Keyboard Shortcuts
Likes
Search
QCX keying issues with full/semi break-in settings. Advice please.
开云体育? Hi. ? Can someone suggest why one of my three QCX units produces a different CW waveform shape depending on whether or not the full or semi break-in option is selected. ? Testing and commissioning these units has had its teething problems but for now I would like to focus on why one board produces different keying characteristics according to the keyer settings. ? ? ????????????????????? ?????????????????????????????????????? ? The 18Mhz board produces a waveform with some “ringing” between characters when set to semi break-in. This is clean when set to full break-in with sharper cleaner output. However this has the side effect of producing key clicks in the audio chain. This may be RF getting into the audio stages whilst on the work table and exposed but I could do with some reassurance that this should not be a functional issue before finally screwing down the covers, especially with the number of recorded fatalities Q6 and final stages ?have had due to early instability problems. ? Does this indicate instability or switching problems and if so how could this be corrected? I look forward to your response. ? For those starting out I would advise posting in block capitals NOT to key up without a dummy load or well matched antenna and do not connect and test a? GPS module unless the keyer is set to “practice” mode! Use a current limiting or fused power supply and thoroughly test the unit before proceeding to box it up. This in itself can increase the chance for errors and should be seriously considered before going down this route. ? I hope this final issue can be resolved and the units can be used to demonstrate how neat these QRP rigs operate on air. A great project to get everyone building and learning more about QRP ?radio architecture. ? ? ? ? Construction ideas: A QCX-17 and QCX-20 in the final stages of construction. ? 73 Derek G4VWI ? ? ? Sent from for Windows 10 ? |
|
Derek
well if you don't use semi-break-in keying maybe you don't have issues.? I would not think of something easily measured as instability, it could relate to actual capacitor values or even wire arrangement.? (I see you have a lot of wires that can sniff RF and put it someplace - I have only a few things mounted off the board).? I confess I have not examined the C22 value change - I don't remember my unit being objectionable in QSK.? you already have the correct recipe to avoid damage to Q6 - it was my folly in not reading the entire manual before plugging in the GPS.? decide when its good enough to install into the case (I perceive its not necessarily permanent then - I think mine has been out of case 3 times already) - and enjoy a fine rig!? Curt |
|
Many thanks for the replies from Steve and Curtis. Much appreciated.
Since purchasing the first board and parts many issues have been raised and modifications made to solve most of them. Some components appear to have been changed from the original release including the ferrites used in the band-pass filter. Certainly there appears to be an improvement in performance over the original toroids used on my first 20Mtr. unit. So far the suggestion to sort this keying issue out is to replace C22 with a 0.1uF cap. I will certainly look into this and get back to you. As far as the enclosure is concerned I have used? separate screening around the circuit board and mini-ferrites to minimise noise but it is not finished yet due to uncertainties about the keying.? 73 Derek G4VWI. I will look for the thread on this topic again to see what others have done. Any other suggestions would be welcome. |
开云体育So what are you saying, Derek?? Are you saying the core supplier or supplied cores have changed and not just changed turns count?? In other words, do the later kits have different supplied cores? Lee, w0vt On 1/11/2018 1:33 PM, derek wrote:
|
|
Derek,
some details of the QSK click issue and the solution can be found here: /g/QRPLabs/message/20540 Regarding the keying of the PA I had a problem, which may or may not be related to what you described. In semi QSK mode the RF power of my QCX was not turned off completely between dits and/or dashes with R41=470 Ohm as per manual. Obviousliy, in the pauses Q6 was still conducting. Adding a second 470 Ohm resistor parallel to R41, thereby reducing the base-to-emitter resistance to 235 Ohm, did the job. The scope screen shots, taken with Vcc=10V, show the details. In the same context, but not necessarely relevant in your case: Some of the Q6 failures might be contributed to the relatively small bias current of Q6 set by R42. According to the data sheet: the MPS2907 current gain for Ic=500mA is in the order of 50. Hence a bias current of at least 10 mA is needed to turn Q6 completely on. If the base of Q6 is at 10V R42=1k is just small enough to provide 10mA. If manufactuing tolerances yield a lower current gain, the base current might not be sufficient to turn Q6 on completely, resulting in a significant emitter-collector voltage drop and a rather high power dissipation in Q6. This excess dissipation could destroy Q6. Reducing the resistance of R42 to 220 .. 470 Ohm should provide sufficient bias current, even for a low current gain transistor Q6. Please note: I did _not_ change R42. In my QCX Q6 is fully switched on with R42=1k, as can be seen in the second screen shot. Therefore I can not verify the above estimates. 73 Wolfgang DK4RW |
|
Lee, Derek There have not been any changes in the cores used in the kits. Suppliers have not changed. 73 Hans G0UPL? ? On Jan 12, 2018 12:20 AM, "Leland L. Bahr" <l@...> wrote:
|
开云体育OK, good to hear, Hans.? Thanks for the update and setting this straight. Lee, w0vt PS:? Your up early, Hans! On 1/11/2018 10:51 PM, Hans Summers
wrote:
|
|
Many thanks Wolfgang.
Your reasoning and accompanying scope traces make very good sense. They much appreciated. I would like to know what scope you use. It produces excellent traces. I decided to try out your modifications the the 17 Metre board first. I added a 470R resistor in parallel with R41. I also changed C22 to 0.1uF electrolytic to sort out the audio key click problem. Positive outcomes: The audio clicks have gone completely. The CW waveform is much sharper in rise and fall times producing a much better cleaner waveform overall. Using full break-in there is no sign of ringing between the dit and dah character formations. This works very well. The output transistors only get a little warm when keying for more than a minute or two continuously.? Issues to resolve: I still have issues with a "ringing" CW tone when using semi break-in! There is still a persistent albeit low level trace shown on the scope between the individual character formations when using semi break-in. The lack of clean switching results in a slight "ringing" CW tone when monitored on a receiver. Strangely the 20Metre board does not display ANY of these characteristics and Q6 in this board has a lower measured DC gain than the one in the 17Metre board which exhibits all of the problems you mention. Any more thoughts on solving the semi break-in issue would be most welcome Wolfgang.? Your time and effort is much appreciated. I can supply some spectrum scope graphs if required. 73 Derek G4VWI. HNY. |
|
The "Ringing" |
|
Derek, Jerry,
the low level signal between the character elements is most likely caused by the TX signal fed to the PA in semi-BK mode. On a breadborad test setup of the QCX PA for 40m I observe a leakage signal about 45 dB (= 7..8 S-units) below the output of the keyed PA. Probably the leakage increases with increasing frequency. I would expect a similar leakage in the QCX. There is, however, still a possibility, that Q6 is not completely turned off, even with a reduced base-emitter restor R41. Derek, what is the leakage voltage on your QCX compared to the output with the PA keyed? 73, Wolfgang DK4RW |