Micro UK101 Build

GARTHWILSON · Post by **GARTHWILSON** » Wed Apr 06, 2011 6:46 pm

You're thinking of A/D (analog to digital), but the PC's A/D converters are meant for audio and are way too slow for what you want. I suppose a parallel port could be used for a logic analyzer of sorts though.

See my debugging tips starting at viewtopic.php?t=342&start=35 (tips #35, 36, and 37).

micro_brain · Post by **micro_brain** » Wed Apr 06, 2011 7:05 pm

Gosh more reading to do! OK, thanks! Luckily a copy of "Programming the 6502" landed on the doormat this morning, so I'm set. Been a while since I hand assembled machine code. Anyone know the UK10o1's entry point at cold start? Must be same as 6502.

Second EEPROM burn failed with an error. I checked the chip was successfully wiped back to 0xFF before starting. Only difference was a delay setting I put in the burner software. Hmm..

@Lee : I found a radio and have heard interference on LW that comes and goes with the application of power to the circuit, so I guess the clock is OK.

[edit] : More info - the EEPROM programmer keeps bailing out at 0x2000, says there is a write error. I think the chip is borked. The RS232 cable I am using is NULL MODEM (with TX/RX swapped) so I have wired the MAX232 outputs as Pin 2 RX, Pin 3 TX and Pin 5 is GND. No more testing until I can get the EEPROM sorted...

digidice · Post by **digidice** » Wed Apr 06, 2011 8:13 pm

you could always try this to test the board

http://6502.org/mini-projects/nop-gen/nop-gen.htm

micro_brain · Post by **micro_brain** » Thu Apr 07, 2011 10:27 am

EEPROM reprogrammed, still nothing coming out of the serial port.

I think I will have to recheck all the wiring. CPU is getting warm, so it must be up to something. All the other chips are at ambient (or close enough to make no difference).

In the meantime, I bought the bits to builds Lee's NOP generator (including a piezo sounder and LEDs so I can perform the testing suggested). I'm actually quite interested in burning a test program into the rom; with a suitable loop to slow things down I should be able to detect signals on the address / data lines....

micro_brain · Post by **micro_brain** » Thu Apr 07, 2011 12:44 pm

By the way, what sort of scope would I need? I'm scanning ebay looking for a cheap model - what specifications should it have?

HiassofT · Post by **HiassofT** » Thu Apr 07, 2011 1:17 pm

micro_brain wrote:

By the way, what sort of scope would I need? I'm scanning ebay looking for a cheap model - what specifications should it have?

I'd say 2-channel 100MHz (or faster).

I have an old Tektronix 465B here and I'm very happy with it. Well, a 2465 would be great, but they cost a lot more... A Tek 475 (similar to 465, but with 200MHz) would also be a nice option.

so long,

Hias

micro_brain · Post by **micro_brain** » Thu Apr 07, 2011 3:12 pm

Am looking at a Tektronics 2213 - should be OK I think, but it is only 60Mhz..

HiassofT · Post by **HiassofT** » Thu Apr 07, 2011 5:07 pm

micro_brain wrote:

Am looking at a Tektronics 2213 - should be OK I think, but it is only 60Mhz..

Yes, this should be OK too.

These old analog Teks perform a lot better than most of the new, cheap digital scopes. Digital scopes might be nice to analyze aperiodic and/or single-shot events, but when doing microprocessor stuff the 2 or 4 channels of a scope are quite limiting and you'd be better off using a logic analyzer (with 8-32 channels) instead.

So: a nice analog scope and maybe a logic analyzer some time later is a very good idea IMO.

BTW: be sure to get X10 probes with your scope, X1 probes are only useful if you want to do millivolt low-frequency audio stuff (again, IMO).

so long,

Hias

GARTHWILSON · Post by **GARTHWILSON** » Thu Apr 07, 2011 5:23 pm

For this kind of troubleshooting, I've used 20MHz dual-trace triggered 'scopes at home and they have been adequate for that (although for a few things I'd like to have something much faster). If you buy a new one, I expect those start at 40MHz or higher today, with a cheaper price than I paid 20 years ago. It is very important that you get a pair of good x10 probes though, which are not cheap if you have to buy them separately. x1 probes will turn a 100MHz 'scope into a 2MHz 'scope!

micro_brain · Post by **micro_brain** » Thu Apr 07, 2011 7:28 pm

Confused. x1? x10? I thought the probes were just wires to connect to the scope's inputs with appropriate connectors. This one comes with two sets (one for each channel), although a tip is missing from one of them Even so, it looks like pretty good value at (currently) £11.50 (+£20 postage). Let's hope no one else bids on it!

micro_brain · Post by **micro_brain** » Thu Apr 07, 2011 8:03 pm

What do you make of this:

http://cgi.ebay.co.uk/ARM-DSO-Nano-Pock ... 2933wt_824

Probably a waste of money but I thought I'd ask..

GARTHWILSON · Post by **GARTHWILSON** » Thu Apr 07, 2011 8:14 pm

In the x1 setting, the circuit is heavily loaded by the capacitance and transmission-line effects of the probe cable. These not only give invalid readings but also upset the operation of the circuit you're trying to measure. Correcting the problems without introducing other problems requires quite a science in making good probes. There's a very short primer at http://www.probemaster.com/pages.php?pID=14&CDpath=4 .

HiassofT · Post by **HiassofT** » Thu Apr 07, 2011 8:28 pm

micro_brain wrote:

http://cgi.ebay.co.uk/ARM-DSO-Nano-Pocket-Sized-Digital-Oscilloscope-New-/280629466420?pt=UK_BOI_Electrical_Test_Measurement_Equipment_ET&hash=item4156d1d534#ht_2933wt_824

Probably a waste of money but I thought I'd ask..

I'd say this is a cheap toy. Max sample rate of 1MS roughly translates to 100kHz bandwidth. This may be fine for some audio stuff, but definitely not digital microprocessor electronics...

so long,

Hias

GARTHWILSON · Post by **GARTHWILSON** » Thu Apr 07, 2011 8:37 pm

micro_brain wrote:

What do you make of this:

http://cgi.ebay.co.uk/ARM-DSO-Nano-Pock ... 2933wt_824

Probably a waste of money but I thought I'd ask..

It's only 1MHz even in analog. As a 1MSPS DSO, it will be less than half that much. The bandwidth should be a minimum of five or ten times as much as the highest-frequency square wave you want to watch. Just the probes alone would cost about that much for what you want to do. Remember that to show a non-sine wave, it has to be good for the relevant harmonics. A perfect square wave has an infinite number of odd harmonics, ie, sine waves of frequency multiples summed in the right amplitudes and phases to add up to a square wave. As you get to higher and higher harmonics, their amplitudes decrease enough that the limitations of affordable equipment are acceptable; but you do still need the bandwidth to be at least five to ten times the frequency of the square wave. There's a pretty good graphical representation here

micro_brain · Post by **micro_brain** » Thu Apr 07, 2011 9:04 pm

I thought it would be pretty useless! Only noticed the 1Mhz rating after I posted. So I guess my best hope is going to be this 60Mhz Tektronics scope. Fingers crossed...