RX-1 DDS troubles, help appreciated

General discussion

RX-1 DDS troubles, help appreciated

Postby OH2JBE » 13 Oct 2014 12:47

Hi! I finally got around to solder together my Juma RX-1 kit I'd purchased a couple years before. Everything connected, I noticed that my 12 volt power source was broken, so tried to boot the radio up with a 9 volt battery. And it worked: I got the greeting CW on the speaker and welcome message followed by frequence & S meter screen on the LCD. Frequency selection worked, too - or so I assume; the number on screen changed. I didn't have an antenna connected so I just heard noise. I disconnected the battery and left the kit on my desk, planning to attach it to a dipole the next day.

However, when I turned the receiver on the next day, the display remained dead, and I could hear no CW from the speaker. I could hear the static, though, and turning the pot changed the loudness. Dumbfounded, I checked the battery voltage - same as yesterday. I found a battery holder for eight AA batteries (I only had NiMH batteries, so got only ~10 volts, but that should be enough, shouldn't it). However, same result: just noise on the speaker, nothing on display, no CW heard.

I checked if there was some misplaced conductive material somewhere on the circuit boards, blown in - couldn't find anything. All the soldered connections appeared good, including connections between the microcontroller and LCD. And checked by multimeter, they were good. Also, there was 5 volts after the DDS card regulator in all the places there should be, including the LCD board and the correct leg of the microcontroller.

So now I'm a bit at a loss. I can hear noise on the speaker, I'm guessing the RX card mostly works. But the microcontroller on the DDS card doesn't seem to boot up. It did boot the day before, and I turned it on and off multiple times, then spent the day on my desk, battery disconnected, nobody touched it, come next morning - dead. Any ideas what to try next?
OH2JBE
 
Posts: 5
Joined: 13 Oct 2014 09:19
Location: Finland

Re: RX-1 DDS troubles, help appreciated

Postby 5B4AIY » 14 Oct 2014 07:07

Hi,
I can only really speak about the TRX-1, but since the RX-1 uses essentially the same logic and circuitry, at least for the receiver section, then there should be commonality. From your description, it would appear that the DDS is not producing any output, or, if it is, then it is random noise rather than a pure carrier. The fact that your display does not come up is significant. You will not get any receiver signals if the microcontroller is not working, because it is the microprocessor which sends the desired frequency word to the DDS to get it to generate a carrier, hence, no microprocessor, no carrier. So the first thing is to diagnose why your microprocessor is not working. You seem to have checked the obvious things, but it would be a good idea to actually check from the microprocessor's pins to the tracks on the printed circuit board to be certain that there are no open or short circuits here. When you measure the voltages, be certain you are checking them actually at the pins of the IC itself, you should have +5V between pins 5 and 14. If this is OK, then you need to check that the main oscillator is running. For this you will need preferably an oscilloscope, but, if you do not have one, then a short-wave receiver tuned to the crystal frequency should allow you to hear a really strong stable signal. If you do have an oscilloscope, and the oscillator is running, using a x10 probe check for activity on all the pins, and that these signals look sensible, that is, the voltage is 5V, the edges are sharp, that the signals transition between ground and +5V. Failure to initialise the display means that the microprocessor is not even completing its boot sequence. Let me know how you get on, and I'll try and walk you through some other tests.
73, Adrian, 5B4AIY
5B4AIY
 
Posts: 160
Joined: 13 Nov 2011 09:22
Location: Cyprus

Re: RX-1 DDS troubles, help appreciated

Postby OH2JBE » 14 Oct 2014 20:11

Hi, and thanks for answering!

5B4AIY wrote:actually check from the microprocessor's pins to the tracks on the printed circuit board to be certain that there are no open or short circuits here


Yeah, I did precisely that. The processor is in its socket and I made the measurements from its pins to other parts of the circuit board. No open or short circuits anywhere. Resistances between certain pins that have resistors between them were as expected.

5B4AIY wrote:When you measure the voltages, be certain you are checking them actually at the pins of the IC itself, you should have +5V between pins 5 and 14.


OK, I checked. 5.00 volts between those pins.

5B4AYI wrote:If this is OK, then you need to check that the main oscillator is running. For this you will need preferably an oscilloscope, but, if you do not have one, then a short-wave receiver tuned to the crystal frequency should allow you to hear a really strong stable signal. If you do have an oscilloscope, and the oscillator is running, using a x10 probe check for activity on all the pins, and that these signals look sensible, that is, the voltage is 5V, the edges are sharp, that the signals transition between ground and +5V.


I might be able to use a friend's oscilloscope, or maybe an oscilloscope at a local hacklab, so this is probably the next thing to try. (My short wave receiver is currently on loan.) Though...

5B4AYI wrote:Failure to initialise the display means that the microprocessor is not even completing its boot sequence.


...the display really doesn't get initialised, so I'd guess this was the case. :(

5B4AYI wrote:Let me know how you get on, and I'll try and walk you through some other tests. 73, Adrian, 5B4AIY


Thanks! And 73!
OH2JBE
 
Posts: 5
Joined: 13 Oct 2014 09:19
Location: Finland

Re: RX-1 DDS troubles, help appreciated

Postby 5B4AIY » 15 Oct 2014 06:45

OK, if you can get the use of an oscilloscope, this is what you should look for, and what to expect. The 'scope needs to have a minimum bandwidth of at least 40MHz, preferably 100MHz, and you should use a standard x10 probe to minimise capacitive loading.

The very first step is to check that pin 4 of the microprocessor is at a steady +5V. This is the master clear pin, and if it is at ground potential, then the microprocessor will be held permanently in the reset state. If it is not at +5V, then check R21.

The main crystal oscillator is the master clock for both the microprocessor and the AD9853 DDS chips. To check it is working, touch the probe on pin 15 of the microprocessor and check that you have a 20MHz square wave there transitioning between ground and at least +4.5V. Then, if you have master clock there, check on both sides of R12 (200R) and, if you are using a modern micro-instrument probe, pin 5 of the AD9853 to make quite sure that the DDS chip is receiving clock signal.

Failure of the master clock can be traced to a crystal problem, or some problem with the capacitors associated with this oscillator circuit, or the microprocessor itself.

If you have clock signal, then check that you have activity on pins 6/7/8/9/10/11/12/13 of the microprocessor. These are the main I/O pins, and during the software main loop cycle these pins will be constantly being used to send data to the LCD, the DDS, as well as read inputs from the encoder, and so there should be lots of activity on all of these pins. Note, all the pins, any pin that is permanently high or low is a fault.

If the I/O lines are active, but the display is still blank, then verify that you have +5V on pin 2 of the display, and ground on pin 1. Then check that the display Read/Write line on pin 5 is at ground. The software does not read the display's busy/idle status, and so only writes to the display, and uses delay routines to ensure that no data over-run occurs.

If none of this works, then it would be a good idea to try another microprocessor chip. These are not expensive, and are easily programmed using the PicKit-3 programmer.

73, Adrian, 5B4AIY
5B4AIY
 
Posts: 160
Joined: 13 Nov 2011 09:22
Location: Cyprus

Re: RX-1 DDS troubles, help appreciated

Postby OH2JBE » 15 Oct 2014 21:16

Hi!

Meh, the oscilloscope I got to use has very low bandwidth, much less than the clock frequency. I could only confirm that pin 4 appeared to be at a more or less steady 5 V. But obviously I couldn't see the clock signal at pin 15 - at this bandwidth, I just got a seemingly stable 2.25 V. Which would probably be what should be expected if there was an actual signal transitioning between ground and 4.5 V at a frequency too high for the oscilloscope bandwidth, but I can't confirm that, and of course the frequency remains unknown. There was the same possible signal on both sides of R12 and at pin 5 of the DDS chip.

Same with the data pins: too low bandwidth to see whether there was an actual signal anywhere. None of the pins was constantly at ground and none showed a value of 5 volts; at this bandwidth, I got seemingly stable values between 4.2 V and 4.7 V. ("high" would be 4.99V..5.00V).

I did have +5V at pin 2 of the display and ground at pin 1. And pin 5 was at ground.

So, the processor isn't at a constant reset state. The clock probably generates some signal, which seems to be present at all the right places, but I don't know at what frequency. And there might be something going on at the data pins, but I don't know what, and can't be sure if there actually is. Should probably find a better oscilloscope. Or just directly try programming a new chip (i.e. purchase the programmer).

Thanks again!
OH2JBE
 
Posts: 5
Joined: 13 Oct 2014 09:19
Location: Finland

Re: RX-1 DDS troubles, help appreciated

Postby OH2JBE » 13 Nov 2014 14:35

Got a PicKit-3 and a couple new microcontrollers. Programmed a microcontroller with the provided hex file and according to MPLAB IPE, everything went fine. Installed it into the socket, and still the same behaviour: display not booting up, no welcome beep on the speaker but just noise. So it has to be in one of the peripheral components or connections. Going to go through all the soldered parts with a magnifying glass next (I already tried with a multimeter)... And I really should find someplace where I could use an oscilloscope with a higher bandwidth.
OH2JBE
 
Posts: 5
Joined: 13 Oct 2014 09:19
Location: Finland

Re: RX-1 DDS troubles, help appreciated

Postby 5B4AIY » 16 Nov 2014 07:41

Now that you have programmed another chip, and the programming was OK, you have eliminated this as a causal factor. As you say, it has to be a fault on the board itself, and my first thought as always, would be to verify that the main oscillator is actually running, is stable , and at the correct frequency, 20MHz. At least see if you can detect it on another short-wave radio or amateur transceiver. If the oscillator is not running, then nothing will work. With a blank LCD this almost certainly means that there is a clock oscillator fault, as the LCD is initialised very early on in the boot sequence, so if it displays nothing, then it suggests that the microprocessor is not even executing its initialisation code. Have a careful look at C2, C3, C20 and R11.
73, Adrian, 5B4AIY
5B4AIY
 
Posts: 160
Joined: 13 Nov 2011 09:22
Location: Cyprus

Re: RX-1 DDS troubles, help appreciated

Postby OH2JBE » 14 Nov 2016 10:06

Finally had time to look at this again. Last time I tried to fix my RX-1, I had disconnected the DDS board altogether. I started by reconnecting it to the RX board, and lo and behold, it booted. I got the greeting, CW on earphones, and everything seems to work. So I assembled the thing, and it still seems to work...

...except I have yet to receive anything. I don't have an antenna that I know would work; I did connect a long random wire to the antenna connector, but all I could hear was noise. Granted, the location was far from optimal, so I'll have to continue trying...
OH2JBE
 
Posts: 5
Joined: 13 Oct 2014 09:19
Location: Finland

Re: RX-1 DDS troubles, help appreciated

Postby 5B4AIY » 17 Nov 2016 09:59

Hi,
OK, now you have the microprocessor working it sounds as if the DDS is not working. If you connect any sort of antenna to the input of the receiver, it is sufficiently sensitive that you should hear lots of signals. If all you hear is noise, this suggests that the synthesiser is not working, or is only producing a noise-like output.

In order to fault find this, you really need a good wide-band (40MHz) oscilloscope. However, it is possible to verify if the synthesiser is working by using a local short-wave receiver. Since the LCD is working, then this indicates that the RB0 through RB7 data lines are working. If you can change the frequency in a consistent manner with the rotary encoder, then the RB4/5/6/7 data lines are working, and if you can sense the push-button switch on the encoder that will also confirm it.

You need to place a coupling loop alongside the DDS chip and tune your short-wave receiver to the same frequency as indicated on the LCD. If the synthesiser is working you should be able to detect a very strong CW signal, that tunes as you turn the encoder knob. I suspect that the synthesiser is not working. Is it mounted correctly? Are all the pins properly connected? I had to re-solder mine in order to make it work because of a faulty joint.

Adrian, 5B4AIY
5B4AIY
 
Posts: 160
Joined: 13 Nov 2011 09:22
Location: Cyprus


Return to Juma Discussion

Who is online

Users browsing this forum: No registered users and 2 guests

cron