From dls@haystack.mit.edu Mon Nov 16 09:10:43 1998 Date: Mon, 02 Nov 1998 12:55:32 EST From: Dan Smythe To: Ralph Spencer Cc: Ari Mujunen Subject: Re: Triple-cap at 18 Mb/s Dear Ralph, The triple-cap head stacks are fine at the highest data rates, but the Spin-physics stacks are marginal for writing at that rate. I am planning to write a Mark IV Memo on that subject. A quick summary of this work is that Metrum stacks are marginal writing at 18 Mb/s using the current design of the read/write and write-only interface boards. I have changed the drawings to show the 1K chip resistors in series with each head changed to 2.2K. With the 1K, the Spin Physics head stacks are useless at 18 Mb/s; with the 2.2K they are marginal, but usable. Any new read/write and write-only interface boards should be built with 2.2K, and any old ones must be changed before using a Spin Physics head stack to write at 18 Mb/s (320 ips). Also, I discovered that the EMU write-only interface boards were built with the read transistors installed. These transistors must be removed in order to write with any head stack at 18 Mb/s. I have added the following note to the drawings (4712S00X, X=5,6,7,8): 2. Delete C1, C2, R19, R20, and Q1 thru Q18 for write-only interface. I have also changed R1 thru R18 from 1K to 2.2K on these drawings. Cheers, Dan From dls@haystack.mit.edu Mon Nov 16 09:11:01 1998 Date: Wed, 04 Nov 1998 8:59:53 EST From: Dan Smythe To: Paul Burgess Cc: res@jb.man.ac.uk, dls@haystack.mit.edu, afreihold@mpifr-bonn.mpg.de, p655miw@mpifr-bonn.mpg.de, amn@kurp.hut.fi Subject: Re: Triple-cap at 18 Mb/s > Query of not from Dan Smythe: >> >> The triple-cap head stacks are fine at the highest data rates, but >> the Spin-physics stacks are marginal for writing at that rate. > >PB: > Does this mean Metrum triple-cap heads are fine at highest rates, > but Spin Physics triple-cap heads are marginal? > It's got nothing to do with triple-cap vs. stepped. Electrically, T-C are identical to stepped. The only difference is geometry. HOWEVER, the Spin Physics stacks are different from Metrum, and have more mutual coupling between heads, causing problems at 18 Mb/s. With the existing write electronics, with the 1K resistors, ANY Metrum stack is marginal (stepped or T-C), and ANY Spin Physics stack is useless. Please stop confusing head stack geometry with the manufacturer. Either geometry is available from either manufacturer. And where did that rumor about Metrum taking over Spin Physics come from???!!! NOT TRUE. Metrum does now own Penney & Giles, but that's another story. ... Dan From dls@haystack.mit.edu Mon Nov 16 09:11:28 1998 Date: Tue, 10 Nov 1998 10:22:22 EST From: Dan Smythe To: Ari Mujunen Cc: dls@haystack.mit.edu, res@jb.man.ac.uk, jr@kurp.hut.fi Subject: Re: Triple-cap at 18 Mb/s >On Mon, 2 Nov 1998, Dan Smythe wrote: > >> Also, I discovered that the EMU write-only interface boards were >> built with the read transistors installed. These transistors must >> be removed in order to write with any head stack at 18 Mb/s. > >Dan, does this mean that only the write-only head will be able to write at >18Mbit/s/track, i.e. do the transistors on the read/write head interface >board interfere with 18Mbits/s/track writing? > The transistors on the read/write head interface board do interfere somewhat with 18 Mb/s/track writing, but not enough to make it impossible. >Or is it just those "dummy" transistors on write-only board which >interfere with 18Mbits/s/track writing? > Those "dummy" transistors on the write-only board cause big problems with 18 Mb/s/track writing, because they get forward biased, and write current flows through them. >> I have >> added the following note to the drawings (4712S00X, X=5,6,7,8): >> >> 2. Delete C1, C2, R19, R20, and Q1 thru Q18 for write-only interface. >> >> I have also changed R1 thru R18 from 1K to 2.2K on these drawings. > >Jouko, please fetch the updated drawings (from dopey.haystack.edu?) and ftp://dopey.haystack.edu/pub/mark4/recorder/4712/autocad/4712s00?.dwg Best regards, Dan Smythe