|
This MEMS 245M is a faithful recreation of the 245 with the addition of a front panel PCB for wiring of all the panel components. MEMS chose an alternate illuminated switch since the Dialco are expensive and rare. MEMs also designed SMT replacement modules for the obsolete and hard-to-find ICs.
The pushbutton switch LEDs will operate directly from 24V.
Construction
As with many of my MEMS builds, these are prototype boards that may require some minor modifications. I send this information to MEMS for improvements. As such your PCB images may be slightly different. This 245M panel PCB simplifies the component wiring and contains the +15V to +24V DC-DC converter.
The panel PCB is assembled and wired to the panel.
The 245 design uses a number of obsolete components. These are their replacement modules using CMOS logic. There are a few discrete parts in 0603 size that take careful soldering.
I bought round PCB pins from Amazon for the headers. There are a number of different header pins available.
10pcs 2.54mm Connector Round Hole Row Needles 140 Round The Needle Male Head 1x40P Gold Plated Single Row
Brand: ElectronicNova
That there are two pads on the IC7 SP380 that need to be soldered terminate the unused CMOS inputs.
Since the replacement modules have round pins, I used machined sockets for those and dual wipe sockets for the ICs. There are a few not-used components with these replacement modules in the lower right corner. I always build with 1% resistors since there is little price difference.
There is a lot of panel wiring so I chose my order to wire from the bottom up on the panel PCB to minimize interference.
Modifications
Don chose to use the start signal from IC2 pin 3 through R20 to bias Q6. There is a negligible amount of current through this when stopped which is not an issue for an incandescent bulb. However, it will dimly light the more efficient Start LED. I added a 4K99 resistor in parallel with the LED to bypass this current so the Start LED is extinguished when stopped.
The Start and Stop LED operate on 24V. The panel PCB also has optional resistors in series with them. I chose 3K3 in series to dim them just a bit.
The desired Analog In voltage levels are subject to some debate. This note is from the original schematic. The 245 analog in has no adjustment and the stage voltages are fixed.
The stock values result in this set of voltage thresholds. The input is attenuated by R113/115 and then used for the comparators. Note all tables assume 1% resistors at nominal value.
Stage | Comparator Resistors | Comparator | Analog In V | Mid Voltage | Delta | ||
Min | Max | Min | Max | ||||
None | 0.00 | 0.26 | 0.00 | 0.70 | 0.4 | ||
1 | 820 | 0.26 | 0.57 | 0.70 | 1.55 | 1.1 | 0.8 |
2 | 1000 | 0.57 | 0.88 | 1.55 | 2.41 | 2.0 | 0.9 |
3 | 1000 | 0.88 | 1.19 | 2.41 | 3.26 | 2.8 | 0.9 |
4 | 1000 | 1.19 | 1.51 | 3.26 | 4.12 | 3.7 | 0.9 |
5 | 1000 | 1.51 | 15.00 | 4.12 | 15.00 | 4.5 | |
R124 | 43200 | ||||||
R113 | 82000 | ||||||
R115 | 30000 | ||||||
Attn | 2.73 |
Some people want a 10V range for the analog input. Replacing R116 with 1K and R124 with a 18K2 resistor results in this set of voltage thresholds which are close to 1.75V apart.
Stage | Comparator Resistors | Comparator | Analog In V | Mid Voltage | Delta | ||
Min | Max | Min | Max | ||||
None | 0.00 | 0.65 | 0.00 | 1.77 | |||
1 | 1000 | 0.65 | 1.29 | 1.77 | 3.53 | 1.47 | 1.77 |
2 | 1000 | 1.29 | 1.94 | 3.53 | 5.30 | 4.42 | 1.77 |
3 | 1000 | 1.94 | 2.59 | 5.30 | 7.07 | 6.19 | 1.77 |
4 | 1000 | 2.59 | 3.23 | 7.07 | 8.84 | 7.95 | 1.77 |
5 | 1000 | 3.23 | 15.00 | 8.84 | 15.00 | 9.72 | 1.77 |
R124 | 18200 | ||||||
R113 | 82000 | ||||||
R115 | 30000 | ||||||
Attn | 2.73 |
I personally like octave settings, so on 1.2V intervals. Replacing R116 with 475R and R124 with a 29K4 resistor results in this set of voltage thresholds very close to 1.2V/Oct.
Stage | Comparator | Comparator | Analog In V | Mid | % of | Delta | ||
Resistors | Min | Max | Min | Max | Voltage | Range | ||
None | 0.00 | 0.21 | 0.00 | 0.57 | ||||
1 | 475 | 0.21 | 0.65 | 0.57 | 1.79 | 1.18 | 100.9% | 0.57 |
2 | 1000 | 0.65 | 1.10 | 1.79 | 3.00 | 2.39 | 100.9% | 1.21 |
3 | 1000 | 1.10 | 1.54 | 3.00 | 4.21 | 3.60 | 100.9% | 1.21 |
4 | 1000 | 1.54 | 1.98 | 4.21 | 5.42 | 4.81 | 100.9% | 1.21 |
5 | 1000 | 1.98 | 15.00 | 5.42 | 15.00 | |||
R124 | 29400 | |||||||
R113 | 82000 | |||||||
R115 | 30000 | |||||||
Attn | 2.73 |
Two more options for the 245M would be to set them
the same as the first 5 stages of the 246 (they are close) or to have the full
15V voltage range of
the 246. For comparison, this is the voltage thresholds for the 246 calibrated
to 14.5V at stage 16.
Stage | Comparator Resistors | Comparator | Analog In V | Mid Voltage | Delta | ||
Min | Max | Min | Max | ||||
None | 0.00 | 0.28 | 0.00 | 0.75 | 0.38 | ||
1 | 820 | 0.28 | 0.61 | 0.75 | 1.67 | 1.21 | 0.83 |
2 | 1000 | 0.61 | 0.95 | 1.67 | 2.59 | 2.13 | 0.92 |
3 | 1000 | 0.95 | 1.28 | 2.59 | 3.50 | 3.04 | 0.92 |
4 | 1000 | 1.28 | 1.62 | 3.50 | 4.42 | 3.96 | 0.92 |
5 | 1000 | 1.62 | 1.95 | 4.42 | 5.34 | 4.88 | 0.92 |
6 | 1000 | 1.95 | 2.29 | 5.34 | 6.25 | 5.79 | 0.92 |
7 | 1000 | 2.29 | 2.62 | 6.25 | 7.17 | 6.71 | 0.92 |
8 | 1000 | 2.62 | 2.96 | 7.17 | 8.09 | 7.63 | 0.92 |
9 | 1000 | 2.96 | 3.29 | 8.09 | 9.00 | 8.54 | 0.92 |
10 | 1000 | 3.29 | 3.63 | 9.00 | 9.92 | 9.46 | 0.92 |
11 | 1000 | 3.63 | 3.96 | 9.92 | 10.84 | 10.38 | 0.92 |
12 | 1000 | 3.96 | 4.30 | 10.84 | 11.75 | 11.30 | 0.92 |
13 | 1000 | 4.30 | 4.64 | 11.75 | 12.67 | 12.21 | 0.92 |
14 | 1000 | 4.64 | 4.97 | 12.67 | 13.59 | 13.13 | 0.92 |
15 | 1000 | 4.97 | 5.31 | 13.59 | 14.50 | 14.05 | 0.92 |
16 | 1000 | 5.31 | 12.00 | 14.50 | 15.00 | 14.75 | 0.71 |
R124 | 27000 | ||||||
R124A | 1900 |
Trimmer adjustment to 14.5V at stage 16 |
|||||
R113 | 82000 | ||||||
R115 | 30000 | ||||||
Attn | 2.73 |
Calibration
There is just one calibration for the internal time range. I adjust it so that at maximum it reaches 10 seconds to agree with the panel legend. This is the Reference signal and Pulser output.
At 10 seconds, you can't see the trigger portion of the pulser output. This zoomed scope image shows indeed it is there.
Calibrated for 10 seconds, the minimum time is 1.8 mS, a bit longer than the panel legend of 1 mS. How nice it is to see full 15V waveforms.
This photo shows the four CV outputs with stepping down from 15V with each channel cascaded.