• Please review our updated Terms and Rules here

Cbm 8032 capacitor

I notice the picture shows the board has tantalum capacitors. These are notorious for going short circuit, sometimes even explosively!
One interesting thing is that if a Tant capacitor shorted out to a very low resistance, it would not heat up much at all and something else would, like the tracks leading to it.

The maximum heat that can be evolved in the leaking Tant capacitor occurs when its resistance has dropped to a similar value as the internal resistance of the power supply feeding it. This would occur say with a 12V supply, where the leakage current from the Tant had forced the supply voltage output down (and the voltage across the Tant's terminals) to around 6V and then the Tant would really be smoking.
 
These large-value caps are frequently beyond the range of many capacitance meters. But you can still test them "manually" by measuring the time required to discharge the cap through a known resistance. I used this method to measure the value of the 95,000uF 10V part in my Altair 8800, and it worked just fine - the cap was found to be close to 120,000uF, on the upper end of the +20% tolerance.
 
These large-value caps are frequently beyond the range of many capacitance meters. But you can still test them "manually" by measuring the time required to discharge the cap through a known resistance. I used this method to measure the value of the 95,000uF 10V part in my Altair 8800, and it worked just fine - the cap was found to be close to 120,000uF, on the upper end of the +20% tolerance.
That works.

Another way is to scope the voltage across the capacitor and see if it troughs down between the ripple peaks to around 7 to 7.5 v or below, it is indicative that it has lost capacity (unless the current consumption was much higher than it should be). That is what happened to mine, in my PET an ripple appeared in the 5V output. Even like this the ESR was extremely low and as you say, they are outside the range of most capacitance meters.The voltage regulators generally need about 2 to 2.5v minimum difference between their input and output voltages because they are not LDO types. If the ripple does appear at the regulator's output, it significantly upsets the VDU's vertical scan.

One interesting thing about these very large caps, when they have been discharged to zero with a bleeder resistor, even overnight, if you take the resistor off and look with the DVM you will see their terminal voltage start to climb, it is a version of dielectric absorption and caused by their chemical electrolyte nature. So it pays to store these big ones with the terminals insulated or a bleeder resistor across the terminals.

These massive uF value capacitors have become more and more rare now. I collected some spares for my SOL-20. These days because of high frequency switching supplies, there is very little need for them and they are not a common new part, especially the large computer grade ones over 50,000 to 100,000 uF.

22,000uF ones for the PET are easy to get.

I have a question about the PET 64K memory expansion board, and it says in what I have read about this that in the 32k PET you end up with 96k, but the boot up still reports 32k (or 31k available).

With the address range available, how is it possible to address that much memory , how does that system work ?
 
Hello Hugo

I'm not exactly sure how it's done with the PET expansion module, but there aren't really many options available. So they very likely use some variant of block mapping or moving as with the (C64) GEORAM or REU carts.

GEORAM takes up a 2-byte register and a 256-byte block in the main address space. The register contents can be set to select which 256-byte chunk of GEORAM is currently mapped into the addressable 256-byte space at any given time.

The situation with REU is similar in that its RAM is not directly accessible, though I believe it transfers & exchanges larger blocks between the expansion RAM and main RAM rather than just remapping 256-byte chunks. Don't quote me on that, though. Consulting one of the REU owner's manual would probably clear matters up.
 
Last edited:
You might want to checkout the manuals at: http://www.zimmers.net/anonftp/pub/cbm/manuals/pet/MemExpansion/index.html.

(Assuming this is the correct card of course).

The standard PET ROMs know nothing of this card, so still report 32K on start-up.

You need to use additional software, POKEs etc. to drive the card to make use of the additional memory.

Basically, the screen, I/O and ROMs can be overlaid with RAM.

Dave
 
And some people told me that the 0.1v at the capacitor can be caused by the resistors and regulator on the board
 
And some people told me that the 0.1v at the capacitor can be caused by the resistors and regulator on the board
You could listen to what "some people" told you, or you could try to figure it out for yourself.

To do this requires a very basic understanding of transformers, rectifiers, power supply filter capacitors and voltage regulator devices.

It is not all that hard, you simply have to read around the topic.

Clearly if the power supply has failed, the regulator failed, no voltage will be delivered to the pcb, or if there is an open circuit or high resistance connection, little voltage would be measured at the pcb. But also, if there is a short on the board, the voltage regulator will go into shut down mode and there will be a low voltage that way.

Generally a simple volt meter is likely all you will need. In some cases it might be helpful to measure resistance or the load current. Most meters these days are "multimeters" and will measure voltage current and resistance.

But of course to make any sense of the readings from the meter you have to understand what voltage, current and resistance are.

Voltage is an electric force field which does work on an electron current, and pushes that current around a circuit. (It is somewhat analogous to pressure which can force a flow of fluid through a pipe). In the old days it was called "electromotive force" or emf. The units of it are Joules per Coulomb (a coulomb is a large number of electrons or electric charges). Charge generally has the symbol Q. In other words this field imparts energy (joules) to the electrons.

So if one Coulomb of electrons is pushed past two terminals of a voltage source of one volt, the work done is one Joule. The work done is always Volts x Charge, or VxQ.

In electronics & electricity though, we are interested in the current or Coulombs per second, passing by some point in a circuit, we call this current I.

In any circuit, there are losses, as current flows through wires, connectors or other devices, mainly these losses occur in the circuit resistances. You can think of resistances as a device which wastes energy as heat. The units are called "Ohm's" but this is shorthand for V/I which has units of (Joules/Coulomb)/(Coulomb/second), which is Joule.Seconds/Coulomb^2

Therefore there are a set of equations which link everything together for circuits where direct currents are flowing , it is called Ohm's law, in its basic form is V=IR (voltage = current x resistance)

The power loss (heat dissipation in the resistance R) can be defined as I^2R or V^2/R, in all cases the units of power are Joules/second if you check. We call Joules per second Watts. 746 Watts is said to be one Horsepower.

Rectifier diodes for practical purposes only conduct current in one direction.

Capacitors are devices which store charge. The equation for the capacitor is Q=CV, where Q is the charge in coulombs stored in the capacitor, V is the voltage across its terminals an C is the capacitor's value in Farads. A Farad is a very large value capacitor so most of the capacitors you meet are in some number of micro-farads or uF, which is 1 millionth of a Farad.

In general, for many vintage computer power supplies, a transformer is used. Silicon rectifier diodes connected to the transformer's secondary windings, conduct on the peaks of the AC waveform to charge a large value capacitor. The voltage across the capacitor's terminals is fairly stable between peaks, but sags down a little as the capacitor discharges. This voltage is then often sent to a voltage regulator device, which maintains a stable output voltage of 5V, even if its input voltage fluctuates. Most of these voltage regulator devices, will shut down though, if their outputs get shorted out into a low Ohmic resistance, for example this could be a shorted tantalum capacitor on the pcb.
 
Last edited:
First things first.

Have you located the schematics and component layout for your particular machine yet?

Without these, you will get nowhere.

Second, can you update your profile with your location, so we can see (roughly) where you are in the world (if you are able to).

Dave
 
I have a commodore pet 8032-32b and i know there is schematics
And for my location im in france
 
If there remains little voltage across the large capacitor terminals there are only a few things it could be:

There is a break in the wiring connections from the transformer's secondary windings, could be the connector where they pass to the pcb where the diodes are.

The diodes could be open circuit.

The transformer is not energized (no line voltage on the primary) or fuse blown.

The transformer has gone open circuit.

The large capacitor has shorted out and also if there would have to something open circuit too or there would be a lot of heat evolved.

combinations of the above problems.

The thing to do is to use the meter first set to measure AC volts and check that there is the correct voltage appearing on the secondary of the transformer at the connector pin 1 and pin 5 with respect to ground , pin 3 or pin 6, and check that this is making it to the in the bridge CR2 diodes on the main board via the connector there. There should be 9 to 10V DC voltage at the diode outputs on pin 4.

The transformer secondary is center tapped (connected to gnd) and feeds to pins 1 & 5 of the connector J8. This is so it can produce a + and - power supply. For the positive supply that charges the large off board filter capacitor, the positive output of two of the diodes in the bridge CR2 connects to pin 4 of the connector. And the big capacitor is shown connect + to pin 2 and - to pin 6 (ground). Therefore, for the capacitor to be connected it relies on pin 4 and pin 2 of the connector, to be connected together somehow, this could be at the capacitor's + screw connection with wires leading to pin 4 and pin 2, or possibly via the connector pins and the pcb tracks at the connector J8, all worth checking. For some reason I cannot see the link between pin 2 and pin 4 drawn on the zimmers schematics.

Page 4 of this Pet Shack document shows how pin 2 and pin 4 get connected, It is at the capacitor's + terminal:

 
Last edited:
The bulk of the schematics for your machine are here: http://www.zimmers.net/anonftp/pub/cbm/schematics/computers/pet/8032/index.html.

However, you MAY have a different logic board to what is in that directory...

Checkout the assembly drawing at: http://www.zimmers.net/anonftp/pub/cbm/schematics/computers/pet/8032/8032030-6.gif to make sure your board is the same as the drawing.

If it is, as Hugo has stated, the first thing is to check the AC voltage from the secondary of the transformer. I would disconnect the transformer from the PET logic board for this test by unplugging the connector that has the cables from the transformer and external capacitor. This is identified as J8 on the assembly drawing above.

Next, look at the schematic for this connector at: http://www.zimmers.net/anonftp/pub/cbm/schematics/computers/pet/8032/8032051-3.gif

You need to set your multimeter to read AC volts (say 20V full scale) and measure the voltage between the following pins of the connector:

8 and 9.
1 and 3.
3 and 5.
1 and 5.

Report the voltage readings back.

The pins are numbered from 1 to 9, but pin 7 is missing - so this should give you a clue as to which end is pin 1 and which end is pin 9...

Try this first then we will see where this leads us next.

Dave
 
I have a clue the previous owner was selling it for parts and told me it worked one time and then stopped working after the first time he power it up
 
That, I am afraid, is how all of the faults behave!

The machine works, and then it doesn't...

Was the failure accompanied by a crackling sound, a burning smell or a small amount of smoke I wonder? This would be the most useful of information.

You need to follow a logical path through the machine from the input (the AC mains power), through to the low voltage DC rails and then work your way through the logic.

What test equipment do you have?

I see you have had a lot of information in the previous posts, so you need to stop guessing, roll up your shirt sleeves, and start to investigate the issue with your test equipment.

Dave
 
Last edited:
Thanks for all your help i will start sooner as i can i jave a good mutimeter and i know how to solder so all i need to do is start investigate
 
Absolutely no problem! Your English is much better than my French!

I have two (2) multimeters. Both of them are quite cheap to be honest. I have both a modern digital and an old analogue multimeter. I find it easier to use the analogue multimeter for some jobs and the digital for other jobs.

I can purchase a 'reasonable' multimeter for less than 10 GBP each in the UK.

You may also find a logic probe an invaluable tool to have.

Don't worry about soldering just yet...

Dave
 
Last edited:
Back
Top