2017 Sansui Serial Number Database

QS-800, S32040723, Coolerking

Coolerking, can you double check this serial number? They don't typically have an alpha character in them. That leading "S" is probably an 8.

Jomark, the serial number digits follow the correct serial number pattern but I don't have another B-2302 on file to compare it to.

Harvey/Ga, Robdunn, I have yours too.

- Pete
 
Z-3000X,603119857,pronoob101

In case it matters, I also have a list of sansui authorized service centers, a program time sheet, and a users manual (both the z-3000x and z-5000x), as well as the box and packaging materials
 
D-70BW,596081585,Andrew L

Just got this dual cassette deck from my father this weekend thought it was lost but he found it in his basement tucked away.
 
I was looking over the dates for the TU-717's in your database and I don't believe this one is correct:
848110005 November, 1988
Shouldn't it be 1978?
 
Interesting. That actually turns up a potential flaw in the model record that tripped up the date detection coding. That serial number is tied to a TU-719 that was entered into the database in April of 2005. When the TU-719 was originally set up in the database it was assigned a "decade reference" of 1979, meaning the earliest production was in 1979. The program looks at the decade reference. If the third digit is lower, it assumes the production was in the next decade. In this case there are several TU-719 produced in 1980.

If that serial number is correct, this unit was produced in 1978 and is the only example produced that year. The next ones on file were produced in January of 1979 so it is the earliest one on file and with a production number of 0005 it could quite possibly be the 5th one produced.

Thanks for pointing it out. I'll revise the decade reference so any other TU-719s produced in 1978 will have the production date decoded correctly.

- Pete
 
I completely missed that those were TU-719's. I thought they were 717's. I think it would be easier to read the report if you included the model on each line, or at least have it included on the top of each page (something like "TU-719 (continued)"). Rather than having to scroll up several pages to see what section you're in, you can just glance at the top of the page. Just a suggestion.

I suspect you have the same problem with a few TU-7900's as well: 026080253 August, 1986

I also don't think this can be correct either: TU-919, 215010048 January, 1985
 
Last edited:
Definitely the same problem with the TU-7900. The Decade reference was set to 1977. Anything built in 1976 would show up as 1986. There also looks like there are some duplicates in that group caused by dropped leading zeros. I had troubles with that when I set the current database tables. The TU-919 is likely an error in the serial number itself. The third digit is reported as a 5. Both 1975 and 1985 are outside the known date range of manufacture of a TU-919. That puts it into the category of "invalid date code." (UPDATE) I used the AK search function and found the original post from 2008. It's actually a TU-9500! The date makes perfect sense now.

I'm fixing these types of errors as they turn up. Many of them are legacy issues with the old files I used to recreate Paul's database. He did a fantastic job setting up the original database starting around 2002/3. At that time there wasn't a lot of information readily available on most of these models so he had to guess on the decade reference. He got most of them right. Even now, every month, new models come up and have to be added to the model list and I often find conflicting manufacture date information on line.

I've been updating the decade references in the Model Name table as I come across these errors. In some cases, like the RA-500, I still have to pin down first production. Then I can run the whole master serial number table through the date extraction macro to correct all the decade reference errors. Although that would fix all these problems, I'm scared shitless of what new problems that could introduce. There are more than 6000 serial numbers on file and countless "invalid date codes." Maybe it's better to do it one at a time.

Please keep reporting errors as you find them. Thanks for your help improving the database.

- Pete
 
BA-3000 22508825

CD-10 210810053
Rob, I'm processing serial numbers today and these two bounced. The BA is a digit short. The fourth digit in the CD-10 serial number can't be an 8, it has to be either a one or zero.

- Pete
 
Last edited:
I completely missed that those were TU-719's. I thought they were 717's. I think it would be easier to read the report if you included the model on each line, or at least have it included on the top of each page (something like "TU-719 (continued)"). Rather than having to scroll up several pages to see what section you're in, you can just glance at the top of the page. Just a suggestion.

It turned out to be pretty easy to modify the report to include the model at the top of the page when it carries over from the previous page. The internet is a wonderful tool. The August report will have that feature. Thanks for the suggestion.

I also considered repeating the model name on each line but I think that would make the report too cluttered.

- Pete
 
9090DB, 817063917, Juan9568

It was my uncle's, purchased in Buenos Aires, Argentina but I don't know the date. It needs a major restoration..will be posting pics soon
 
7070, 217011668, tyella

The last two numbers are pretty worn down, could be either 6 or 8. The above number is my best guess.
Thanks for putting this list together! I have a AU-7700 integrated that I'll get the number off of next time I pull it from the shelf. It's doing preamp duty right now.
 
Back
Top Bottom