Re: Measuring the output impedance of a 20m push-pull power pre-amp
#transformer
Wow, what a wealth of information you guys have shared! A really big Thankyou! I've included some images to show my circuit and setup in more detail and I'll try to summarize what you guys have told me so far:
1) Don't bother trying to measure the output impedance using a Nano VNA. It's likely to be very inaccurate with no RF input applied and potentially fatal for your VNA should you try to apply an input.
Comment: That's certainly what my instinct would suggest, but I'm intrigued by the correlation between the real part of the impedance being near 50 ohms and the demonstrable increase in power output that follows. If I move away from 50 ohms real, the power output falls (in either direction, obviously).
2) A better method might be to "tune" the circuit to output a power peak into the 50 ohm load, then disconnect the transformer primary from the collectors. If I then attach the VNA looking into the transformer primary, I should measure the primary impedance as transformed by the impedance transformation properties of the transformer and according to Manfred, I should see around 200 ohms (theoretically - assuming a 1W output and an impedance transformation that is working as it should!).
Comment: I like this, it makes total sense and I will try this later and report back.
Donald, I think you reiterated that measuring the output impedance as I've attempted to do makes no sense. By the way, RF Man Channel on youtube has a video on this very topic and if I understand his approach correctly, he's doing exactly as I have done?? ()
But, I think Manfred has hit the nail on the head. This amplifier simply isn't giving me the output power I expected and I've read all the passages from EMRFD and multiple other sources on how things are "supposed" to go. I understand the first few paragraphs of Manfred's response and this is certainly the approach I took at the start. However, when I tried impedance matching the collectors to the 200 ohms or so required to output 1 watt, I got an output that fell disappointingly short of that. Since then I've been in experimental mode, swapping transistors (I had tried BD139s, two in parallel for 4 in total), introducing and reducing feedback and changing the winding ratios of the transformers.
@Manfred I understand your point about the low-pass effect of the transistor capacitance, but what I don't understand is how other experimenters seem to generate much more power from similar circuits (like this one: )?? I just get the sneaky feeling that I'm doing something incorrectly and my suspicions are converging on the output transformer.
Just a couple of additional notes: the collector base feedback network was something I'd planned into my PCB, but hadn't used until I started playing with the 2SC2078's. Without that network, the circuit oscillates like crazy. I've been pushing up the value of the feedback resistor (thereby reducing the feedback), but I'm going to suggest that even if that feedback wasn't there, it still wouldn't produce the output power it should.
The output transformer windings fall far short of Manfred's (and others!) suggested rule regarding inductance > 5-10 times load at lowest freq. The thing is, I've tried quite a few winding ratios and turns at this stage and the current ratio seems to give the best output. I find this very strange, but there you go. I'll try upping the turns later, just to double check that.
|
Once again my thanks. Formatting is very simple once you're told how. I used W10 tools for a practice run but also downloaded the recommended software to do the new 32GB card when I get my hands on it.
|
Oh, and if you don't have one you will need a USB SD Card Reader. A google search will give you an idea of what is available.
|
This is a good utility to format SD cards:
|
Thanks for the help! I was using a 64G card but will switch down to a 32GB. I will have to figure out how to format the card, have never done that so a new learning experience. I'm using the 1.0.70 FW. The slot on the case and the slot on the board are not very well aligned so first pass the card went directly into the interior of the case. Had to open the case to extract the card. I used that opportunity to see which way to insert the card. Have marked the case for the future when I forget which way.
|
On Tue, Dec 14, 2021 at 02:29 PM, N5SE wrote: Roger,
Downloaded en.stsw-stm32102.zip from:
Unzipped en.stsw-stm32102.zip to six files:
readme.txt
VCP_V1.5.0_Setup_W7_x64_64bits.exe
VCP_V1.5.0_Setup_W7_x86_32bits.exe
VCP_V1.5.0_Setup_W8_x64_64bits.exe
VCP_V1.5.0_Setup_W8_x86_32bits.exe
version.txt
Since I run windows 7 32bit, I installed:
VCP_V1.5.0_Setup_W7_x86_32bits.exe
Went through install procedure Got install confirmation. Plugged in nanovna,
no bing! Turned on nanovna, no bing! No indication at all anything happened!
Go to device manager and find:
STMicroelectronics Virtual COM Port (COM15)
General: Port_#0003.Hub_#0004
Is this what I am looking for. Is everything oi?
Yes the driver for your NanoVNA is installed. Try installing NanoVNA Saver and see if you can connect to it on COM15 Roger
|
Re: Measuring the output impedance of a 20m push-pull power pre-amp
#transformer
There is no point really in measuring the actual output impedance of a power amplifier stage, because you should not be trying to match it with an equal load impedance.
What you need to determine is what the correct load impedance between collectors has to be, to get correct operation of that stage.
And that's pretty easy to calculate. Instead of giving you a simple formula, I will explain step by step:
Each of your transistors can pull its collector voltage pretty close to ground, but not fully. The closer the instant drain voltage gets to ground, the higher becomes the transistor's internal capacitance, so there is a soft limiting effect. In practice, you can simply assume that a transistor like that can pull its collector down to roughly 1V, but the actual value will depend on the load, the frequency, and the amount of distortion you can accept.
The opposite peak voltage is as much above the supply voltage, as this 1V level is below the supply. So, if you happen to use a 12V supply, that high peak would be 23V, and this means that the peak-to-peak voltage on each of your collectors is 22V.
Since it's a push-pull circuit, the collector-to-collector voltage is twice that, so, 44V p-p.
The RMS voltage, important for power calculation, is the peak-to-peak voltage divided by 2 and then divided by the square root of 2, so it's 15.6V.
If you want 1W output, divide 1¡Â15.6, and you get 0.064A. That's the RMS current that needs to flow in your primary winding.
Now apply Ohm's Law, and it tells you that the load resistance that needs to appear between collectors is 244?. This means that you need to transform the impedance DOWN to 50?, not UP! The 1-3? value mentioned by Dave is typical for high power transistors, but not for a 1W amplifier. In any case the correct value has to be calculated.
If you use a simple 2:1 turns ratio, then a 50? load on the secondary will cause a 200? load between collectors. Having 200? instead of 244? isn't a problem. It simply means that when delivering 1W, the collectors will swing over a slightly smaller voltage range than 22V p-p, and that the RF current will be slightly higher than the 64mA calculated above. So the efficiency of the amplifier will be somewhat lower, while at the same time giving you a little more headroom, and slightly lower distortion. So, 2:1 turns ratio is what I would use. Remember that this assumes a 12V supply! If you are using a different voltage, the values will all change.
You need to use enough turns so that the inductance of the primary is at least 3 to 5 times the collector load value, at the lowest frequency of operation. That means roughly 1k? reactance, in this case. 12 turns will get you close to this. The secondary, of course, will then have 6 turns. You don't need a separate feed choke in this case. You can simply center tap the primary, and apply the supply there. The bifiliar feed choke becomes necessary when you have a single-turn primary, in a high-power, low impedance situation, because a single turn cannot be center-tapped. A connection to its midpoint would result in no coupling between the halves. With 12 turns, you don't have that problem, so, save that FT-37-43!
After winding and installing this transformer, your amplifier should work decently at low frequencies. As the frequency rises, you will run into trouble coming from leakage inductance in the transformer, capacitances everywhere, and speed limitations of the transistors. The 2SC2078 has 45pF output capacitance at 10V. Add stray capacitances in the windings and wiring, and you probably get 60pF or so, total. Two of these in series, for the pushpull, make 30pF across your primary. To have negligible effect, the reactance of this capacitance should stay above 3-5 times the load value, at the highest frequency of operation. With a load of 200-244?, that's close to 1k?, and 30pF will limit you to operate up to about 5MHz!
That's the problem with broadband circuits. The parasitics often make it hard to cover the whole frequency range desired. You can improve matters by using tricks like absorbing the transistor output capacitance into lowpass filter sections, but it all has its limits.
Broadband work gets easiest if the impedances stay in a range close to 50?. The 2SC2078 can easily deliver 1W alone, rather than using a pushpull pair. The required load impedance would be very close to 50?, so you could use direct coupling, without any output transformer! But you would need to run it in class A, to get sufficiently low distortion. That means a constant dissipation of about 4W in that transistor. Easy to do, since that transistor can handle a lot more, but not as efficient as class AB. The lower load impedance reduces the effect of the transistor's capacitances, pushing the upper frequency limit up, but not nearly to 30MHz. It should work fine at 7MHz, and acceptable at 14MHz, but will drop a lot on higher bands.
If you need all efficiency you can get, and thus want to keep the push-pull design, you will have to move to smaller transistors. Also if you want flat response to 30MHz or beyond. My secret tip would be the 2N5109. Although pricey these days, it's a very good little transistor, and its output capacitance is fifteen times lower than that of the 2SC2078, essentially removing transistor capacitance as a limiting factor at HF. For the same reason it will generate far lower distortion. This has to do with the fact that much of a transistor amplifier's distortion comes from its voltage-modulated internal capacitances. Smaller capacitances, smaller distortion.
2N3866 is a cheaper alternative. Unfortunately these days there aren't a lot of suitable transistors in this power class to choose from.
The FT82 core is huge! You can use a much smaller core for a 1W amplifier. I always prefer twin-hole cores over toroids, because they give a better ratio between inductance and wire length, and that allows better bandwidth,
To measure with the nanoVNA whatever transformer you build, I suggest doing it outside the circuit, with suitable load resistors. You can add the transistor capacitances by adding suitable small capacitors, and see what happens. Most important is to measure the leakage inductance, and optimize your transformer to get it very low, because leakage inductance most often ends up being the final limiting factor in bandwidth. When you cannot get the leakage inductance of a conventional transformer low enough, you have to move on to a transmission-line transformer.
43 material can be used throughout the HF range, but is more lossy than 61, specially in the center of the range, where 61 excels. Between those two, you always need to weigh costs and benefits. When you need lots of inductance, use 43. When you need to handle some DC, use 61. When linearity is important, use 61. When loss is important, 61 of course. When cost is important, specially with large cores, 43. When the cores might get very hot, use 61. I often end up using small 43 cores in the driver stages, and large 61 cores in the high power stages.
Bifiliar/twisted winding gives you much lower leakage inductance, but much higher interwinding capacitance. So, twisted windings are optimal at low impedances, while more separated windings work better at high impedances. The limit between "low" and "hIgh" might be roughly 100?, but depends on the situation. The nanoVNA is your friend when optimizing these transformers!
Manfred
|
Roger, Downloaded en.stsw-stm32102.zip from: Unzipped en.stsw-stm32102.zip to six files: readme.txt VCP_V1.5.0_Setup_W7_x64_64bits.exe VCP_V1.5.0_Setup_W7_x86_32bits.exe VCP_V1.5.0_Setup_W8_x64_64bits.exe VCP_V1.5.0_Setup_W8_x86_32bits.exe version.txt Since I run windows 7 32bit, I installed: VCP_V1.5.0_Setup_W7_x86_32bits.exe Went through install procedure Got install confirmation. Plugged in nanovna, no bing! Turned on nanovna, no bing! No indication at all anything happened! Go to device manager and find: STMicroelectronics Virtual COM Port (COM15) General: Port_#0003.Hub_#0004 Is this what I am looking for. Is everything oi? Thanks Roger, 30, Billy ________________________________ From: [email protected] < [email protected]> on behalf of Roger Need via groups.io <sailtamarack@...> Sent: Monday, December 13, 2021 19:02 To: [email protected] < [email protected]> Subject: Re: [nanovna-users] Confused Newbie Billy, I understand your frustration so I will help get you started. The NanoVNA uses 2 drivers to communicate with the PC. One is when you want to exchange data with a terminal or application program like NanoVNA Saver. The second is when you want to update the firmware by loading a "DFU" file. In this post I will only cover the first. Windows 7 does not have the driver installed for exchanging data so you have to install it manually. The procedure to download it from ST is cumbersome and requires registration for some reason. So I have placed it on my Box clouid drive for you to easily download. Here is the link: Once you have successfully downloaded it you need to unzip it and install the driver Win 7. There are 32 bit and 64 bit depending on what type of CPU you have in your PC. Do not plug in the NanoVNA until after you have installed the driver. When you plug in the NanoVNA after installation you will hear Windows "bing" and you will set it in the Device Manager as a COM port. Next step is to install NanoVNA Saver. You will find it at this link. Download the Windows 32 or 64 bit version, unzip and install. Then run it and patiently wait - it often takes 10 seconds or so to load because it is a Python program and it unpacking the support routines. Then in the bottom left hand corner it should show the COM port of the NanoVNA (Rescan if necessary). Click "Connect to Device" and it should plot in the graphs. This will get you started. In the Wiki for this group and on Youtube there are tutorials on how to use NanoVNA Saver. Roger Roger
|
Re: Measuring the output impedance of a 20m push-pull power pre-amp
#transformer
I don¡¯t think I know how to read impedance either on a nanoVNA, but it lets you read resistance R on one output and reactance X on another, and the impedance would simply be R + jX.
toggle quoted message
Show quoted text
On Dec 14, 2021, at 2:43 PM, David Spaedt <dspaedt@...> wrote:
A quick question on impedance. I'm using a NanoVNA H4 for simply checking
SWR on antennas across the various amatuer HAM bands.
It seems to work great. However, I've also been curious about the
Impedance. In the menu I select "SWR" to be displayed. I do not see a
selection for Impedance.
How do I configure the H4 to display the Impedance that is being measured?
Keep in mind this is just a simple case of a HF antenna and 25' - 75' feet
of coax hooked directly into the H4. I'm just curious as to what the
impedance is and I can't puzzle out how to see that value.
Sorry for the rookie question.
On Tue, Dec 14, 2021 at 2:27 PM W0LEV <davearea51a@...> wrote:
1) The impedance of the primary, the collectors, is highly dependent on
drive level from the previous stage. Looking into the 50-ohm port with the
VNA even with the PP output stage properly biased but with no RF on the
bases, the impedance will not be that when the PP stage is properly biased
AND RF in the circuit. The measurement you are attempting is highly
questionable for that reason. A better and more rigorous method would be
to optimize power output into a 50-ohm load. Then disconnect the
collectors and look into the "network" from what would be the collector
side with the "network" output terminated with a 50-ohm load.
2) The broad band matching transformer on the output should have its turns
ratio adjusted to transform something in the vicinity of 1 to 3 ohms (real)
to 50-ohms - "nominally" a 5:1 turns ratio. Most of the BJT RF power
transistors reside on the extreme left side of the Smith chart with some
reactance as well. The data for the Smith chart plots from the suppliers
are usually made at full rated power and not a couple of milliwatts (from
the VNA).
Dave - W?LEV
On Tue, Dec 14, 2021 at 3:52 PM picoguy via groups.io <s1ig0=
[email protected]> wrote:
Hi,
I'm in the process of building a BJT transistor based push-pull amplifier
that I hope will develop 1 watt to drive a final power amp stage.
The design is relatively common. It uses a pair of impedance matching
transformers at the base and collector of the power transistors. The input
transformer is an FT37-43 with a biflar wound centre tapped secondary fed
from an active biasing circuit and the output transformer is an FT82-43.
The output transformer primary is connected across the transistor
collectors, with the DC (12V) supplied via a biflar wound centre tapped
FT37-43. The transistors are 2SC2078's, which should have no problem
supplying the power I need at this stage: 1 watt.
I've had some issues with the output transistor, where my experiments
would suggest it isn't acting like an "ideal" impedance transformer. The
toroid is confirmed genuine and the type 43 material should be good for
14MHz.
Anyway, that prompted me to try looking for the output impedance of the
push-pull arrangement. I figured I could use the NanoVNA by having it feed
a signal through the output port with a bias applied (but no input signal
obviously!) The experiment sort-of works...let me explain. I find that
the impedance (resistance + reactance) is not quite what I expected. The
reactance is very inductive and my experiments would suggest that this is
certainly influenced by the secondary winding inductance. If I put 10
turns on the output secondary, I get xx uH. If I reduce the turns to 3 or
4, I get nH's. However, and importantly for what I'm looking to achieve,
when the real/resistance part of the impedance is near 50 ohms I can see
that the amp reaches a power peak.
So, while I may not have figured out what the output impedance of my amp
is, I can at least see a figure of merit on the VNA and I can tune my setup
to achieve a 50 ohm resistance. I'm still puzzled by the inductive
reactance and I'm not sure I'm doing any of this right!? Any advice from
an old hand who has done all this before would be greatly appreciated!!
Regards.
--
*Dave - W?LEV*
*Just Let Darwin Work*
|
Re: Measuring the output impedance of a 20m push-pull power pre-amp
#transformer
A quick question on impedance. I'm using a NanoVNA H4 for simply checking
SWR on antennas across the various amatuer HAM bands.
It seems to work great. However, I've also been curious about the
Impedance. In the menu I select "SWR" to be displayed. I do not see a
selection for Impedance.
How do I configure the H4 to display the Impedance that is being measured?
Keep in mind this is just a simple case of a HF antenna and 25' - 75' feet
of coax hooked directly into the H4. I'm just curious as to what the
impedance is and I can't puzzle out how to see that value.
Sorry for the rookie question.
toggle quoted message
Show quoted text
On Tue, Dec 14, 2021 at 2:27 PM W0LEV <davearea51a@...> wrote: 1) The impedance of the primary, the collectors, is highly dependent on
drive level from the previous stage. Looking into the 50-ohm port with the
VNA even with the PP output stage properly biased but with no RF on the
bases, the impedance will not be that when the PP stage is properly biased
AND RF in the circuit. The measurement you are attempting is highly
questionable for that reason. A better and more rigorous method would be
to optimize power output into a 50-ohm load. Then disconnect the
collectors and look into the "network" from what would be the collector
side with the "network" output terminated with a 50-ohm load.
2) The broad band matching transformer on the output should have its turns
ratio adjusted to transform something in the vicinity of 1 to 3 ohms (real)
to 50-ohms - "nominally" a 5:1 turns ratio. Most of the BJT RF power
transistors reside on the extreme left side of the Smith chart with some
reactance as well. The data for the Smith chart plots from the suppliers
are usually made at full rated power and not a couple of milliwatts (from
the VNA).
Dave - W?LEV
On Tue, Dec 14, 2021 at 3:52 PM picoguy via groups.io <s1ig0=
[email protected]> wrote:
Hi,
I'm in the process of building a BJT transistor based push-pull amplifier
that I hope will develop 1 watt to drive a final power amp stage.
The design is relatively common. It uses a pair of impedance matching
transformers at the base and collector of the power transistors. The input
transformer is an FT37-43 with a biflar wound centre tapped secondary fed
from an active biasing circuit and the output transformer is an FT82-43.
The output transformer primary is connected across the transistor
collectors, with the DC (12V) supplied via a biflar wound centre tapped
FT37-43. The transistors are 2SC2078's, which should have no problem
supplying the power I need at this stage: 1 watt.
I've had some issues with the output transistor, where my experiments
would suggest it isn't acting like an "ideal" impedance transformer. The
toroid is confirmed genuine and the type 43 material should be good for
14MHz.
Anyway, that prompted me to try looking for the output impedance of the
push-pull arrangement. I figured I could use the NanoVNA by having it feed
a signal through the output port with a bias applied (but no input signal
obviously!) The experiment sort-of works...let me explain. I find that
the impedance (resistance + reactance) is not quite what I expected. The
reactance is very inductive and my experiments would suggest that this is
certainly influenced by the secondary winding inductance. If I put 10
turns on the output secondary, I get xx uH. If I reduce the turns to 3 or
4, I get nH's. However, and importantly for what I'm looking to achieve,
when the real/resistance part of the impedance is near 50 ohms I can see
that the amp reaches a power peak.
So, while I may not have figured out what the output impedance of my amp
is, I can at least see a figure of merit on the VNA and I can tune my setup
to achieve a 50 ohm resistance. I'm still puzzled by the inductive
reactance and I'm not sure I'm doing any of this right!? Any advice from
an old hand who has done all this before would be greatly appreciated!!
Regards.
--
*Dave - W?LEV*
*Just Let Darwin Work*
|
Re: Measuring the output impedance of a 20m push-pull power pre-amp
#transformer
picoguy, I infer that you have a class B push-pull amplifier you are measuring.
S-parameters are by definition characteristics of a linear small-signal device. Any non-Class A power amplifier will exhibit different impedance characteristics depending upon the drive level at its input.
Measuring the output impedance of an un-driven, "cold" amplifier will tell you little about its behavior under drive, or "hot".
Measuring it WITH drive is risky and not for the unexperienced nor faint of heart and will almost certainly result in a destroyed nanoVNA. I have been making VNA measurements since the '70s and have always been able to avoid making such a measurement.
Google "hot S22" and you will find many references.
73, Don N2VGU.
|
Re: #general SD card function on H4
#general
Some earlier versions of FW had timing issues. Use the latest HUGEN or DiSlord FW for the H4 and you should be OK.
toggle quoted message
Show quoted text
On Tuesday, December 14, 2021, 02:32:10 p.m. EST, <rseverns@...> wrote:
I have a new H4 with a micro SD port.? The port polarity is not marked on the case but with a little care I was able to insert it correctly.? However, when I try to save a screen plot I get "fail write".? What have I missed?
Rudy N6LF
|
Re: #general SD card function on H4
#general
Hi Rudy!
Be sure to have the unit fully charged, and use a SD card max 32GB formatted with FAT32. Turn unit OFF, insert the card and then ON again. If you try to put it the wrong way, there is some kind of lock, you just "feel" it. If correct, you feel the spring. If you use a little force, it's possible to insert it the wrong way. Be careful!
I had the same expirience with another unit!
Karl
|
#general SD card function on H4
#general
I have a new H4 with a micro SD port. The port polarity is not marked on the case but with a little care I was able to insert it correctly. However, when I try to save a screen plot I get "fail write". What have I missed?
Rudy N6LF
|
Re: Measuring the output impedance of a 20m push-pull power pre-amp
#transformer
1) The impedance of the primary, the collectors, is highly dependent on drive level from the previous stage. Looking into the 50-ohm port with the VNA even with the PP output stage properly biased but with no RF on the bases, the impedance will not be that when the PP stage is properly biased AND RF in the circuit. The measurement you are attempting is highly questionable for that reason. A better and more rigorous method would be to optimize power output into a 50-ohm load. Then disconnect the collectors and look into the "network" from what would be the collector side with the "network" output terminated with a 50-ohm load. 2) The broad band matching transformer on the output should have its turns ratio adjusted to transform something in the vicinity of 1 to 3 ohms (real) to 50-ohms - "nominally" a 5:1 turns ratio. Most of the BJT RF power transistors reside on the extreme left side of the Smith chart with some reactance as well. The data for the Smith chart plots from the suppliers are usually made at full rated power and not a couple of milliwatts (from the VNA). Dave - W?LEV On Tue, Dec 14, 2021 at 3:52 PM picoguy via groups.io <s1ig0= [email protected]> wrote: Hi,
I'm in the process of building a BJT transistor based push-pull amplifier
that I hope will develop 1 watt to drive a final power amp stage.
The design is relatively common. It uses a pair of impedance matching
transformers at the base and collector of the power transistors. The input
transformer is an FT37-43 with a biflar wound centre tapped secondary fed
from an active biasing circuit and the output transformer is an FT82-43.
The output transformer primary is connected across the transistor
collectors, with the DC (12V) supplied via a biflar wound centre tapped
FT37-43. The transistors are 2SC2078's, which should have no problem
supplying the power I need at this stage: 1 watt.
I've had some issues with the output transistor, where my experiments
would suggest it isn't acting like an "ideal" impedance transformer. The
toroid is confirmed genuine and the type 43 material should be good for
14MHz.
Anyway, that prompted me to try looking for the output impedance of the
push-pull arrangement. I figured I could use the NanoVNA by having it feed
a signal through the output port with a bias applied (but no input signal
obviously!) The experiment sort-of works...let me explain. I find that
the impedance (resistance + reactance) is not quite what I expected. The
reactance is very inductive and my experiments would suggest that this is
certainly influenced by the secondary winding inductance. If I put 10
turns on the output secondary, I get xx uH. If I reduce the turns to 3 or
4, I get nH's. However, and importantly for what I'm looking to achieve,
when the real/resistance part of the impedance is near 50 ohms I can see
that the amp reaches a power peak.
So, while I may not have figured out what the output impedance of my amp
is, I can at least see a figure of merit on the VNA and I can tune my setup
to achieve a 50 ohm resistance. I'm still puzzled by the inductive
reactance and I'm not sure I'm doing any of this right!? Any advice from
an old hand who has done all this before would be greatly appreciated!!
Regards.
-- *Dave - W?LEV* *Just Let Darwin Work*
|
On Mon, Dec 13, 2021 at 05:51 PM, N5SE wrote:
Came back in and checked eMail one more time. I have gone to both URL's and
downloaded en.stsw-stm32102.zip and NanoVNASaver.x86.zip. Since I run Windows
7 32 bit, that looks right. will get after this Wednesday. Gone this time.
Will let you know via private eMail rain or shine. Thanks a million.
Hope it now works for you. Please post any replies to this group instead of private email. That way others that search this topic later will have the full conversation. Roger
|
Re: Newer version firmware
Here is the latest firmware for the H/H4 posted on the manufacturers (Hugen) github.
DiSlord is the major developer of new firmware and his improvements/changes are verified by Hugen and then posted on his github for buyers of his products.
Roger
|
Re: Measuring the output impedance of a 20m push-pull power pre-amp
#transformer
Apologies, my original mail should have read: "I've had some issues with the output *transformer*, where..."
|
Measuring the output impedance of a 20m push-pull power pre-amp
#transformer
Hi,
I'm in the process of building a BJT transistor based push-pull amplifier that I hope will develop 1 watt to drive a final power amp stage.
The design is relatively common. It uses a pair of impedance matching transformers at the base and collector of the power transistors. The input transformer is an FT37-43 with a biflar wound centre tapped secondary fed from an active biasing circuit and the output transformer is an FT82-43. The output transformer primary is connected across the transistor collectors, with the DC (12V) supplied via a biflar wound centre tapped FT37-43. The transistors are 2SC2078's, which should have no problem supplying the power I need at this stage: 1 watt.
I've had some issues with the output transistor, where my experiments would suggest it isn't acting like an "ideal" impedance transformer. The toroid is confirmed genuine and the type 43 material should be good for 14MHz.
Anyway, that prompted me to try looking for the output impedance of the push-pull arrangement. I figured I could use the NanoVNA by having it feed a signal through the output port with a bias applied (but no input signal obviously!) The experiment sort-of works...let me explain. I find that the impedance (resistance + reactance) is not quite what I expected. The reactance is very inductive and my experiments would suggest that this is certainly influenced by the secondary winding inductance. If I put 10 turns on the output secondary, I get xx uH. If I reduce the turns to 3 or 4, I get nH's. However, and importantly for what I'm looking to achieve, when the real/resistance part of the impedance is near 50 ohms I can see that the amp reaches a power peak.
So, while I may not have figured out what the output impedance of my amp is, I can at least see a figure of merit on the VNA and I can tune my setup to achieve a 50 ohm resistance. I'm still puzzled by the inductive reactance and I'm not sure I'm doing any of this right!? Any advice from an old hand who has done all this before would be greatly appreciated!!
Regards.
|
Hi ,
Sometimes I have tis problem as well, mostly it helps when I delete the .ini file in NanoVNA-App's directory? restart the application and set my preferences again.
Just my 2 cents,
Joss.
Op 14-12-2021 om 14:11 schreef Antonio (EA7HJ) via groups.io:
toggle quoted message
Show quoted text
Hi
Thank you, Larry Rothman for your help.
I'm sorry, but after trying everything, I give up.
1) installed the firmware version that you indicate to me (V1.0.69), also v1.0.71 and V1.1.
2) Uninstalling all software and controllers on the PC.
3) I make a clean installation of VCP_V1.5.0_SETUP_W8_X64_64BITS.EXE and DFUSE_DEMO_V3.0.6_SETUP.EXE.
4) Terratem installation. (The nanovna is connected and disconnected from the terminal, as well as Teraterm perfectly recognizes the serial port with the installed driver STMicroelectronic).
5) Nanovna-App works in DFU mode, he installed the different firmware, but it does not recognize the Nanovna in serial port mode.
6) Back to original firmware V.0.5.0 and everything returns to normal.
So far my experience. I hope that one day you see the light. LOL
I am very grateful for your help.
Antonio - EA7HJ.
--
Dit e-mailbericht is gecontroleerd op virussen met Avast antivirussoftware.
|
Roger Need
N5SE
Donald S Brant Jr