|
Hello, let me put a hypothetical question: If it was possible to add to the 50W PA an extra board containing LP filters for 80m, 60m, 40m 20m to match capabilities of QMX (and maybe add 160 meter band, just for case...) would the following process be viable?
- feed PA output to the extra multi-band LP board internally by a coaxial cable, e.g. by adding a SMA connector to the main board and a coaxial jumper made of RG316 or similar cable?
- band switching would be handled only on the extra multi-band board, manual or automatic.
- to support design without relays, switching by 1N4007 and a small SMPS giving ~ -300 volts to shut down 1N4007 would be necessary
- main question: is there available enclosure that would have the same footprint as the current enclosure for the 50 W PA, but higher, so that the extra board would fit in?
Maybe it's a stupid idea given compact and low-cost design of the QCX 50 W PA. OTOH, I plan to use QMX potentially in emergency communication situations. If 5 watts are enough, good. If, however, we needed a little bit more power to ensure reliable communication, 50 W would be welcome. Unfortunately, we have three bands available, 80, 60 and 40 meters, and there are situations, when only one of them is usable, or at least one of them is unusable, so building a single band version would probably not help much (in my country the distances would always be not more than some 300 km, so the most likely scenario is NVIS communication during daylight hours).
|
- main question: is there available enclosure that would have the same footprint as the current enclosure for the 50 W PA, but higher, so that the extra board would fit in?
Not available from me... all my enclosures are custom made in a CNC and cutting and laser-printing factory and I don't get the pre-drilled versions here. Only when they're drilled, CNC'ed, laser etched, cut, anodized blah. But there should be options.?
Aluminium extrusion molds are a generally expensive process, much like it would be if you wanted injection molded plastic parts. You need high volume. Accordingly the practical and economic route for a small volume operation is to use existing parts where possible. In our case I select existing extrusion profiles then design my board dimensions to fit, and choose a cut length appropriate to whatever I need to put inside. I have a factory do all the customizing mentioned above, but the crucial point is, the aluminium extrusion is an existing one, not exclusive to QRP Labs. Therefore you would probably be able to find enclosures with similar profiles on AliExpress, eBay, Amazon etc.?
73 Hans G0UPL
Hello, let me put a hypothetical question:
If it was possible to add to the 50W PA an extra board containing LP filters for 80m, 60m, 40m 20m to match capabilities of QMX (and maybe add 160 meter band, just for case...) would the following process be viable?
- feed PA output to the extra multi-band LP board internally by a coaxial cable, e.g. by adding a SMA connector to the main board and a coaxial jumper made of RG316 or similar cable?
- band switching would be handled only on the extra multi-band board, manual or automatic.
- to support design without relays, switching by 1N4007 and a small SMPS giving ~ -300 volts to shut down 1N4007 would be necessary
- main question: is there available enclosure that would have the same footprint as the current enclosure for the 50 W PA, but higher, so that the extra board would fit in?
Maybe it's a stupid idea given compact and low-cost design of the QCX 50 W PA. OTOH, I plan to use QMX potentially in emergency communication situations. If 5 watts are enough, good. If, however, we needed a little bit more power to ensure reliable communication, 50 W would be welcome. Unfortunately, we have three bands available, 80, 60 and 40 meters, and there are situations, when only one of them is usable, or at least one of them is unusable, so building a single band version would probably not help much (in my country the distances would always be not more than some 300 km, so the most likely scenario is NVIS communication during daylight hours).
|
To keep it simple in an emergency situation you could just use outboard lowpass filters, swapping them manually as needed.? ?Put your highest-band filter in the box (always inline) and add lower-band ones when required.??
Instead of "...and a small SMPS giving ~ -300 volts to shut down 1N4007..."? ?you could just tap and rectify/voltage-double the RF output.? This would guarantee always having sufficient voltage to bias your diodes off.
73, Don N2VGU
|
On Fri, Jan 5, 2024 at 06:25 AM, Donald S Brant Jr wrote:
To keep it simple in an emergency situation you could just use outboard lowpass filters, swapping them manually as needed.? ?Put your highest-band filter in the box (always inline) and add lower-band ones when required.??
Does this always work? I wondered about this elsewhere and was told:?The input impedance of LPF's is designed to be 50 ohms, but this is not the case over all input frequencies. Typically it is "close" to 50 ohms only for a small portion of the HF band below the cutoff frequency. So adding a lot in series that aren't designed for the operating frequency can mess up the output SWR of the PA (whose output transformer is designed to match the 50 ohms of the - correct - LPF).
-Nate
N8BTR
|
It is common practice to use a low-pass filter that is common to all bands. In my homemade transceiver, a low-pass filter with a cutoff frequency of 32 MHz is installed on the PA board. The remaining filters are located separately. They need to be configured individually. And don't believe that your output stage has the greatest output at a 50 ohm load. This value can simply be unknow
--
73, Олег, UY2ZA
|
Perhaps a 3D-printed "plug" which mated with both top and bottom could be designed to increase the height.? New, taller end plates should be simple to fabricate, from a PCB shop or by hand with snips, a drill and file if necessary.
73, Don N2VGU
|
I like Don's idea of the 3D printed side extensions....
Oleh, interesting point about top end filters...... Makes me tempted to buy some of the easy LPFs from Pacific and play with using my own 20m amp on other bands.?
Those are set up to just be plugged in after the amplifier or transceiver with BNCs on board.?
-Nate
N8BTR
|
Hi?
I played with this idea
I made an 50w amp with no LPFs inside
try to use an external switchable LPF that I bought from aliexpress
result: No Go :(
why? for each band you'll need not only different LPFs but also different bias settings
73
muhsin TA1MHS
?
|
On Fri, Jan 5, 2024 at 03:07 PM, Muhsin TA1MHS wrote:
I played with this idea
I made an 50w amp with no LPFs inside
try to use an external switchable LPF that I bought from aliexpress
result: No Go :(
why? for each band you'll need not only different LPFs but also different bias settings
Muhsin,? ? ?What do you mean, No Go? What was the problem? Lower output on other bands? Are you sure the LPF was fine?...I've seen some not good data about some of the cheap pre-made LPFs out there.? On my 20m amp, at least into a dummy load, power was fine when I put 40 and 30m through it. 80m looked odd, but I'm told I was seeing harmonics...and the power was fine, for that matter. -Nate N8BTR
|
On Fri, Jan 5, 2024 at 09:28 PM, <natereik@...> wrote:
What do you mean, No Go? What was the problem?
why? for each band you'll need not only different LPFs?but also different bias settings
73
muhsin TA1MHS
|
A couple of things, 3D side extensions and end plates are pretty easy. I made some for my QMX. And aluminum extrusions are really cheap but the MOQ is one metric ton from my supplier who specializes in small runs. Large complex heat exchanger dies were very cheap compared to the ton of aluminum.?
--
Colin - K6JTH?
|
Hi Muhsin
I'm interested in your observations.?
Theorietically different bias settings aren't needed. Bias setting is a DC adjustment in the absence of any RF (of any band), described on page 66 of the assembly manual. Turn the trimmer clockwise just enough to start an observable current flow.?
During the adjustment the 50W PA has no incoming RF of any sort, the poor IRF510s have no idea what frequency you are about to send them, nor any idea of what LPFs you have connected downstream.?
Therefore the correct bias setting is band independent. Absolutely by definition.?
On the other hand you may have chosen to adjust for maximum smoke (power output). Which may very well change the adjustment per band; but this is not the correct recommended adjustment procedure, and you will probably have produced more power at the expense of PA efficiency, which could cause overheating. In other words by that method you push the amp too hard; and tuning for maximum smoke, sometimes actually results in real actual smoke...?
Overall I'd say the idea of a 50W PA for multiple bands with external low pass filtering is sound.
73 Hans G0UPL
On Fri, Jan 5, 2024 at 09:28 PM, <natereik@...> wrote:
What do you mean, No Go? What was the problem?
why? for each band you'll need not only different LPFs?but also different bias settings
73
muhsin TA1MHS
|
QDX has excess output power to work with PA. In my copy, with a power supply of 22 Volts and a maximum current consumption of 3 Amperes, the required input power, depending on the band was: 80 m - 0.4 Watt; 40 m - 1.2 Watts; 30 m - 1.5 Watt; 20 m - 2.2 Watt. I did not change the resistor values of the input attenuator. At the same time, the amplifier had an output power of about 40 watts.
--
73, Oleh, UY2ZA
|
I was browsing Aliexpress for quite some time and did not find even the original enclosure that is used for PA 50 W :) and the potential extension would require approximately twice the height to make sense. Plus I read your heat dissipation analysis and it seems that for harsher handling, such as linear regime or long FSK transmissions it would require a bigger heatsink.
Jindra
toggle quoted message
Show quoted text
On Fri, Jan 5, 2024 at 10:13 AM, Hans Summers wrote:
- main question: is there available enclosure that would have the same footprint as the current enclosure for the 50 W PA, but higher, so that the extra board would fit in?
Not available from me... all my enclosures are custom made in a CNC and cutting and laser-printing factory and I don't get the pre-drilled versions here. Only when they're drilled, CNC'ed, laser etched, cut, anodized blah. But there should be options.?
?
Aluminium extrusion molds are a generally expensive process, much like it would be if you wanted injection molded plastic parts. You need high volume. Accordingly the practical and economic route for a small volume operation is to use existing parts where possible. In our case I select existing extrusion profiles then design my board dimensions to fit, and choose a cut length appropriate to whatever I need to put inside. I have a factory do all the customizing mentioned above, but the crucial point is, the aluminium extrusion is an existing one, not exclusive to QRP Labs. Therefore you would probably be able to find enclosures with similar profiles on AliExpress, eBay, Amazon etc.?
?
73 Hans G0UPL
Hello, let me put a hypothetical question:
If it was possible to add to the 50W PA an extra board containing LP filters for 80m, 60m, 40m 20m to match capabilities of QMX (and maybe add 160 meter band, just for case...) would the following process be viable?
- feed PA output to the extra multi-band LP board internally by a coaxial cable, e.g. by adding a SMA connector to the main board and a coaxial jumper made of RG316 or similar cable?
- band switching would be handled only on the extra multi-band board, manual or automatic.
- to support design without relays, switching by 1N4007 and a small SMPS giving ~ -300 volts to shut down 1N4007 would be necessary
- main question: is there available enclosure that would have the same footprint as the current enclosure for the 50 W PA, but higher, so that the extra board would fit in?
Maybe it's a stupid idea given compact and low-cost design of the QCX 50 W PA. OTOH, I plan to use QMX potentially in emergency communication situations. If 5 watts are enough, good. If, however, we needed a little bit more power to ensure reliable communication, 50 W would be welcome. Unfortunately, we have three bands available, 80, 60 and 40 meters, and there are situations, when only one of them is usable, or at least one of them is unusable, so building a single band version would probably not help much (in my country the distances would always be not more than some 300 km, so the most likely scenario is NVIS communication during daylight hours).
?
?
|
With a supply voltage of 22 V and a current consumption of about 3 A, the existing radiator copes quite well with RTTY or FT. This confirms my experience. But I prepared a 4x4 cm fan and a 55-degree thermostat for installation; they will come in handy in the summer. For linear mode, it is better not to modify this mind. He's beautiful just the way he is.
--
73, Oleh, UY2ZA
|
On Sat, Jan 6, 2024 at 12:20 AM, Muhsin TA1MHS wrote:
why? for each band you'll need not only different LPFs?but also different bias settings
73
muhsin TA1MHS
I understood that,...I was trying to ask what the effect was of NOT changing the bias settings in a different band. I'm assuming power output...but was curious how much different. 73 -Nate N8BTR
|
On Sat, Jan 6, 2024 at 06:29 AM, <natereik@...> wrote:
I understood that,...I was trying to ask what the effect was of NOT changing the bias settings in a different band. I'm assuming power output...but was curious how much different.
73
-Nate
N8BTR
I disagree with Mushin on this. I built a multiband QRP-Labs amp for the QDX and it worked fine. I used the QRP-Labs ultimate low pass filter kit for holding the filters and do the switching. If you don't need it automated, I would suggest building external LPF modules.? I posted quite a bit out it here somewhere.? Joshua
|
Lastly, if you do automate this process, be sure to build a band select buffer between the relays and the QMX. I killed two QDX microcontrollers in my experimenting. The relay current was traveling through the band select pads to ground on the microcontroller pins killing that I/O.? I have a simple circuit built up in Eagle to show what the buffer looks like.?
I sold my entire QDX amp combo to W6LPM a year ago if you would like to contact him and see what he thinks about it.?
Joshua
|
Hi
50 Watt, 20 Volt Multiband??
Best you can get 12-13.8? Volt 20 Watt multiband maybe
After you smoke some poor IRF510s we can come to the same page :)
73
muhsin TA1MHS
|
Hi Muhsin, all
The 50W PA was originally designed for CW use. For the much harsher conditions of high duty cycle mode operations it should be appropriately derated. I suggest 20-25W. Which is roughly what you will get from roughly 12-13.8 supply voltage.?
However this is a separate question to whether it can be used multi-band or needs band by band bias adjustment; if the correct Class C manual recommendations for setting bias are followed, then this bias setting won't be band dependent.
73 Hans G0UPL
Hi
50 Watt, 20 Volt Multiband??
Best you can get 12-13.8? Volt 20 Watt multiband maybe
After you smoke some poor IRF510s we can come to the same page :)
73
muhsin TA1MHS
|
Donald S Brant Jr
Colin Kaminski
Joshua KJ7LVZ