Building a replacement MTX PSU

[gunrock]

Actually Mark, I retested this morning the diode reverse polarity protection and the reading was 0.03 to 0.04V. I used the LM2577 dialled to 9.8V and repeated, the reading was the same. Must have been my eyes late last night

Using the higher voltage meant that my load test was much better, that 6V led really shone!

Now I have to start wiring up. This is the bit I'm most nervous about, because as great as all the photos you have produced are, there's a lot of shrink tubing, which obscures things a bit (obviously *not* a criticism). So I need to refer to the schematic more and I'm an almost complete noob with electronics, thus I'm concerned I'll misread it.

I also know I'm going to forget to put shrink tube on something and have to unsolder one end to fit it on!

Steve

[gunrock]

Ok, so I have a question It seems I have foolishly bought the wrong type of LED; instead of a Kingsbright L-53gd I bought a Kingsbright L-7113id-5v, which is a 5v LED with a built-in resistor.

I can't work out how on the one you specified, there's need for a 1.2K resistor and that's because I can't understand the scematic with my dunce level of electronics knowledge. How much current feeds the LED? 16V from the LM2577, to the center tap (series joined dual secondary) which is rated what? The datsheet says 2.0 forward voltage and 10mA forward current, but TBH I'm lost

I'm trying to work out if I can use my LED with built-in resistor and whether I will need extra resistance still. I've learnt a lot the last few days (center tapping dual secondary windings, for one!) but this is twisting my melon, man!

Please help.

[1024MAK]

LEDs are light emitting diodes. Like a normal diode, once the forward voltage gets to the ‘turn on’ value, they start to conduct current. With a silicon diode, this is about 0.6V. With a LED, it’s between about 1.8V to 3.6V depending on the type/colour.

By forward voltage, I mean a positive voltage on the anode (A) compared to the cathode (K) terminals/leads.

Once a diode is conducting, a very small increase in voltage results in a VERY BIG increase in current. With a normal diode this is not normally an issue, as it feeds some other load.

But with a LED, normally we only want the LED to light up, and there is no other load. So we have to find another way to limit the current. The easiest and cheapest way is to include a resistor in series with it. Either wired to the anode terminal, or to the cathode terminal. Hence this is what I did.

So what happens if no resistor is used? Well, once the voltage across the LED gets to it’s forward voltage (in this case let’s say 2V), it will try to maintain about 2V across itself. But here the supply voltage is 16V. So the current will increase and increase and increase until something goes pop! In practice the LED will glow brightly for a very brief instant and then it will never work ever again.

To calculate the value of the resistor, you use this formula:

(Vs - Vf) / If = R

Where Vs is the DC supply voltage, Vf is the forward LED voltage, If is the forward LED current that you want, and R is the value of the resistor that you need.

Alternatively you can use this formula to work out the current with a particular resistor:

(Vs - Vf) / R = If

With a standard brightness green LED, the forward voltage is about 2V, so Vf = 2V. The recommended forward current should be 20mA or less, so this gives the maximum If.

And yes, the DC supply voltage in the MTX PSU circuit for the LED is 16V DC.

So the formula becomes:
(16V - 2V) / 1200Ω = 11.6mA

A 5V LED often just contains a built in resistor. Now, it’s unlikely that the manufacturer will tell us the value of resistor that they used. But they do give a typical current. In this case, I believe it’s 13mA.

So I work out that they have used a resistor of about 220Ω.

If you wire the 5V LED in place of an ordinary LED in this circuit along with the 1.2kΩ resistor, the forward current will be about 9.9mA. So the LED will still work, and it should still be a reasonable brightness.

So no need to panic!

The configuration of the transformer secondary windings that I am using is to use the two separate secondary windings to form a centre tapped secondary. I have done this because it’s the arrangement that the power supply circuitry in the MTX requires. The middle / centre tap is effectively the MTX 0V/GND.

I hope this helps.

Mark

[gunrock]

Hi Mark,

That is very helpful, yes, thank you. Normally, in the case where I hit a knowledge gap that I can't seem to fill by some research, I try to make an educated guess and carry on (or obtain the right component!) but this being a PSU, I'm being way more cautious.

I don't want to create a dangerous device nor kill my new Memotech, so I'm being very careful.

I'm back at work now, so progress might be (even) slower, but will keep you posted.

Thanks
Steve

[gunrock]

And we're done:

20210207_195128.jpg (154.98 KiB) Viewed 28873 times

I have performed the tests that you prescribed on earlier pages, Mark.

16.0V DC on the nose, from the yellow wire. Combined AC from both 7.5V lines is 17.1V. The AC reading from the 16V DC is 34.4mV.

Without load, the 7.5V AC reading is around 8.4V.

I tested with a 12V bulb on the 7.5V AC with success:

20210206_233140.jpg (106.69 KiB) Viewed 28873 times

I soldered the connector and as I was concerned about the quality of that work, I tested each pin again and got the same readings.

So far, so great! However, when I powered up the MTX, the fuse blew (F3). There was no picture, but the start of a beep before the power is cut. Turns out, I had the wrong size fuse. I replaced it with the right one and:

20210207_224757.jpg (117.49 KiB) Viewed 28873 times

The keyboard doesn't seem to work well, this is what QWERTYUIOP looks like:

20210207_225219.jpg (20.7 KiB) Viewed 28873 times

Anyway, not sure how to start troubleshooting, I'll do some research, maybe I have to build the diagnostic board.

Thanks for all your help and such a great design!

[Dave]

Well done - great result!

The diagnostic board probably won’t help you with the keyboard problem, but see . . . .

http://www.primrosebank.net/computers/ ... yboard.htm
http://www.primrosebank.net/computers/ ... arren2.htm
http://www.mtxworld.dk/memorum/viewtop ... 7d55051207

Regards
Dave

[gunrock]

Thanks Dave. Only one month (32 days) in the making

Actually, I feel quite proud. That might seem a bit daft to a lot of the folks here, for whom specifying and making your own add-ons (CFX, NFX, ReMemorizer, et al.) and whole computers (MTXplus, Z180 etc.) is like second nature. All had to do was follow a schematic and look at lots of detailed photos and information provided by Mak and others and not screw it up.

Yet I feel quite proud, as I have never built an electronics project before. Full disclosure- I've installed two simple console modchips and a pin header install in an original Xbox, made an LED with resistor on an extension cable for a Commodore 64 last year and finally, I built a LED Christmas tree kit in December to test my soldering skills prior to attempting this PSU build. That's it!

I have no prior electronics training or experience and it showed. I made lots of mistakes and my build looks nowhere as professional as Mark's example photos. However, I saw it through, rectified (or lived with) the mistakes, learnt a lot on the way about taking on a project like this, electronics theory and putting it into practice. And best of all, it bloody works! Plus I have the start of a supplies bin for future projects.

Even my wife, who was dubious about what I was doing and how long I was taking on it and whether or not I might electrocute myself, said she was impressed when she saw the finished item and the general quality of the work inside.

So thanks for your support, Dave and to Mark for the design and feedback. I'm now on the way to having rescued another MTX from obscurity.

Cheers,
Steve

[Martin A]

Your keyboard's problem could be as simple as a misaligneded/damaged keyboard cable. If you have an old 40 way IDE cable you can use that as an alternative (just avoid the side with the filled in pin). They're phyically stronger than the original MTX part too.

[1024MAK]

And we're done

Well done!

The keyboard doesn't seem to work well, this is what QWERTYUIOP looks like

But at least every key you pressed did something! Funny enough, as I’ve had “off by one” problems with connectors back in the distant past, I always carefully inspect most including both pin header and other similar connectors (as well as DIL/DIP chips/sockets/headers) before powering things up - saves the dogs barking and family yelling when the smoke detector goes off after the magic smoke escapes! So I don’t actually know how a MTX would react if the keyboard connector was misaligned by one, because I’ve never had this with a MTX.

Does pressing both unmarked keys at the same time (either side the space bar) reset the machine? If not, do check that the keyboard connectors are correct aligned.

Mark

[Bill B]

That is certainly a keyboard cable displaced by one:

Q (Drive 3, Sense 0) -> W (Drive 2, Sense 1)
W( Drive 2, Sense 1) -> 4 (Drive 1, Sense 2)
E (Drive 3, Sense 1) -> R (Drive 2, Sense 2)
R (Drive 2, Sense 2) -> 6 (Drive 1, Sense 3)
T (Drive 3, Sense 2) - > T (Drive 2, Sense 3)
Y (Drive 2, Sense 3) -> 8 (Drive 1, Sense 4)
U (Drive 3, Sense 3) -> I (Drive 2, Sense 4)
I (Drive 2, Sense 4) -> 0 (Drive 1, Sense 5)
O (Drive 3, Sense 4) -> P (Drive 2, Sense 5)
P (Drive 2, Sense 5) -> ~ (Drive 1, Sense 6)

It does look as though all those keys are working