Keyboard Shortcuts
Likes
Search
Low Power - Final Amplifier on through - hole kit
|
Greetings all, I built the through-hole kit from Sunil Lakhani. The receiver and exciter are working quite well. My only problem is that I am getting very little power out of the PA. The schematic of the PA is attached. I am using the IRF 510 with power coming from a 24v Lithium Polymer battery. (Believe me, the battery can source LOTS of current. That is not an issue.) Into a 50 ohm restive load, I am getting only 24v pk-pk. Which translates to about 1.5 watts. I checked that my DIY dummy load is in fact a restive load with an inline SWR meter. 1:1 I have re-wound my transformers tighter, no change. Tried a different IRF510, no change. I removed the inductor, and two capacitors between Q1 (the driver) and the IRF510. Slightly more power. About 2 watts. Still far from the 6 watts advertised for the kit, and the 25 watts people are getting out of the SMD Bitx. Any suggestions?? I chose the through-hole bitx for the journey of building it. So far I have enjoyed it. The exciter is working, and the receiver sounds great and is quite sensitive. The PA however has been rather frustrating. |
|
Do you know for sure that the exciter has enough output? Maybe find somebody with a scope, measure the AC and DC voltages at Q1 base and at Q2 gate, see if Sunil (or somebody here) sees trouble in any of those figures. You could do away with the 3 C's and 2 L's of the output filter if driving a dummy load, just in case the filter was cooked up wrong. ? |
|
The exciter board is producing 2v pk-pk. I do have a scope. The fastest it will scan is 0.5uS/div, so it is very hard to get a good picture of the waveform. But amplitude is pushing 2v pk-pk. I just removed the low pass filter from the output of the IRF 510. Now seeing 30v pk-pk (2.2w) at the dummy load. Either the filter is wonky, or there is a lot of harmonic power. Correct? Thanks for the help! On Thu, Jan 19, 2017 at 11:29 PM, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:
--
Mark Baldridge 608.561.3853 |
|
2v pk-pk sounds about right to me, though I don't really know. Output pk-pk voltage should be about the same with or without the filter and that seems to be pretty much the case, you have trouble elsewhere. I assume you have the bias voltage into the gate of that IRF510 set correctly, what is your current draw when the transmitter is quiet? I'd be really curious what the DC and AC voltages are on the IRF510 gate. With the output filter and the filter at the IRF510 gate both removed, this amp should be very broadband. Maybe create a source of a couple MHz at 2v pk-pk and feed that into the amp, that might let you better see what's going on with a slow scope. Might be able to use your VFO, adding a bunch of caps to lower the frequency. Weird that you can tell the amplitude "is pushing 2v pk-pk" if the scope is otherwise too slow to display it, a simple diode RF probe might be a more suitable instrument to measure RF voltages. ? ? Hopefully somebody out there more familiar with Sunil's design can give you some better pointers. Jerry, KE7ER |
|
Yes, the bias of the IRF 510 is set to 50mA per the instructions that came with the kit. I will take a look at the gate's AC & DC voltages tonight. Maybe it is a distortion issue of some sort in the exciter. Jerry, my scope is an old 20MHz analog thing probably right out of the 80s. It was free so I don't complain. :) Since it is analog with a scanning beam, I am able to see the amplitude of the waveform, but the individual periods are quite close together and it is hard to see the shape of the waveform.? Thanks guys! ~Mark, KC9DZT On Fri, Jan 20, 2017 at 9:17 AM, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:
--
Mark Baldridge 608.561.3853 |
|
When testing the modulation envelope the scope sweep speed should be set around 1ms per division so the modulation envelope and not the usb frequency will be displayed. With the scope set at around 1 usec and no modulation, the carrier balance would be set to minimum out. You should end up with less than 1 volt displayed. Most of the ones I've worked on was between 250 mv and 500mv. Leonard |
|
Yes, could look at the modulation envelope. ?I was concerned that if the max sweep speed was 0.5us/div, the bandwidth would likely be down around 5 MHz. ?But if he really has a 20 mhz scope, then he's all set. What's the gain of the Q1 stage work out to? ?With a 2v pk-pk signal going in, anyone care to guess what ac signal he should see at he gate of Q2? Jerry, KE7ER |
|
Leonard raises a good point. ?I assume Mark is supplying audio at an appropriate level when measuring power out, and not just measuring residual carrier. ?The standard Raduino firmware should generate an audio tone on one of the digital outputs so we have a known audio level to conduct tests like this with. |
|
Yes, about the audio: I injected a 100mV pk-pk signal at the mic input. The source was Audacity on my PC, decoupled with a capacitor, fed through a 10k pot as a voltage divider, and then applied to the exciter board. I turned the pot all the way down, turned on a 1KHz audio signal, and then turned the pot up until 100mV was at the mic input. From there I increased the mic gain until RF was generated at the output of the exciter board.? There is one curious issue with the exciter board however. With a 1KHz tone at 100mV pk-pk at the mic port, and the mic gain all the way down, there is no RF out of the exciter as expected. When the mic gain is increased, the RF level out of the exciter appears quite suddenly, and then increases with mic gain. When the mic gain is less than 1/4 of the way up, no further RF amplitude is gained. I can turn the gain up more and more but the RF maxes out at 2v pk-pk. Almost as if an amplifier is getting saturated somewhere along the line. Could this then cause distortion, and then out-of-band RF energy that is then wasted in one of the filters? Thanks everyone! On Fri, Jan 20, 2017 at 11:58 AM, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:
--
Mark Baldridge 608.561.3853 |
|
This video is from an older version 3 linear
|
|
I would be real suspicious of T1 or T2 not phased right if the problem is in the pa. Also, a mike input of 25-50 mv's is plenty. With 25-50 mv's into the mike jack you should get 1 volt or so at the input to the pa when the 2 boards are connected together. I would guess you are over driving the mike input trying to correct for a different problem. I have a series of pictures at the common points for the pa board but I'll have to put them into web page format so it will take me a couple of days. In the mean time, try checking the linear as shown in the video. You should get meaningful information with your scope. Leonard |
|
On Fri, Jan 20, 2017 at 10:30 am, kc0wox Leeper wrote:
Rule of thumb says the scope bandwidth should be about 10 times the signal displayed I suppose I could mostly agree with that, though a 20 mhz analog scope should be good enough to diagnose this problem. I assume his transmitter is running at 14 mhz. If you want to check signal shapes to see if there's a harmonic present at 14*3 or 14*7 mhz, then yes. ?If you just want to see if a 14 mhz signal is present and get a fair idea of the amplitude, then a scope with a bandwidth of 14 mhz or more is sufficient. ?A 14 mhz signal through a device with 14 mhz of bandwidth will be 3db down. At 3db down the displayed voltage on the scope will be down by sqrt(2), if it was really 1.414 volts pk-pk the scope would show 1 volt pk-pk. ?Good enough.? Now if we're talking a cheap digital scope claiming a 20 mhz sample rate, then yes, likely not good for looking at a signal much faster than 2 to 4 mhz. ?Sample rate and bandwidth are two different things. ?A good quality digital scope will sample at 10 times the bandwidth of the vertical amplifier, the cheaper ones might limit the bandwidth to 5 times the sample rate. Jerry, KE7ER ? |
|
? ?Thanks Leonard for finding this page I was needing it and I need to know also the voltages in Q1 (2N3866) if you know please pass me ok? ? ? Jorge PY2PVT _._ 2017-01-20 17:17 GMT-02:00 kc0wox Leeper <kc0wox@...>:
--
? 73 de Jorge PY2PVT
? Campinas SP ? GG67MD |
|
Greetings all. I started tracing my PA with Leonard's documentation.? The first discrepancy I noticed right away.? He had 30v pk-pk on the collector of Q1 (2n2218) and I only had 15.94v pk-pk. And the waveform is quite different. My guess is that the problem is around here. I double-checked all the component values, orientation of the 2n2218, and my solder joints. I'm still stumped. Attached is a capture from the scope.? My voltages at Q1 were as follows:? Base: 1.79v Collector: 10.97v Emitter: 1.14v And yes, I am running the entire board from a 20vDC regulated bench power supply (HP3630A), and I am feeding in 2v pk-pk at 14.00MHz from the function generator (Tektronix AFG3251). Any insights would be much appreciated! On Sat, Jan 21, 2017 at 6:09 PM, Jorge Luiz Fenerich <py2pvt@...> wrote:
--
Mark Baldridge 608.561.3853 |
|
One quick question before I do some more tests tonight: Looking at this version of the Bitx PA: The resistors at the emitter of Q1 are 22E and 6.2E. Why were they changed to 10E and 10E on the Bitx 3B? (See PDF attachment on original post.) Wouldn't changing from 6.2E to 10E reduce the gain of Q1? Thanks all. On Mon, Jan 23, 2017 at 4:17 PM, Mark Baldridge <marktbaldridge@...> wrote:
--
Mark Baldridge 608.561.3853 |
|
Alright guys. The big trick here is MEASURE EVERYTHING WITH A x10 PROBE. That was my mistake. Per Leonard's documentation here: I re-measured everything. With 2v pk-pk in, I was getting 30v at the collector of Q1, and with 22v in my 6 cell LiPo on the IRF510, I was getting 54v pk-pk into a 50 ohm load. This is 7-8 watts. The PA is performing as expected. The next step for me is to figure out why the exciter isn't driving the PA as it should. (I should re-measure that with a x10 probe even!) Lesson learned. Thanks for all your insights and patience everyone. :) ~Mark On Tue, Jan 24, 2017 at 3:43 PM, Mark Baldridge <marktbaldridge@...> wrote:
--
Mark Baldridge 608.561.3853 |
|
Hi Mark There are two ways to measure with a scope as outlined in EMRFD. First way is with a x10 probe hung on a 50 ohm load. Load is on the o/p. Second way is with a x1 probe.The probe is plugged into a tee. The other tee connection has a 50 ohm load on it. Then the entire tee is plugged into the scope. The probe is on the o/p. I have tried both ways and both work. 73 Ken VA3ABN On Wed, Jan 25, 2017 at 12:25 PM, Mark Baldridge <marktbaldridge@...> wrote:
|
|
If using a scope, it's important to know how probes work. According to EMRFD chapter 7.4, the BNC on a scope is typically terminated at 1 megaohm in parallel with about 20pf. ?And a 1x scope probe is apparently just a piece of coax from probe tip into that BNC. ?A 10x probe has a 9 meg resistor in parallel with a typically 5pf cap, placed out in the probe tip. ?So the DC loading of the circuit you are looking at is now 10 megs. ?More importantly, the (often adjustable) 5pf cap emphasizes high frequency edges, allowing us to compensate for the 20pf at the scope and the capacitance of the coax which would otherwise form a low pass filter. ?This works no matter what the impedance of the circuit we are looking at is, so long as it is not up in the megaohms. If the signal you are looking at is coming out of a source that can drive a 50 ohm load, we can do much better by simply running 50 ohm coax from that source into the scope and terminate that coax with a 50 ohm resistor at the scope. ?For example, for a really good look at what is coming out of that IRF510, we could unhook the antenna and instead drive a 50 ohm resistor placed right at the scope BNC connector through a piece of coax. ?Ken suggests mounting ?a coax tee coupler on the scope BNC, driving one side of the tee with your 50 ohm coax, and mounting a 50 ohm resistor on the other side of the tee. ?(That 50 ohm resistor had better be be several watts if driving it from the Bitx40 PA, and the scope had better be able to deal with whatever RF voltages are present.) ?There are not many other signal sources in the Bitx40 that are driving into a 50 ohm load, so this trick is not generally useful here. ?But if, for examle, the bidi amps were all 50 ohms in and out as suggested in post 20657, we could remove the connection from one of those amps out to the following stage and instead have it drive a piece of coax directly into our scope, where it is terminated into 50 ohms. ?(Must remove the connection or the amp will be doubly terminated and see a load of 25 ohms.) ?This doesn't work for sources that can drive something other than 50 or 75 ohms, as that's typically all we have lying about in the way of coax. ?Though I suppose you could cook up some sort of ladder line designed for the impedance of interest, and then terminate at the scope with an appropriate value resistor. ?But that's a lot of work, and might be tough to route that ladder line in such a way that nearby objects don't interfere. OK, enough of that. ?What's an "o/p"? Jerry, KE7ER On Wed, Jan 25, 2017 at 09:35 am, Ken wrote:
? |
|
Output 73 Ken On Wed, Jan 25, 2017 at 1:54 PM, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:
|
