Microdrive Unit Design
Re: Microdrive Unit Design
Hi,
I agree with `Byteman', the measuring current using ordinary Multimeter will add up some remanent magnetism to the head corpus material. It is questionable if we changed the polarity during measurement may correct it in one way, and worsen it in the other.
I have consulted Louis Seidelmann (fellow country QL'er) about our discussion, also studied a student's thesis on magnetic recording principles. As we have MD technology here, using direct currents for read/erase/write operations this will cause the head be magnetised just by regular use. Measuring current used in numeric MM seems negligible and will not make it any different.
This remanent magnetism may or may not with enough many passes through the already magnetised head's gap decrease the recording's intensity. But this may correspond to the life of media, so adhering to MD usage guidelines, to keep master backups, should be no issue.
Magnetised heads may cause audible noise in playback of Hi-Fi recordings, frequently discussed among audiophils. Audio Hi-Fi systems overcome this by using alternating currents for erase operations.
Having no exact means to measure remanent values, we can just probe that it is present, with a tiny needle hanged on a length of a very thin thread, the needle is attracted to the head by the magnetism.
Even if we demagnetised the head, (a simple DIY solution could be a coil for transformer pistol iron, see picture), the head will become magnetised again just by ordinary use.
Tomas
I agree with `Byteman', the measuring current using ordinary Multimeter will add up some remanent magnetism to the head corpus material. It is questionable if we changed the polarity during measurement may correct it in one way, and worsen it in the other.
I have consulted Louis Seidelmann (fellow country QL'er) about our discussion, also studied a student's thesis on magnetic recording principles. As we have MD technology here, using direct currents for read/erase/write operations this will cause the head be magnetised just by regular use. Measuring current used in numeric MM seems negligible and will not make it any different.
This remanent magnetism may or may not with enough many passes through the already magnetised head's gap decrease the recording's intensity. But this may correspond to the life of media, so adhering to MD usage guidelines, to keep master backups, should be no issue.
Magnetised heads may cause audible noise in playback of Hi-Fi recordings, frequently discussed among audiophils. Audio Hi-Fi systems overcome this by using alternating currents for erase operations.
Having no exact means to measure remanent values, we can just probe that it is present, with a tiny needle hanged on a length of a very thin thread, the needle is attracted to the head by the magnetism.
Even if we demagnetised the head, (a simple DIY solution could be a coil for transformer pistol iron, see picture), the head will become magnetised again just by ordinary use.
Tomas
Last edited by tcat on Tue Oct 31, 2017 11:13 am, edited 1 time in total.
Re: Microdrive Unit Design
I'm sorry, I cannot write detauled post just now about that, I will describe some things some days later. I'm tightly working with analog tape recorders and will tell some things about it.
Re: Microdrive Unit Design
Hi Byteman,
I am not the one in the know here, I am just learning about it, so will be glad to also learn from your own experience.
The reason I suggested measuring the head is to exclude faulty erase coil, very common MD problem I have learnt. Here 3 of 4 MDs faulty! So one does not have to mount it into QL to find that out. Another common fault could be ULA.
Most of the mechanical aspects can be rectified if not correct from prime production.
You may see my quest here, thanks for your help.
http://qlforum.co.uk/viewtopic.php?f=2&t=1381&start=50
http://qlforum.co.uk/viewtopic.php?f=2&t=1511&start=30
P.S. I am now at magnetic hysteresis, and linear pre-magnetic correction, my next post to come.
P.P.S. I should possibly apologise for stealing this thread from RWAP
Tomas
I am not the one in the know here, I am just learning about it, so will be glad to also learn from your own experience.
The reason I suggested measuring the head is to exclude faulty erase coil, very common MD problem I have learnt. Here 3 of 4 MDs faulty! So one does not have to mount it into QL to find that out. Another common fault could be ULA.
Most of the mechanical aspects can be rectified if not correct from prime production.
You may see my quest here, thanks for your help.
http://qlforum.co.uk/viewtopic.php?f=2&t=1381&start=50
http://qlforum.co.uk/viewtopic.php?f=2&t=1511&start=30
P.S. I am now at magnetic hysteresis, and linear pre-magnetic correction, my next post to come.
P.P.S. I should possibly apologise for stealing this thread from RWAP
Tomas
Re: Microdrive Unit Design
Byteman wrote:It is more correct to measure not only resistance, but inductance too.
Code: Select all
MAGNETIC HEAD COILS INDUCTANCE (mH)
=========================================================
| STRAIGHT-EDGED | STANDARD 1 | STANDARD 2 |
_______|________________|________________|_______________|
| MDV1 : MDV2 | MDV1 : MDV2 | MDV1 : MDV2 |
|-------:--------|--------:-------|-------:-------|
| : | : | : |
ERASE | 12.2 : 12 | 11.5 : 12 | 11.3 : 12.4 |
| : | : | : |
| : | : | : |
UPPER | 0.8 : 0.8 | 0.9 : 1 | 1 : 1 |
| : | : | : |
| : | : | : |
LOWER | 0.76 : 0.8 | 0.85 : 0.94 | 1 : 1 |
_______|_______:________|________:_______|_______:_______|
Re: Microdrive Unit Design
Hi,
I am not able to interpret inductance measured, the values seem in spectrum of +/- 0.2 mH (mili henry?), which may be the precision of the inductance-meter used?
My toys here are that of a hobbyist, transformer pistol iron and a simple multimeter (not very precise).
I am now learning about magnetic hysteresis, it seems a welcome quality of magnetic materials, absorbing some inaccuracies we see here. EDIT
Some pre-magnetisation current is needed during writing. It's purpose is to build up some magnetics in the material in the linear (green) part above. The pre-magnetisation current is typically set to frequency between 60-120Khz not to interfere with the recording signal carrying the actual information. The theory has it, that without premagnetics, it won't be quite possible to read recorded information back.
Now looking in the QL Service Manual, chapter 4 on MD shows some `waveforms', I can see 100Khz frequencies. Can someone explain the process of writing? What frequency carries "1", and what "0", and what frequency is the actual premagnetics?
It would be nice to capture `waveforms' using numeric oscilloscope, and post here, but I have none so far, and I would not know how to use it anyway.
Tomas
I am not able to interpret inductance measured, the values seem in spectrum of +/- 0.2 mH (mili henry?), which may be the precision of the inductance-meter used?
My toys here are that of a hobbyist, transformer pistol iron and a simple multimeter (not very precise).
I am now learning about magnetic hysteresis, it seems a welcome quality of magnetic materials, absorbing some inaccuracies we see here. EDIT
Some pre-magnetisation current is needed during writing. It's purpose is to build up some magnetics in the material in the linear (green) part above. The pre-magnetisation current is typically set to frequency between 60-120Khz not to interfere with the recording signal carrying the actual information. The theory has it, that without premagnetics, it won't be quite possible to read recorded information back.
Now looking in the QL Service Manual, chapter 4 on MD shows some `waveforms', I can see 100Khz frequencies. Can someone explain the process of writing? What frequency carries "1", and what "0", and what frequency is the actual premagnetics?
It would be nice to capture `waveforms' using numeric oscilloscope, and post here, but I have none so far, and I would not know how to use it anyway.
Tomas
Re: Microdrive Unit Design
Hi,
I have read some posts from Gert, also looking further at MD `waveforms' in the QL Service Manual.
MD uses direct current for premagnetisation, and FM modulation for data, 50kHz ~ "0", 100kHz ~ "1".
I do not know what value the premag current is, in the picture shown as 30mA.
Tomas
I have read some posts from Gert, also looking further at MD `waveforms' in the QL Service Manual.
MD uses direct current for premagnetisation, and FM modulation for data, 50kHz ~ "0", 100kHz ~ "1".
I do not know what value the premag current is, in the picture shown as 30mA.
Tomas
Last edited by tcat on Sat Nov 04, 2017 9:37 am, edited 3 times in total.
-
- Aurora
- Posts: 950
- Joined: Sat Oct 25, 2014 9:53 am
Re: Microdrive Unit Design
Looking at the Iss6 schematic, looks like a 180 ohm resistor in series switched by transistor at 5V, so naively, I'd say 5V/180 = 28mA
Re: Microdrive Unit Design
Hi,
Thank you for the encouragement, now comes difficult part for me, that is `Frequency Modulation' (FM), I do not quite understand.
Consider `1 0 0 0 1 1 1 1' byte to be stored to one tape track.
How does ULA amplify and decode FM signal to TTL "0", and "1" we read in MD h/w data register?
EDIT
I borrow this definition from Gert's, he has managed to reenginer MD data line format.
Thank you for the encouragement, now comes difficult part for me, that is `Frequency Modulation' (FM), I do not quite understand.
Consider `1 0 0 0 1 1 1 1' byte to be stored to one tape track.
How does ULA amplify and decode FM signal to TTL "0", and "1" we read in MD h/w data register?
EDIT
I borrow this definition from Gert's, he has managed to reenginer MD data line format.
This possibly explains that, except how ULA is synchronised for half square waves not having the clock? See block diagram. Tomas
To read a bit I have to wait for an edge, then wait 1 usec to make sure the data level is stable after the edge, sample the level of DATA1, wait 6 usecs to sample the level half a bit further and compare that level with the level sampled before.
If it has changed it is a '1' else it is a zero.
Last edited by tcat on Fri Nov 03, 2017 7:43 am, edited 4 times in total.
Re: Microdrive Unit Design
Hi,
ULA block diagram above shows, in my opinion, only reading data from two tracks, outputting two channels DATA 1-2 (A/B).
But how is data written to tape, we know we need two frequencies 50/100kHz and some direct current?
Tomas
ULA block diagram above shows, in my opinion, only reading data from two tracks, outputting two channels DATA 1-2 (A/B).
But how is data written to tape, we know we need two frequencies 50/100kHz and some direct current?
Tomas
Last edited by tcat on Sat Nov 04, 2017 9:38 am, edited 1 time in total.
-
- Aurora
- Posts: 950
- Joined: Sat Oct 25, 2014 9:53 am
Re: Microdrive Unit Design
Hi Tomas
As I understand it, writing involves the ULA 'switching' its inputs to outputs and driving the same (stereo) pair of coils with the 50/100KHz signals.
The direct current is entirely independent and flows just through the single erase coil - which has a larger 'gap' - and which is offset physically to effectively de-polarise the tape shortly _before_ that same section of tape reaches the RD/WR gaps. This de-polarisation is effectively placing white-noise on the tape, which the ULA (or the 8302?) can distinguish from the polarised data and is used for the sector GAPs in the MDV timing protocol.
Also, if my understanding is correct, whilst the ULA outputs serial binary when reading, what it actually reads off the tape (during the DATA sections) are peaks in current that represent the transitions between low and high. The actual 'levels' that were recorded during writing become irrelevant - only the timing between these current surges at each transition. Thus, it's not quite pure FM, but more like (but not quite) 'Manchester' encoding - where the relative timings between transitions in a given 'bit-cell' time-window determines the bit-state. In such encoding schemes, the 'CLOCK' is actually encoded within the data stream. I may be wrong here...
I've found a good analogy described in the DIYTK 'FlexiNET' replacement driver for QLAN.
As I understand it, writing involves the ULA 'switching' its inputs to outputs and driving the same (stereo) pair of coils with the 50/100KHz signals.
The direct current is entirely independent and flows just through the single erase coil - which has a larger 'gap' - and which is offset physically to effectively de-polarise the tape shortly _before_ that same section of tape reaches the RD/WR gaps. This de-polarisation is effectively placing white-noise on the tape, which the ULA (or the 8302?) can distinguish from the polarised data and is used for the sector GAPs in the MDV timing protocol.
Also, if my understanding is correct, whilst the ULA outputs serial binary when reading, what it actually reads off the tape (during the DATA sections) are peaks in current that represent the transitions between low and high. The actual 'levels' that were recorded during writing become irrelevant - only the timing between these current surges at each transition. Thus, it's not quite pure FM, but more like (but not quite) 'Manchester' encoding - where the relative timings between transitions in a given 'bit-cell' time-window determines the bit-state. In such encoding schemes, the 'CLOCK' is actually encoded within the data stream. I may be wrong here...
I've found a good analogy described in the DIYTK 'FlexiNET' replacement driver for QLAN.