HL2 Weak Voice Power Output
435 views
Skip to first unread message
Ed Stroh
Oct 1, 2023, 5:50:42 PM10/1/23
to Hermes-Lite
Hi all,
I'm a new HL2 user and have only been playing with it for about a week. I realized yesterday that I'm getting low output power on voice transmissions compared to when I use the tune function or run digital modes.
For example, I can get in excess of 5W when tuning at full drive via Thetis, SDR Console, or other SDR software or by pushing a tune or any digital modulation from, say, WSJT-X over VAC. This is 100% consistent every time I test. However, when I use SSB I get at most 2W, AM gives about 1W, and FM gives about 2.5W. I don't own a keyer so can't test CW.
I tuned my audio/DSP chain according to this YouTube video and all meters are maxing at appropriate values per the poster's instructions.
I also tried plugging a mic directly into the HL2 (I have the HL2+ expansion board) instead of sending audio to/from my PC over the network but got the same results.
Is this normal or is something amiss here? Please let me know what additional info, if any, is needed to help troubleshoot. Thanks in advance!
Currently running the latest MI0BOT Thetis v2.10.0 HL2-Beta1 on Windows 11 x64.
73,
Ed KS7ROH
Ron Lewkowicz
Oct 2, 2023, 10:18:41 PM10/2/23
to Hermes-Lite
I don't use Thetis but I have a plugin host that I insert between my Mic capture device and the TX audio sinks on the SDR software. These are LV2 or VST signal chain plugins for studio use. When speaking normally and watching the power meter on my antenna tuner the level fluctuates but is generally swinging up close to 5W on SSB. If you push it so hard to have it constantly hitting 5W I expect it would sound rather distorted.
radi...@mail.com
Oct 3, 2023, 11:48:37 AM10/3/23
to Hermes-Lite
Ed, I used to feel exactly this too when I first had my HL2 but of course it's entirely normal to see low average powers with voice. If you have a PEP (hold-peaks) power meter available you will almost certainly see the peaks are 5W but the average power is only around 2W. That is to be expected with no audio processing and a "flat" mic.
Many of us were brought up with analogue SSB rigs that had ALC action and the general guide was to increase mic gain until the ALC just starts to kick on peaks. With SSB, the overall effect of this is somewhat analogous to audio compression, that is, main body of voice waveform is lifted and peaks are flattened by the ALC action reducing the PA gain on peaks.
If you use SDR Console the audio input has most excellent audio input monitors and compression tools. Open up the audio scope in the TX DSP panel and you will be able to readily see the audio peaks. Add maybe 40% compression, then adjust audio input gain until you see the audio on the scope just beginning to flatten. Actually you can allow them to flatten quite a bit with no significant audible degradation. Simon, author of SDRC, has some excellent algorithms in there. The audio input can take a considerable amount of abuse until it becomes audible in any negative way in the audio output signal from the TX DSP section.
Also on SDRC you can monitor the output of the TX DSP (you need to do this in headphones of course) so you can also add a little TX EQ to tune the audio response if you wish.
Finally there is the audio profile of the mic to be considered. Unlike analogue SSB days where comms mics were to be favoured (most of us would never have considered mic EQ in those days so the EQ profiling tended to take place in the mic audio response) IMHO you are these days much better off with a USB audio interface and a decent quality studio-style microphone (unidirectional dynamic, in the Shure SM58 style). These are both to be obtained these days at astonishing value. Both for sale on Amazon, I have the Behringer UMC202HD audio interface and the Behringer XM8500 mic. Both are amazing quality for very low price. I have 30 years in UK state broadcasting so I like to think I'm a reasonable judge of audio quality. With these sending a good base level of audio quality to the SDRC audio input, and addition of the excellent SDRC input compressor you will be able to achieve much improved voice power output but a very nice quality audio and a clean signal.
I don't know Thetis at all so I imagine it may have equivalents, but I know SparkSDR also offers audio compression for the mic input.
73
Max
Many of us were brought up with analogue SSB rigs that had ALC action and the general guide was to increase mic gain until the ALC just starts to kick on peaks. With SSB, the overall effect of this is somewhat analogous to audio compression, that is, main body of voice waveform is lifted and peaks are flattened by the ALC action reducing the PA gain on peaks.
If you use SDR Console the audio input has most excellent audio input monitors and compression tools. Open up the audio scope in the TX DSP panel and you will be able to readily see the audio peaks. Add maybe 40% compression, then adjust audio input gain until you see the audio on the scope just beginning to flatten. Actually you can allow them to flatten quite a bit with no significant audible degradation. Simon, author of SDRC, has some excellent algorithms in there. The audio input can take a considerable amount of abuse until it becomes audible in any negative way in the audio output signal from the TX DSP section.
Also on SDRC you can monitor the output of the TX DSP (you need to do this in headphones of course) so you can also add a little TX EQ to tune the audio response if you wish.
Finally there is the audio profile of the mic to be considered. Unlike analogue SSB days where comms mics were to be favoured (most of us would never have considered mic EQ in those days so the EQ profiling tended to take place in the mic audio response) IMHO you are these days much better off with a USB audio interface and a decent quality studio-style microphone (unidirectional dynamic, in the Shure SM58 style). These are both to be obtained these days at astonishing value. Both for sale on Amazon, I have the Behringer UMC202HD audio interface and the Behringer XM8500 mic. Both are amazing quality for very low price. I have 30 years in UK state broadcasting so I like to think I'm a reasonable judge of audio quality. With these sending a good base level of audio quality to the SDRC audio input, and addition of the excellent SDRC input compressor you will be able to achieve much improved voice power output but a very nice quality audio and a clean signal.
I don't know Thetis at all so I imagine it may have equivalents, but I know SparkSDR also offers audio compression for the mic input.
73
Max
Ed Stroh
Oct 3, 2023, 1:11:37 PM10/3/23
to Hermes-Lite
Hey Max,
Thank you for the reply.
I wish you were right about the PEP but as it happens I do have a wattmeter that has a PEP mode and when I activate it I see even less than the 2W I see when in avg mode. More like 1-1.5W. And that's with all the DSP processing, leveling, compression, mic gain, etc. that Thetis or SDRC have to offer.
I posted a big email chain in this thread last night that was apparently deleted by the mods for some reason but it contained a back and forth between Simon (SDRC dev) and I troubleshooting using SDRC. We looked at the spectrum readout and cranked all the proc and gain settings but still resulted in the same low output power. He suggested continuing to seek help here, particularly from Steve and Jim, as he was unable to provide any additional troubleshooting suggestions.
As to the mic, I'm already using a studio grade mic plugged into a USB audio interface (Focusrite Scarlett 2i2) and the gain knob on it is set such that the audio level isn't clipping but is using the max dynamic range.
Eager to hear if anyone else has additional input on this situation. As far as I'm concerned, the HL2 is faulty, but I'd love to be proven wrong.
73,
Ed KS7ROH
Simon Brown
Oct 3, 2023, 1:14:06 PM10/3/23
to Ed Stroh, Hermes-Lite
Ed - wasn't posted to this list.
From: herme...@googlegroups.com <herme...@googlegroups.com> on behalf of Ed Stroh <stro...@gmail.com>
Sent: 03 October 2023 18:11
To: Hermes-Lite <herme...@googlegroups.com>
Subject: Re: HL2 Weak Voice Power Output
Sent: 03 October 2023 18:11
To: Hermes-Lite <herme...@googlegroups.com>
Subject: Re: HL2 Weak Voice Power Output
--
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 on the web visit https://groups.google.com/d/msgid/hermes-lite/837c17ec-5428-4e5c-acde-b51581d89362n%40googlegroups.com.
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 on the web visit https://groups.google.com/d/msgid/hermes-lite/837c17ec-5428-4e5c-acde-b51581d89362n%40googlegroups.com.
Steve Haynal
Oct 3, 2023, 1:47:06 PM10/3/23
to Hermes-Lite
Hi Ed,
Since you report:
"For example, I can get in excess of 5W when tuning at full drive via Thetis, SDR Console, or other SDR software or by pushing a tune or any digital modulation from, say, WSJT-X over VAC. This is 100% consistent every time I test. However, when I use SSB I get at most 2W, AM gives about 1W, and FM gives about 2.5W."
There is nothing wrong with your HL2 hardware. Tune sends the same type of IQ data as SSB to to the radio and you are seeing 5W. There is nothing different the HL2 hardware does for SSB.
Please check the relevant software sections of:
73,
Steve
kf7o
Steve Haynal
Oct 3, 2023, 2:09:19 PM10/3/23
to Hermes-Lite
Also, the AK4951 companion board does not bypass software. All microphone audio is sent back to the host for processing via software and then sent back again to the HL2 as IQ data. See the protocol wiki page. The AK4951 is nothing more than an expensive ethernet audio card with low latency hardware monitoring, which most other off the shelf sound cards have too.
73,
Steve
kf7o
Steve Haynal
Oct 3, 2023, 2:16:56 PM10/3/23
to Hermes-Lite
HI Ed,
Looking at www.qrz.com, you only live ~5 miles away from me! If needed we can meet and/or drop of the HL2 for testing.
73,
Steve
kf7o
Ed Stroh
Oct 3, 2023, 2:30:32 PM10/3/23
to Hermes-Lite
Hey Steve!
Thanks for the replies. Since we're so close, is there any chance we can arrange a visit to my home QTH for some testing? I just got off the phone with Tim and he suggested trying a third PEP wattmeter if I could find one. Do you have one handy? Could you maybe bring over a known good HL2 to compare and contrast with? I'm available most days, especially after about 3:30PM. Let me know what you think and what your availability is. Feel free to reach out directly via email to coordinate.
73,
Ed KS7ROH
Steve Haynal
Oct 3, 2023, 4:01:18 PM10/3/23
to Hermes-Lite
Hi Ed,
Sure, but I am 99.9% certain that your HL2 is okay if you can send a tune or wsjt-x signal at 5W. The HL2 just receives IQ samples from software and doesn't know anything about SSB, WSJT-X, FM, etc. I may just be your microphone audio levels. It would have to be something very surprising and unexpected to be a hardware problem. (I am the engineer who designed most of the HL2.)
73,
Steve
kf7o
Ron
Oct 3, 2023, 4:23:44 PM10/3/23
to herme...@googlegroups.com
I was confused by this as well but once I finally realized the peak/avg
switch on my MFJ Tuner is wired backwards things came together quickly.
On 2023-10-03 2:01 p.m., Steve Haynal wrote:
> Hi Ed,
>
> https://github.com/softerhardware/Hermes-Lite2/wiki/No-Transmit <https://github.com/softerhardware/Hermes-Lite2/wiki/No-Transmit>
> <https://youtu.be/LYpYS9rYSm0?si=IKJT5GoowcV3kfN7> and all meters are maxing at appropriate values per the poster's instructions.
> You received this message because you are subscribed to a topic in the
> Google Groups "Hermes-Lite" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/hermes-lite/eFb1zP_yCrg/unsubscribe
> <https://groups.google.com/d/topic/hermes-lite/eFb1zP_yCrg/unsubscribe>.
> To unsubscribe from this group and all its topics, send an email to
> hermes-lite...@googlegroups.com
> <mailto:hermes-lite...@googlegroups.com>.
switch on my MFJ Tuner is wired backwards things came together quickly.
On 2023-10-03 2:01 p.m., Steve Haynal wrote:
> Hi Ed,
>
> Sure, but I am 99.9% certain that your HL2 is okay if you can send a
> tune or wsjt-x signal at 5W. The HL2 just receives IQ samples from
> software and doesn't know anything about SSB, WSJT-X, FM, etc. I may
> just be your microphone audio levels. It would have to be something very
> surprising and unexpected to be a hardware problem. (I am the engineer
> who designed most of the HL2.)
>
> 73,
>
> Steve
> kf7o
>
>
> On Tuesday, October 3, 2023 at 11:30:32 AM UTC-7 stro...@gmail.com wrote:
>
> Hey Steve!
>
> Thanks for the replies. Since we're so close, is there any chance we
> can arrange a visit to my home QTH for some testing? I just got off
> the phone with Tim and he suggested trying a third PEP wattmeter if
> I could find one. Do you have one handy? Could you maybe bring over
> a known good HL2 to compare and contrast with? I'm available most
> days, especially after about 3:30PM. Let me know what you think and
> what your availability is. Feel free to reach out directly via email
> to coordinate.
>
> 73,
> Ed KS7ROH
>
> On Tuesday, October 3, 2023 at 11:16:56 AM UTC-7
> softerh...@gmail.com wrote:
>
> HI Ed,
>
> Looking at www.qrz.com <http://www.qrz.com>, you only live ~5
> tune or wsjt-x signal at 5W. The HL2 just receives IQ samples from
> software and doesn't know anything about SSB, WSJT-X, FM, etc. I may
> just be your microphone audio levels. It would have to be something very
> surprising and unexpected to be a hardware problem. (I am the engineer
> who designed most of the HL2.)
>
> 73,
>
> Steve
> kf7o
>
>
> On Tuesday, October 3, 2023 at 11:30:32 AM UTC-7 stro...@gmail.com wrote:
>
> Hey Steve!
>
> Thanks for the replies. Since we're so close, is there any chance we
> can arrange a visit to my home QTH for some testing? I just got off
> the phone with Tim and he suggested trying a third PEP wattmeter if
> I could find one. Do you have one handy? Could you maybe bring over
> a known good HL2 to compare and contrast with? I'm available most
> days, especially after about 3:30PM. Let me know what you think and
> what your availability is. Feel free to reach out directly via email
> to coordinate.
>
> 73,
> Ed KS7ROH
>
> On Tuesday, October 3, 2023 at 11:16:56 AM UTC-7
> softerh...@gmail.com wrote:
>
> HI Ed,
>
>
> I also tried plugging a mic directly into
> the HL2 (I have the HL2+ expansion board)
> instead of sending audio to/from my PC over
> the network but got the same results.
>
> Is this normal or is something amiss here?
> Please let me know what additional info, if
> any, is needed to help troubleshoot. Thanks
> in advance!
>
> Currently running the latest MI0BOT Thetis
> v2.10.0 HL2-Beta1 on Windows 11 x64.
>
> 73,
> Ed KS7ROH
>
> --
> I also tried plugging a mic directly into
> the HL2 (I have the HL2+ expansion board)
> instead of sending audio to/from my PC over
> the network but got the same results.
>
> Is this normal or is something amiss here?
> Please let me know what additional info, if
> any, is needed to help troubleshoot. Thanks
> in advance!
>
> Currently running the latest MI0BOT Thetis
> v2.10.0 HL2-Beta1 on Windows 11 x64.
>
> 73,
> Ed KS7ROH
>
> You received this message because you are subscribed to a topic in the
> Google Groups "Hermes-Lite" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/hermes-lite/eFb1zP_yCrg/unsubscribe
> <https://groups.google.com/d/topic/hermes-lite/eFb1zP_yCrg/unsubscribe>.
> To unsubscribe from this group and all its topics, send an email to
> hermes-lite...@googlegroups.com
> <mailto:hermes-lite...@googlegroups.com>.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/hermes-lite/6ff5c38f-5985-4413-b649-2950fc8fda5cn%40googlegroups.com <https://groups.google.com/d/msgid/hermes-lite/6ff5c38f-5985-4413-b649-2950fc8fda5cn%40googlegroups.com?utm_medium=email&utm_source=footer>.
Reply all
Reply to author
Forward
0 new messages