my DAC

[mstechca]

After all the nice answers regarding my DAC, I think I'll stick with a few hundred Kiloohm resistors at the outputs.

Now here is the interesting part. I made myself a transmitter that makes one tick a second. I turned on my receiver (with the DAC circuit in it), and I am actually picking up the signal on one station (or binary state). Not bad, considering NO ONE TAUGHT ME IN SCHOOL (sorry, but I had to yell that out.)

It seems that now I might have to increase the number of bits in the DAC just to accompany for the stations that fall in between one binary state, and the next.

I'm wanting to cram my entire radio into a 5cm by 3.5cm space and I can do it with a 4-bit DAC. If I do an 8-bit DAC, then the IC could present me with problems.

If I were to avoid any trim capacitors, what is the minimum number of bits most radios use in their DAC's when selecting stations?

I don't want to use too many bits, because it will take longer for a scan,and it will take more board space. I don't want to use too few bits because it will skip too many stations inbetween.

oh, and by the way, I determined that the 2N3904 transistor makes the BEST varicap diode for me.

 

[audioguru]

Hi MStechca,
Digital tuners don't use a DAC. They use a crystal controlled frequency synthesizer with exactly 200kHz between frequencies on the FM band.

Instead of using jerky DAC tuning steps, why don't you make your receiver tune with a smooth voltage ramp from a charging capacitor, like the scanning FM radios in the Dollar Store?

 

[mstechca]

Instead of using jerky DAC tuning steps, why don't you make your receiver tune with a smooth voltage ramp from a charging capacitor, like the scanning FM radios in the Dollar Store?

Digital is flexible.

and wouldn't me switching over to your idea require extensive modifications, or extensive error checking?

 

[Roff]

Hi MStechca,
Digital tuners don't use a DAC. They use a crystal controlled frequency synthesizer with exactly 200kHz between frequencies on the FM band.

Instead of using jerky DAC tuning steps, why don't you make your receiver tune with a smooth voltage ramp from a charging capacitor, like the scanning FM radios in the Dollar Store?

I agree, except... how do you do an "infinite" sample and hold (or stop and hold) on the ramp? Seems like that's just a way for the PLL to acquire the next signal. Maybe they use AFC (see below).
Mstecha, digital is flexible but discrete. You will never be able to stop it on the exact center frequency of all stations, unless you have something like automatic frequency control . BTW, I have no experience with this technique - I just know that it exists.

 

[audioguru]

Hi Ron,
Philips have many new radio ICs. This one **broken link removed**
is for FM stereo and RDS data and is crystal controlled with a frequency synthesizer. It can use I2C from a microcontroller to search for a station if you want.
I couldn't find their old scanning FM radio IC.

 

[Roff]

Hi Ron,
Philips have many new radio ICs. This one **broken link removed**
is for FM stereo and RDS data and is crystal controlled with a frequency synthesizer. It can use I2C from a microcontroller to search for a station if you want.
I couldn't find their old scanning FM radio IC.

Hi Audioguru (or is it Scrooge?),
Yeah, I sorta kinda understand synthesizers. I was just wondering about the function of the analog ramp in the Dollar Store receivers.

 

[mstechca]

You will never be able to stop it on the exact center frequency of all stations...

unless I use alot of bits (probably at least 20 or so) and the correct varicap is used.

 

[mstechca]

I have a comment.

With the DAC, I think I was stupid to remove the (680K resistor) voltage divider I initially planned to use with my DAC. why?

because:
a) I can't figure out the voltage that is fed into the varicap when the bits are all 0, because all resistors are grounded, and there is not one that is connected to VCC.

and

b) I can't figure out the voltage that is fed into the varicap when the bits are all 1, because all resistors are connected to VCC and there is not one that is grounded.

therefore, I can't mathematically determine the voltage coming through the resistor. Also, I think voltage stability is improved when voltage dividers are used.

by the way, I think my DMM is dead.

 

[Roff]

Sure you can. If you have binary-weighted resistors, and they are much higher in value than the output resistance of the counter (so that the output resistances are insignificant), then if you cycle the counter through all 16 states from 0 through 15, you will get a 16-step staircase from 0 volts to VCC volts. VCC volts is not a problem. As I pointed out in an earlier post, 0 volts could cause the oscillations to fwd-bias the "varicap" on negative peaks. The output resistance of the DAC is constant, so you can simply add a resistor from VCC to the DAC output, and you will get a staircase that goes from, say, 1 volt (or whatever you choose - use Ohms law, as you are fond of saying ) to VCC.

 

[Roff]

You will never be able to stop it on the exact center frequency of all stations...

unless I use alot of bits (probably at least 20 or so) and the correct varicap is used.

20 bits - that will require a an expensive commercial DAC. You can't build that out of 5% resistors - or even 0.1% resistors, for that matter.
I realize that you may be engaging in hyperbole.

 

[mstechca]

20 bits - that will require a an expensive commercial DAC. You can't build that out of 5% resistors - or even 0.1% resistors, for that matter.

why not?
Why is it different than if I used 4 bits?

I definitely do intend to go up to 8 bits.

 

[Roff]

20 bits - that will require a an expensive commercial DAC. You can't build that out of 5% resistors - or even 0.1% resistors, for that matter.

why not?
Why is it different than if I used 4 bits?

I definitely do intend to go up to 8 bits.

As the number of bits go up, the precision of the resistors has to go up, or you get places where the output is nonmonotonic. You will increase the count by one step, expecting the voltage to increase, and it will go down instead, or vice-versa. Here is an example of an 8 bit DAC made from +/- 5% tolerance resistors, being driven by an 8 bit counter. The actual results might look better or worse than this, depending on individual resistor error.

 

[mstechca]

As the number of bits go up, the precision of the resistors has to go up,

I didnt want to hear that

I guess I will have to do alot more math, and adjust the resistors, and maybe 5% will work in my favour.

thanks