Jump to content

Recommended Posts

Posted

Hello everyone,

Have been in the process of building my midibox SID for years now, but never seem to be able to get
around it. Always concerned that once its built, it could be better.

So I've been looking and gathering information about the best projects for midibox SID Models,
different SID configurations, different front panel options, and most important, different power supplys.

Currently trying to figure out the best way to supply my SID circuit, and this seems to be the best 2 options ou there:

  • C64 transformer suplying a 9V linear regulator and a 5 Volt line directly from transformer to the SID Vcc pin
  • Using a 12V-15V power supply to supply both a 9V linear regulator, and a 5V buck converter in parallel.

Both of this options leave me a bit weary. Connecting the SID Vcc directly to the transformer would probably prevent me from
having a good night sleep ever again, but also I'm not sure how this buck converter works out when bad stuff happens.
I imagine that if the unlikely event of the Mosfet on the buck converter would blow up, it could act as a short circuit and
expose the Vcc rail to whatever input voltage is receiving(Please correct me if I'm wrong).

I've seen lots of Posts in the forum regarding this topic, and was wondering if
I could get an engineer's input on the following circuit(no pun intended).  

A Google search for 'discrete over voltage protection circuit' will reveal this link as one of its first results:

https://www.maximintegrated.com/en/design/technical-documents/app-notes/7/760.html

The article from Maxim describes how the circuit works in great detail, but here's an overview:
This simple protection circuit is composed by:
- 3 resistors
- 2 PNP transitors
- 1 Diode
- 1 Zener

 

The circuit provides a power supply line with adjustable over-voltage protection, meaning that
once the voltage at the input exceeds a certain level, the circuit will shut down its output power.
The maximum current can also be programmed to certain extent by carefully selecting the transistors bias resistors.
A simple reversed bias diode is used as a weak moderate protection against power line reverse voltages.

I've simulated the circuit and carefully selected the components so that they meet the 8580 Chip
Power limits specifications. I actually took this info from the 6581 datasheet, so not really sure
if the same values apply, but good enough to use as a case study.

6581 Vdd pin = 9V  25-40 mA
6581 Vcc pin = 5V  70-100 mA

I'm using 2 separate circuits in my simulation, one as a protection for the SID Vdd pin, the other for the Vcc pin.

Here is the outcome of the following tests:

 

Test 1: Ideal conditions - sim001_stable.jpg
I feed the expected voltages(in range) into the circuits, using some moderate loads (around 200 ohm) to pull some current  and everything works as expected.
The pass transistor Q2 passes the amount of current required by the 200 ohm load, but also imposes a small voltage drop on the line so this should be taken into consideration.

sim001_stable.thumb.jpg.1704f983d4bfa898

 

Test 2: Over current protection in action - sim002_heavy_load.jpg
I started to simulate faulty conditions. First I decreased the resistance of the loads progressively until I reached their smallest possible value (around 1 ohm), so that maximum current is pulled from the pass transistors Q2.
After the load resistances goes lower then a certain value, the current through the pass transistor Q2 is kept the same, and does not go higher as it is limited by resistors R2, R3 and transitor Q1 which controls the current flow in Q2.

sim002_heavy_load.thumb.jpg.48dd15dbe372

 

Test 3: Extreme limits near tip off - sim003_heavy_load_over_volt.jpg
I started raising both Voltages at the circuit inputs until near their tip off points.The load resistance level was kept the same, so that maximum current flow is expected.
In this scenario we can see the maximum current that the circuit will allow to pass, just before its tip off point.
For Vcc, when there's a slight over voltage condition, the maximum allowed current will be around 20 mA more then its rated max limit of 100mA.
For Vdd, when there's a slight over voltage condition, the maximum allowed current will be arounf 10 mA more then its rated max limit of 40mA.
This upper limits are carefully set as to provide some wiggle room for the IC current requirements.

sim003_heavy_load_over_volt.thumb.jpg.b4

 

Test 4: Over voltage protection in Action - sim004_over_volt_shutdown.jpg
I continued to raise the input voltages above their tip off points (6.2V for Vcc and 11V for Vdd).
Once the input voltage exceeds the tip off voltage, the circuit cuts off the pass transistors shutting down the power lines to the IC pins.
The transistor is kept off until the input voltage level is set lower then its tip off point.

sim004_over_volt_shutdown.thumb.jpg.b4c1

Although I seem confident about this circuit's performance, I'm not an engineer, so it would be really great if someone with more expertize

could give me their input on how great(or great disaster) would this circuits be as a power protection for the SID power pins.

Not so sure this circuit would provide a nice output impedance for the SID chip to work as naturally as possible, without any restraints.

What do you guys think?

 

Also, does anyone out there knows the Absolute Maximum Ratings for the 8580 chip?

Kind regards,

analog_mo

 

 

 

Posted

Isn't the transistor topology similar to a linear voltage regulator anyway? 

You might avoid the issue in the first place by starting with a more reliable power supply. Users here often also employ a crowbar circuit, so those might give you some possible ideas or directions.

For your simulation, consider that the transistors should be properly specced to handle the current loads, also real components might have tolerances or voltage drops that push your limits.

Posted

Hi Analog_mo,

I don't have enough expertise in this area to assess your circuit proposal, but can give you an alternative advice:

There is a new C64 PSU available which seems to be safer:

I use it with my MBSID meanwhile because of the same worries that you have...

Best Regards, Thorsten.

Posted

Hello everyone,

thank you for your replies.

 

@latigid on Regarding your initial question, the answer is Yes, this is a simmilar topology to the linear voltage regulator circuit, but it presents slight differences in its operation.

Without getting into many details, the presented circuit acts by cutting off the voltage/current instead of making an effort to regulate it.

You're also correct about considering the transistors used as the current passing devices. In the maxim article is noted that for the current pass transistors, the following characteristics are crucial for a successful operation of the circuit:

  • VceSAT as this will give us the voltage drop when the transistor is saturated;
  • The transistor speed will dictate the response speed of the circuit;

I'm now looking at crowbar circuit examples, and realizing the over voltage detection operation is simmilar with the one in the circuit I presented.

Some differences in how the over voltage situation is handled though.

 

My original idea was to apply a couple of this circuits for each SID, not to the power supply itself. Following this scheme I think I could overcome any 'voltage dropping' issue:sid8580_power_diagram.thumb.jpg.ed1f33ae

But I think I need to do some more research on the crowbar circuit as this might definetly be a more complete fail-safe solution.

 

Hello @TK. thank you for your input on this topic.

It is great that new C64 Power supplys are available, I currently only posess 1 left out of the 3 original C64s I owned, the remaining 2 died along the way =(

This is actually one of the reasons for my concerns.

I would still prefer to power my SIDs using the original C64 Power Supply I have left (I think the transformer itself is a beautifull piece and I like to keep it original)

IMG_20191221_180429.thumb.jpg.cf51d5a9f3

but because I've already seen 2 of them failing while operating the actual C64 circuitry, I wouldn't do this without any extra layer of protection, even with the new ones.

Also, I think having some flexibility by providing the option for a single voltage power supply is a big big plus, as it might lead the path for a MBSid battery pack(just a fetish).

 

So I guess I'll be digging into crowbars tonight.

Thanks for your help once again, I'll be posting whatever I come up with at the end.

If anyone out there could help me out with the answer for: what is the Absolute Maximum Ratings for the 8580 chip it would be great.

 

Kind regards,

Analog Mo

 

 

 

 

Posted

Hello @Zam thank you for your reply,

I think all the circuits described here go a few steps beyond the simple resistor/zener circuirt in what it comes to achieving over voltage protection.

Having a biased zener will only output constant voltage when the current is also constant, and this is not the hypothetical scenario were protecting our IC's from.

Most linear regulator IC's out there apply this type of circuit for voltage rectification, so having a linear Voltage Regulator IC implies having an improved "resistor/zener" circuit.

After researching a bit about crowbars and Switching Voltage Supplies I found this book by Martin Brown, "Practical Switching Power Supply Design" where is stated:

Quote

Semiconductor power transistors and MOSFETs always almost fail in the short-circuited condition when they do fail. This results in the input being short-circuited to the output load.(...) The designer must add an overvoltage crowbar circuit to the output of the power supply and a fuse in series with the input.

Also found this post about applying a crowbar circuit to the MB6502 power supply

I think this is a good overvoltage protection circuit to use as an example, and I will try to apply its topology using a 5V Switching Voltage Regulator.

Still need to get a better understanding of SCR specs though.

 

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...