RE: JamMan memory problems, Part 2

["Hogan, Greg" <GHogan@lexicon.com>]

Good explanation Robin,

JAMMAN does not test the DRAM at power up.  You can run a DRAM test as   
follows:Start with the power off.  Hold down the "RESET/Bypass" and   
"Function" buttons and turn the machine on.  You should get a "d" in the   
display.  Release the buttons.  Press and release the "RESET/BYPASS"   
button.  You should know have a number between 0 and 15 which corresponds  
 
with the position of the Mode encoder.  Turn the mode encoder to "PUNCH   
IN LOOP 4."  You should now have a "6" in the display.  Press and release  
 
the "RESET/BYPASS" button to run the test.  If the test fails you will   
have a hexidecimal number between 1 and F flashing in the display which   
will indicate which DRAM is failing as follows:1=U20, 2=U21, 3=U20 & U21,  
 
4=U22, 5=U20 & U22, 6=U21 & U22, 7=U20, U21 & U22, 8=U23, 9=U20 & U23,   
A=U21 & U23, B =U20, U21 & U23, C=U22 & U23, D=U20, U22 & U23, E=U21, U22  
 
& U23, F= U20, U21, U22, & U23.

I hope this helps.

Please let me know if you have any questions or if there is anything that  
 
I can do for you.

Best regards,

Greg Hogan
Lexicon Customer Service
Phone 6172-280-0372
FAX 617-280-0499

 ----------
From:  robin.b2[SMTP:robin.b2@ukonline.co.uk]
Sent:  Monday, August 18, 1997 12:03 AM
To:  Loopers-Delight
Subject:  Re: JamMan memory problems, Part 2

 ----------------------------------------------------
> On the second permutation, I got the desired effect: 32 seconds of
> glorious, distortion-free looping.
>
> So, now I'm really curious about how this can work this way. I've   
talked to

Well it sound to me like you *have* got a duff chip there but it's now
in a position to cause distortion that's inaudible. Say you had the most
significant bit of your sixteen stuck on (or random) you'd have either a
massive DC offset on your signal (half the full swing ) or large amounts
of noise. Move this duff chip to the least significant bit and all of a
sudden your noise of offset is only 1/65535th of the total output swing
possible... hardly noticable, in fact there are some intersting
techniques for hiding encrypted data (which looks like a random
bitstream) in digital audio this way.. simply string out all the bits in
you data bytes and use them as the least significant bit in your WAV
file or whatever... similar things are done to hide digital copyright
data in CD's etc but I digress!

All the above is rendered a bit invalid if the jamman runs a memory test
on powerup as it should spot this and complain... though it might not
spot an error knocked out  address bits, this would mean that data got
replicated inblocks throughout the RAM and in it's least noticeable form
would be like halfing the sample rate...hmmm some jamman inside knowlege
needed here! Bob?

Anyway the long and the short of it is that there are plenty of possible
explanations for a faulty chip making your unit behave as it does so I'd
say ask them for one more, swap the ram back to a noisy config, then
swap the replacement in  one at a time until you find the culprit.

Hope this helps a bit,
                    Robin.