QL's external MDV port

[tcat]

Hi Silvester,

Thank you for your help.

I will try to measure the tape speed (~ motor RPM) by a reference cartridge, measuring time when a random sector comes by again. So I believe, should I need to calibrate motor RPM, I just flip off the small black cover over the adjustment hole, at the bottom of motor corpus to get reach to the potentiometer.

Meanwhile I have completed the adapter schematics.

MD PORT

Remaining questions>>
EC1 edge connector numbering 1A-8A is TOP, 1B-8B is BACK, or vice versa?
Substituting ZTX551 for BC640, and BC184 for BC547B, I hope it is alright?


Many thanks so far.
Tomas

[Outsoft]

Hi Silvester,

Thank you for your help.

I will try to measure the tape speed (~ motor RPM) by a reference cartridge, measuring time when a random sector comes by again. So I believe, should I need to calibrate motor RPM, I just flip off the small black cover over the adjustment whole, at the bottom of motor corpus to get reach to the potentiometer.

Meanwhile I have completed the adapter schematics.

mdport.png

Remaining questions>>
Edge connector numbering 1A-8A is TOP, 1B-8B is BACK, or vice versa?
Substituting ZTX551 for BC640, and BC184 for BC547B, I hope it is alright?

If we can adjust the motor speed...we can use also the Spectrum microdrives on the QL I think

Speccy MD's seems more fast than the QL ones, right?

Great project Tomas!!!


Many thanks so far.
Tomas

[Bloodnok]

Hi Tomas,

Belatedly replying to an earlier post on using Spctrum microdrives on the QL. In looking into why they don't appear to support formatting but can be read OK I set about adjusting the motor speed as Dilwyn mentioned to see if this made a difference (I've checked through the circuits and nothing else jumps out as obvious to me - the erase current is different but this doesn't appear to be critical). The good news is that I have one of my Spectrum drives (unmodified in any way) which formats fine on the QL at a speed of 37-38 rpm. Anything under or over this value then the same drive would fail to format. The drive works fine on the Spectrum with the same speed setting. The bad news is that I've only managed to get the one drive to work. I have (quite) a few other drives so will look into it further and report back.

Regards,

Charles

[tofro]

The Spectrum drive I once had formatted, read and wrote perfectly on the QL - Without anything done do it.

Tobias

[tcat]

Hi,

Provisional board design, as I am not much into electronics, this is my second ever design in KiCad, next to Simple RAM.

I struggle a little with transistors, there seem variants PNP, NPN, having different voltages, currents and gains, they also differ in pinouts BCE, EBC etc.

So I hope I picked suitable substituting types to the original QL design.

MD PORT, mPORT board design

P.S. I used components reference as per HEAD BOARD 2, so can be checked against

Tomas

[Silvester]

tcat wrote:
I just flip off the small black cover over the adjustment hole, at the bottom of motor corpus to get reach to the potentiometer.

Just push small screwdriver through existing hole. There is a rubber membrane in base of motor which has already been punctured to allow factory set up.

The MDV connector on right hand side of QL is, looking into PCB slot:

8A 7A 6A 5A 4A | 2A 1A
====================
8B 7B 6B 5B 4B | 2B 1B

(verified on unpopulated issue 6 QL PCB. Also, unused WR PROT pin 2B has no track connection)

There is a missing link on your diagram (shown in red), it would work but the LED would not light. Everything else checks out. (Have to look up transistor data, haven't got it to hand.)

I would be very careful checking you wiring - the ZX8302 ULA is very easy to damage with momentary short-circuit. I did so once when putting a probe on COMCTL (pin 29) to monitor IPC link.

tofro wrote:
The Spectrum drive I once had formatted, read and wrote perfectly on the QL - Without anything done do it.

I agree, on the occasion I tried a ZX drive (on backwards) it worked OK.

AFAIK it is not a ZX/QL motor speed difference, the fixed data frequency is just faster (IIRC about 11/8 faster).

Unfortunately at one fell swoop they made ZX cartridges unreadable and QL ones less reliable. I'd have traded the extra space for reliability and portability.

[tcat]

Hi Silvester,

Thank you for checking my circuit diagram, I will add the missing link, and redo the routing.

I am also glad for your EC1 pinout confirmation, I had measured 9V voltage between VIN (4B) and GND (8A 7A 6A 8B 7B 6B), so I was almost certain then.

As for transistors I picked these types based on their specs.

Original QL specs:
==================

TR6 / TR7 :
ZTX551 / Zetex Semiconductors
PNP 60V/ 1,0A/ 1,0W TO92 C B E
hfe 50..150

TR4 / TR5 :
BC184 / Fairchild Semiconductors
NPN 30V/ 100mA/ 350mW T092 C B E
hfe 100..130

Found close specs:
==================

TR6 / TR7 :
BC640-BLK / Fairchild
PNP 80V/ 1,0A/ 0,80W TO92 B C E 7c
hfe 40..160

TR4 / TR5 :
BC547"A"-BLK / Diotec
NPN 45V/ 0,10A/ 0,50W TO92 E B C 7a
hfe 90..120

Please NOTE
There are other variants "B","C", but "A" has closer hfe gain to the original.

Tomas

[1024MAK]

As for transistors I picked these types based on their specs.

Original QL specs:
==================

TR6 / TR7 :
ZTX551 / Zetex Semiconductors
PNP 60V/ 1,0A/ 1,0W TO92 C B E
hfe 50..150

TR4 / TR5 :
BC184 / Fairchild Semiconductors
NPN 30V/ 100mA/ 350mW T092 C B E
hfe 100..130

Found close specs:
==================

TR6 / TR7 :
BC640-BLK / Fairchild
PNP 80V/ 1,0A/ 0,80W TO92 B C E 7c
hfe 40..160

TR4 / TR5 :
BC547"A"-BLK / Diotec
NPN 45V/ 0,10A/ 0,50W TO92 E B C 7a
hfe 90..120

Please NOTE
There are other variants "B","C", but "A" has closer hfe gain to the original.

Hi Tomas

Both transistors are just being used as switches, so selecting replacements is not particularly critical.

For TR4 / TR5, any small signal NPN type that has a gain greater than 90 should be okay. A higher gain figure does not matter. So a BC547B or BC547C or indeed BC548, or BC549 plus many others should all work okay.

With TR6 / TR7, they are being used to switch the motor current. So the requirement here is the current rating. So we need a transistor rated at 1A at least (a higher rating does not matter). The gain should be 30 or more (again, a higher figure does not matter). The power rating should be 0.5W (500mW) or greater.

Note that the overall gain (multiply the gain of the two transistors together) should be greater than 2538 (so for example 90 x 30 = 2700). And in fact, in the design process, engineers prefer to have a large margin because real word transistor gain varies with temperature and operating current.

As this is a 9V nominal system, the transistors voltage ratings don't really matter, as very few discrete transistors have low voltage ratings.

Mark

[1024MAK]

About the BC184 transistor, while working on another project, I wondered why I could not see these listed, on the Fairchild website they say "Obsolete as of 14-Dec-2011".

Be aware that there are two different transistor packages (cases) with different lead-outs. Don't confuse a BC184 and a BC184L!

Mark

[Outsoft]

The Spectrum drive I once had formatted, read and wrote perfectly on the QL - Without anything done do it.

Tobias

Me not but I would like that the Spectrum MD can be totally compatible with my QL.

I don"t know If Tomas is making an adaptor for let the Speccy drive compatible with the QL one but It will be great to adapt an internal QL md unit too