Vintage Motorola PLL chip

Post new topic   Reply to topic

View previous topic View next topic Go down

Vintage Motorola PLL chip

Post by VE7DES on Fri Jan 24, 2014 12:38 am

Thanks for posting the PLL articles re: MC14515x series.  May I present some memories?

My first job after leaving engineering school in 1981 was to implement a UHF-FM PLL TX using a MC145156 chip.  Early devices had no "dash" suffix.  Our client was the Toronto bus system and the application was voice/data/vehicle location.  The chip was very rare and expensive and we had only 2 samples.  I got into a lot of trouble when I crossed the 12 and 5 volt supplies and burned out one out!  The sample chips were in a MIL gold and ceramic package.  Production chips (in epoxy) cost many $CDN each.

We could not make the tri-state phase detector work well, so we used the phi-V/R outputs and a CA3140 op amp.  The 3140 was selected for its wide common-mode range (down to -Vsupply), good slew rate and low noise.  I still have some of the CA3140 chips, and I am using one in a LF signal booster for NDB listening.  The 3140 is still in production in Asia.  Nice chip.

The PLL topology was dual-modulus /64/65 using MC10178 and MC12011 ECL prescalers.  Later we replaced these clunky chips with a Plessey 8718, which used much less power and came in an 8-pin DIP.

Our VCO used a ~1.2mm-wide serpentine microstrip lambda-over-8 resonator on 1.6mm PTFE laminate with a 2N5397 JFET.  The better known U310 was too noisy for us.  This VCO was invented by the man who founded Nexus Engineering (who made TVRO earth station receivers).  Kvco was 0.5 MHz/volt and the varactor was a Motorola BB-series device.  Output was approx 5 dBM into a Motorola MWA MMIC, followed by 2 more MWAs for isolation and driving a Motorola MHW720 PA module.  Coarse tuning and Kcvo setting were done with Johannsen gold/sapphire trimcaps, and tuning range was 406-512 MHz.  We used surface mount capacitors in this design and had problems with hand soldering.  Someone pointed out that you are supposed to use silver solder for SMT work so a fellow from the purchasing department was detailed to obtain some.  He went straight into the Canadian Tire auto parts jobber next door and came back with a coil of acid-core silver solder.  Suitable for jewelry repair but not electronics.  Many VCOs were ruined until we figured out what had happened.  Despite such mistakes, and not to mention a lot of microphonics, the VCO was reliable.

The 12.8 MHz reference was implemented with an MC12061 ECL chip, and temperature compensated to +/-5 ppm over -30/+60 C.  Although the client used standard 16F3, the PLL had to step 12.5 kHz because the government allocated odd channels.  We had great difficulties controlling reference sidebands from the PLL loop filter.

One of our technologists built a box with 20 toggle switches and a CMOS loading circuit for use with in MC145156 protoyping work.  Flip the switches and watch the PLL step up and down.  Fun.  Another technologist, whose task was to adjust new VCOs from the production line, asked me my the TX frequency did not change whenever he adjusted the sapphire coarse tuning trimcap on his VCO.  He was seeing the tuning voltage go up and down and that puzzled him.  He had forgotten to open the PLL loop.

After modifications to the loop filter (to slow it down -- considerably), the 145156 TX design was successfully used to transmit 9600 BPS GMSK data in a long-distance UHF link in Manitoba.  Very fast data for the time (c.1986).  Our Toronto client eventually settled on 4800 GMSK, and for this application we invented a totally new (and more complex) PLL using a sampler topology.  We built over 2000 of the new radios for the bus and tram fleet in 1988.  They are still in operation but are at the end of their service life.

We also played with the 145152 and made some board-level UHF TX strips.  I am not proud of them!  I used our 145152 prototype as an emergency downlink for my club's 2 metre repeater telephone interconnect.  Could be the only 145152 chip to make telephone calls in the history of the world.

Also we tried the 145155 (?) which was different in that it was BCD serial programmable.  This may be the wrong number and/or not the same chip sold today.  My memory fails here.  We put this chip on some UHF-FM boards and sold them to a Rafael Garcia who transferred them to Cuba.  Garcia introduced us to Lacetel Ltd, a Cuban company that made crystals for mobile radio.  My firm was absolutely desperate to get its costs down and was not hesitant about buying parts with soft currency!

Our products relied on an 8085 MCU to program the 145156.  Because the 8085 was not always built into our products, I invented a stand-alone circuit for loading N/A/R data from a 2716 EPROM.  It could not continuously load the data without putting noise on the VCO.  So when the PLL locked, loading was inhibited and the transmitter would be quiet.  I used the same EPROM concept to build some automatic CW identifiers for amateur repeaters.

Another application of this MC145156 PLL design was a UHF-FM link for an advanced ROV submarine.  The prototype sub went out of control and sank in a fjord, and all the radios were exposed to the ocean at 300 m.  The sub was salvaged by another ROV and we got the radio boxes back.  I saved the radio boards and they are still in my desk drawer.  The PCB, the MC145156, the ECL and other components are corroded but still operate.  The submarine's mini UHF duplexer was also given to me, although it was full of water.  I drained it, cleaned it up, and installed it in the 2-metre telephone interconnect mentioned above.

I left the firm in 1989 due to redundancy.  I accepted 3 months' pay as severance and never looked back.

Dan of Canada (E.E., retired)

VE7DES
Guest


Back to top Go down

View previous topic View next topic Back to top

- Similar topics

Post new topic   Reply to topic
 
Permissions in this forum:
You can reply to topics in this forum