UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Computers

Notices

Vintage Computers Any vintage computer systems, calculators, video games etc., but with an emphasis on 1980s and earlier equipment.

Reply
 
Thread Tools
Old 18th Jan 2021, 9:09 pm   #161
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

Quote:
Originally Posted by Vintage_RC View Post
Quote:
Originally Posted by ScottishColin View Post
Hi Mark - how do I do that please.

Both; I'm looking at this - do either of you have any experience with one of these cheaper scopes? Thanks.

https://www.amazon.co.uk/DSO-Shell-O...&rps=1&sr=8-19
I have one of these scopes, quite useful subject to its limitations. The most severe limitation is the upper frequency limit of 200KHz which would rule it out for most of the measurements you will need to do.
Thanks - I'm off to look for a 20Mhz scope now.
ScottishColin is offline   Reply With Quote
Old 18th Jan 2021, 9:53 pm   #162
Mark1960
Hexode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 285
Default Re: Non-working Commodore PET 3016

Quote:
Originally Posted by ScottishColin View Post
With regard to pins 9/10/11 of the 6502, please can you tell me where the black probe should be?
Black on pin 1 or 21 of the 6502,

Then measure frequency with red on pin 9 of 6502 with board powered on.
Mark1960 is online now   Reply With Quote
Old 19th Jan 2021, 12:07 am   #163
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

For the measurements on the 6502 pins the black probe can go on any known 0V point but earlier, Mark did single out pins 1 or 21 as 0V pins on the 6502 (post #148).

The resistance readings you are getting on the new meter are completely weird. When you have the IC in the breadboard do you have it in the centre of the slab with the legs on either side of the centreline gap? Putting it into the breadboard anywhere else or with the body oriented east-west will join IC legs together.

The measurements with the old meter on the chip only sound completely sensible - infinite reading when on 200Ω just means that the resistance is in excess of 200Ω and when you rightly went up-range you discovered that the real reading was in the MΩ range, which is to be expected on a healthy chip.

Given the above, from now on please use the old meter ONLY when making resistance measurements, you can use either of them for voltage measurements and of course the new meter for frequency measurements, although there is still a question mark, Mark, over why It can't 'see' the 1Mhz out of UG5 pin 7? Let's see if it can see the lower frequencies we expect on 6502 pins 9, 10, 11.

Another offhand question for you Colin, do you have some loose resistors of various values?

..And crossed with Mark again.
SiriusHardware is online now   Reply With Quote
Old 19th Jan 2021, 4:17 pm   #164
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

Right. I'm frustrating myself now, so I will go a bit slower this end. As you correctly surmised, I put the IC in the wrong place on the breadboard and took measurements there. I will do that again, along with the measurements in post #162 from Mark and come back to you.

I have ordered a scope and will let you know when it arrives. Let's hope I don;t have to pay extra taxes on it when it does...

I have plenty of resistors lying around of lots of values.
ScottishColin is offline   Reply With Quote
Old 19th Jan 2021, 4:38 pm   #165
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

Well, one thing you can do to while away the time is to try measuring the values of some of your resistors with the new meter to see what values it says they are.

Brief aside: On my first ever day at work someone set about trying to teach me the resistor colour code. I said I already knew it (which was true). My job for the rest of the first day was sorting trays full of hundreds of jumbled up resistors into the correct drawers...
SiriusHardware is online now   Reply With Quote
Old 19th Jan 2021, 5:36 pm   #166
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

That'll teach you. Luckily mine are in marked bags ranging from 1Ω to 1MΩ and all points in between.

In terms of readings, I can get no frequency readings from the 6502 pins 9,10 and 11.

I have performed a powered-off continuity check from pin 1 to pin 21 and confirm continuity and also a powered off continuity check from pin 1 and 21 to the chassis and get continuity so the earth seems OK to me.

Resistance reading between UG5 pin 7 and UG5 pin 8 with the IC out of the PET and not even on a breadboard with the old meter on 200 ohm range is 1 (I think that means infinity - am I right with that).

With the meter set at 2K, I get a pretty solid reading of 0.76. With the meter set at 2M, I get a pretty solid 1.4

Would it be possible to explain to me why I'm getting differing readings, depending on which selection I make with the dial on my old meter? I had expected the same.

Thanks.
ScottishColin is offline   Reply With Quote
Old 19th Jan 2021, 6:03 pm   #167
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

Ohm Meters work by passing current through the item or circuit being measured. The meter probably uses different amounts of current on each range - If you try measuring a 'linear' item like an 18K resistor you should get about 18K whether you measure it on the 20K range or the 200K range. On the 20K range the reading will be at a higher resolution - more digits after the decimal point.

On a 'non-linear' item like a semiconductor device or junction the measured resistance will very likely vary from range to range. You will find that if you measure between two IC pins one way around and then measure between the same two pins on the same range with the probes the other way around, you will see a different resistance even then. That is normal, and nothing to worry about.
SiriusHardware is online now   Reply With Quote
Old 19th Jan 2021, 6:14 pm   #168
Mark1960
Hexode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 285
Default Re: Non-working Commodore PET 3016

What you are measuring is the parasitic diodes in the IC, this is how the transistors in the IC are isolated from the silicon substrate by reverse bias pn junctions which is why you have to keep the input and output voltages on an IC between the supply voltage rails, otherwise you can damage those pn junctions.

When you change the range switch on your multimeter, it changes the current supplied to the device under test, then measures the voltage across the device. The different resistance measured on each range is because the diode is is not linear, at different current the meter will convert the different voltage across the diode to a different resistance value.

Your meter probably has a diode switch position, this is just so you have a voltage above 0.6v to drive the diode past the switch on point, useful for checking polarity of diodes with unknown polarity markings.

Edit:this time Sirius beat me
Mark1960 is online now   Reply With Quote
Old 19th Jan 2021, 6:51 pm   #169
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

Picking up on something you mentioned earlier, Colin, can you try this for me:

Take two resistors, say both 1K or both 10K and connect them in series (one above the other, inline) between +5V and 0V on the PET PCB. Connect the black lead of the meter to the centre connection between the resistors, and then use the red lead to look around for 1MHz on UG5 pin 7 and on the 6502 pins 9, 10, 11.

The hypothesis here is that maybe the frequency-meter feature on the meter only works well with AC input, in which case the input voltage on the red lead needs to swing Plus and Minus relative to the black lead.

The waveforms we are asking you to look at are pulse-DC, alternating between 0V and +5V, and maybe the meter does not like that.

By connecting two resistors as described you are making a half-supply-voltage point (2.5V) to connect the meter black lead to. From the meter's point of view the clock and data lines will then look as though they are swinging between -2.5V and +2.5V rather than 0V and +5V.

It's not difficult to try and if this dodge allows the meter to work reliably that will be a good step forward.

As to why the meter does manage to measure 2MHz and upwards, I think that at those high frequencies maybe the waveforms are overshooting just enough to provide a small below-zero spike for the meter to work with.

Attached, a quick sketch of what I'm asking you to do.
Attached Thumbnails
Click image for larger version

Name:	Frequency.jpg
Views:	40
Size:	18.0 KB
ID:	224952  
SiriusHardware is online now   Reply With Quote
Old 20th Jan 2021, 2:06 pm   #170
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

About to get onto this - is there a good place I should get +5V from?
ScottishColin is offline   Reply With Quote
Old 20th Jan 2021, 3:31 pm   #171
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

The unpopulated pads of UG4 are a good place to pick up both +5V and 0V. Clear the indicated pads, solder one end of each resistor into the +5V pad and 0V pad respectively and then twist or solder the other ends of the resistors together to make your centre connection.
Attached Thumbnails
Click image for larger version

Name:	Pet_UG5_etc.jpg
Views:	29
Size:	83.8 KB
ID:	225007  
SiriusHardware is online now   Reply With Quote
Old 20th Jan 2021, 4:06 pm   #172
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

Once you have that, system power on, meter in frequency mode, black lead on centre connection between the two resistors and red lead on UG5 pin 7. Do you see 1MHz?
SiriusHardware is online now   Reply With Quote
Old 21st Jan 2021, 1:15 pm   #173
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

re #172

This feels like the first positive thing I've replied with for some time.....yes, I do get a nice solid 1.000Mhz @ 47% reading.

Couple of things:

1. I took readings from pins 8,9 and 16 on UG4 before I soldered the resistors in and got the following:
8 0.32V
9 0V
16 5.35V

I used pin 9 as it was 0V

Why might pins 8 and 9 be different?

2. I found out that I can only get a 1Mhz reading as above after I tap the red and black probes together, otherwise I get no reading. Is this just best practice with a meter anyway?

Thanks. feels like a step forwards.

Colin.
ScottishColin is offline   Reply With Quote
Old 21st Jan 2021, 1:32 pm   #174
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

Quote:
Originally Posted by ScottishColin View Post
re #172
This feels like the first positive thing I've replied with for some time.....yes, I do get a nice solid 1.000Mhz @ 47% reading.
That is good news. The fault you seemed to have, with 8MHz, 4Mhz and 2MHz all present but no 1MHz out even after replacing the chip just seemed impossible and then your confirmation that you had ~2.5V (not 0V or 5V) on pin 7 seemed to suggest that the 1MHz was really there after all.

As to your question about pin 9 - looking at your photos of both sides of the PCB earlier I can see that there is no connection to the pin 9 pad of UG4 so it's not surprising that you read 0V there - there is no voltage on that pad for your meter to read.

What this means is that you actually have the 'bottom' end of your resistor divider going to the wrong place. Move it from the pin 9 pad to the pin 8 pad, and then it will be connected to 0V and you may find you no longer have to do the 'probe short trick' to get the meter to read.

There is a fine distinction to be made between a point which has no voltage on it (pin 9) and a point which is connected to 0V (pin 8). What you need in this case is the one which is connected to a known 0V (that's why I asked you to use that pad, and not pin 9).
SiriusHardware is online now   Reply With Quote
Old 21st Jan 2021, 1:46 pm   #175
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

Ah - I had assumed pins 8 and 9 were connected.

I will move it to pin 8 and retry.
ScottishColin is offline   Reply With Quote
Old 21st Jan 2021, 1:56 pm   #176
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

Moved to pin 8 and a nice consistent 1.000Mhz @48% without having to touch the probes together as you thought.

Thanks.
ScottishColin is offline   Reply With Quote
Old 21st Jan 2021, 2:31 pm   #177
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

That's excellent, so from now on when we ask you to measure frequency always put the black lead on that centre point of those two resistors, rather than on a known 0V. When we ask you to measure voltages, still put the black lead on a known 0V point such as UG4 pin 8, where the leg of the 'bottom' resistor now goes to.

I'm sorry it has taken so long to get to this point but I have used quite a few meters which had a 'frequency' function and all of them were as happy to read the frequency of a pulse-DC signal as they were to read the frequency of an AC signal. At least we've finally worked out that your meter only really likes AC input signals when measuring frequency and we now have a workaround for that.

Moving on, can you now do Mark's suggested check for any frequencies coming out of the 6502 pins 9, 10, and 11?
SiriusHardware is online now   Reply With Quote
Old 21st Jan 2021, 2:46 pm   #178
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

There's absolutely no need for you to be sorry. I'm very grateful for your help.

I've checked the frequencies and have the following:
Pin 9 - meter stays at 0Hz
Pin 10 - meter stays at 0Hz
Pin 11 - meter stays at 0Hz

Colin.
ScottishColin is offline   Reply With Quote
Old 21st Jan 2021, 2:56 pm   #179
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

OK, still with power on, black lead to resistor centre point, meter in frequency mode, can you see if you have frequencies on the 6502 pins 37, 3, and 39?
SiriusHardware is online now   Reply With Quote
Old 21st Jan 2021, 3:06 pm   #180
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

I get 1.000Mhz on each of those pins at between 44%-48%.
ScottishColin is offline   Reply With Quote
Reply

Thread Tools



All times are GMT. The time now is 9:11 am.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2021, vBulletin Solutions, Inc.
Copyright ©2002 - 2021, Paul Stenning.