Tweet
I can't actually see on the schematic what the power supply rail is for the CD4024 etc ... is it 10V or 12V ? I see it's split with a virtual ground generator to +/-5V or 6V for the opamps.
-
-
Also, your oscillator module design is not going to work. If you power it from a low voltage, like 3V3 etc, its output will be low voltage ( 3V3 / 0V ), which can't be simply connected to another logic IC powered from , say 12V, because both 3V3 and 0V will both be seen as "zero Volts" .
Taking a quick scan over the thread, it looks like the power rail is +/- 8V or +/- 7.5V , in which case the CD4060 should be OK with a 6.7MHz crystal.Comment
-
CD4060 does not have Q0, Q1, Q2 and Q10.
For 52740Hz (26370x2) = 1042300 ohms and 22nF in Q3 = 6592,5Hz
I'll try now....
Comment
-
"CD4060 does not have Q0, Q1, Q2 and Q10."
Correct, and this is not a problem. If you use the 6.7xxx MHz crystal, you take the Q7 output and connect to the CD4024 input. We are trying to generate ( ideally ) 26600 Hz. This gets divided down by 4 to drive the TX circuit, and divided further to drive ( non-critical ) audio stages, and to synchronise the smpsu oscillator. If you use the 420 / 429kHz ceramic resonator with the CD4060, you use the Q4 output, and connect it to the CD4024 input.
Furthermore: you can change which CD4024 outputs you use. The outputs used are Q2, Q4, Q5, Q6 . These could be changed to Q3, Q5, Q6, Q7, and you input 53200 Hz to the IC. This gives you options with you additional divider / oscillator. You can also use outputs Q1, Q3, Q4, Q5, and input 13300 Hz to the divider. These options may be useful, they may not, but they exist.Comment
-
I have tried with the oscillator of the 4060 and I have not gotten a stable frequency .....
I have ordered the 25600 Hz crystal and I will wait to receive for connect one to the 4024, which I think is the best option.
Adapting the coils to 6400 Hz is not difficultComment
-
Visual Analyzer + Windows 10
Oscilloscope, Spectrum Analyzer, etc ... perfect for tuning coils.
https://www.sillanumsoft.org/
Comment
-
Probes for Visual Analyzer -> PC Sound In
Comment
-
Change the 1k0 resistor to 1k1, you will get better accuracy on the x100 range ( and a little bit better on the x10 range, too. )Comment
-
Thx Skippy, I will do.
Do you know someone who has built the tm808 and it works well?
The more I learn, the more doubts I have...
Comment
-
"Do you know someone who has built the tm808 ?"
I know almost nothing about the 808. I contributed to the thread because I had some workable ideas about the unobtainable crystal problem, and I thought it would be good if you included them in a new PCB layout. Then others don't have to mess about making "daughter boards", or programming PIC micro's or MEMS devices to generate the clock waveforms.
My only interest in two-box machines involved making a two-box add-on for my Fisher F75. This has very good non-motion mode performance, so it should be a good match for a two-box. But I also thought a very large ( 60cm ) DD coil would also work equally well.
Lack of time, and too many other projects means it could be ages before I do anything on this.Comment
-
Regarding that scope input attenuator: If you want to get it "perfect", the 1k0 should be changed to a 1k2 in parallel with a 15k. This will give the correct 1.11111k value; but 1k1 is just in error by 1%, so it's close enough.Comment
-
I am using the Spectrum Analyzer directly without the attenuation since the coils work with millivolts.
With the Analyzer I have verified that the Cscope Metadec works at a frequency of 16,480 kHz and the Minelab Excalibur at 3,050 Hz !!
For me, it's one of the best detectors I've ever had.
I was thinking how the TM808 would work at about 3,000Khz.... Comment
-
Excal's work with standard BBS waveforms, so nominally 25 kHz and 3.125 kHz, a ratio of 8 to 1. There are 8 cycles of 25 kHz, followed by one cycle of 3.125 kHz, then repeat. The frequency shift options vary across the BBS machines, I think +/- 8% is possible for some of them, I don't know what shift the Excal is capable of.
If you're considering a change of frequency ( lower ) for the 808, I suggest halving the freq to 3.325 kHz (nominal ). This can be done by using a different divided output from the crystal / divider circuit, but leave the audio tone generation part alone. ( Obviously, many other aspects of the circuit will need changing, too ... )Comment
-
Yes !, BBS Excalibur Spectrum, from 1500Hz, 3000, 4500 ... 21Khz (Sound card no give more..) :
Comment
-
That's quite funny, and shows how troublesome sound cards are ... with few exceptions, unfortunately. Your scope trace clearly shows the full cycle of 3.125k, as quarter-cycle; half-cycle; quarter-cycle. Then inbeween is absolutely nothing, where in fact there are 8 cycles of 25k. If I recall, it's a quarter-cycle, then 15 half-cycles, then the final quarter-cycle. Your trace just shows "ringing". It's not ringing, of course. If you take a square wave, and remove the 7th, 9th etc harmonics, you are left with a waveform that only has fundamental, 3rd harmonic and 5th harmonic. This looks at first glance like terrible ringing .... but notice it starts "ringing" before the big 'edge' , not just after one.
Many PC's today have "super" sound cards, that go to higher frequencies. Usually called High Definition Sound Cards, or HD Audio, they can sample audio up to 48 kHz, (I think).
What is super rare is audio inputs that go down to DC. They normally have very sharp cut-off digital high-pass filters at about 16 Hz, that can't be 'turned off', plus electrolytic caps to AC couple signals, for example from the pre-amp to the main ADC input.
I have an external USB sound box that apparently goes to DC with some minor modifications .. I just haven't tried it out, yet.Comment
Comment