mc12017 and mc145152

[radium98]

Happy new year 2017 .
i just came on an old project that i did not have finished yet it is 1w pll fm free radio berkely anyone know it ?
hope so and can help me i will put some pictures my problem cannot find prescaler mc12017 scale by 64/65 so i am trying to replace by u664 any help .VCO work but have no locks prescaler also work
any help

[Analyser]

You need to check the oscillator is near the frequency you need, and also check what the output of the LF351 op-amp is doing.
Have you got an oscilloscope? It would be useful to check the output of your new prescaler as well.

I don't really understand the placement of the 5-30pF trimmer in the circuit, in it's current place it'll vary the coupling the PLL has on the oscillator, so not a great choice if you want to use it to change the VFO frequency. I would make that trimmer a fixed value, say 22pF, and adjust the 100nH coil to make sure the circuit locks correctly across the FM band.

[radium98]

Analyser thank you for the quick reply to help
oscillator is working as i can tune the variable coil to the desired frequency i tune to mid band here 98.00 mhz for test
lf351 at its output i have nearly some volts
i the near moment i dont have in my hands an oscilloscope but i can see the prescaler working well in other txer transmiter and also i can check the scaling factor on frequency counter and it its ok .
i will try to remove the trimmer as you said
but as a question mc12017 scale by 64/65 and have pin 1 MC u664 scale by 64 also but have no MC will that matter a lot ? i dont think so isnt so what could be the problem .
(i dont have another mc145152 but i have an mc145151 in my hands ) but they are not compatible to test.
so what to do ?

[Analyser]

I'm not really familiar with either of those prescalers, but if the original on was divide by 64/65 there must be some way of selecting 64 or 65.
For example, if the original circuit is using the divide by 65 and you new IC is dividing by 64 then the circuit won't lock or may lock to the wrong frequency. Have a look at the data sheet for the original prescaler and see if you can work out which mode it's in.

[Analyser]

Have a look at the MC pin of the MC145152, if it's 5V then your prescaler should work. I had a quick look at the datasheet for the MC12015 and pin 1 controls the divide ratio, 0V for 65 and 5V for 64.

[Albert H]

The U664 has a completely different pin-out to the MC12017. You can't just plonk the wrong IC in the board and expect it to work!

[Analyser]

The U664 has a completely different pin-out to the MC12017. You can't just plonk the wrong IC in the board and expect it to work!

Look at the bottom photo Albert.

[Albert H]

Missed that! Bit of a kludge, and bearing in mind the frequencies involved, random lengths of wire aren't going to do much to help the cause!

Prescalers are often a source of problems. I used to use the SAB664 for a UHF prescaler and then the manufacturers made some subtle changes to it (supposedly to improve its sensitivity) and they broke it! It loaded my VCO with lots of extra capacitance, and the switching noise that it put on to the 5V rail was diabolical!

I abandoned the use of the low frequency Motorola PLLs at that point (I sold my stock of them on Ebay!) and went over to the TSA5511, which is happy to work to over 1400MHz, and is easily set with a PIC. The other advantage of the 5511 is that I can have a wide varicap voltage range, which allows the use of "unconventional" varicaps. I've successfully used 1N4003 rectifier diodes and even LEDs as varicaps (the green ones seem to work best). "Real" varicaps are getting expensive!

Just for grins, I decided to throw together a "minimal" 1W exciter, to see how simple I could possibly make a repeatable design, and how cheap it could be. I abandoned the use of PLL ICs - they're too expensive - and went back to discrete logic. A 74ALS74 will divide a Band II signal by 4, to bring it into the range of the 74HC logic series. I didn't even bother with a 5V regulator - a resistor and a zener were cheaper, and almost as effective.

The oscillator uses a single-gate FET (BF244B) and a BF199 in a Hartley configuration (5 turn coil tapped at 1 turn). This gives a surprisingly robust signal, and was almost enough to drive the output device directly, but I decided that I wanted to make sure that the output device received plenty of "push", so I used an un-tuned pair of BF199s. The final device posed a bit of a problem - I could have used a BFG135 (since I have a tube of them that I got for 20p each), but that would be cheating. I considered using an IRF740 FET - I've done it before and it's very cheap, but it's nasty, needs some ferrite, and wouldn't necessarily be repeatable. I tried a BFY51 and was amazed to get about 550mW out of one, but it ran very hot, and other examples of the device didn't work as well. In the end, I bit the bullet, and used a 2N4427. This is the single most expensive part in the whole circuit!

The PLL logic is a 74HC4040 with diode programming to select the frequency, a 74HC4060 for the reference oscillator and divider, a phase comparator using a 74HC74 and a 74HC00, and a passive loop filter feeding a single transistor DC amplifier to control the varicaps.

The results out of the end of the lowpass filter were surprisingly good. Carrier noise was low, a 12V supply gave exactly 1 Watt into 50Ω anywhere in the band, and the modulation sensitivity was quite consistent across the band. The first one was on a piece of PCB material about 4½" × 2½", but a fully double-sided etched PCB could be a lot smaller. The programming diodes are below the board, and the top side has no adjustment points at all. I made a screen to separate the digital part from the RF, and then decided to build a copperclad box for the whole thing. It's now about 4½" × 2½" × 1".

Total parts cost was about £7 if you had to buy stuff in (and you didn't go to Maplin). I'll finish designing a PCB for it and make the board available from OHSPark or one of the other cheap Far Eastern PCB companies. This could be a very viable solution for cheaper rigs - it also has the advantage that the average numpty rig thief won't have a clue about how to change its frequency - especially if the exciter box is completely filled with epoxy resin!

[Shedbuilt]

as a question mc12017 scale by 64/65 and have pin 1 MC u664 scale by 64 also but have no MC will that matter a lot ?

Yes. It does matter - unless I'm severely misunderstanding the design.
The MC145152 "N" and "A" counters control the Dual Modulus prescaler via the "MC" line.
With a Dual Modulus prescaler, this gives a divide ratio of (N x 64) + A.
The "R" counter is dividing by 256, to divide the 5.12MHz crystal down to 20KHz.
Using the FRB switch settings for 88.1MHz (1010100100), the "N" counter value is set to 68 (N6 is "NC", therefore permanently pulled high by a pull-up resistor), and the "A" counter value is set to 53 (A0 is permanently pulled high). So our divide ratio is (68 x 64) + 53, or 4405 (which is the divide ratio necessary to divide 88.1MHz down to 20KHz).
I believe the U664 is a single modulus /64 prescaler, and I don't think this design is feasible with a single modulus prescaler.

[Shedbuilt]

If you just want to get this to lock, and prove that it works - rather than do anything useful with it, you could try the following. Connect the A0 pin (pin 23 of the MC145152), to ground. Then set the switches to 0000011010. If everything is connected correctly, and working, I "think" you should be able to get it to lock to 96.0 MHz.
The "A" counter, tells the PLL chip, how many cycles (of the total divide cycles), to tell the prescaler to divide by 65 (the remainder of cycles being implemented as divide by 64). So by setting all of the "A" pins to 0 (which - according to the datasheet, is a valid value), the PLL chip will only expect it to divide by 64.
The downside, is that - with a single modulus /64 prescaler, and a reference frequency of 20 KHz, your frequency steps are 1.28 MHz. This only ends up on an exact multiple of 100 KHz, every five increments (6.4 MHz). So for exact frequencies in Band II, your choices would be 89.6, 96, or 102.4 MHz. I picked 96 MHz, because it's the closest to the centre of band which you mentioned before. If my maths, and quick reference to the schematic and MC145152 datasheet worked out right, the above should give an "N" counter value of 75. With "A" counter value of 0, this should give a divide ratio of 75 x 64, or 4800. Multiplying 4800 by the reference frequency of 0.02 MHz, should give 96 MHz.

[shorty]

The MC12017 is available from this guy on ebay in Germany, the ic's are now obsolete, either not available or if you can find them they are expensive.

http://www.ebay.co.uk/itm/MC12017P-Oper ... SwImRYKgu2

I've never built the frb design you have above, but i have used the MC145152 with a Plessey SP8793 to devide by 40 that works well for band 1 50Mhz, band 2 broadcast band upto band 3 200Mhz.

I used a 6.4Mhz crystal with pins 4, 5, 6 set to devide by 1024 (1 0 1) giving 6.25Khz, so my lowest dip switch is 6.25Khz and the highest 128Mhz, then just work my way down from the highest frequency either on or off.

You will probably have to slightly redisgn the prescaler ciruit layout if you don't go with the mc12017, if you do that you may aswell alter circuit to use a 6.4Mhz as these are inexpensive, try the guy in Germany first.

[radium98]

thanks for all youre suggestions i will try to get first the prescaler on aliex...
but one question is there modifications i can do to try to lock on any frequency with the u664 ?
please ref to schematic above i need some help

[Shedbuilt]

thanks for all youre suggestions i will try to get first the prescaler on aliex...
but one question is there modifications i can do to try to lock on any frequency with the u664 ?
please ref to schematic above i need some help

Hi radium, My last post (above shorty's), was a suggestion, which I think is worth trying. If it works, it should lock on 96.0 MHz. Does that help, and does the post make sense ?
I haven't looked at your re-wiring for the U664, but I think Analyser did before. If it's correctly wired, I think what I've suggested, should work with a single modulus divide by 64 prescaler (eg the U664).

[radium98]

Shedbuilt yes surely it make sence i will try .Also my wiring is ok ground and vcc and 1 input to ground via capacitor and second input to signal in from vco and the two outputs connected together

[Shedbuilt]

Shedbuilt yes surely it make sence i will try .Also my wiring is ok ground and vcc and 1 input to ground via capacitor and second input to signal in from vco and the two outputs connected together

I can't find a datasheet for the U664, but I'm not sure that connecting the outputs together is correct. Isn't one an inverted version of the other (not sure) ?
Was it used this way in the other design ?

[radium98]

[img]

719_freque3k.gif

[/img]look here for datasheet and give me ure advice
https://www.sendspace.com/file/0130aw original datasheet for dip package and the sip package is in that schematic

[Shedbuilt]

[img]

719_freque3k.gif

[/img]look here for datasheet and give me ure advice
https://www.sendspace.com/file/0130aw original datasheet for dip package and the sip package is in that schematic

It does look like the inputs and outputs are differential, and in the SIP package schematic, they have only used one of each. Pin 6 (input), is grounded by a 10nF capacitor, with the input on pin 5 driven. Output is taken from pin 3, with pin 2 unconnected. I would do this as a starting point. From what you said, I think the only thing you did differently, is to connect the two outputs together. Don't forget to ground the A0 pin of the MC145152. This is normally left unconnected in the FRB design, but needs to be pulled down to 0 for this exercise (along with the DIP switch settings in my earlier post).

[shorty]

When i googled u664 this is the pin out i found for that device, i wouldn't tie two pins together either on the u664, i know it's shows pins 2 & 3 tied together on the mc12017 but the u664 doesn't have the same pin out, i'd try the pin out in this schematic.

[Shedbuilt]

When i googled u664 this is the pin out i found for that device, i wouldn't tie two pins together either on the u664, i know it's shows pins 2 & 3 tied together on the mc12017 but the u664 doesn't have the same pin out, i'd try the pin out in this schematic.

I think radium has the 6 pin SIP package shorty, and I think has worked out the pinouts, and mostly hard wired it above the board. Radium mentioned one input to ground - via a capacitor, and the other capacitor coupled to the signal.

Also my wiring is ok ground and vcc and 1 input to ground via capacitor and second input to signal in from vco and the two outputs connected together.

It was the part about both outputs connected together which jumped out at me.

[yellowbeard]

It's not really a solution, but Dónberg have that part - its goddam expensive though - 35 Euro ex VAT.

Code: Select all

http://www.donberg.ie/descript/m/mc_12017.htm