|
I looking to buy some capacitors to replace the LPF caps on my 20m and 40m rig. I look for ceramic disc caps on EBay (50v) but they look a lot larger than what I already have in this kit. What are the caps that Hans is using for C25 - C33 in these kits?
syd/wt1v
|
Here are some of them from an order i placed at Digikey.... You'll be able to find others in the series.
| ? |
|
BC1016CT-ND
CAP CER 180PF 50V C0G/NP0 RADIAL |
| ? |
Available
Quantity |
Backorder
Quantity |
Unit
Price |
Extended
Price |
| ? |
| 10 |
0 |
0.16900 |
$1.69 |
|
| ? |
|
|
BC1020CT-ND
CAP CER 390PF 50V C0G/NP0 RADIAL |
| ? |
Available
Quantity |
Backorder
Quantity |
Unit
Price |
Extended
Price |
| ? |
| 10 |
0 |
0.20800 |
$2.08 |
|
| ? |
|
|
BC1018CT-ND
CAP CER 270PF 50V C0G/NP0 RADIAL |
| ? |
Available
Quantity |
Backorder
Quantity |
Unit
Price |
Extended
Price |
| ? |
| 10 |
0 |
0.20800 |
$2.08 |
|
| ? |
|
|
BC1025CT-ND
CAP CER 1000PF 50V C0G/NP0 RAD |
| ? |
Available
Quantity |
Backorder
Quantity |
Unit
Price |
Extended
Price |
| ? |
| 10 |
0 |
0.34600 |
$3.46 |
|
| ? |
|
| ? |
-- Julian, N4JO.
|
Yes... please do NOT buy NP0 capacitors from eBay... if you do, all bets are off.
Digikey, Vishay or TDK, or probably anything that says NP0 or C0G on it, and you will not go wrong.?
BUT... bear in mind, what are you doing?this for? There was an occasional problem with the 20m capacitors on early QCX kits. Anything in the last 3 years - which means all QCX+ PCB revisions, all QCX-mini PCB revisions, and all QDX PCB revisions, as well as all Ultimate3S over the same timeframe, all that uses Vishay capacitors from Digikey US. Even before that, there weren't any problems with 40m capacitors. It was just 390pF capacitors that had a few bad ones that were lossy in the LPF application.?
EVEN if you are looking at a 20m early version QCX, which you think has a low power problem... changing capacitors should still be the last place to look. Read? and read? and look at? and only in the LAST resort, consider the capacitors...?
On Tue, Nov 2, 2021 at 7:13 PM Julian N4JO < n4jo@...> wrote: Here are some of them from an order i placed at Digikey....
You'll be able to find others in the series.
| ? |
|
BC1016CT-ND
CAP CER 180PF 50V C0G/NP0 RADIAL |
| ? |
Available
Quantity |
Backorder
Quantity |
Unit
Price |
Extended
Price |
| ? |
| 10 |
0 |
0.16900 |
$1.69 |
|
| ? |
|
|
BC1020CT-ND
CAP CER 390PF 50V C0G/NP0 RADIAL |
| ? |
Available
Quantity |
Backorder
Quantity |
Unit
Price |
Extended
Price |
| ? |
| 10 |
0 |
0.20800 |
$2.08 |
|
| ? |
|
|
BC1018CT-ND
CAP CER 270PF 50V C0G/NP0 RADIAL |
| ? |
Available
Quantity |
Backorder
Quantity |
Unit
Price |
Extended
Price |
| ? |
| 10 |
0 |
0.20800 |
$2.08 |
|
| ? |
|
|
BC1025CT-ND
CAP CER 1000PF 50V C0G/NP0 RAD |
| ? |
Available
Quantity |
Backorder
Quantity |
Unit
Price |
Extended
Price |
| ? |
| 10 |
0 |
0.34600 |
$3.46 |
|
| ? |
|
| ? |
--
Julian, N4JO.
|
?? I went to Digikey and bought 5 of all of the caps in the LPF area just to be sure. Yes I've never found a bad cap either!? Maybe I'll screw around with prototyping some toroids and these caps, and then measure the prototype filter with my tracking generator. The installed caps all measure the correct value according to my LCR meter.? I also rewound L1 - L4 and then adjusted the turns using my LCR meter to make sure all the inductance values meet the values given in your tables. I also measured the PA output pp and can see 32Vpp which is a little less than what is shown in the troubleshooting manual, but not enough, I? believe, to get the output power down to the 0.5W that I've been getting. I've seen your 3 links above and they are very instructive.
?? My main problem is that no matter which coil I adjust, I never see any output change, so I don't know whether to add or subtract turns. On L1 I took out 2 turns and got no response, so I added in 4 turns to go in the other direction and still got no change to the power.? Now that I know the inductance values are pretty accurate, I'll probably go back and try the compressing and stretching the turns once again.
|
I increased my supply of disk ceramics from digikey when I got into the QRP Labs kits. Mostly so I could modify the band I was building for, etc. I got an assortment of toroids as well.
|
Syd, adding turns to a inductor used in a LPF will NEVER result in increased output. Take a look at the response curve for a LPF, there is a knee in the curve, below which there is full output, and above which the output reduces. All that happens when you add turns is the knee frequency decreases, when you remove turns the frequency increases.? -- > I finally got it together...now I can't remember where I put it< VY 73, Wes
AE6ZM
Sierra Vista, AZ
?
|
Revised comment: When you change the circuit to move the frequency where the knee occurs higher, you also reduce the signal suppression at the harmonic frequencies. Need to be careful there. -- > I finally got it together...now I can't remember where I put it< VY 73, Wes
AE6ZM
Sierra Vista, AZ
?
|
Hi Wes ? Syd, adding turns to a inductor used in a LPF will NEVER result in increased output. Take a look at the response curve for a LPF, there is a knee in the curve, below which there is full output, and above which the output reduces. All that happens when you add turns is the knee frequency decreases, when you remove turns the frequency increases.?
Respectfully I disagree... What?really happens in practice is that quite often you find that increasing the inductance does increase the output power. I see this all the time. *IF* an LPF is perfectly designed and perfectly?constructed with perfect components, and with perfect 50-ohm source and load, then yes the frequency response will be a flat line until there's a knee and at that point there will be an increasing attenuation with frequency; the 3dB point is called the cut-off frequency.?
If the inductances are anything off perfect, then you can easily end up with a response which has a peak in it, and a few dB insertion loss at lower frequencies. If that peak of the response is not aligned with the operating frequency then you see low power. If the inductances are too low for the design, then increasing them to the design value can definitely increase the output power.?
What happens in QCX is even more complex because there is normally not a perfect impedance match between the PA and the LPF input. When you tweak inductance values the output power can be increased if the impedance match is improved.?
Practically speaking, when tuning up QCX LPFs I have often found that increasing the inductance of L2 improves everything. I know it is slightly counter-intuitive but this is the reality.?
Generally I think the correct approach is to see what happens to the output power when you squeeze the turns and when you spread the turns. If the power increases as you squeeze the turns, this means that increasing inductance is increasing the output power. If the maximum output power occurs when you bunched together the turns as much as you could, then the tuning range available from turns squeezing is not enough; so in that case try adding on another turn or two, then try again. Conversely if the maximum power occurs when all the turns are spread out evenly around 100% of the toroid core, then this means that the adjustment range is not enough in the other direction - and in that case try removing a turn or two, then try again.?
73 Hans G0UPL
|
Even if the filters are perfect, the passband is not perfectly flat. If we were using Butterworth low pass filters it would be, but most of our ham designs are Chebyshev or Cauer filters that have some passband ripple. With ideal components the ripple would be only a fraction of a dB, but it could be larger if the component values are not perfect. Optimizing the filters could be?moving the ripples around so that the desired output frequency is at one of the peaks.
Hi Wes ? Syd, adding turns to a inductor used in a LPF will NEVER result in increased output. Take a look at the response curve for a LPF, there is a knee in the curve, below which there is full output, and above which the output reduces. All that happens when you add turns is the knee frequency decreases, when you remove turns the frequency increases.?
Respectfully I disagree... What?really happens in practice is that quite often you find that increasing the inductance does increase the output power. I see this all the time. *IF* an LPF is perfectly designed and perfectly?constructed with perfect components, and with perfect 50-ohm source and load, then yes the frequency response will be a flat line until there's a knee and at that point there will be an increasing attenuation with frequency; the 3dB point is called the cut-off frequency.?
If the inductances are anything off perfect, then you can easily end up with a response which has a peak in it, and a few dB insertion loss at lower frequencies. If that peak of the response is not aligned with the operating frequency then you see low power. If the inductances are too low for the design, then increasing them to the design value can definitely increase the output power.?
What happens in QCX is even more complex because there is normally not a perfect impedance match between the PA and the LPF input. When you tweak inductance values the output power can be increased if the impedance match is improved.?
Practically speaking, when tuning up QCX LPFs I have often found that increasing the inductance of L2 improves everything. I know it is slightly counter-intuitive but this is the reality.?
Generally I think the correct approach is to see what happens to the output power when you squeeze the turns and when you spread the turns. If the power increases as you squeeze the turns, this means that increasing inductance is increasing the output power. If the maximum output power occurs when you bunched together the turns as much as you could, then the tuning range available from turns squeezing is not enough; so in that case try adding on another turn or two, then try again. Conversely if the maximum power occurs when all the turns are spread out evenly around 100% of the toroid core, then this means that the adjustment range is not enough in the other direction - and in that case try removing a turn or two, then try again.?
73 Hans G0UPL
|
Thanks Hans and Shirley, for the information. Looks like I need to put my new VNA to work, and learn some stuff beyond the theory. -- > I finally got it together...now I can't remember where I put it< VY 73, Wes
AE6ZM
Sierra Vista, AZ
?
|
From Hans's UTube tutorial on aligning the LPF I might make mention to that part of the tutorial where he specifically tells the viewer to add/remove turns to the toroid according to whether one gets a power increase by compressing or decompressing the turns on the toroid!? If one sees an increase in power when one compresses the turns, then more inductance is needed hence add 2 turns because compressing turns adds inductance. And decompressing turns says you need to remove turns to get? the power up to snuff.
syd/wt1v
|
All this is nice and does make a lot of sense. Unfortunately for me, after constructing the kit and then trying to adjust the turns on L1 - L3, I don't get a response to the power output when I compress the turns or when I decompress the turns. I did try removing 2 turns and got no output power change! All I ever see is 0.6W!? So I took all the inductors out and now can see the cap values: they are correct!? I measured them with my LCR meter: ditto!? The peak to peak waveform out of the PA is a little over 32V which is a little low, but I assume it will go up after I fix my LPF problem. So now I rewound all the toroids and rechecked the inductance values with my LCR meter:? the values measured pretty close to what is published in the table, but I adjusted the turns to make the inductances match that given in the table. I have not gotten around to reinstalling the inductors yet.
syd/wt1v
|
Syd.?
Can you bypass the filter with a wire link straight to the antenna socket then check the RF output? It should confirm if the problem is in the filter or not.?
toggle quoted message
Show quoted text
On 5 Nov 2021, at 9:37 am, Syd via groups.io <nhuq1@...> wrote:
?All this is nice and does make a lot of sense. Unfortunately for me, after constructing the kit and then trying to adjust the turns on L1 - L3, I don't get a response to the power output when I compress the turns or when I decompress the turns. I did try removing 2 turns and got no output power change! All I ever see is 0.6W!? So I took all the inductors out and now can see the cap values: they are correct!? I measured them with my LCR meter: ditto!? The peak to peak waveform out of the PA is a little over 32V which is a little low, but I assume it will go up after I fix my LPF problem. So now I rewound all the toroids and rechecked the inductance values with my LCR meter:? the values measured pretty close to what is published in the table, but I adjusted the turns to make the inductances match that given in the table. I have not gotten around to reinstalling the inductors yet.
syd/wt1v
|
Not even slightly.? The filters are resonant, meaning that they both reduce harmonics and boost the fundamental. You could consider that the filters take the energy off the edges off the square wave and add it to the center of the waveform to boost the middle to make a sinewave of greater peak amplitude than the original square wave. There's no way you'll get thirty-odd volts peak to peak from a 12 volt supply without a resonant filter circuit, and I don't even want to think about the splatter of harmonics all the way up to daylight that you'll be emitting. If you want to measure the initial drive to the LPFs, you could measure the waveform at the drains of the output transistors, but leave the LPF in circuit if you're connecting anything to the output, even a dummy load.
--
Julian, N4JO.
|
Hans
? What do you do if you get 0.5W output, then compress the turns and still get 0.5W, and then when you spread the turns you get 0.5W again!!!? Checking the caps, they are all the correct value. I rechecked the turns on the coils and measured with an inductance meter:? looking pretty good. Continuity from PA to RF connector OK.? Now I adjusted the turns to measure exactly the inductance values? (using my LCR meter) given in the table, and soon I will put the coils in again and go through the alignment procedure again.? Also I'm in the process of prototyping the LPF on a breadboard with a 50 ohm output termination and try a 100K input termination (just a guess) and then look at it with my tracking generator and NanoVNA. Maybe I'll find something but it will be nice to experiment with these devices.? Other than that I am baffled!
syd/wt1v
|
If we're talking about a QCX (any version), output of 0.5W is very low unless you're operating the rig at a low supply voltage. If that power measurement is accurate, your problem is almost certainly something other than the low pass filter.?
My guess for the first place to look is the key shaping circuit. That consists?of Q4 (the fourth BS170) and Q6 (MPS751) and is connected to the PA through inductor L4. The #1 problem with QRP Labs builds is bad connections to toroids, so have a look at L4. Make sure the insulation is properly removed and you have continuity through it. If not, the key shaping circuit is not properly coupled to the PA and the drains of the PA MOSFETs are not getting pulled up to?+12V. Problems elsewhere in the key shaping circuit could also cause trouble.
If you don't find the problem there, then next place to look would be the driver stage, IC3A and IC3B (two of the four parts of the 74ACT00 quad NAND gate) or the PA itself (the three BS170 transistors Q1 through Q3). If you have access to an oscilloscope, the points you want to check are pin 4 of IC3B (the CLK2 output of the Si5351A), pin 3 of IC3B (the CLK2 output after passing through two NAND gates), and the output of the PA before the low pass filter (the drain of the three BS170 transistors).
Hans
? What do you do if you get 0.5W output, then compress the turns and still get 0.5W, and then when you spread the turns you get 0.5W again!!!? Checking the caps, they are all the correct value. I rechecked the turns on the coils and measured with an inductance meter:? looking pretty good. Continuity from PA to RF connector OK.? Now I adjusted the turns to measure exactly the inductance values? (using my LCR meter) given in the table, and soon I will put the coils in again and go through the alignment procedure again.? Also I'm in the process of prototyping the LPF on a breadboard with a 50 ohm output termination and try a 100K input termination (just a guess) and then look at it with my tracking generator and NanoVNA. Maybe I'll find something but it will be nice to experiment with these devices.? Other than that I am baffled!
syd/wt1v
|
Related to the excellent advice Shirley just gave, check the voltage at the top end of L4 during transmit: it should be less than a half volt down from supply voltage (measured on the cathode of the input protection diode D33, that is, not on the supply side).
--
Julian, N4JO.
|
In the past I have purchased a NP0/C0G assortment from eBay and have been very pleased with them. The specific ones I received were manufactured by Fujian Torch Electron Technology Co. They were green and have the torch symbol on each cap. Torch is a large multi product company that has a full line in various dielectric materials of radial and axial values with the common spacings. The issue I have found is finding a distributor for 100pc. qty. If you find any please share with the group.
|
That is a very nice idea!!!? I will definitely give it a try! I guess the obvious is so easy to see that you can't see it.
Syd/wt1v
|
Ian
On second thought, don't see how that will give me any more information than an ohm meter measuring continuity from C29 to the bnc RF output connector which I have already done and found it to be OK! Taking the output this way will only show the waveform at the output of the 3 PAs, which I have already measured the Vpp with a scope which I found to be 28 Vpp.
syd/wt1v
|
Julian N4JO
Chuck ke9uw
Wes WB7BR
Shirley Dulcey KE1L