Hermes-Lite2 beta3 boards have arrived!

[Steve Haynal]

Hi Group,

The Hermes-Lite2 beta3 boards have arrived! They look pretty good. The swapped resistors that Taka caught appear to be fixed. Attached are photos. There is a serial number 1-5 written in pencil near the copyright on each side to distinguish the boards. I have to get back to work but will start testing one tonight. Pictures of all boards exceeded the file size limit for Google groups so I just attached 2. I will send all pictures to the 5 beta testers.

So far I've been very impressed with www.elecrow.com. I submitted all the files for their bid on July 5. They responded within 2 days. I placed the order for 5 boards on July 10. I have them in hand on July 31, three weeks after placing the order. This was entirely turnkey. They fabricated the boards, sourced all the parts and assembled the units within 3 weeks. I am still waiting for a quote from the local assembly house down the street... Although I plan to eventually sell some assembled Hermes-Lite2 boards, it is also possible for others to organize small group buys from Elecrow. You just need to provide the appropriate design files.

73,

Steve

KF7O

 

[Steve Haynal]

Hi Group,

Good news! My Hermes-Lite2 beta3 board is up and running with receive. I will try and test TX tomorrow. This is a relief that there do not appear to be any major problems (counterfeit FPGA, etc.) with this build. The radio is currently skimming WSPR and JT with Spark SDR on 40,30 and 20M. There doesn't appear to be any obvious noise issues. In fact, it feels a little less noisey than my beta2 build.

73,

Steve

KF7O

[Alan Hopper]

Steve,

that's great news, I can't wait to get mine.

73 Alan M0NNB

[in3otd]

Thanks for the great news!
I'll prepare the bench to do some measurements as soon as the board arrives then, hi.

73 de Claudio, IN3OTD / DK1CG

[Sid Boyce]

Thanks Steve,
Great news for the tremendous amount of work you have done.
Many of us are just waiting in the wings.
73 ... Sid.

On 01/08/17 06:03, Steve Haynal wrote:
> Hi Group,
>

> Good news! My Hermes-Lite2 beta3 board is up and running with receive.
> I will try and test TX tomorrow. This is a relief that there do not
> appear to be any major problems (counterfeit FPGA, etc.) with this
> build. The radio is currently skimming WSPR and JT with Spark SDR on
> 40,30 and 20M. There doesn't appear to be any obvious noise issues. In
> fact, it feels a little less noisey than my beta2 build.
>
> 73,
>
> Steve
> KF7O
>
>

> Â

>
> On Monday, July 31, 2017 at 12:01:40 PM UTC-7, Steve Haynal wrote:
>
> Hi Group,
>
> The Hermes-Lite2 beta3 boards have arrived! They look pretty good.
> The swapped resistors that Taka caught appear to be fixed.
> Attached are photos. There is a serial number 1-5 written in
> pencil near the copyright on each side to distinguish the boards.
> I have to get back to work but will start testing one tonight.
> Pictures of all boards exceeded the file size limit for Google
> groups so I just attached 2. I will send all pictures to the 5
> beta testers.
>
> So far I've been very impressed with www.elecrow.com

> <http://www.elecrow.com>. I submitted all the files for their bid

> on July 5. They responded within 2 days. I placed the order for 5
> boards on July 10. I have them in hand on July 31, three weeks
> after placing the order. This was entirely turnkey. They
> fabricated the boards, sourced all the parts and assembled the
> units within 3 weeks. I am still waiting for a quote from the
> local assembly house down the street... Although I plan to
> eventually sell some assembled Hermes-Lite2 boards, it is also
> possible for others to organize small group buys from Elecrow. You
> just need to provide the appropriate design files.
>
>
>
>
> 73,
>
> Steve
> KF7O
>
>
>
>
>

> Â
>
> --
> 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
> <mailto:hermes-lite...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.


--
Sid Boyce ... Hamradio License G3VBV, Licensed Private Pilot
Emeritus IBM/Amdahl Mainframes and Sun/Fujitsu Servers Tech Support
Senior Staff Specialist, Cricket Coach
Microsoft Windows Free Zone - Linux used for all Computing Tasks

[Glenn P]

From the pictures, it appears some of the ic' s are displaced to the right.

Obviously working but something to keep an eye out for in new batches.

Glenn
Vk3pe.

[Takashi K]

Hi Steve,

I have a question.

After we beta 3 testers receive the board,
Do we need to check from smoke test ? or you have already done ?

Also Do you have any recommended test procedure (item) ?

73, Taka  ji1udd

[Steve Haynal]

Hi Taka,

I only did a quick visual inspection and took a few pictures of the other 4 boards. You will be the first to apply any power...

We should start a more detailed beta3 build and test wiki page in the next few days, but I did the following:

1. Closer visual inspection with magnification (10-15 minutes)

2. With an Ohm meter, verified that no power planes were shorted to ground or each other. There are six different power sources.

3. With an Ohm meter, measured the precision resistors to set switching and other regulator voltages. Since the resistors are in circuit, you do not measure the actual value. I compared the measured resistance with what I see on my beta2 board. I will add a table to the wiki of what I measure so others can know if they are at least in the correct range.

4. Add the power connector and connect to a power supply. Turn on briefly and check for heat or smoke. It is better to use a current limited power supply if you have one.

5. Connect the programmer and program the full beta3 firmware. I still need to post that. You don't want the unit on for long periods (>5 minutes) with out firmware programmed as a fast 100 MHz clock may be supplied to the AD9866 by default. I haven't had problems but just want to avoid any potential problems with overheating the AD9866 with a fast clock.

6. Add TR relay and main RF connector.

7. Receive!

8. Add low power RF connector.

9. Test transmit and Vop amp at low power.

This is as far as I've gotten. Low power TX is working. I haven't run into any problems. My progress is a bit slow due to lack of time lately. I will start the wiki and post a few TX spectrum plots, hopefully before the weekend.

73,

Steve

KF7O

[Steve Haynal]

I forgot step 4.5, check with DVM all voltages are correct. If any of these steps fail, we will have to add more procedures!

73,

Steve

KF7O

[Takashi K]

Hi Steve,

Thank you for your explaining kindly.

I understood the board situation and the test procedure roughly.

> I will add a table to the wiki of what I measure so
> others can know if they are at least in the correct range.

That's good idea. very helpful.

> We should start a more detailed beta3 build and test wiki page in the next
> few days

Thanks.

I hope that you describe about "TX inhibit", "rffe_rfsw_sel" and "PA bias" settings.

73, Taka  ji1udd

[Steve Haynal]

Hi All,

I've started notes for hardware bring-up and final assembly of the HL2 beta3 at https://github.com/softerhardware/Hermes-Lite2/wiki/Build-Notes-2.0beta3

Please feel free to edit and add to this wiki page. I will add more details over the weekend as I finish my bring-up.

I have started releasing both beta2 and beta3 versions of the firmware: https://github.com/softerhardware/Hermes-Lite2/tree/master/firmware/bitfiles/20170729

73,

Steve

KF7O

[Steve Haynal]

Hi All,

I worked a little this weekend on bring up of the PA, but not much as I have family in town, yet my wife is out of town, plus my day job is demanding lots of time.

Low power output from the preamp appears to be working okay. I was seeing 20 dbm instead of 17 dbm. The beta3 schematic and BOM specify 75 Ohm for R55, but that was supposed to be changed to 120 Ohms as it is in my beta2 build. My mistake. Testers should change R55 to 120 Ohm to not over drive the PA, or else be careful to run the TX DAC output at a lower level.

I noticed that the preamp takes about 1 minute to reach full output when starting with a cold board. I don't know what is required to heat up. I went back and saw the same behavior with my beta2 board.

I added T3 but on power up the LDMOS devices quickly became very hot. I think I removed power soon enough to not fry them. The current firmware has VPA always on so the default bias on the LDMOS devices must not be low enough. When it is not so late at night that I am prone to make mistakes, I will turn VPA off and check the bias carefully and try to figure out what is going on. Any beta3 testers should beware.

73,

Steve

KF7O

[Steve Haynal]

Hi Group,

This evening I discovered that J32 is not stuffed on my board. This probably explains why Q4 was quickly heating up for me. From the pictures, J32 is stuffed on all the other 4 boards.

The I2C address is different for U15, and the range values are reversed from the beta2. I did test and see bias voltages ranging from 2.5 to 3.5 V as designed. Software will have to be updated, and I'd like some mechanism to distinguish between beta2 or beta3 so softwae can support both. Perhaps we can increment the revision number to 50 and higher for beta3.

The AD9866, FPGA and ethernet IC on my beta3 feel like they are running much hotter. Perhaps this is because the weather is very hot here, or the thermal pads are not soldered as well, or the board is not heat sinked in an enclosure currently. I still need to measure and compare current consumption. This is something for testers to keep an eye on.

I haven't powered up the PA again as I am still carefully comparing measurements against the beta2. Maybe tomorrow night.

I received the 4 enclosures from China. They are 150mm in length not the 100mm ordered and have a coating applied on the inside. They ignored my instructions and just sent me a note saying customization requires a minimum order of 20 units. I should never have paid for the order without confirmation that they would customize. The enclosures are still usable but may require a little sanding to remove the coating.

73,

Steve

KF7O

 

 

On Sunday, August 6, 2017 at 10:59:20 PM UTC-7, Steve Haynal wrote:

Hi All,

[Steve Haynal]

Hi Group,

This evening I updated my local Quisk to work with both beta2 and beta3 board bias adjustment. I confirmed that bias adjustments can be saved on to the digital pot. I soldered on J32 and connected up Vpa again. Q4 is not heating up quickly. I'm almost ready to adjust the bias but need to make sure the current calculations in the software are correct. I also updated the firmware with a few minor changes and started using firmware code revisions >=60 for beta3. <40 is Hermes-Lite 1.x, >=40 and <60 is beta2. I will release firmware and changes to Quisk 4.1.9 to testers in the next few days.

Jason has received his board and power supply tests are passing. The other 3 boards, which are international, have all reached their destination country but are in customs.

73,

Steve

KF7O

[Alan Hopper]

Steve,

I received my board today, followed the steps up to receive and it is happily reporting wspr as I type with no fixes needed and nothing getting too hot.  It all looks very professional, great job. 

I did have to power cycle it a few times to get receive data even though it was discovered every time, I'll try and investigate this.

I'm off sailing for the next couple of weeks so don't know how much further I'll get, if the weather is bad I'll add the tweaks to Spark once I see what you do to Quisk.

73 Alan M0NNB

On Wednesday, August 9, 2017 at 7:03:05 AM UTC+1, Steve Haynal wrote:

Hi Group,

[Takashi K]

Hi Steve and all,

My HLv2 beta3 board has arrived today. Now I'm listening 40M radio stations using PowerSDR.

No trouble by now. Great!

I was suprised that the programming time is very short (using altera USB-Blaster). I felt write error at first, but it was OK.

The operating current is around 190mA at 13.8V.

Next I will check Low Power output.

73, Taka  ji1udd

[Steve Haynal]

Hi Alan and other HL2 beta3 testers,

FB on RX working. Please see the other thread for details on a TX issue.

Since others may be testing this weekend, I wanted to share the latest software I am using for testing. I am using Quisk 4.1.9 but have made modifications to the two attached files. If you are running on Windows with Quisk installed from the .msi installer, you can copy these files into C/Python27/Lib/site-packages/quisk/hermes to update your 4.1.9 install. In Linux, I always build and run from the distribution directory (no install to /usr) and then you can replace the files in the hermes subirectory of your distribution directory. These software changes enable setting and storing the PA bias as well as accurate bias current measurements for the PA. I am using firmware 20170809 which is on github and reflects all the changes to the RTL.

Regarding the PA bias current, the software reports ~20mA for no bias. This is because there is always some voltage offset at low levels for the sense amp. I have not investigated yet if this is an offset that should be used for correction across the entire range, or just some garbage that should be ignored and we assume we can't measure currents <50 mA accurately. I set my PA bias for 110mA on each transistor and the PA appears to work when it sees RF. Since both transistors are measured with the same probe, we can be certain that they are balanced even if the current measurements are off by 5 mA. Current measurements are accurate when transmitting at full power, and it is neat to be able to compute the efficiency of the PA.

73,

Steve

KF7O

[Alan Hopper]

Steve,

I've made the two mods on the wiki.The low power tx output is fine. I've put your Quisk tweaks into my local copy of Spark and the current and temp outputs look believable. A very quick and tentative play with the pa looks ok. Hopefully tomorrow I'll get a chance to add better heatsinking and test properly.

73 Alan M0NNB

[Steve Haynal]

Hi Alan,

Great to hear that your HL2beta3 is working fine so far. There is more bare area under the LDMOS devices for heat sinking and I'm curious if/how others might use this.

73,

Steve

KF7O

[Alan Hopper]

Steve,

as I don't have a case I put a 30x12x6mm ali block under that area and screwed it into a very oversized heatsink I had.  I set the current using the sensors and the supply current looked about right. The pa seems to work and a hurried and rough measure of power on 20m was a bit over 5w. Great job Steve and the board house.  The current measure seems rather noisy, I'll add some smoothing in software but will wait till I've done the op amp mods.  It all stayed cool in a short test.

I found a bug in the current versions of spark that causes ptt to be momentarily  set to receive when changing the bias, this makes a rather frightening noise from the relay so I don't recommend using the released versions of spark to set bias, it only works for Beta 2 anyway. I'll release a new version next week.

73 Alan M0NNB