Meteor
Stern Electronics, 1979Mark, when I connected the meter to the 24V output - it worked but the meter was bouncing around too much. Quench suggested it was because of the ripple for the zero crossing. He suggested connecting the sence wire after the diode. That worked but now the meter reads too high. I've tried adjusting the meter but it won't go down enough. It reads about 36V - I want it to read the output volts - 24 volts.
Bob
OK, that's where I thought this discussion was headed. Then you need to add a resistor and a capacitor on the sense line to smooth out the ripple. Doesn't affect the 24V output. Just need a time constant longer than the ripple. Got any 1/4W resistors around 10K and maybe a cap 20uF? Values aren't critical. Bigger is better.
I never thought a meter would be this complicated. I thought adding the 24V to 12V converter would solve my problems. Building the entire power supply wasn't this difficult.
Nah, not complicated...the meter is just struggling a bit since you are trying to read something it's not quite designed to measure. I edited my commit on what to try above.
The resistor and cap simply add a time constant much longer than the ripple frequency...smoothing it out and making the meter happy.
Quoted from mbwalker:Then you need to add a resistor and a capacitor on the sense line to smooth out the ripple. Doesn't affect the 24V output.
Where? If I add it before the diode (where I need to measure the voltage) won't it affect the ripple on the output (like the converter does)?
If I add it after the diode will it change the voltage reading on the meter?
So the right side goes to the meter. Come to think of it (brain drain since it late)...omit the diode and that side goes to the 24V. Sorry about that. I'll redraw.
What voltage is the cap going to charge to? Isn't it just going to cause the same problem with the sense voltage being too high?
Why not just use two resistors after the existing diode? Say a 4.7k and 2.2k with the 4.7k on the lower side? Bob your resistor schematic is correct.
Quoted from Quench:What voltage is the cap going to charge to?
Why not just use two resistors after the existing diode? Say a 4.7k and 2.2k with the 4.7k on the lower side? Bob your resistor schematic is correct.
I simply set 1 time constant to 200mS (10K and 20uF) which should be much longer than the ripple frequency. No issue with your method, I was simply headed towards giving the meter a filtered DC to measure. 
Plus I like to chime in at 2 AM half asleep and derail the postings. JK. Wife is out of town, so I don't have anything else to do....
Quoted from mbwalker:Plus I like to chime in at 2 AM half asleep and derail the postings. Wife is out of town, so I don't have anything else to do.
Thanks Mark. I got a big laugh out of that - something I needed after a long and "not productive" day in the shop. This morning when I received the meters and converters I thought I'd slap them in and be done. Boy was I wrong.
I found some 22uF electrolytic caps. I think I have some 10K resistors. (I gotta sort out my component collection).
The one thing I was trying to avoid was having a bunch of components just stuck in there (like my old power supply). Now it looks like I may have to add resistors, diodes, capacitors, etc. I thought I might just add those to a seperate little PC board. Then I thought I might as well omit the converter and add a small PC board with the diode, zener, resistor and filter cap that you both suggested back around post 229.
I'm going to have to think about this for a while.
Thanks
Bob
Seems to me (for what little I know), that adding the resistor and capacitor (without the diode) like Mark shows would have the same affect as the converter in smoothing out the zero crossing output.
If I go the 'diode, zener, resistor and filter cap' route, how would that be wired? What values should I use.
Thanks
Bob
Quoted from oldschoolbob:The one thing I was trying to avoid was having a bunch of components just stuck in there
Don't sweat it. The components will be out of sight with very little chance of you ever needing to open it up again once you're done.
Quoted from oldschoolbob:If I go the 'diode, zener, resistor and filter cap' route, how would that be wired? What values should I use.
This won't solve the problem with the meter showing unsteady readings because you'll still be sensing unregulated DC.
For the voltage sense, just add the resistor/cap or two series resistors as we mentioned.
Quoted from mbwalker:Wife is out of town, so I don't have anything else to do....
Doing your best to keep out of trouble hey? 
Quoted from Quench:What voltage is the cap going to charge to? Isn't it just going to cause the same problem with the sense voltage being too high?
Why not just use two resistors after the existing diode? Say a 4.7k and 2.2k with the 4.7k on the lower side? Bob your resistor schematic is correct.
Quench,
Sorry didn't mean to derail your voltage divider discussion. I was thinking (out loud) with a very large resistor, it would not effect the 24V output going back to 0V. But now that I paused and thought about it, even with a very large resistor - the cap will still pull up the low end. Especially w/o a load.
Yeah, nix that idea.
Sorry Bob - led you on a wild goose chase.
Quoted from Quench:Doing your best to keep out of trouble hey?
Hah, she has nothing to worry about. My idea of an exciting Saturday night is hanging out w/you and Bob and making stupid suggestions. I tried beating her high score on Houdini, but that was serious fail too.
Quoted from mbwalker:Sorry didn't mean to derail your voltage divider discussion.
Don't be sorry, it's just sharing of ideas and different approaches. 
I was initially concerned the capacitor would end up getting charged to near peak voltage resulting in a high voltage meter reading. But I didn't expect what your simulation showed (I'm impressed) with the loss of zero crossing even with a 10k resistor.
Carry on..
Quoted from Quench:...But I didn't expect what your simulation showed...
LTSpice...freebie. Very nice library comes with it too. Granted, it's LT's parts but you can import spice models if needed.
Quoted from Quench:Say a 4.7k and 2.2k with the 4.7k on the lower side?
The 4.7k on the negative side?
I'm going to try this first today.
If that don't work I'll try the 'diode, zener, resistor and filter cap' route. Anyone have a schematic for that?
Thanks
Bob
Quench, your resistors were right on the money, meter shows 24.5V - DMM at the output shows 24.57V. I'm impressed.
I'm not going to tell you that I misread the numbers and used a 47k instead of 4.7k the first time.
I just need to check for amps then clean this up a bit.
Thanks
Bob
Much better day today. First I finally got the meter to show the proper voltage. Then I tried the amps. I connected two 12 volt car bulbs in series - the voltage went down a little (2 volts) and the meter showed 0.22 amps. But by this time my workbench looked like colorful spaghetti. I took several photos and made a sketch to help me remember how it goes together.
I figured there's no way I'm going to stuff all that back into the box. So I made a little printed circuit board.
Then I attached everything related to the meter to the board. Now all I have to do is connect everything up and test it out one last time. Everything has worked before (it even booted up an old MPU - all seven flashes) so it should work after I get it all connected. I have some errands to do but hopefully I'll have it all together in the next couple of days.
Thank you all for your help and suggestions.
Bob
IMG_3127 (resized).JPGQuoted from oldschoolbob:I connected two 12 volt car bulbs in series - the voltage went down a little (2 volts) and the meter showed 0.22 amps.
In this scenario can you take some waveform snapshots of the input and output of the 7812?
I suspect the 7812 dropout voltage is part of the problem.
Quoted from oldschoolbob:(it even booted up an old MPU - all seven flashes)
Are you hooking up to the MPU board with alligator clips, or have you wired up a Molex connector to plug directly onto MPU-J4 ?
Quoted from mbwalker:LTSpice...freebie. Very nice library comes with it too. Granted, it's LT's parts but you can import spice models if needed.
Interesting, thanks. I had a bit of a play. The rotate/mirror functions are greyed out so I couldn't rotate the diodes to match your orientation. Plus for some reason the capacitor is killing the lower half of the unregulated DC (zero crossing).
Quoted from Quench:I suspect the 7812 dropout voltage is part of the problem.
The 7812 doesn't play here. This is unregulated 24V for the zero crossing.
I haven't made up a molex connector yet - I'll probably do that soon. The down side is most of the time when I work on MPUs the connector headers are removed or unreliable so I couldn't use the molex connector anyhow.
I tried that LTSpice a couple of weeks ago. Way over my head. You guys are amazing.
Thanks
Bob
Quoted from Quench:Interesting, thanks. I had a bit of a play. The rotate/mirror functions are greyed out so I couldn't rotate the diodes to match your orientation. Plus for some reason the capacitor is killing the lower half of the unregulated DC (zero crossing).
[quoted image]
One thing I noticed when I first modeled it was the diode, without a model associated with it, caused the circuit grief. Just right click the diode, and select 'Pick New Diode'. There you will find models that have more realistic parameters defined. I just grabbed a 1N4148 - no particular reason.
To rotate a part, select it the hand symbol. Then left click on the part once. Ctrl R rotates.
pasted_image (resized).png
I redrew your schematic using a 1N4148, and labeled the nodes the same as yours. Mine's below.
Sorry Bob, hijacked the thread for a few. I think you'd really would like using the simulator, it actually fairly easy to use. I'd be more than happy to help.
pasted_image (resized).png
pasted_image (resized).png
Quoted from oldschoolbob:... I connected two 12 volt car bulbs in series - the voltage went down a little (2 volts) and the meter showed 0.22 amps...
Bob[quoted image]
Bob, what current is the 24V transformer rated for? And what part number is the bridge? Might be a bit of a stretch, but a diode junction voltage varies with current. I doubt that is the problem, but it might play a little bit of a role.
Just measure the VAC at the bridge input, with and with out a load. That tells us a lot.
Quoted from oldschoolbob:The 7812 doesn't play here. This is unregulated 24V for the zero crossing.
Sorry about that, got my wires crossed.
Quoted from oldschoolbob:The down side is most of the time when I work on MPUs the connector headers are removed or unreliable so I couldn't use the molex connector anyhow.
I'm the opposite, I don't start testing until I've replaced the headers - working on too many corroded boards
Quoted from mbwalker:To rotate a part, select it the hand symbol. Then left click on the part once. Ctrl R rotates.
Got it, rotation works thanks.
Using diodes 1N4148 or RR2LAM4S rectifiers changed the zero crossing signal a little but only when I put a resistor "load" (even crazy high like 100M) across it did the full ripple come into effect. Which is probably what really happens since there's zero current when D5 is not conducting (input load of measuring device creates a current path).
Quoted from Quench:Sorry about that, got my wires crossed
Got it, rotation works thanks.
Using diodes 1N4148 or RR2LAM4S rectifiers changed the zero crossing signal a little but only when I put a resistor "load" (even crazy high like 100M) across it did the full ripple come into effect. Which is probably what really happens (input load of measuring device).
[quoted image]
Ah...I now see what you were after (the zero crossing). If you think about it (edit: oops, you mentioned it), the isolation diode (D5) will shut off before the bridge does, leaving the the ZC_Output node without a DC path to ground. Hence, the simulation struggles. In real life, parasitics, part Q, or even the input impedance of a voltmeter, insignificant as they are, can provide that path. This issue is likely due to the diode model being so-so (I'd have to check). The 'voltmeter', in this case, is truly ideal. Infinite input impedance.
Simulations are tricky and fool a lot of people. I do RF designs that are always based on simulations. Crazy impedance, frequencies up to around 10GHz, power levels above a 1 KW. Understanding parasitics is the key. I even make my own models since many are sub-par or not designed for my application.
Quoted from mbwalker:Simulations are tricky and fool a lot of people. I do RF designs that are always based on simulations. Crazy impedance, frequencies up to around 10GHz, power levels above a 1 KW. Understanding parasitics is the key. I even make my own models since many are sub-par or not designed for my application.
Indeed, must have taken you years to get on top of it all.
Now you need to build some gloves with artificial intelligence so you can beat your wifes highest score on Houdini 
Quoted from Quench:Indeed, must have taken you years to get on top of it all...
Ages of getting wrong results, then figuring out why.
At least I'm kicking her butt on AS. And JD.
Quoted from mbwalker:Bob, what current is the 24V transformer rated for?
I don't know much about the 24V transformer - I got it from Radio Shack a while back. They don't list it anymore.
Quoted from mbwalker:And what part number is the bridge?
Quoted from mbwalker:Just measure the VAC at the bridge input, with and with out a load. That tells us a lot.
I'll do that tomorrow - I'll have more time.
Quoted from Quench:I'm the opposite, I don't start testing until I've replaced the headers - working on too many corroded boards
I guess that's my confidence level. I don't want to invest in headers until I know it's going to work.
More good news coming. Stay tuned.
Bob
I was a hurry to get into the shop tonight because I knew I was close. Getting everything connected wasn’t easy but my sketch and photos helped. Also it’s beginning to get cramped in there. But I got it all connected and it works great! I checked the voltages at the output terminals with my DMM and the 12 volt was just a little low (maybe 0.2V). Tomorrow I’ll adjust the 12V meter, scan the outputs with the oscilloscope and see if it will boot a MPU. Then I can close up the case. I think I’ll call this my “PinSide Tester”.
Parts should be here for the adjustable power supply by the end of the week. That gives me a couple of days to clean up my bench – looks like a bomb went off. With all I’ve learned here, the adjustable power supply “hopefully” will be easier.
If anyone is interested in photos of this build, I’ve uploaded a ton of photos here:
https://photos.app.goo.gl/LcDp6v6B4QzjBjFP7
Thanks again for all your help.
Bob
IMG_3142 (resized).JPGIMG_3144 (resized).JPGIMG_3149 (resized).JPGIMG_3151 (resized).JPGQuoted from oldschoolbob:But I got it all connected and it works great!
...
I’ve uploaded a ton of photos here:
Well done, it looks great! Hope it helps you repair many more boards
You should be pretty proud of yourself Bob, nice supply!
I went thru all the pics...is there one final schematic of everything? They way it's currently 'as built'? I saw schematics, bits and pieces of schematics, but not familiar enough to know. If not, you really need that for later.
I really like the LED meters! Looks like a million bucks.
Now...what's the next project? The adjustable supply? I'm going to be board (ha, pun intended). Wife doesn't come home until Sat eve. I'll need something to do at 2AM Fri night!
What's the regulator on the adjustable supply?
Getting back to one of my other questions...the voltage drop across the bridge: I looked up the bridge (PBL405), The diode drop delta from pretty much no load to 0.22A is ~0.2V Then x2 since you go thru 2 diodes. So the diode drop delta from no load to 0.22A is ~0.4V
pasted_image (resized).pngThis is the completed power supply. It really looks nice setting on my workbench shelf - ready for the next MPU.
I'm not sure I have an 'as built' schematic. I found a PC board layout with the components but I know we added two resistors to boost up the 12V a little. I can't find any revised schematic showing the added resistors. I agree that it's something I may need in the future. I'll work on that.
Next up is the adjustable power supply. Parts should be here tomorrow.
We gotta find something for Mark to do for the next few days - you know what they say about idle hands.
Thanks again everyone - couldn't have done it without you.
Bob
Quoted from oldschoolbob:Mark, I couldn't find a final schematic. This is the PC board layout with the components and the added two resistors.
Bob[quoted image]
OK, not a big deal. Just trying to save you a headache down the road (if it breaks). Everything is still fresh in you brain...couple of years down the road, it will turn into "Why's this part in there and what does it do?" Hah, been there, done that plenty of times.
Bob,
Any chance you would like to use a LT108x series of regulators instead of the LM350? Pretty much the same thing as the LM350, but there's a model for it in LTSpice. No biggie if not. But it would allow us to help you better if you want to change things up a bit. We can just simulate and give you the correct answer.
The LT1085 is good for 3A, the LT1084 is good for 5A, and the LT1083 is good for 7.5A
BTW, you should add a resistor on the output (below). You need 10mA of load current for regulation. If you don't have anything hooked up, it won't regulate properly. Adding a resistor always keeps it in regulation.
Data sheet: https://www.analog.com/media/en/technical-documentation/data-sheets/108345fh.pdf
You caught me just in time - I was about to etch the board. I widened the trace to pick up the resistor. What size do I need to add?
I think I'll stay with the LM350K. The LT108's look like TO 220's. I'm partial to the TO 3's - old school I guess. And besides I have the LM350 already.
I'll probably start assembly tomorrow.
Thanks
Bob
IMG_3202 (resized).JPGIMG_3203 (resized).JPGFor those who have been following for the Meteor re-setting problem, I think the problem is fixed. I’ve played over 100 games without one re-set.
A brief review – The game was re-setting every once in a while. It had a Rottendog solenoid driver board and the 5 volt regulator was getting very hot (maybe that’s normal). I replaced the driver board with an original board. I tested it for a few games and it re-set again. Next I removed the lower right flipper coil to replace the diodes. I found the coil sleeve was stuck so I replaced the coil and diodes. Then I replaced the diodes on the upper flipper. This seems to have fixed the problem. I think the problem was the coil was getting hot, the winding would expand and the coil would short out. Then it would cool down and the short would open again.
Thanks to everyone who made suggestions.
Bob
Bob, what's the lowest and highest voltage you want out of the supply? If you go between 5v and 24v, then 500 ohms, 2W will work.
The 108x is available in T-03 case, but the 350 reg is fine. You can always swap out for the 7.5A version and just change pwb part values, if needed.
P.S. Beat the wife's high score on Houdini! (and I'm not telling her).
Update on the adjustable power supply. I got the board assembled yesterday. Today I worked mainly on the front panel. Tomorrow I'll letter it. If you were making this would you letter it (where the ???? are):
front panel a (resized).png
DC ADJUSTABLE
ADJUSTABLE DC
VARIABLE DC
DC VARIABLE
While waiting for the paint to dry on the panel I connected the board. I don't have the meters connected so I used my DMM. It works!!! 1.25V to 26V. I'll connect it to my oscilloscope tomorrow and see what the wave looks like but so far I'm impressed. Just think 6 months ago I couldn't have done this - I wouldn't have even tried. You guys have taught me a lot - thank you all.
IMG_3222 (resized).JPG
IMG_3229 (resized).JPG
Mark, your wife has just been playing with you. As soon as she returns she'll set a new high score.She just wanted to give you something to shoot for while she's away.
Thanks
Bob
Quoted from oldschoolbob:If you were making this would you letter it (where the ???? are):
DC ADJUSTABLE
ADJUSTABLE DC
VARIABLE DC
DC VARIABLE
Of the four, I kind of think "ADJUSTABLE DC" is most technically correct.
Are you going to print 1.25V and 26V (or whatever maximum is with a minor load) at the minimum and maximum sweep of the rotary control?
Quoted from Quench:Are you going to print 1.25V and 26V (or whatever maximum is with a minor load) at the minimum and maximum sweep of the rotary control?
No but I would like to put on a sweep symbol like this:
potentiometer-icon-vector-450w-1241247724 (resized).jpg
Don't know how to make it.
Bob
I knew it was too good to last. Looking at the PC board in post 294 above, there are two blue terminal blocks. The one on the right connects to the potentiometer - the one on the left comes right off the rectifier for the 24V to 12V converter. Yesterday when I checked the left block it read 24 volts. And it didn't vary with the output (turning the pot). Today I connected the converter and the output read 12 volts. Then I connected the volt meter to the converter (and the sence wire to the output) It read fine and showed the variable output. Then I connected the Amp meter. When I turned it on both meters lit for a second then went dark. I disconnected the converter and tried the power supply again. Nothing. I found the 3 amp fuse from the transformer to the rectifier was blown. I replaced the fuse and tried the power supply and it works again. But when I checked the terminal block for the converter it now reads 36 volts.
Do you think the rectifier is blown? What would cause it to blow?
Thanks
Bob
I did some more investigating. Using my DMM - 26.3VAC at the transformer (AC input). 35.06 VDC at the bridge rectifier output. It was 24VDC yesterday.
Then I connected my oscilloscope The first scan is at the AC input. The next scan is at the rectifier output.
Then to be sure I went back and scanned again. The AC input looked the same but the rectifier scan looks like crap. I've never seen a waveform like that.
Could this be the problem?
This is the rectifier I'm using:
I just ordered 14 different rectifiers from Arrow. They should be here in a couple of days.
Thanks
Bob
Well I seem to be getting more lost everyday. First I checked the meters to see if they still work. I connected each one (volts and amps) to my Pinside tester (on the 5V output). They still light up fine. Then I checked the 24V to 12V converter by connecting it to the zero crossing output. 24 volts in - 12 volts out - it's fine. Nothing blew up yesterday - good news.
Then I checked the adjustable power supply. 26 VAC in and 34.75 VDC out at the rectifier. Then I connected it to my scope. The AC in was exactly the same as yesterday (see first photo above). When I checked the 'after rectifier' I got almost exactly the same as photo 2 above. When I checked again I got the same messed up wave as photo 3. That's when I noticed I wasn't making good connections. I added a couple of wires to get a good connection and now the scope shows a steady output (but still high).
I'm concerned about putting 35 volts into the converter may blow it up. Could that be what blew the fuse? I still don't know what caused that.
Thanks
Bob
IMG_3235 (resized).JPGafter br monday (resized).jpgWanna join the discussion? Please sign in to reply to this topic.
Great to see you're enjoying Pinside! Did you know Pinside is able to run without any 3rd-party banners or ads, thanks to the support from our visitors? Please consider a donation to Pinside and get anext to your username to show for it! Or better yet, subscribe to Pinside+!
This page was printed from https://pinside.com/pinball/forum/topic/meteor-resets-with-rottendog-solenoid-driver-board/page/6 and we tried optimising it for printing. Some page elements may have been deliberately hidden.
Scan the QR code on the left to jump to the URL this document was printed from.
Collinsville, IL