Palomar Digicom 100 by CPI

[nomadradio]

Someone must have made custom chips at some point

You mean, like, a substitute for a 1975 part made from more-modern devices?

Good luck with that.

This chip stores 128 8-bit words. The stored pattern comes from the photo mask that's used to lay down the chip's internal wiring pattern. They took the desired bit patterns and put the zeros and ones on the chip when it was etched, along with all the other transistors, resistors and diodes that make up its internal circuitry.

Pretty much the picture of "custom". But all it does is to translate. The binary outputs of the digit switches feed into the MM5221. The output is the translated 8-bit binary code stored inside. This is what feeds the actual programmable-divider chips IC5 and IC6. This is where the binary channel selector would connect to a single-chip PLL in a newer-design type radio. No need to translate if the selector is custom-made to deliver the desired forty binary codes.

This radio isn't exactly steampunk, but it's about as close as you'll get.

73

 

[TM86]

You mean, like, a substitute for a 1975 part made from more-modern devices?

Good luck with that.

For a "I had no friggen choice" modern replacement you could try a pic or arduino. Might need to do some level conversion depending on what parts you used and it would most likely be ugly as hell, but it could work.

 

[nomadradio]

Thirty-plus years ago we were hand-wiring an adapter for the Browning Mark 4 transmitter. It used a primitive mask ROM like this one to do the same job, translating from two 4-bit BCD digits to a binary code needed by the transmitter's PLL.

The original part was built with what's called "Depletion mode metal-gate PMOS" transistors inside. Like the 5221 it needs more than one separate supply voltage, and one of them is negative polarity. This technology had stability and reliability issues.

The state of the art for ROMS in the early 80s was an erasable part called the 2716. It ran from a single positive 5-Volt supply and would not change data if the temperature drifted up a bit. Was made with "Ehancement mode N-channel silicon gate" transistors on the inside. This eliminated the need for multiple power supply voltages and made it much more temperature-stable. The pinout, however is totally different. We figured out the stored data pattern and burned it into a 2716. A perf board with a 24-pin socket with a flat cable leading to a matching DIP plug took a bunch of wiring to build. The plug went into the radio's ROM socket and an appropriately-programmed 2716 went into the adapter board's socket. One advantage was that the regulator for the negative 9 Volts could be removed from the receiver altogether, along with the heat it produced.

Didn't have to build that many to get tired of it. Eventually we gave up on the Mark 4's primitive two-and-a-half dozen chip channel selector/PLL forever. Even with the flaky "poof" chip's problems eliminated it was an unreliable flaky train wreck of a design.

Getting custom printed circuits made 35 years ago had a much-higher upfront cost than it does these days, so there was never any chance of that toy ever becoming a product.

But if push came to shove, that's what I'd do if this radio's ROM were to go over a cliff.

73

 

[357]

I wonder if a chip reader could copy the info?

Thirty-plus years ago we were hand-wiring an adapter for the Browning Mark 4 transmitter. It used a primitive mask ROM like this one to do the same job, translating from two 4-bit BCD digits to a binary code needed by the transmitter's PLL.

The original part was built with what's called "Depletion mode metal-gate PMOS" transistors inside. Like the 5221 it needs more than one separate supply voltage, and one of them is negative polarity. This technology had stability and reliability issues.

The state of the art for ROMS in the early 80s was an erasable part called the 2716. It ran from a single positive 5-Volt supply and would not change data if the temperature drifted up a bit. Was made with "Ehancement mode N-channel silicon gate" transistors on the inside. This eliminated the need for multiple power supply voltages and made it much more temperature-stable. The pinout, however is totally different. We figured out the stored data pattern and burned it into a 2716. A perf board with a 24-pin socket with a flat cable leading to a matching DIP plug took a bunch of wiring to build. The plug went into the radio's ROM socket and an appropriately-programmed 2716 went into the adapter board's socket. One advantage was that the regulator for the negative 9 Volts could be removed from the receiver altogether, along with the heat it produced.

Didn't have to build that many to get tired of it. Eventually we gave up on the Mark 4's primitive two-and-a-half dozen chip channel selector/PLL forever. Even with the flaky "poof" chip's problems eliminated it was an unreliable flaky train wreck of a design.

Getting custom printed circuits made 35 years ago had a much-higher upfront cost than it does these days, so there was never any chance of that toy ever becoming a product.

But if push came to shove, that's what I'd do if this radio's ROM were to go over a cliff.

73

Could something like this be used to read the chip?
http://www.compendiumarcana.com/eprom/

Also do you know what the stud mount final is?

thanks

 

[unit_399]

From what I can dig up, Sams 103 lists the final as a 2514 which subs to a Motorola HEP S3006. Trouble is, this transistor's frequency is shown as 175mhz. What happened to all of the photos of the rig that you posted ?? The post is shown as "NA."
Can you check the final and see if the designation stamped on it is actually 2514.
Anyway here is the data for the S3006.

- 399

 

[BJ radionut]

What happened to all of the photos of the rig that you posted ?? The post is shown as "NA."

Still here scroll UP

 

[nomadradio]

I wonder if a chip reader could copy the info?

Only if MM5221 appears on its list of supported devices. Or if the software includes a utility to "roll your own" custom configuration file. Just finding a programmer than can place a negative-polarity ten Volts on the chip may prove a struggle.

73

 

[tramd]

Years ago on the Yahoo CPI group there was a lot of discussion about this and no one was ever able to come up with a programmer that was able to read or write to any of those proms or eproms that were used in the Digicoms or CPIs.

 

[AudioShockwav]

Maybe someone across the pond might have an idea, in the UK using EPROMS was much more popular than here.

73
Jeff

 

[nomadradio]

no one was ever able to come up with a programmer that was able to read or write to any of those proms or eproms

We reverse-engineered the binary codes for the Browning Mark 4. Never owned a programmer old enough to work with a 1302 or 1702. Built one in 1983/84. Never set it up for any part that needed two or three voltages just to read the data. Was no need to by then. None of the computers we supported used a ROM that old.

The PLL binary part on the ROM's output side will be easy once you know the code for channel 1, then channel 2. This will tell you if the binary sequence counts up with the channel number, or down. Just remember to skip one code for each RC channel.

The input side will be a brute-force kind of thing, simply checking the logic state of each address pin on the ROM at each setting of the two channel knobs. Once a pattern emerges from that, you can probably predict the rest of the codes without explicitly checking and recording the state of each bit on every frequency.

Had to hand-enter each memory location in hex to come up with our 2716 retrofit EPROM. Wasn't that bad. Only have to do it once, then save it as a file.

Sometimes "impossible" and "pain in the a$$" are in fact the same thing.

Depends on who you're asking.

73

 

[357]

thanks for the info folks.

I'm glad my chip works and has the extras.
I assume there weren't many of these radios made?

 

[unit_399]

Not many out there at all. Once the FCC withdrew the type acceptance, production (and sales) ceased.
-399

 

[unit_399]

When I worked at TRIGON ELECTRONICS in SoCal, we made Z80-based entry systems for hi-rise apartments and gated communities. I had to dig deep through my old files, but I found that we used a ProLog M980 prom programmer to program our Eproms. I did a quick search and found them on NewEgg and Ebay. Also lots of National Semiconductor MM5203Q blank Eproms available. We would put our master eprom and a blank in the machine and it would make a copy. Most of the early arcade machines (Pacman, Galaga, etc.) were Z80-based and there are a lot of arcade resto companies that could read your chip, copy it, and/or help you with custom programming. That CPI 100 is worth a bundle, and it might be worthwhile to develop some custom chips to unleash its potential.

- 399

 

[tramd]

I know this is a bit old but 357 was wondering about channel skips. I have a CPI 300 that has a switch on the front like that Digicom 100 does and it is for 5k skips. Some one added it to it. Mine also has a different chip that allows it to go to channel 79. I read an explanation a while back from Dave Wisherd who was one of the founders of CPI and he stated that they knew the FCC was going to add channels in 1977 but no one knew how many so they designed the Palomars and CPIs to go to 99 channels by using the two selectors. When they found out that the limit was only 40 they used a single selector in the new CPI 2000 base radio like most other companies did but continued using two selectors in the CPI 400 as it was basically the same as the 300 but was now a 40 channel radio and they didn't want to redesign it.

 

[357]

I know this is a bit old but 357 was wondering about channel skips. I have a CPI 300 that has a switch on the front like that Digicom 100 does and it is for 5k skips. Some one added it to it. Mine also has a different chip that allows it to go to channel 79. I read an explanation a while back from Dave Wisherd who was one of the founders of CPI and he stated that they knew the FCC was going to add channels in 1977 but no one knew how many so they designed the Palomars and CPIs to go to 99 channels by using the two selectors. When they found out that the limit was only 40 they used a single selector in the new CPI 2000 base radio like most other companies did but continued using two selectors in the CPI 400 as it was basically the same as the 300 but was now a 40 channel radio and they didn't want to redesign it.

I think it goes 5kc steps after channel 30. 38 is ch75
the dial on the left has two 1's and two 0's and is missing 4 and 5.
I remember seeing this radio in cb books when I was a kid and it always said it was waiting type approval and futuristic channel map.
Glad I got one