|
On Sun, Jul 14, 2024 at 08:50 AM, ajparent1/kb1gmx wrote:
I make it a point of using separate relays to get better isolation and often its
a mechanical convenience as well.
Thank you for that very practical tip, given how low cost these little relays are, it's probably best to carry out that seperation as you suggest - I;ve never built an RF amplifier before, so all these tips are most helpful Thank you, 73 Peter - GM8JCf
|
One area is that input and output must not be close as in using the same relay.
If isolation is not enough oscillation can be a unwanted result.
I make it a point of using separate relays to get better isolation and often its
a mechanical convenience as well.
--
Allison
------------------
Post online only,?
direct email will go to a bit bucket.
|
On Sun, Jul 14, 2024 at 07:20 AM, Timothy Fidler wrote:
Put colour coding via Red and blue heat shrink on the Two cables which have the common BNC connectors and a red and blue spot adjacent to the bulkhead connector at the amp. Viz J2 and J5.
Useful tip, thank you Peter - GM8JCF
|
On Sun, Jul 14, 2024 at 07:20 AM, Timothy Fidler wrote:
Use a contact from Relay one ?as a the input controller for the RF use a contact from Relay 2 as the output controller for RF out. ?Currently you have RF in and RF out in the same relay. Some ?Chinese boards do that (RF in and RF out in the same relay housing ) but it is not good practice.
? Tnx for the tip Peter - GM8JCF
|
On Sun, Jul 14, 2024 at 08:22 AM, Barbaros WB2CBA wrote:
Here is my take of QRPL linear 10W PA which I use with my usdx trx. I thought this might be a helpful guide for you.
Hi Barb Thank you very much indeed, really detailed instructions, very much appreciated. 73 Peter - GM8JCF
|
Hi Peter,
Here is my take of QRPL linear 10W PA which I use with my usdx trx. I thought this might be a helpful guide for you.
--
?
73, Barb
|
my back of envelope calcs show that if you use four pole filters and chose the correct break point for the 5 Mhz filter at about 8.5 Mhz you can put in a no switched filter set for 60- 40- 30 mtrs and get about -17 dB suppression for all of those three bands The basic LP filter design for the 7 Mhz filer would be as per the Qmx 40/30 mtr filter. However if you had only 2 poles for the 7 Mhz filter you'd get only about -13 dB of suppression on the 14 Mhz harmonic which may be/ would be light on once the primary signal had gone up to 55.4 dBm
his concept is on the basis that 60-40-30-20-17-15 MTRS is a complete band set that is acceptable . Spinning the cat by the tail that way may make it possible to have only one set of DPDT Relays (a pair) providing filter selection over that entire range. This the post 10 W amp filter I refer to. THis is based on harmonic plots I have seen published for the bare Hermes Lite.
|
Use a contact from Relay one ?as a the input controller for the RF use a contact from Relay 2 as the output controller for RF out. ?Currently you have RF in and RF out in the same relay. Some ?Chinese boards do that (RF in and RF out in the same relay housing ) but it is not good practice. ?Some will burst into oscillation as result ?I'm sure.
?
Detailed build item -
Put colour coding via Red and blue heat shrink on the Two cables which have the common BNC connectors and a red and blue spot adjacent to the bulkhead connector at the amp. Viz J2 and J5.
TEF
|
On Sun, Jul 14, 2024 at 05:36 AM, ajparent1/kb1gmx wrote:
One note if you go direct to antenna (not using the BLA-amp) from the
QRPL amp a set of low pass filters are needed.
Thank U Allison, it is tempting to add a set of the QRP Labs low pass filters so as to make a fully standalone 10W Linear Amp; something to think about indeed. Once again, thank you Peter - GM8JCF
|
That should do what you desire.
One note if you go direct to antenna (not using the BLA-amp) from the
QRPL amp a set of low pass filters are needed. Driving the amp that
is not required.
--
Allison
------------------
Post online only,?
direct email will go to a bit bucket.
|
Hi
I've redrawn the system schematic to reflect using the HL2 0~17dBm RF output, please see below
And in case the quality is insufficient, there is also a .PDF attached.
Could I please ask that you all take a close look and find errors, suggest improvemnts.
I would really like to thank everyone for their ideas, and suggestions in helping me to get this right, but special mention must go to KB1GMX, thank you Allison.
Than You, Take Care All, 73
Peter - GM8JCF
|
On Sat, Jul 13, 2024 at 06:53 AM, ajparent1/kb1gmx wrote:
So the QRPL amp with the HL2 17dBm output is a better match to the task.
If that ends up being too much drive a small low power (50mW) 2-4db
attenuation is far easier to do if needed.
Hi Allison I rigged up a 50ohm BNC dummy load and connected it to the DSO, (Siglent SDS1102CML+ 100MHz Dual channel oscilloscope), and life is suddenly V good, see table below !!!
| Fmeasure= |
5398.5 |
KHz |
Zload= |
50 |
ohm |
| ? |
? |
? |
? |
? |
? |
| dB Output |
Vpp |
Vrms(mV) |
V^2 |
P(mW) |
P(dBm) |
| -7.5 |
1.77 |
580 |
0.3364 |
6.73 |
8.28 |
| -7.0 |
1.80 |
620 |
0.3844 |
7.69 |
8.86 |
| -6.5 |
1.90 |
660 |
0.4356 |
8.71 |
9.40 |
| -6.0 |
2.00 |
700 |
0.4900 |
9.80 |
9.91 |
| -5.5 |
2.10 |
740 |
0.5476 |
10.95 |
10.39 |
| -5.0 |
2.24 |
780 |
0.6084 |
12.17 |
10.85 |
| -4.5 |
2.36 |
820 |
0.6724 |
13.45 |
11.29 |
| -4.0 |
2.52 |
880 |
0.7744 |
15.49 |
11.90 |
| -3.5 |
2.66 |
920 |
0.8464 |
16.93 |
12.29 |
| -3.0 |
2.80 |
980 |
0.9604 |
19.21 |
12.83 |
| -2.5 |
2.96 |
1040 |
1.0816 |
21.63 |
13.35 |
| -2.0 |
3.16 |
1160 |
1.3456 |
26.91 |
14.30 |
| -1.5 |
3.36 |
1180 |
1.3924 |
27.85 |
14.45 |
| -1.0 |
3.54 |
1280 |
1.6384 |
32.77 |
15.15 |
| -0.5 |
3.76 |
1320 |
1.7424 |
34.85 |
15.42 |
| 0.0 |
4.00 |
1400 |
1.9600 |
39.20 |
15.93 |
So the RF1 17dBm output is indeed fully variable, and a simple fixed 3dB attenuator between it and the QRPLabs 10W PA should bring down the drive level to around 25mW. The "dB Output" data are the values from the "Drive Control" in the Thetis SDR s/w, ie 0dB is full power out Thank you ever so much for your suggestions, and prompting me to try the 17dBm output, this certainly makes for a cleaner/simpler design both philosophically and practically. Thank You once again, 73 Peter - GM8JCF
|
On Sat, Jul 13, 2024 at 11:36 AM, G8JCF wrote:
The 17dBm output is very attractive, but I'm not sure if it's variable, eg if the "drive control" in the various SDR s/w can control the drive level, or if the drive level is fixed at 17dBm
I've taken a look at the HL2 schematics, RFFrontend.sch and PA.sch - and the drive to the 17dBm connector, "RF1", and the drive into the 5W PA come from the same pins on the FPGA, so it looks like the RF1 RF level should be variable. Looks like I'll just have to hookup the 50ohm dummy load and 'scope the signal at RF1 and hopefully find out that it's fully variable, now that would be a neat solution. 73 Peter - GM8JCF
|
On Fri, Jul 12, 2024 at 11:02 PM, Steve G4EDG wrote:
I reduced the attenuation and was able to drive the amp to the full 100w with 5w.
Hi Steve That would mean voiding the warranty, and I'd rather not risk damaging GBP 1,000+ kit if I can do a solid solution with QRPLabs ultra-stable 10W Linear Amp BTW, do U know where I can get my hands on a BLA-350 plus schematic so that I can add it to the documentation of my BLA-350 plus ? Thank You Peter - GM8JCF
|
On Sat, Jul 13, 2024 at 06:15 AM, John AE5X wrote:
And here is another solution from EU1SW. Costly but offers 14-bit, 6m and 160m and power to run your amp to full output:
I've already got the HL2, the EU1SW is basically an HL2 with a 20W Linear in an enclosure for about USD400 more than what I paid for the HL2 itself. The QRPLabs 10W linear is about USD70 delivered inc taxes etc, so all I need to do is add some relays, an enclosure, 5 connectors, and of course build the QRPLabs kit, which all in all should come in at less than USD150 (assuming that I buy all new parts) Thanks for the suggestions Peter - GM8JCF
|
On Sat, Jul 13, 2024 at 05:45 AM, ajparent1/kb1gmx wrote:
So far you have given a bunch of suggestions that does not get him
that 10W and a few that can barely do what he already has. So stop
polluting the topic with wild and inapplicable ideas.
Thank You Allison, I was beginning to feel like Alice in Wonderland !! The 17dBm output is very attractive, but I'm not sure if it's variable, eg if the "drive control" in the various SDR s/w can control the drive level, or if the drive level is fixed at 17dBm - I've read lots of manuals but so far haven't found anything which states that it is indeed variable, and it always seems to be referred to as the "fixed 17dBm output" The variable element is required because here in UK for example there is a 100W PEP max license condition for 60m, 32W PEP max license condition for most of 160m, and of course the general principle to not use more power than is required, so it would be much more convenient if I could just adjust the "drive control" in eg Thetis SDR s/w, rather than fiddling about with different attenuators, although the 17dBm level would be just within the safe operating range of an electronically controlled RF attenuator eg  not as convenient as the on-screen "drive" control, but doable. Once again, thank you Peter - GM8JCf
|
Don,
Its 350W and 1db compression is likely above that.
if he gets 260 hes less than 3db down. If the QRPL amp
is run at 15-18V it will give you more I usually see 11W
at 13.6V and at 16V more like 15W. Also the amp at
20V will produce 25+W. So there is margin there.
That and most devices for that power level might tolerate
20W before becoming fuzes. The BLA350 has a input
sensor and will trip if over 13W.
Myself I'd check the BLA for input attenuator as the apparent
gain is about 14db. Two cases for that, bipolar devices that
would be typical, or Mosfet with attenuator. Looking at the
schematic it has Mosfets and a input attenuator.
--
Allison
------------------
Post online only,?
direct email will go to a bit bucket.
|
On Sat, Jul 13, 2024 at 08:45 AM, ajparent1/kb1gmx wrote:
a nice BLA-350 that with his 5W gives him 130W. It needs 10W for
full power (greater than 350W).
Again, the numbers do not add up. Unless the amplifier has a lot of gain expansion (compression is more likely) 10 W will only give 260W, far short of the desired 350-400W. I suspect that something like 15+ watts is going to be needed, likely more if the gain compresses. I also suspect that 400W is going to be elusive. If this were my project, I would beg-borrow a 100W transceiver and use it to characterize the amplifier over frequency and input power so that the OP knows what the actual situation is. Otherwise it is all just conjecture. 73, Don N2VGU
|
>>>Linear Tech's AD8009 highspec VIdeo amp is your friend here - costs about 5 Benjamins each from Chicom dealers . THese have been successfully used as driversfor RD 15HH Mosfets in class A to get 7 W linear out. Will give up to 24.7 dBM LINEAR power into a 50R single ended load at 12 V supply, without sweating the silicon. that's 7.7 dBm over your start position.<<<
Not useful at all for this application.
First he has 37dbm (5W) already from the HL2.? So there is no step up there as
24.7 DBM is 350mW and 7 W (using RD15) is still short of the 10W desired.
Plus a lot of construction to make it work.
Since he has 5W already its only +1.4db increase for that 7W.
So the QRPL amp with the HL2 17dBm output is a better match to the task.
If that ends up being too much drive a small low power (50mW) 2-4db
attenuation is far easier to do if needed.
--
Allison
------------------
Post online only,?
direct email will go to a bit bucket.
|
And here is another solution from EU1SW. Costly but offers 14-bit, 6m and 160m and power to run your amp to full output:
https://www.ebay.com/itm/256147932051
--
John AE5X
|
Barb WB2CBA
Timothy Fidler
Donald S Brant Jr