Tracking Adjustment with Oscilloscope

BennehBoy · 31. Oktober 2022

Hi,

Apologies for this being in English - I don't think Google translate does a great job and I don't speak German.

After having had reasonable success with adjusting tracking on my 3x1541's using just World of Jani's 1541 diagnostic cartridge, I still have some titles that won't load on 1 of the drives. All have very close to 300rpm spindle speed.

I thought I would use my Hantek DSO to try the procedure described here -> 1541: Wartungs- und Einstellarbeiten

The instructions call for setting the scope to AC, voltage at 5mv per division & 200us time division - but this simply results in a screen full of noise (even at 50mv division)

Any advice?

Retro-Rentner · 31. Oktober 2022

Sorry for my bad english.

Based on the website from your link i didn´t assume a rectangular waveform as in a Logik-Analyzer.

The better the tracking, the higher the amplitude/voltage of the signal.

Did you used shielded signal-probes/wires? Properly grounded?

How does your Hantek-Screen looks if you measure a different floppy drive?

The one you are working on maybe damaged/bad.

So your screen could be because of bad signal from your "bad" floppydrive or from improper measurement.

BennehBoy · 31. Oktober 2022

I'll try another drive.

I don't believe the stock probes are shielded - happy to order some new ones if anyone has a recommendation.

At the end of the day it's a cheap scope so I may be expecting too much from it.

Zitruskeks · 31. Oktober 2022

I'd guess the persistance of the analogue scopes screen does a bit of averaging. (look at that more zoomed in screen of the scope in the thread you linked)

You could play with the persistance setting of your scope or with the averaging function. Or use a faster timebase.

I used a digital scope to align the drives in my 3040 with an original commodore disk, too, Just wrote me a litte basic program that constantly read certain tracks. It didnt look to far off from your images, I just used a different timebase to push the spikes closer together.

BennehBoy · 1. November 2022

My image was just the baseline noise across the probes - when I push data through the head it's indiscernible because of that noise floor.

The drive works pretty well, but a couple of original games won't load - I suspect those disks are probably a bit 'weak' / use copy protection so need the mechanism to be spot on.

Not got round to checking another drive yet - busy life Thanks for the responses.

BennehBoy · 4. November 2022

OK, so I get the same noise with another drive - even when the drive isn't powered on so it's definitely RF interference on the probe cables. I'll see if I can get some decent shielded ones.

Zitruskeks · 4. November 2022

Are you using the described test points for the probe and also the ground clip on the probe? Is the probe set to 10x?

A scope probe has a coax cable, so there arent no "unshielded" ones.

The Hantek surely isnt a toy and even the 50Mhz ones come with good, 150Mhz Probes.

Zitruskeks · 4. November 2022

PS: thats an 1541 you have, no 1541II?

I'll dig out one and test it here, with my hacked Hantek 1054 and my Tek314.

BennehBoy · 4. November 2022

Yeah I've tested 2 1541 short board drives (I have 3, a Master-41, & a 1571).

The problem here seems to be the scope and probes.

Zitruskeks · 4. November 2022

Oh, shortboard, did you measure at the correct pin? The Text says "Track position: At pin 16 of the hybrid module. Pin numbering from left to right" but thats a thing, its NOT "from left" if the drive front points towards you, like it normally does. Better use the visible tracks/via as a marker. Out of luck i took my shortboard to try it and stumbled about that, too.

Otherwise, it should work but its also a bit what I said earlier, the CRT hides outliers way better as the intensity differs due to the persistance.

Here with my cute little TEK 314 10MHz Scope

and with my 1054Z, "hacked" to have 100Mhz Bandwitdh

you can see a lot more of the noise, but you can change quite a lot to see it better:

1: change the brightness! On 100% brightness the high amplitude parts nearly hide everything

2: change display from vector to dots. this also keeps the focus away from the outliers.

3: use the average function:

hides the high amplitude outliers really good.

PS: i goofed myself extremely today. set up my Tek and there was a 1 second high intensity pulsing on the signal. even without probes. BANG BANG... i searched and searched. After 20 mins i finally realized i had placed the scope directly on my logitech powerplay mousepad. Yes, its a wireless charger.

BennehBoy · 5. November 2022

Sorry, I meant the intermediate length board, so not like a 1541II board.

I'll try those settings to see what it looks like after, thanks.

BennehBoy · 5. November 2022

Well altering the settings as you described has really helped, but I'm now even more confused.

According to the linked instructions, a good peak to peak measurement would be 25-30 mV. Assuming 5mV division per grid.

I'm seeing 80-100mV peak to peak.

Zitruskeks · 5. November 2022

First: its not really about absolute measures, we are just trying to find a peak.

Also: 10x probes. While digital probes adjust the reading on the screen according to the setting, a analogue scope (usually) dont.

Thats why my screens (5mV/div on the Tek, 50mV/div on the Rigol) still are equal.

PS: Funny thing: the 314 came originally with a P6149 probe. Thats a fixed 10x probe. Still its calibrated for a 1x input.

BennehBoy · 5. November 2022

OK got it, so probably just the fact I'm a scope noobie that's the issue here. Thanks for all of your input, I really appreciate it.

Zitruskeks · 5. November 2022

Im still a noob, too, no worries xD