ND2E's Blog

For the past few weeks I have been assembling new household computers. For the past year I have been using an older Dell Dimension GX-Series as the main computer for radio and web development. It is an old-style Socket 478 Pentium running at 2+ GHz. HRD and background tasks often ran the machine into the 100% realm and HRD’s performance suffered. However I had not planned on upgrading machines or even replacing the machine since I could use my Dell Studio laptop for web development and for the radio.

I also do some video editing for our local congregation so that the weekly sermon is available on the church web site. In doing the weekly edits, I found that even the Studio with its T6500 2.1 GHz processor was not really as fast as I wanted. It was taking too long to create the clip I needed. So, I needed ; ) a faster machine.

I opted to build a new Socket 775 machine with a 2.9 GHz Pentium Dual-Core (E6500) processor using a 32-bit XP operating system. The machine worked very well and cut the time down to a more reasonable area but I figured I could do better if I overclocked the CPU. That worked, the time was cut down to roughly 75 minutes to load the .VODs and another 90 to write the final clip. This was more reasonable but still not fast enough to get things done on a Sunday afternoon after church. 165 minutes was still better than I had been doing with the Studio laptop but it was still too long.

After using the new computer for a couple of weeks, I decided to investigate replacing the Pentium with a Quad. What could I expect in performance gains if I moved to a Quad? And which Quad would be a better choice? I searched and researched the available 775 Quads and determined that the best would be a Q9550 but the best price I could find was more than I really wanted to pay. My second choice was the Q8400 but it was a 2.55 GHz chip – slower than the E6500 so many tasks other than video editing would suffer and run at roughly 85% of what I had grown used to.

I still don’t know if HRD is multi-processor aware but I did find that many hardware hackers had pushed the Q8400 to 3.4 Gz with little problem other than heat. If my chosen motherboard could handle overclocking the Q8400 to roughly 3 GHz without problems, I would not suffer any execution slowdowns. I tested the overclock capabilities of the MB by setting the overclocking on automatic and going for a 15% OC on the E6500. That worked so I knew that the MB could provide some OC capabilities for the new processor.

All that remained was the heat issue. The E6500, when OCed to 3.2 GHz during a video creation would hit 61 Degrees C. Not really good but under the 70 deg C limit for the processor. That was for a 65 Watt cpu, what would happen when I put a 95 Watt cpu under the same load. Would it break the 70 degree mark? Being chicken, I opted for a non-stock cooler for the quad. I chose a massive cooler that was supposed to be able to handle a 130 Watt cpu at full load. I figured that the OCed quad would be hitting about 130 to 150 watts at full load so the cooler would probably work. It does…under 90% load at 3.1 GHz, the cpu never gets over 42 degrees C. The new processor and DDR3 memory cuts the overall load and create time to roughly 1.5 hours and normal tasks work at roughly the same speed as with the E6500.

So what does this mean to the old 478 Pentium machine? It means that I now have a spare E6500 and memory. I need a MB, case, and power supply to create a fast machine for the radio. I purchased a less expensive case, bought another MB like the first one, and bought a new 500W PS. I reused a CD burner and hard drive I had from an earlier machine so I basically got by with only having to buy the case, MB, and PS to create the new machine. In both instances I used mATX MB so I didn’t have to buy expensive Radeon display boards. I can add them later if I find I need to upgrade the video.

Yesterday I finished the radio machine except for a new after-market cooler which should be in today. It works well and doesn’t slow down when other tasks are running.

It has been a long dry summer, I design and build Joomla-based web sites for several organizations. I am really quite new at web site design and operation so it should not be a shock when one of the sites was hacked. It has taken me a long time to understand how it was done and what was needed to slow down the attacks. The version of Joomla that I was using was really easy to hack, as was this blog. So to cut down on things I was forced to turn off comments and do several things that made several sites less useful. I am slowly returning things to normal after rebuilding, modifying, and adding security to the sites. The web security activities have not left me with much time to do Ham radio things.

I was able to work Field Day from home. I think I did better this year than the last two years. Hopefully I have learned a bit and become a better operator.

Just before Field Day I noticed that my hexbeam was not looking right. It was slouching off at a strange angle and the beam was not facing the same direction as the rotator. I managed to get through Field Day without everything falling down but my main antenna definitely needed a lot of work.

So this month I have made it a priority to get the radio things in order for the winter season.

I lowered the antenna and found that the fiberglass mast I was using had been eaten by the mast clamps on the rotator. I had expected a bit of wear but the clamps had removed enough material that the antenna was almost falling out of the rotator.

I modified the clamps and mast by epoxying an aluminum protective skin to the clamping surface. I also shimmed the spider to the mast and epoxied it in place.

So far the modifications seem to work. We have gone through one major wind storm and the antenna stayed facing the proper direction and there was no apparent wind forced movement of the various parts. Hopefully, that will solve the problem for a couple of years. I expect that the antenna will need to be rebuilt in a couple of years but until then I think it will survive.

There has been some recent solar activity and there is a chance that the higher bands will come back this coming year. Thinking that 6meters might be a useful band again, I built a 6meter dipole to be added to the short mast that carries my NVIS 40meter hamstick dipole.

I also want to finally get my 40ish foot vertical working in time for this winter. I had been using that aluminum mast as the 2m/70cm mast for the J-pole. I was having doubts that the vertical and the J-pole could co-exist without putting a lot of RF back on one or the other of the feed lines, so the J-pole had to come down.

The solution was to get my four-band mobile vertical out of storage and mount it on top of the short mast. I knew that the 2m/70cm sections would work as they were complete antennas but suspected that the 6 and 10meter sections would not be really effective since they were 1/4-wave sections. I was hoping that the 20ish foot mast would be enough of a ground for the 6- and 10meter sections to be somewhat useful but knowing that I had 10-meters available on the Hexbeam and a new 6-meter dipole gave be a chance to play with the verticals. I am running out of feed lines to the antenna farm so I put the 6meter and 40meter dipoles on the same feed with no attempt to phase them. So far both tune up well using the autotuner in the K3.

The aluminum vertical is now up but has no feed lines and the SGC coupler has not been mounted nor have the radials been run. If we have a couple of sunny days, I will get that done before the winter rains start.

Hopefully, all the antennas will survive the winter and I will have time to play with the radio.

The USB Serial device that has given me a problem is the IOGEAR USB to 2-Port Serial RS-232 Converter, Model GUC2322. I reinstalled it and reinstalled the driver on their web page.  Again I noted the major increase in the System Process CPU usage. I removed it and placed an old USB 1.1 Prolific adapter in and the CPU usage for the System task went to 7% on the peaks and stayed down at zero most of the time.

I will eventually contact IOGEAR and Moschip to find out why.  But for the moment the Dual Port device is not usable in my system.

Last year when I lost my big Dimension 8400 to a series of power faults, I started using my old Dell 900MHz laptop. It was horribly slow and was really limping when running the Ham Radio Deluxe V4 package. To fix the problem I bought an older 2.2 GHz Pentium 4 Dell GX270 figuring that it would be much faster than the Dell laptop. Since is was supposed to be so much faster I upgraded to the latest HRD V5 Beta package expecting blindingly fast operation. I was disappointed and started looking for the reason. I somehow got the machine into a state where it was so slow that it took 20 minutes to boot. I finally fixed that problem so I was back to the original complaint that when running HRD it would slow to a stop  for several seconds. I couldn’t run another background tasks without significantly impacting the QSO.

I have spent several hours off and on looking for the cause of the slowdown. I have finally located one of the reasons slowing my computer when using Ham Radio Deluxe.

According to the info on the HRD website, HRD is a very light load on the computer. I concur that it is a very small load but whenever I ran HRD/DM780/HRDlog my computer bogged down. I spent several hours tracking down what was causing the slowdown. Finally, I noticed that the System process was consuming valuable realtime. I searched the web for System process information and came across several articles that indicated USB-connected peripherals such as USB-Serial adapters can cause the System process to balloon.

I changed the interface point to my only serial port and the load presented by the System process dropped from 50-70% to 0-5%.  That is a really significant change that may allow the HRD package to run without hesitation or slowdown that have plagued me since I brought the Dell GX270 on line.

Hopefully this info will be of use to others.

73 de ND2E

This has been a really nice week, the sun has been shining and I haven’t been stuck inside except by choice. Monday before last we also had a bit of sun and even though it was a cold day I used a bit of it to hang the 40 meter hamstick dipole. It is sitting at 12 feet so it didn’t want to tune down to the CW/PSK end of the band. I used a bit of copper wire off the ends to move the resonance down to where my AT-897 could handle the SWR. It isn’t pretty but it is working. I have even made a couple of contacts with it.

I have also been working on getting caught up on the QSLs. I have a batch ready to go to the bureau and have mailed several US QSLs. Having lost the old database, I can not be sure of covering all of the offshore QSLs, so If you need a paper copy for DXCC use the contact form on my web site to let me know. Same for folks needing paper for WAS. I apologize for the delays but things have just been out of hand for a while here.

In the last couple of months, several of my friends have had significant computer problems and I have been called in to fix the problems. In all but one case it appears that the problem has been anti-virus malware causing the problems. Compounding the problems has been dirty heat exchangers/cpu coolers. They have to be cleaned every now and then.

I will be sending my ts-2000 in for service shortly to ensure that it is working properly before I sell it. I want a Elecraft K3 but need to sell off a bunch of equipment that I don’t use.

73 de ND2E