Taming the Paper Tiger - Tackling the QSL Backlog at KC1XX
"No job's finished until the paperwork is done." No, I'm not referring to sending in the contest log and summary sheet, but rather sending QSLs to the stations who you worked. After we've simulated terrain and stacking heights, assembled the antennas and raised the towers, set up the radios and amplifiers, configured the network, then sat down for 48 hours straight trying to work the world, who wants to think about answering QSLs? Especially with the nice summertime weather during the lull between the CQ WPX contests in the spring and the CQ WW contests in the fall, wouldn't you rather be outside, dreaming about next year's stack?
Judging by the number of cards KC1XX receives through the bureau (see Figure 1), the art of QSLing is very much alive. This year, we resolved to address the backlog of cards that have been building since we first operated the 1995 CQWW SSB M/S from Matt's new house in Mason, NH. Since then, we have operated every CQWW (both CW and SSB), starting out Multi-single but running Multi-multi the last few years, most ARRL DX (both CW and SSB), WAE and the occasional WPX contest. This is not to mention the times Matt sits down before going to work in the morning, or after dinner in the evening, running "the boys" on whatever band happens to be open.
Table 1 shows the KC1XX QSO statistics by band and mode, only for the contests for which we could still find electronic log files (we used CT Version 9 to log all our QSOs):
Table 1 -- QSO totals for KC1XX - 1995 - 2000
As you can see, the total is just about 100K QSOs in just 5 contest seasons!
We had several problems to solve:
At various times in the past, one of the team members took a stab at running the current CT log through QSL9 (the companion QSLing program to CT) and generating labels. However, after labels had been applied to cards for a while, the job got boring, or we ran out of cards, and the project was pushed aside, never to be returned to again.
This year, after much "motivation" from our friends overseas, we were going to really do it!
We decided to just send everyone a confirmation for each time they worked us on a new band/mode combination. Thus, the maximum number of QSOs confirmed per station would be 12 (6 bands 160-10, CW and SSB). This way we would never have to open the cards from the QSL bureau. The only problem with this (besides sending cards to people who may not even want them) is that if we logged a callsign incorrectly, the busted call will get the card, not the correct call. We hope eventually that people will rework us if they still need the card but are not in the log.
We started by importing all our logs into QSL9 to build a QSO database. QSL9 is a CT accessory program that keeps track of which bands and modes have been confirmed for each callsign. Using the program over and over again, we would only send cards when we worked a station on a new band or mode. Ken Wolff, K1EA, the author of CT, tried to format the output in such a way that it could be printed onto standard Avery-5160 labels (10 rows, 3 columns to a sheet) using a laser printer. However, we could never get the format quite right, and each label would only be able to hold a few QSOs at most.
We also began to get quotes from some of the QSL card printing services. First, we thought 20K cards would be enough, then 30K, but as we started adding more and more logs to the QSL9 database, it looked like we'd be sending out close to 40K cards.
Not only would we have to pay for cards but we would have to pay for labels to put on each card. And there was the labor of affixing the labels to the cards. If we couldn't stay interested enough to put labels on a couple thousand cards, how were we going to do 10 or 20 times that?
Dave Pascoe, KM3T had heard about a program called BV that Bernd, DF3CB wrote. This was a windows-based program, had some of the QSL9-like features (one QSO confirmed for each band/mode), and printed output was formatted for any printer you could attach to Windows, most importantly a laser printer. But the best part was that we could design a QSL card format and print the QSOs directly to the card -- 4 cards per page, with up to 6 confirmed QSOs per card. There would be no additional expense for the stickers, but best of all, no additional labor! Also note that with up to 6 QSOs confirmed per card, we would never need to print more than two QSL cards per station. Figure 2 shows the design of the plain-paper QSL.
Now, we just needed a way to print them!
Steve Gilbert, K1SG, is the founder and president of the Newton Copy Shop. We approached him about printing some color cards for the direct QSL backlog. Matt KC1XX took some digital photographs of his towers, which Bill Myers, K1GQ formatted on his Macintosh computer into a QSL card design (see Figure 3). This design was reviewed by the team members via E-mail. Bill then gave the final design to Steve, who ran off 500 cards. It did not take long to use most of them up!
In addition to the expense of printing the cards, we would also have to pay postage for the cards to be delivered, either via the ARRL bureau, or directly to each country's incoming QSL bureau. Matt and some of the boys would be doing a presentation at the Friedrichscafen Hamfest in Germany at the end of June. This was a perfect opportunity to deliver all the German QSLs to the DARC bureau with no charge for postage!
The clock was ticking! About two weeks before the hamfest, the batch of cards for Germany (3923 cards) were printed to floppy disk and taken to Steve at Newton Copy. There, the file was read in, printed to plain paper, then copied onto card stock. A week later, Charlie, N1RR had picked up the cards and delivered them to Matt -- all 16 pounds (7.3 Kg) of them -- to take with him to Germany. Figure 4 shows Matt dropping some cards off at the bureau.
It would have been cheaper to print the cards directly to card stock ourselves, and let the copy shop cut them into QSL-sized cards (3.5" by 5"). However, the card stock would not work reliably through our printer, and the printer could not hold very many sheets of card stock at one time. Therefore, we decided to print to plain paper, and have the copy shop copy onto card stock and cut the cards down to size. This was a little more expensive than doing all the printing ourselves, but much less expensive then the shop doing all the printing for us from disk.
To print 40K cards with 4 cards per sheet, at 8 pages per minute, it would take almost 21 hours of non-stop printing, not to mention the down time while the printer was waiting for more paper. You can see why it is important to be able to put as many blank sheets into the printer as possible at one time!
Jim went out and bought 2 cartons of paper (10,000 sheets total, enough for 40K QSOs) and began to print some other countries. Italy was first, followed by the United Kingdom, then Japan, Belgium, the Czech Republic and Slovakia. All the printed pages were dropped off at Newton Copy for another run - this time about 10K cards were printed. The cards were kept separated by country, so they could be boxed up and mailed to each country's incoming QSL bureau when they were done.
While printing cards for Finland, the printer ran out of toner! Another expense! Operations stopped for a few days while new toner was ordered on-line.
Also while printing the cards for Japan, Jim encountered a filtering bug in the BV program, which DF3CB promptly fixed. This made it much easier to print cards for countries that had multiple, consecutive prefixes (i.e. JA-JS for Japan, or PP-PY for Brazil).
After having printed about 3/4 of the cards, Steve's shop installed a new high-speed, large-capacity printer. So we paid a few extra cents per card to have the last 5000 cards printed from floppy disk directly onto card stock using the new printer. The output quality was excellent! The last batch of cards was picked up on July 17. Figure 5 is a picture of all the printed cards (except Germany) on Jim's dining room table.
Now for the final step - mailing the cards. Jim solicited advice from Martin Cook, N1FOC at the ARRL outgoing QSL bureau. For many of the higher-volume countries (a pound or more of cards), it was going to be cheaper to mail the cards ourselves directly to their incoming QSL bureaus than to mail them via the ARRL outgoing bureau. Martin also gave us special shipping instructions for certain countries (like Yugoslavia). Jim downloaded the guide to International Postage Rates and Fees (Publication 51) from the United States Postal Service web site and studied it to find the cheapest rate. Surface rate printed matter (non books) was slow, 4-6 weeks delivery, but was the cheapest way (about $2.50 per pound). Besides, if people have been waiting since 1995 for cards, what difference is another 4-6 weeks going to make? On the weekend of August 11-12, 23,815 cards were boxed up and mailed at the local post office for a total of $354. To send these same cards via the ARRL Outgoing QSL Bureau, assuming 150 cards per pound, and $6.00 per pound (members fee), this same shipment would have cost $953, or more than twice as much!
After we had printed cards for the high-volume countries (DL, I, JA etc.) there were many countries left which would only get a few cards each. These were to be printed and sent to the ARRL outgoing QSL bureau. However, among these cards were cards for stateside stations - about 2500 cards! These were mostly from several WPX contests which Matt operated. We have NO WAY to distribute these cards cheaply - there are no stateside-stateside QSL bureaus, and the W0-W9 call area incoming QSL bureaus won't take them! At $0.22 per card (post card rate), you can see that we would have DOUBLED our mailing expenses, not to mention the time and effort of finding addresses for all those cards. So we still have a box of cards for USA stations, in the hope that one day we can find an inexpensive way to deliver them.
In total, 33,066 cards were printed and delivered to various QSL bureaus. This represents a total of 66,312 QSOs confirmed, out of about 100K total QSOs worked. Table 2 shows how many QSLs were sent to countries that had 1,000 or more QSLs, and Table 3 shows a summary of the total expense of this project. It is expected that we will need to send many fewer QSL cards in the future, if we keep up with the job!
Table 2 -- QSL totals by country
Table 3 -- Expenses
Like any big project, this one had a lot of help. All of the operators at Matt's place chipped in either time or money to help get the project done. Steve Gilbert did a great job for us printing all the cards. We got a lot of valuable advice from Martin Cook at the ARRL. Finally, we would not have been able to make this a manageable task without the fine work of Bernd DF3CB and his BV program. Also, a special thanks to the "X" sorter for the W1 QSL bureau (you know who you are) to probably gets more cards for Matt than all the other "1X" callsigns combined!
If you are curious about your cards, you can visit the KC1XX Log Search page on the internet at http://www.kc1xx.com/logs/. If you are a stateside station looking for a card, check here first. If you're in the log, you're probably also in the box! Follow the instructions on the web page if you want a QSL.
Please give us a call in the next contest -- QSL 100% via bureau!
CT by K1EA can be found at http://www.k1ea.com/