Changing minimum DRIVE level below -7.5dB

[Eugen Mihai Trifu]

Hello there,

My XPA125b amplifier delivers too much power to attack Acom 1000, even when the level of DRIVE is set to -7.5dB.

Is it possible to set another minimum level for the DRIVE cursor? For example -12db ...

Thanks,

73 de YO9VA

Eugen

["Christoph v. Wüllen"]

The hardware is restricted to -7.5 dB, beyond that you have to
use IQ scaling. This is implemented in some software and not
in other, since there have been mixed feelings towards IQ
sampling in this forum.

piHPSDR has full-range drive control via IQ scaling, but in
your case a 6dB attenuator that can swallow few watts
would bemore appropriate.-

> --
> You received this message because you are subscribed to the Google Groups "Hermes-Lite" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/a099c11b-094d-4005-b36a-ad0eac6df45cn%40googlegroups.com.

[James Ahlstrom]

Please do not use I/Q scaling which may be available in SDR software. Use an attenuator instead.

Jim

N2ADR

[Eugen Mihai Trifu]

OK, thank you for clarifications and suggestions, I will buy a 50W Neptune that I will surely be able to manage its power much easier.

[Scott Strehle]

A 3db or 6db attenuator would be cheaper. I don't like the idea of making power then destroying it. But it is a simple fix. Another option is to bypass an attenuator in the acom if there is one in it. I drive my (4) mrf300 amplifier directly with the hl2.
Scott
Kj7rhp

To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/0078d996-3781-4b10-b432-bdd8ab588caan%40googlegroups.com.

[Eugen Mihai Trifu]

Thanks Scot, but as I specified above I already purchased a 50 Watt HF/6m power amplifier with automatic band switching from www.60dbm.com.

In the future, I am thinking of purchasing a kit with 4xMRF300 from DX WORLD-E if I am satisfied with the way HL2 works.

73 ,

YO9VA

Eugen

[Scott Strehle]

If you do go that route with the DX world option.  My hl2 drives the dxworld board at .5 to 2 watts usually producing 700 to 900 watts peak depending on the band.  It may not be driving it to full potential but the signal is clean and there is pretty much no way to over drive the amplifier. Watch out for stray RF in the amplifier build though. I think that was the culprit for making smoke in a couple of mrf300's.
Scott

To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/5f5a32bc-bb57-4ff2-8c28-4a22fa782a78n%40googlegroups.com.

[Lukasz Olszewski]

On Fri, 28 Feb 2025, 16:34 Eugen Mihai Trifu, <eugen...@gmail.com> wrote:

Thanks Scot, but as I specified above I already purchased a 50 Watt HF/6m power amplifier with automatic band switching from www.60dbm.com.

In the future, I am thinking of purchasing a kit with 4xMRF300 from DX WORLD-E if I am satisfied with the way HL2 works.

I've purchased the mrf1k50 kit from dx world as well, but I've yet to put it together.  I'll have to do it before the summer as my current tube amp will turn my shack into a sauna if used in the peak of summer. I'm very pleased with the quality of the boards, but I'd much prefer the software was open source.

My current tube amplifier requires 200mW (that's mili watt) peak drive. So my HL2 also is way to powerfull. Additionally my amps input is not 50 ohms so different drive levels are needed for different bands (more like 75mW at 80m and closer to 200mW on 10m).

Ive built a simple voltage attenuator by essentially sending the signal to a dummy load , then there is a resistor (switchable between 200ohm up to 16k) from the hot lead to the amp input.

As long as the amplifier is voltage driven (like tubes and FETs are) you can use this extremely simple method). But if course buying an amp that matches you rig is even better :-)

73,Lukasz 

To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/5f5a32bc-bb57-4ff2-8c28-4a22fa782a78n%40googlegroups.com.

[Ron Lewkowicz]

Wondering what is the problem with I/Q scaling?

[Clifford Heath]

On 1 Mar 2025, at 6:50 am, Ron Lewkowicz <ron.le...@gmail.com> wrote:
> Wondering what is the problem with I/Q scaling?

Every 3dB scaling in IQ loses a bit of precision, which produces higher (relative) spurious content that must be filtered out.

Clifford Heath

[ron.ni...@gmail.com]

 If running an amplifier with a tap and using adaptive pre-distortion, then not scaling the IQ signal, and instead using an attenuator, would be the better practice.

The quantization noise of a 12-bit ADC is over 65 dB below a full swing signal.  So there's at least 20 dB of headroom before the difference gets close to being 43 dB or less.  Furthermore, proper noise shaping or dithered rounding of the quantization noise will spread that noise across the entire 48 kHz of Protocol 1 IQ bandwidth, dropping the spurious quantization noise down another 10 dB as measured within typical SSB bandwidths, and even farther down compared to a CW or FSK carrier.  If one is using an amplifier without adaptive pre-distortion, one is already generating more spread out skirt noise than that from the properly noise filtered quantization of a moderately scaled down IQ signal. 

For generating test signals with my HL2, I've even scaled IQ signals to down to below 1/2 LSB in amplitude, and they show up well above the quantization noise with proper noise shaping of the signal.

73, Ron, n6ywu

[Steve Haynal]

Hi Eugen,

The HL2 only offers 7.5 dB of true analog TX attenuation based on the way the AD9866 IC is used and what it offers. If you need to further reduce the 5W PA output, there are two options: add an attenuator and "burn" the extra RF, or reduce the amplitude of the signal sent to the DAC. The first option preserves the full 12-bits of DAC dynamic range. The second option throws away DAC bits so you have reduced DAC dynamic range. As you can see in this thread, there are two camps regarding the second option. Some say throwing away a few bits doesn't matter, others don't like that.

You might also be interested in disabling the 5W PA on the HL2 and instead use the second low power output which offers about 12 to 20 dBm of power. 

73,

Steve

kf7o

[Gérard F6EHJ]

I faced the same problem knowing that the amplifier used (an old FT450 PA) has a large input power variation (28dB between 160 and 10m). I solved the problem by using a 31.5dB attenuator module (HC472) driven by I2C. The desired attenuation is selected automatically with the band in use.