinion. If you do add the heat shrink, be sure to mark which end is the cathode. Now, you are ready to install your diode combo. The anode end connects to the twisted junction point of R11 and the 1K resistor that stands on end tee pee style at the R11 location. The cathode lead connects through the via hole just below R14 and just above the "S" in RESEARCH. This location is electrically ground. Now you can solder the top connection of R11, the new 1K resistor and the anode end of your series diodes. Photos can be found on page 4.
The diodes are covered in heat shrink to prevent shorting with any other components on PCB /
End of Page 4
The diodes run from the via hole up to the point where R11 and the new 1K ohm resistor meet at the top /Disclaimer: I am not an engineer. I am not recommending that you modify your Time Machine. However, I am sharing the results of my own modification for your consideration should you find you have a similar need. I am not responsible for damage to your Time Machine.
Larry Hendry 02/27/2002 – Revision 2
End of Page 5 – End of document