Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: [1] 2 Next   Go Down
  Print  
Author Topic: W7IEQ Power Meter Design & Build  (Read 13233 times)
G0MGX
Member

Posts: 39


WWW

Ignore
« on: October 25, 2012, 11:07:05 PM »

Hi Guys

First time post here, so please be gentle!

I'm building the project above from the info I found here:

http://www.arrl.org/files/file/Product%20Notes/2012%20Handbook/KAUNE.pdf

Plus some other details that have been delivered with the far circuits PCBs.

I have one question, and one problem I am looking for help with.

First the question, can anyone tell me if the LCD display listed in the parts in an HD44780 compatible device? I need to source an alternative part in the UK and I dont seem to be able to figure this out, the datasheet for the device doesnt really tell me but it does show a block diagram using a NT3881D controller - but I am not sure if this is a HD44780 compatible device or not?

Secondly, I have built my 1st pass at the directional coupler. It's a variation on the theme of the ARRL handbooks of the past "directional wattmeter". I have posted a high res picture of my coupler here:

http://www.qsl.net/g0mgx/images/DSC_0067.JPG

The problem I have is that when I put this in line with a radio/RF source, the SWR increases dramatically on the line. Now, there are quite a few differences in my build to the documented design:

  • I have used T68-2 toroides for the inductors
  • the design calls for RG8, but I have used RG-213
  • the images I see of the design have the screeen of the coax grounded at the same end as the inductors - mine is at the other - does this matter

I would be grateful for any pointer on the display or my directional coupler; there is something clearly wrong with the build, but I'm not too sure what!

I've also got a blog where the most recent images of the boards and PSU build for this project can be seen, here:

http://g0mgx.blogspot.co.uk/

Thanks!

Mark. G0MGX
Logged
KA4POL
Member

Posts: 1978




Ignore
« Reply #1 on: October 26, 2012, 12:08:18 AM »

Most likely the display is HD44780 compatible. I recommend you send a mail to Bill. His address is in the article. I had some questions which he answered right away.
From your picture the solder joint from braid to the grounding wire does not seem to be good. RG8 or 213 does not make much difference. The grounding can not be exchanged at your liking. You need to follow the given design. The use of shorter screws should also be considered  Wink You are dealing with RF here.
Logged
VK2TIL
Member

Posts: 327




Ignore
« Reply #2 on: October 26, 2012, 03:05:23 AM »

Nice work.

The single thing that you have done wrong is to use iron-powder cores; ferrite (#67 in the original design) should be used.

It shouldn't matter which end of the co-ax shield is grounded; as long as one end only is grounded it will act as a Faraday shield.  If both ends are grounded the shield will be a "shorted turn".

RG-8 can be substituted for RG-213.

Change the toroids!



Logged
G0MGX
Member

Posts: 39


WWW

Ignore
« Reply #3 on: October 26, 2012, 06:31:54 AM »

Thanks both for the response.

Here's a remake (I've found some #67 toroids so have used them - the idea for 68-2s came from the ARRL Handbook - they were recommended for directional couplers...

Anyhow, here's the remake:

http://www.qsl.net/g0mgx/images/DSC_0078.JPG

and here the other bits as of now:

http://www.qsl.net/g0mgx/images/DSC_0077.JPG


I will address the length of the screws also!

I tried emailing Bill but havent got a response, so I guess I just try a display I have here and see what happens...

Thanks again,

Mark. G0MGX
Logged
KA4POL
Member

Posts: 1978




Ignore
« Reply #4 on: October 26, 2012, 07:36:43 AM »

Sorry for being such a pain in the ... exactly, there.
The pictrues still don't show a good solder connection for the magnet wires. I usually take a knife and scrape away the protective layer before soldering.
Did the change result in a better SWR?
Logged
G0MGX
Member

Posts: 39


WWW

Ignore
« Reply #5 on: October 26, 2012, 08:58:11 AM »

Yes, the re-make shows no issues when in-line and no SWR increase; I will make super certain of the solder connection however at the same time as I remove the length from the bolts!  Grin

Did you make one of these yourself then? I just wondered if you used a alternate display component or the actual one in the parts list?

I've applied power after testing everything and I am seeing just black squares on the LCD (which I can make go away with the contrast), I'm about to program my own PIC as the one I am using right now came with the boards from far circuits... This is my fist check....

Will keep you posted.

Thanks,

Mark. G0MGX
Logged
KA4POL
Member

Posts: 1978




Ignore
« Reply #6 on: October 26, 2012, 10:36:00 AM »

I put a wattmeter together following the data sheet. It is similar to http://www.webx.dk/oz2cpu/radios/miliwatt.htm It did work fine. I sold it meanwhile. My Bird wattmeter is good enough.

On the display seeing black squares is to much voltage. As you can adjust it this should be fine. Just after you made the squares disappear the figures should become visible. Doesn't the PIC you received work?

Logged
G0MGX
Member

Posts: 39


WWW

Ignore
« Reply #7 on: October 26, 2012, 01:03:23 PM »

Well, All I'm getting is black squares or when I reduce the contrast nothing; I'm using a HD44780 device which I just tested with a Arduino board - the display is working fine just not with this project!

I've programmed my own PIC and the results are identical; I'm rapidly concluding that the software isnt compatible with my display  Embarrassed

I've looked at the assembler source, but my assembly programming skills are about nill; Have done quite a few PIC based projects in the past using Arduino and C++, but never assembler. Interestingly when I load the ASM file into MPASM from MicroChip - it wont assemble with errors associated with "status" and some other names being used in the source file. I'm guessing it was written for a differnet assembler, but I cant find any details on that either....

Not too sure what my next move will be. If I were to source the actual LCD diaplay part from the part list Mouser is the only place I can find and they have a backorder date of mid to late December!

Any bright ideas?
Logged
KA4POL
Member

Posts: 1978




Ignore
« Reply #8 on: October 26, 2012, 09:56:27 PM »

You could order this one: http://www.reichelt.de/Background-green/LCD-202A/3//index.html?ACTION=3;GROUPID=3005;ARTICLE=31658;SHOW=1;START=0;OFFSET=500;SID=326D7XBqwQASAAAD05H@g4bfbb973038482955c40bf8f7c069a1c;CCOUNTRY=447;

However, I'd check the connection from the PIC to the display. You seemingly have no data input for the display. I wouldn't suspect the program. I had some programs that didn't work but always there was some data displayed. The LCD I linked here is without backlight. The one with backlight will only be available from Dec 3  Grin
Logged
G0MGX
Member

Posts: 39


WWW

Ignore
« Reply #9 on: October 27, 2012, 04:10:16 AM »

Well here is the connection details from the board to the display:

http://www.qsl.net/g0mgx/images/DSC_0080.JPG

The display is actually a 4x20, so I should see output on lines 1 and 3; I only see black squares on lines 1 and 3 and nothing on 2 and 4. This display is auto contrast hence there is no connection to pin 3.

If I connect this display to my Arduino board and do a normal LCD initiatiise I can write characters to it easily - the Arduino routines are designed for the HD47780 compatible controllers.

I've had a rummage in the code and the LCD routines are clearly commented as being for a Hitachi module - so thats a HD47780 compatible if ever I saw one, although the initialisation routine doesnt look like anything I've seen before, it's certainly not sending the sequence of numbers I would expect to initialise the display.

There is so little to go wrong! The LCD interface to the PIC is very straightforward and the pins are connected 1-1 2-2 et cetera, I've added the sockets you will see in the photo so I can easily try different displays on this board and on my Arduino boards, but the first thing the code does is write to the display and I'm just seeing nothing!

What I am planning to do is to strip the code down to the bare bones to just display "banana" (or something) on the LCD and then take it from there; I think I still need convincing that the LCD routines are OK with my display. So far they seem not to be...

Mark. G0MGX
Logged
W8JI
Member

Posts: 9296


WWW

Ignore
« Reply #10 on: October 27, 2012, 06:23:30 AM »

A few constructive comments:

1.) If the crossing vertical boards inside are intended to be shields, they won't be. They have a grossly flawed ground connection. They will leak nearly as bad as not being there at all. To be a shield, it has to be "grounded" at all sides and edges. It looks like you only have a ground lug at the bottom on one wall. To be an effective shield it has to be grounded vertically along the sides, horizontally across the bottom, and even at the top.   

2.) Leads appear to be very long if you expect accuracy at 30 MHz or higher.

3.) If you connect this thing without the ports terminated, you are destined to have high SWR. This is because you have added impedance from the current transformers. Current transformers are only current transformers when properly designed, constructed, and properly terminated.

4.) The sample cable length appears terribly long. Remember although that looks like coax to us, it is not coax to a radio signal. It is a single conductor with a "Faraday" shield, it is NOT in coaxial mode. This means it does not have 50 ohms impedance. (If it did have normal shield function, it would not work as a sample line.)

I really think you have too much lead length everywhere, a poor box, and the wrong cores. When I do an HF directional coupler, I try to limit the current sample length to less than one inch.  At about two-three inches there will be significant standing wave, and it will also affect upper HF null and coupling attenuation to the sample detector.

I'm not sure you need an internal shield wall, but if you do, the shield wall you have now would need grounded at the top, base, and ends. They would need bonded at the crossing point. With crossed shielding, that would be at least 5 grounds and a seam bond at the center.

73 Tom
Logged
KA4POL
Member

Posts: 1978




Ignore
« Reply #11 on: October 27, 2012, 08:59:53 AM »

If you check http://www.sprut.de/electronic/lcd/index.htm#0 you can see a difference in row and column addressing of the 2x20 and the 4x20. This could well be your problem. It is a site in German language but the guy is very experienced and the diagrams are selfexplanatory.
Logged
G0MGX
Member

Posts: 39


WWW

Ignore
« Reply #12 on: October 27, 2012, 09:41:21 AM »

Well,

I'll go back and look at the Directional Coupler later..... but thanks for the feedback Tom.

Meantime here's what I have done to try and get the display to work:

I first stripped the software back to just the LCD routines and called the initialisation routine and then the routine to display a string from the main section - as kind of expected nothing - just the black squares I have grown to love!

So; no output means *something* is wrong, so I added a loop in code after the LCD calls to just toggle an output pin every second and connected an LED to it. As hoped the LED flashes when the board is powered so my software is running AND it's not getting stuck anywhere - this is good.

So, I changed the INIT routine to what I glean from the LCD datasheet to be the required initialisation sequence- still no change.

Then I increased the delays used in code between the setting of the ENA line high and low and BINGO - we have output on the display. The code as I downloaded uses a "nop" instruction to implement the delay between setting ENA High and then Low to receive data and commands in the LCD; Not sure how long a single instruction takes on this PIC (could work it out), but clearly it's enough for the display module on the parts list, but not for any one of three that I have tried here! My understanding of "nop" is a null instruction so time wise just one clock tick.

As pathetic as it looks, here is the display with some text on it:

http://www.qsl.net/g0mgx/images/DSC_0081.JPG

So, looks like I am moving in the correct direction, I have also sorted out the assembly errors in the code I downloaded (it wouldnt assemble on the MPLab I have). They were are all due to either capitalisation being incorrect in labels and symbols internally declared or declared within the include file for the processor. Additionally the decimal values being used for iteral values against a "loadw" command were not being interpreted as decimal by default and hence some were too big in value for the operand.

Oh boy, its been tricky so far but I do think I am now getting somewhere.

Mark. G0MGX
Logged
G0MGX
Member

Posts: 39


WWW

Ignore
« Reply #13 on: October 28, 2012, 12:47:38 AM »

Me again,

I thought you might like to see the update I've posted today on my blog:

http://g0mgx.blogspot.co.uk/2012/10/its-becoming-real-power-struggle.html

Regarding the directional coupler, and the comments from Tom, I'm proposing to go get a smaller diecast box today at some point, the original article suggests a box from radio shack, that I can see is 5.25x3.0x2.125" (13.3x7.6x5.4cm), so I will try and find something like that or smaller.

The software seems to be running OK now with my changes to the display routines, I'll keep you posted, but thanks very much for all the suggestions and help so far.

Mark. G0MGX
Logged
KA4POL
Member

Posts: 1978




Ignore
« Reply #14 on: October 28, 2012, 01:27:16 AM »

Tnx for the good news.

We see again, using a computer/PIC saves you half of the time you wouldn't need without a computer/PIC. May be you can eventually get in touch with W7IEQ about those issues.
Logged
Pages: [1] 2 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!