Oscilloscope pong for 1 or 2 players.

[FrankBuss]

I thought a polyfuse is too modern. Looks like commercial polyfuses were available after 1981. Every other part so far was available before I was born

[MK14]

I thought a polyfuse is too modern. Looks like commercial polyfuses were available after 1981. Every other part so far was available before I was born

There were resetable fuses before 1981, using magnetic, thermal and maybe other methods. So it was sort of possible.
Some early AVO multimeters had a cut-out button, using that method.
The Polyfuse adds the concept of "automatically reseting", once it cools down, but otherwise is similar.

I see it more as swapping metal canned Germanium transistors for Black plastic Silicon transistors, than a bad breach of replicas.

But strictly speaking, you are right, they weren't (in that form), available in the old days (1970s).
Unless there were thermistor equivalents ?

[BrianHG]

Tonight I got the power supply section done:

Don't forget the V+&- test points by the trim-pots.

[BrianHG]

Speaking of supply etc., has any effort yet been put into BOM minmisation?

Do you mean the last 2 pages on timb's schematic here:

Here's the updated schematic:

http://timb.us/PDF/Scope_Pong_20170406.pdf

Edit: Fixed link to schematic.

[timb]

Speaking of supply etc., has any effort yet been put into BOM minmisation?

Do you mean the last 2 pages on timb's schematic here:

Here's the updated schematic:

http://timb.us/PDF/Scope_Pong_20170406.pdf

Edit: Fixed link to schematic.

BOM minimization means trimming unneeded components or replacing them with components already on the BOM. (I.e., if your BOM has, say, ten 100k resistors and one 200k resistor, you could replace the 200k part with two 100k parts, that way you don't need another BOM item, stuff like that.)

To answer Mike's question: No, not yet. That's generally something I don't do on a first pass anyway.

[mikeselectricstuff]

Speaking of supply etc., has any effort yet been put into BOM minmisation?

Do you mean the last 2 pages on timb's schematic here:

Here's the updated schematic:

http://timb.us/PDF/Scope_Pong_20170406.pdf

Edit: Fixed link to schematic.

Seems like there may be some scope for optimisation - just looking at the BOM list
33 different resistor values
100uf 16v and 25v
two different types of 10u 25V

[timb]

Tonight I got the power supply section done:

Don't forget the V+&- test points by the trim-pots.

Yeah, I've got that on my list of things to add to the power supply section tonight, along with fixing the trim pot connections, adding a connection to the transformer center tap, adding polyfuses, adding additional decoupling caps around the entire board, looking into an on-off switch, changing the LED resistors to 1K (glad Mike caught that, I meant to calculate the actual value but never got around to it) and a couple of other small changes.

[timb]

Speaking of supply etc., has any effort yet been put into BOM minmisation?

Do you mean the last 2 pages on timb's schematic here:

Here's the updated schematic:

http://timb.us/PDF/Scope_Pong_20170406.pdf

Edit: Fixed link to schematic.

Seems like there may be some scope for optimisation - just looking at the BOM list
33 different resistor values
100uf 16v and 25v
two different types of 10u 25V

One 10u is a bi-polar AE Cap, the other is just a standard polarized one used in several places. I could put two standard ones back to back I guess, but I never liked doing that.

The 100uF 16V is Tantalum and *has* to be Tantalum because it's used in an Integrator, so low leakage is needed. The 25V one is used in two places on the power supply: 1) For ripple reduction on the base bias resistors; 2) As bulk decoupling on the output.  Now, I *could* up the 16V Tantalum to 25V and use it all three places, but it would be a lot more expensive. The Tantalum is like $1 vs $0.33 for the AE cap, so I think it's better off having two separate types.

Edit: The 100u cap on the base bias resistors could be reduced to 10u, however output ripple will increase from about 500uV to around 5mv (at a 200mA load).

The output cap may be able to be reduced as well, as we'll have additional (optional) bulk decoupling down the line, however I specced 100uF to be safe until I test an actual finished board.

(By the way, when you look at my schematic, if you zoom in you can see a label to the top right of any non-ceramic cap that tells you what it is; AE for Aluminum Electrolytic, Tant for Tantalum, Film for Metalized Polypropylene Film, etc.)

[timb]

You know, it just occurred to me... I wonder if there would be any issues placing the LED before the 1K resistor that biases the Zener diode in the power supply circuit? Drop the resistor down to 800 Ohms or so and boom. There's plenty of headroom and it seems to work fine in basic simulation.

The idea being were already draining a constant 8mA biasing the Zener, so why not put the power supply indicator LED in-line with the biasing resistor and put that current drain to use. Two birds with one stone.

Edit: Scratch that. Under some situations this can cause a weird oscillation loop during startup, preventing it from reaching the full output voltage. I suspect the voltage drop of the LED causes this. It might be fixable by adjusting the position of the LED in the circuit, but I'll stick with what works.

[timb]

Okay, I added changes to the power supply section:



http://timb.us/PDF/Scope_Pong_Power_V2.pdf

The power switch and transformer connector will be moved a bit further from the regulators when placed in the final location on the board, but otherwise I incorporated all the changes described earlier.

Updated Schematic:

http://timb.us/PDF/Scope_Pong_20170407.pdf

[calli]

Just reply to say that is quite impressive crazy!

 

I am doing some stuff with micros controlling beams but this now seems quite lame after seeing yours.

Carsten

[GK]

He he he............... just wait 'till you see the raster video version............ Am soldering lots of MPSH10 for the digital parts now (BC550C too slow here)...........

[BrianHG]

Will you be adjusting the ball vertical speed & direction depending on where it is hit on the paddle?

[GK]

Will you be adjusting the ball vertical speed & direction depending on where it is hit on the paddle?

The current plan is to differentiate the paddle vertical control signals from the controller potentiometers and sample-and-hold the output at the moment of collision between ball and paddle; the DC potential at the output of the S&H being the balls velocity vector. This way the speed of the ball is determined by how fast the paddle is moving vertically at the moment of impact.

 

[BrianHG]

Don't forget a 50hz/60hz switch.  With this, since there is no color, the mono video would work with most TV and video capture cards around the world.  Only the capture cards with super-strict H&V timing, ie they wont support VCR playback in some cases, will fail to lock onto your video.

[timb]

Today I finished the Function Generator section and started tying together the top board:





http://timb.us/PDF/Scope_Pong_Function_Generator.pdf

Here's a low resolution render (there's so many parts it literally takes 10 minutes to load the 3D view in high quality) of how the top board is coming together. (I'm still positioning all the elements together, but it's a good start!)

[GK]

Looks good. I wonder if you could fit the audio power amplifier in there. A good place for it would be directly in front of the power supply, on its own partitioned ground plane section, star-earthed at the PSU common. That way you won't have the heavy speaker currents (up to several hundred mA) pumping through your board interconnects and ground plane.

[BrianHG]

Timb, very nice.  1 thing you should do is add thermal relief for the GND flood fill.  For hand made boards, it's hard to soldier or repair solid copper fill + for the diodes, it helps to see pad #1, the square pad, for orientation if it is inside the fill.

[BrianHG]

And who is Brian?

I'm the evil overseer 

[timb]

Looks good. I wonder if you could fit the audio power amplifier in there. A good place for it would be directly in front of the power supply, on its own partitioned ground plane section, star-earthed at the PSU common. That way you won't have the heavy speaker currents (up to several hundred mA) pumping through your board interconnects and ground plane.

Yeah, that's a good idea! I think I can fit it behind the power supply. The sound generation circuits might not fit, but the amplifier should no problem.

[timb]

Timb, very nice.  1 thing you should do is add thermal relief for the GND flood fill.  For hand made boards, it's hard to soldier or repair solid copper fill + for the diodes, it helps to see pad #1, the square pad, for orientation if it is inside the fill.

I assume you mean adding thermal reliefs to the top and bottom copper pours? Yeah, I can do that. I just need to go in and switch the top and bottom pours from direct to 4-sided thermals.

I was debating on if I should go direct or 4-sided; I've always preferred direct when possible, as I feel the pads have much, much less, of a chance of lifting during rework. I guess I could always turn thermals on for one set of Gerbers (for homemade boards) and off for another (for professional boards with soldermask). Hmmm...

The only thing that won't have thermals are the solder pins for the TO-220 heatsinks. You'll need a decent iron for that (or a heat gun/hairdryer/hot air pencil to pre-heat the heatsinks while you solder the pins; I've made the pads massively oversized to take the extend heat needed during soldering).

Can you elaborate on your second point about square pads being easier to see inside a fill? I'm not quite sure I understand.

Edit: Never mind, I understand. Home etched boards won't have soldemask to mask out the shape of pads directly connected to a copper pour; if you use thermals on these planes, the shape of the pads will be masked out due to the copper being pulled back around them. Gotcha.

[tautech]

Square pad for diodes signifies cathode. For homemade PCB's I like to also add a small K symbol in the copper.

[timb]

Square pad for diodes signifies cathode. For homemade PCB's I like to also add a small K symbol in the copper.

Yeah, I get that and I'm already using square pads. I think I understand what BrianHG means now: A homemade PCB won't have soldermask, so if a particular pad connects to the ground plane, it'll just be a hole in the solid copper plane (that is, the shape of the pad won't be visible because there's no soldermask to, uh, mask it out). However, if you use thermal reliefs on all pads connecting to the plane, the shape of the pad *will* be visible, because the copper is backed off around the pad (except for the 2 or 4 traces that connect it to the plane).

[bitseeker]

Very good point. I get it now too.

[timb]

Very good point. I get it now too.

Though, when you think about it, does it really make a difference? After all, it's not like both pads of a diode will be connected to a copper pour, right? So you *will* see the shape of one of the two pads; if it's round you know the other is square and vice versa.

That said, the thermals *do* make it easier to solder.