Low Power Output

[Jim Ancona]

Hi,

I received my IO Board a couple weeks ago and have been working with it and on some other projects. I just got back to using the HL2 for transmitting and noticed that the output seemed lower than usual. Instead of my usual 5 or so watts on most bands, I'm seeing between .5 and 1.6 watts, depending on the band, into a dummy load, measured with both Quisk and a Diamond SX-200 wattmeter. SWR into the dummy load is less than 1.2:1 on all bands.

160M  0.4W

80M    0.5W

60M    0.5W

40M    0.7W

30M    1.2W

20M    1.1W

17M    0.8W

15M    0.8W

12M    1.2W

10M    1.6W

Those numbers are with the N2ADR filters engaged. The numbers are somewhat higher if I disable it.

I've had my HL2 since 2019 and I haven't had it out of the case for a couple of years, but I removed it to disconnect the Arduino I had been using to listen on the IC2 bus to do amplifier band switching and instead install the IO board. I've checked software settings and tried both SparkSDR and Quisk with similar results. I've inspected the HL2 board and I didn't see anything that looked out of whack. I've also tried removing the IO board. I reset the bias. I assume I must have messed something up while working with the board. Any suggestions on how to troubleshoot it?

Thanks,

Jim

N1ADJ

[G4ZAL]

Hi Jim,

That's annoying!

My first thought is that it sounds typical of heat shim alignment/damage - that is if you have one fitted? as you say your HL2 is 2019 and might be a bit early for shim included in the kit.

If you do use a shim, have a good eyeball of the components around the underside of the board in the vicinity of the shim.

You can search this forum (online) for 'low power shim' and will come up with several hits.

Hope it's something 'simple'

Nigel

G4ZAL

[Jim Ancona]

There was no heat shim when I bought mine and I've never added one. I have a 55mm case. I've inspected the boards and haven't noticed obvious damage.

I have a voltmeter and a VNA, but no other RF test gear. Are there voltages I should be checking?

Jim

--
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/43c3f93a-da80-47a1-8048-15f05d07a71bn%40googlegroups.com.

[Rick Langford (N8SDR)]

Have you checked the bias? 

[Jim Ancona]

Yes, I tried resetting the bias and I've tried both Quisk and SparkSDR, both with their drive settings at the maximum.

Jim

N1ADJ

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/2750b813-061e-4f9c-b96d-f3c5cd212c23n%40googlegroups.com.

[Kristofer Danner]

I had something similar to this once before and it was the nut for the case screw shorting to one of the MOSFET pads after reassembling. You might take out the nut/screw and see if any change.

KD

[Jim Ancona]

Thanks. I tried with the nut and screw removed but the output remains low. I inspected the board again, but didn't see anything damaged. I checked voltages ast suggested at https://github.com/softerhardware/Hermes-Lite2/wiki/No-Transmit. I measured the following voltages:

Vpa: 7.95

Vop: 10.43

Vbias: 10.01

Vop is a bit higher than the nominal 10.15. Is that significant?

Jim

N1ADJ

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/63f93df3-08ba-453f-a15e-ab93b13fbc18n%40googlegroups.com.

[Stano Arendarik]

Hi Jim,

the voltage levels you mention are OK. Did you checked the voltage on bias0 and bias1 points? It has to be about 3.0V. It can be measured on J31, J32.

Stano OM7ADE.

Dne pátek 3. listopadu 2023 v 15:49:39 UTC+1 uživatel Jim Ancona N1ADJ napsal:

[Jim Ancona]

Thanks for the suggestion. I measured the bias at 2.92 and 2.94 volts. (I measured at L33 and L34 because they were accessible from the top of the board.)

Jim

N1ADJ

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/b2714c68-bd5c-410a-9bb8-3d3a1dbda360n%40googlegroups.com.

[Steve Haynal]

Hi Jim,

It would be helpful to determine if your low power is from the AD9866, preamp or PA. What is the power output from RF1 when the PA is disabled? See step 5 of the no RF power output wiki page for how to measure AD9866 power output with an oscilloscope. Also, please do these experiments with the HL2 removed from the case (don't engage full power for more than 10-20 seconds to avoid heat) and always in CW mode.

https://github.com/softerhardware/Hermes-Lite2/wiki/No-Transmit

73,

Steve

kf7o

[Jim Ancona]

Thanks Steve!

No change in results with the HL2 removed from the case. I don't have an oscilloscope, only a DVM and a NanoVNA. I'll look into borrowing an oscilloscope.

Jim

N1ADJ

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/426e105e-4e20-43a2-90ef-07f06c01e36fn%40googlegroups.com.

[Steve Haynal]

Hi Jim,

You may be able to measure the power in other ways:

https://qrp-labs.com/qcx/rfpower.html

73,

Steve

kf7o

[Kurt V.]

As far as Thetis is concerned, what version are you using? 
If it is Thetis for HL2 and it is one of the latest versions, can you please go to Setup/PA Settings/PA gain. Change the values from the default to 100 on all bands? 
It seems these values now have to be at 100 as it is referring since the latest version to how much it represents in percentage. 

I just try to give it a shot with this one. 

Good luck. 

[Jim Ancona]

I attempted a rough measurement with my SWR/wattmeter. I calibrated it so .5 watt was full scale, then turned off the HL2 PA and hooked the metered dummy load to RF1. With Quisk drive set to full, the meter moved a bit. I estimate that I got somewhere in the range of 2.5 to 5 mw (4 to 7 dBm). The wiki says I should be seeing 17 dBm on RF1, so if my measurement is anywhere near correct I'm at least 10 dB below where I should be. 

Jim

N1ADJ

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/e000b970-630a-4234-855d-89d6b3c108b9n%40googlegroups.com.

[Jim Ancona]

Kurt,

I'm on Linux, so I'm not using Thetis. I've triple-checked the equivalent settings in Quisk and SparkSDR, so I don't think software configuration is the issue. Plus this all worked until I messed around with the HL2 in order to install the IO board.

Thanks for the suggestion. I need all the help I can get!

Jim

N1ADJ

--

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/59d8ae0b-9d10-4447-bc0a-034c8f890efdn%40googlegroups.com.

[Steve Haynal]

Hi Jim,

I assume the low power still exists after removing the new IO board?

Your "10dB below" observation is consistent with the power output readings you shared earlier.

Maybe some component in the preamp was disturbed when you installed the IO board. Can you try finger pressure on U9 and the surrounding components during TX to see if the power goes back to 5W? U9 is the 8-pin IC near Vop in this picture:

 https://github.com/softerhardware/Hermes-Lite2/wiki/Hardware#power-supply-shorts

Also, if you attach a high resolution picture (most phone cameras are okay) of this region, top and bottom, we can compare with working HL2s.

Beyond that, I think you will need a scope or other means to measure the power output of the AD9866 to determine if it is the preamp or AD9866.

73,

Steve

kf7o

[Jim Ancona]

Yes, removing the IO board had no effect.

Pressing on U9 didn't have any effect on the output. Below are photos of the top and bottom of that section of the board. Perhaps you can spot something.

Failing that, perhaps buying a budget oscilloscope is in my future. Any recommendations?

Thanks for your help!

Jim

N1ADJ

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/9e721570-6dd9-423a-ac06-86a341eaa529n%40googlegroups.com.

[Jim Ancona]

Hi Steve,

So I bit the bullet and bought a cheap oscilloscope (an Owon HDS242) which came today. 

Here are some measurements I made (Quisk CWL mode spot, drive 1.00, 50 ohm dummy load):

PA on, ANT output: 14.4V P-P = 5V RMS = .5 W  

This is consistent with what Quisk reports for output, so my oscilloscope is reasonably accurate and I still remember how to apply Ohm's law.

PA off, RF1 output: 4.2V P-P = 1.5V RMS = 45 mW = 16.5 dBm

That's higher than what I thought I measured with my wattmeter, but I'm inclined to trust the oscilloscope. The wiki says RF1 should output 17 dBm, so it's pretty much at spec.

PA off, DB3 pin 3 (AD9866 pin 51): 0.6V P-P = 0.21V RMS  

PA off, DB3 pin 4 (AD9866 pin 52): 0.6V P-P = 0.21V RMS  

I don't know what the expected voltage here is, but since RF1 is at the proper level, this probably is too. 

PA on, DB20 pin 1 (TXPAp): 4.24V P-P

PA on, DB23 pin 1 (TXPAn): 4.6V P-P

I checked these in case there was something going on with a relay. Is the difference in the voltages significant?

I assume these measurements point back to the PA. Are there other things I can check there?

Jim

N1ADJ

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/9e721570-6dd9-423a-ac06-86a341eaa529n%40googlegroups.com.

[Steve Haynal]

Hi Jim,

I need to measure values on a good board to provide you with correct information. Hopefully I can do that this weekend or before. 

If we don't find anything obvious with your board, you can send it to me for repair. Even though it is out of warranty, I am happy to try and repair it given all your great contributions to OpenWebRX.

73,

Steve

kf7o

[Jim Ancona]

Thanks for the generous offer, Steve. I hope I don't need to take you up on it. 

Let me know if there is other troubleshooting I can do.

Jim

N1ADJ

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/bd83cad5-f760-4145-a473-d149982ec34cn%40googlegroups.com.

[Steve Haynal]

Hi Jim,

For DB3 pins 3 and 4, I measure 580mV peak to peak for both pins referenced to ground.

With the PA on, I measure 3.92 V and 3.76 V peak to peak on DB20-1 and DB23-1. I am not too concerned with the differences in voltage here.

It does look like something is wrong with final PA. With PA on, please measure peak to peak voltages referenced to ground for all 3 pins of Q4 and Q3. The pin nearest the silkscreen "Q4" or "Q3" is the drain, the middle pin is the source, and the third pin is the gate. I measure about:

13.7 Vpp on the drain

0 Vpp on the source

3.12 Vpp on the gate

for both. 

Also, please share a high resolution photo of the top and bottom quarter of the HL2 pcb which contains the PA. 

73,

Steve

kf7o

[Jim Ancona]

Hi Steve,

Here's what I measured. Some of the waveforms looked messed up so I included pictures of the oscilloscope trace. This is using Quisk the Spot function, drive at 1.0. Frequency 4.000 MHz. All voltages are P-P relative to ground.

Q3 Gate, 3.8V:

Q3 Drain, 6.24V:

Q4 Gate, 3.32V:

Q4 Drain 4.48V:

I've attached photos of both sides of the PA section of the board.

Thanks again for your help!

Jim

N1ADJ

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/3c6a28ae-a81d-42b6-96d4-923f394373a5n%40googlegroups.com.

[G4ZAL]

Jim,

Looking at the topside of the board, just above the fixing hole between both PA, capacitor B106 looks skewed and left end of the cap missing?

Nigel

G4ZAL

[Rick Langford (N8SDR)]

The solder joint just above C82 also looks cold /bad

Get Outlook for Android

---

From: herme...@googlegroups.com <herme...@googlegroups.com> on behalf of G4ZAL <devon...@gmail.com>
Sent: Monday, November 13, 2023 9:34:08 AM
To: Hermes-Lite <herme...@googlegroups.com>
Subject: Re: Low Power Output

 

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/gIJfBjQAW4Q/unsubscribe.
To unsubscribe from this group and all its topics, send an email to hermes-lite...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/dd47d2c5-13a3-4e6d-bee2-c3e61780d0bcn%40googlegroups.com.

[Jim Ancona]

Nigel,

I agree that B106 looks like it could be the culprit. It's also in the right place for me to have damaged it while removing or replacing the nut. It's also in the right place to explain the input waveform being messed up at the gate of Q4, but not at Q3.

I've never done any surface mount soldering. I have an old Weller WTCPN iron and I see they have some 1/32" tips available. Do I have any chance of doing this successfully? If so, what temperature tip should I go for, assuming I have a choice? Normally they have 600, 700 or 800 degrees.

Is the part I need: https://www.ebay.com/itm/385042605406

Thanks again for your help!

Jim

N1ADJ

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/dd47d2c5-13a3-4e6d-bee2-c3e61780d0bcn%40googlegroups.com.

[Jim Ancona]

Hi Rick,

The one on U19 or on C88? In either case, I think it may just be the photo. They look okay to me in person.

Thanks for looking!

Jim

N1ADJ

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/CH3PR17MB6316E99FF2D2C8E47D7B7E71F6B3A%40CH3PR17MB6316.namprd17.prod.outlook.com.

[G4ZAL]

Hi Jim,

To remove such a component I simply use 2 soldering irons and use them like 'tweezers' and heat both ends and at the same time press toward the middle of the component and lift both irons up with the component.

According to the HL2 BOM, B106 is a 0.1uF 0805 >=6v3 X5R/X7R cap - https://github.com/softerhardware/Hermes-Lite/tree/v2.0

Clean the pads with solder wick and tin one end and slide the new cap into the molten solder, if it looks good solder the other end - simples!

If you were in the UK I'd do it for you.

Anyone in the US able to help Jim if he gets stuck?

Hopefully replacing it will resolve your problem...

Nigel

G4ZAL

[Jim Ancona]

Thanks, Nigel!

I was using this BOM: https://github.com/softerhardware/Hermes-Lite2/blob/master/hardware/hl/releases/hl2p0build8/bom.standard.pdf

For B106, it says: "0.1uF 0603≥X5R≥25V≤20%". That's how I ended up on the EBay part I linked. The specs for that say:

- Capacitance: 0.1uF
- Package Size: 0603
- Tolerance: ±10%
- Maximum Operating Voltage: 50V
- Operating Temperature: -55°C ~ 125°C
- Temperature Coefficient: X7R

Does that look okay?

Thanks for the soldering/desoldering tips. I'll watch some Youtube videos as well.

Jim

N1ADJ

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/53a99064-3719-46bc-b899-18306be4974fn%40googlegroups.com.

[G4ZAL]

Looks good spec to me Jim.

Best of luck

Nigel

G4ZAL

[Steve Haynal]

Hi Jim,

A damaged B106 does look like a good candidate for your problem. I typically use hot air to remove the component. After that I clean the pads with copper desoldering braid. Finally, I solder the component with a regular soldering iron. You should be able to do a functional if not pretty job with a standard soldering iron. I'd go for the middle of the road temperature, but don't think it will matter much.

I can drop several replacement B106 capacitors in the mail if you like.

73,

Steve

kf7o

[Jim Ancona]

Thanks Steve!

I'll give the replacement a try and let you know how it goes. I ordered 100 capacitors from the EBay seller I linked, so I should be all set.

Jim

N1ADJ

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/39121a37-a842-4635-925d-00025adca528n%40googlegroups.com.

[Jim Ancona]

My HL2 is back up and transmitting normally! I installed the replacement B106 without much trouble, except that I clumsily melted the corner of the T/R relay. So when I fired it up I got no output. I tacked a "dummy load" resistor ahead of the relay and everything looked okay, so I ordered a replacement from Mouser. Once I installed it yesterday, everything was back to normal. I still have the Build 8 kit to increase output by rewinding T3 and changing a couple other components. I considered doing that and then caught myself and decided to leave well-enough alone.

Thanks to Steve and everyone else who helped me track this down! The HL2 is a great radio with a great community!

Jim

N1ADJ

[Steve Haynal]

Hi Jim,

I am happy to hear you repaired your HL2. The T/R relay can be painful to remove as there are so many pins and so much PCB ground copper around that area. I've always used hot air and have sometimes had to physically break the relay so I can remove it one pin at a time. 

At one point we had a group purchase of relays to replace the T/R relay as it can go bad. You can check if any are still available:

https://groups.google.com/g/hermes-lite/c/sksq5-31d30/m/BfNL_BbQBQAJ

73,

Steve

kf7o

[J P Watters]

Jim,

I will send you relays as Steve mentioned. Let me know how many you need.

With that Information I would post them to your QRZ address. 

Otherwise let me know the corrected address.

..jpw J P Watters

KC9KKO

Morris, IL USA

--
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/11e788cf-0c79-4fda-823b-4cf495a711a3n%40googlegroups.com.

[Jim Ancona]

Thanks for the offer, but I already bought my relay from Mouser and installed it. I guess I should have checked the group before doing that.

Jim

N1ADJ

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

Hi Jim,

Which part number for the replacement relay from Mouser did you order?

What technique did you use to remove to previous T/R relay?
Thanks & 73, Ron, n6ywu

[Jim Ancona]

Hi Ron,

This is the relay I ordered: https://www.mouser.com/ProductDetail/80-EC2-3NU 

Digikey had them as well. The minimum shipping was a killer, but I couldn't find anything that looked good on Ebay or Amazon unless I wanted to wait for it to ship from China.

I removed the old one without too much trouble with my soldering iron (an ancient Weller WCTPN) and Soldapult solder sucker (another old tool that's paid for itself many times over). I did break a couple pins on the old one, but I wasn't trying to save it for anything. The hardest thing was clearing the holes in order to get the new one in. Neither the Soldapult nor solder wick worked. Perhaps my iron wasn't hot enough. I finally managed to get a small pointed tool through the holes and cleared them with that. After that, soldering the new one in went smoothly.

Jim

N1ADJ

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/92b2f718-3936-426f-9508-649caf822c6bn%40googlegroups.com.

[Ron Lewkowicz]

Steve put a link to some relays on Aliexpress so I assume the quality on those is acceptable.  I find shipping from China is not much different than shipping from USA to Canada and stuff from China comes right into my mailbox.  USA mostly comes via something other than the post office and is a pain to collect.

I've changed similar relays on audio amplifiers.  I think a vacuum de-soldering tool would likely make the task much easier. I don't have one so I just carefully use the iron and a solder sucker.  It takes some patience.  Some have mentioned breaking the old relay apart but I fear cracking the PCB.

[G4ZAL]

I use one of these type of 'cheap' desoldering tools...

https://www.amazon.co.uk/gp/product/B09P4YTYN8

Invaluable on through hole stuff.

I usually first apply some extra decent quality leaded solder to each component leg to be removed before using the pump tool as it works better with a decent amount of solder to 'suck' on rather than just a little.

Most times it clears all the solder in one shot, if not, apply more solder and have another shot.

I have a hot air station too, but generally prefer this tool on relays/through hole components.

Here are some other ideas for desoldering...

https://www.youtube.com/watch?app=desktop&v=Vou2xlJkuoU

Nigel

G4ZAL

[dh1tt]

Desoldering is true art and real craftsmanship.  Anyone who thinks that with a desoldering station all the parts just fall out will be surprised.  Industrial circuit boards in particular are fighting hard to release the components.  This is probably thanks to the lead-free solder with a very high melting point.  I bent various loops out of rigid copper wire, which then absorb the right speed and temperature in a 200(!)W Monster iron.  Especially for relays, with pins that are far apart and in 2 rows.  It works... but also fiddly.  After dismantling a 6-band LP board, you're completely fed up.  My solution for desoldering is bismuth.  As a pearl/waste product of the industry you can get it on eBay for little money.  A 100g bag is enough for years.  Not the expensive stuff from Quickchip!  ..... ChipQuick.... Maybe not for everyone but definitely worth a try.

https://www.ebay.de/itm/122789157335?hash=item1c96cdd9d7:g:6jgAAOSwq5hehese&amdata=enc%3AAQAIAAAA4C%2BUrDupXKU2C415mtNpqfWmrV%2B9NUBMoHiS3u%2Bl1YI%2FDX1udoSYlml9kxx38W1AZWBcT7uv%2FDouk0j11QJAPnC4ldEig3PA1%2FYX2S6sHm43VaT8vBMXwl%2BJhqIkTbBir5drUAt%2FLgzzwsHfhNvDd%2B%2B%2FGVHQopzMfUVp%2ByJpFtftrYPc4hua1htCrimiiM%2Ban3CamTEAdMVV6ZYGR2%2FQ%2B%2Bh%2FOe1UUFPQlJXIFIqaegCqoGTOUrFnpQXes0EPdSa0sCzhBl%2F4v5UnClE0T4oQgGr%2BqYczysJaKzwBBsBKwEn3%7Ctkp%3ABk9SR768nPeDYw

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/da96a81e-f868-4b89-ac8d-9c95f0715c7dn%40googlegroups.com.

[Clifford Heath]

On Thu, 30 Nov 2023, 19:59 dh1tt, <oliver...@gmail.com> wrote:

My solution for desoldering is bismuth.  As a pearl/waste product of the industry you can get it on eBay for little money.

I believe that bismuth is also used for reloading shotgun cartridges by folk green enough to want to reduce lead pollution, but not green enough to stop shooting defenceless creatures.

In any case, I wanted to buy some, because, alloyed with the right amount of 60/40 solder, it forms a eutectic alloy that melts (just) in boiling water. Low melting point metals are very interesting for various fabrication tasks, and this is one of the cheapest. The best are indium alloys, which are hugely expensive.

Clifford Heath