PBW needed

If you are talking about the DCS-RC Watchdog Generators I might be able to help you out in a couple of weeks. My email is in my profile.

rtr12 posted:

If you are talking about the DCS-RC Watchdog Generators I might be able to help you out in a couple of weeks.

I've lost track of where "we" are on the PBW.  Are you gathering raw boards and parts for GRJ's trimpot version? 

Will these be assembled or just a kit with bare PCB and a bag of parts?

For reference, was the Bill-of-Materials (e.g., DigiKey parts list) posted for the trimpot version along with the OSHPark link for anyone to order boards?  It's just that even if assembled boards are offered, the user must still solder the 3-pin square-pin header/socket to their DCSRC.  I haven't checked but I'd think MTH might have discontinued the DCSRC (???) if they are offering the Explorer in "Starter" sets.  So curious is anyone has considered offering a ready-to-go PBW that doesn't require soldering.  

We were actually considering a 555 based version that requires no adjustment, PnP.  It's probably what we should have done before, but better late than never.

Well.  I'm all for PnP but doesn't this still require the user to solder the 3-pin header (or socket) to the DCSRC?

Also, one reason for using the 74HC14 (vs. 555) was the drastically lower current requirement since no information is available on how much power can be "stolen" from the DCSRC to power this add-on.  I suppose one could use the CMOS LMC555 rather than the bi-polar LM555.

So that's why I was asking what the latest thinking is wrt the PBW.  I'd have to hunt it down but I seem to recall an OGR thread where I suggested a $5 cycling timer-module that would perpetually cycle power to a stock (unmodified) DCSRC on for 1 sec, off for 0.1 sec which means the DCSRC would not need to be opened and modified at all.  Problem was this was accompanied by relay clicking which could drive one nuts.  I don't recall what happened but perhaps a ~$1 eBay SSR (solid-state-relay) module could be paired with some kind of timer-module for click-less operation.

Stan, soldering the header to the DCS-RC doesn't require calibration, but tweaking the pot does.  The most questions I ever got were how to accurately adjust the pot without a counter or 'scope. 

I was thinking of the TI CMOS TLC555CP version, it has very low power consumption, easily drives the reset line, and makes the package calibration-free.  Quantity one from Digikey is 84 cents, so the price is reasonable.  Exclusive of the PCB, the parts for this should only cost a couple of bucks from distribution, so the parts cost is not really an issue.  I'd rather spend an extra buck on parts and have a board that can just be simply assembled and is ready to function.

Stan,

I had 2 extra PBWs that someone wanted. My originals from GRJ (the extras) went to the person in need. I wanted to keep one for backup, but was now sitting here with none left. Since I needed to make another one, I got with GRJ about offering some PBWs as kits, he was out of PCBs and parts. He brought up the 'no calibration needed' suggestion and went back to your idea from the original PBW thread mentioning the 555 timer idea and a PBW needing no calibration. I was using a different 555 here for testing, he mentioned your concern for too much power draw...and here we are. So far it is working pretty good on a breadboard (I ran my findings by GRJ and I think he was also satisfied). 

I'm in the process of getting some PCBs and parts for these and was hoping to offer them as unassembled kits. I was going to try this with PCBs and parts for only 10 kits. They are the ones GRJ posted above with the 555 timer, no adjustments needed. The parts I ordered include the 3 pin header that still needs to be soldered to the DCS-RC board. I would prefer offering kits, but suppose I could offer assembled units for those that are really opposed to soldering, but as you say the 3 pin header would still need to be soldered. 

As you point out with the Wifi Explorer, the DCS-RC has most likely been discontinued so demand could be quite low here. With that in mind I imagine there are probably still a ton of the DCS-RC sets out there in need of something to do. Perfect for conversion to PBWs.

I think the no soldering version you mention might also be worthwhile for those that do not want to even think about soldering something. If you could link me to the thread where this is described, I'll see about maybe putting together a few kits of that variety to offer along with the PBWs and see how it goes with those as well. Not sure if these are the proper type, but I have a couple of solid state relays around here somewhere. They are similar to the small ebay modules with the 10amp electro-mechanical relays that have become popular around here.  

I'm not too good at designing this stuff, but I am becoming somewhat proficient as a shipping and receiving clerk. I also kinda like doing this stuff, makes me feel like I am occasionally helping out here and there. So, IMO, adding the 'no soldering required' kit would be a bonus.

It would be cool if there was a 3-pin socket that would press into the PCB and make a secure connection.  Hard to believe something like that doesn't exist, I just couldn't find one.   If that were available, a truly "no solder" option would be available.

Actually, as it turns out, there is a press-in connector for .1" spacing.  Mill-Max 801-43-003-61-001000.  Not exactly cheap, but it does eliminate the soldering. 

Pretty cool, that was a good find!  But, I just put in an order at Digikey earlier today. Price wasn't all that bad, after all it is gold plated.  I'll add a few to the next shopping cart, just to see what they are like. I probably forgot something so I'm sure another order will be needed to finish something.  

I bet that price wouldn't deter someone that was opposed to soldering or would have to buy all the stuff to solder with for that one 3 pin header. 

There may be others, I just found that one first in a search at Digikey.

I'll try looking around for more possibilities. I was looking at this in Digikey when adding to my new order and it is a 'Value Added Item' and went on to explain they make them up when ordered. Still indicated you could get same day shipping though. That's a new one on me. Interesting!

Let's see now, what else can I add to that order to make it worthwhile and hurry things up?

I didn't even notice that it was labeled as such, I never ordered a "value added item" before from them.  Apparently, they chop a long strip of these up to make the size you order, interesting.  I was just looking for the capability and the proper spacing.

I just did some more searching and there seem to only be 3 in a 3 pin configuration. Here's what I found. They show up as 'Press Fit, Solder'. Guess I'll look at the data sheet a little closer. They are staying on the order regardless, curiosity has the better of me now.

They had many more listed with different pin counts. Just change the search results from the link and they should show up, I think. It worked, just uncheck the box for 'Number of Positions'.

Did you confirm the finished (plated) holes on the DCSRC board itself are 0.040" per the socket spec?

Can this socket accept the 0.025" square-pin header?  It looks like it wants round-pins - typically more expensive. 

Stan, this is from the full spec sheet.

The DCSRC holes appear to be approx. 1mm as closely as I can see it with a MM ruler and my old eyes anyway. Pretty close to the 0.040". The pin headers I have on the bench right now are slightly smaller than 1mm, either square or slightly rectangular, as nearly as I can see. I don't have anything to actually measure the 0.025" with, but I think we may be in the ballpark.  

Datasheet says 0.040" +/- 0.003 so perhaps OK.  Do you have a numbered drill-bit set?  #61 bit is 0.039, #60 bit is 0.040", #59 bit is 0.041".   Looks like GRJ confirmed socket can accept the 0.025" square-pins you most likely have.  Yes, if you're only offering a bag of parts requiring soldering then 93 cents for a press-fit socket may not be a good bang for the buck.

You need a good set of calipers.  Here's what I use, very handy as they have both inch and mm readings.  I got tired of the batteries always being dead in the digital ones, and went for these.

SHARS 6" 150mm Inch Metric Dual Reading Dial Caliper MM

stan2004 posted:

Yes, if you're only offering a bag of parts requiring soldering then 93 cents for a press-fit socket may not be a good bang for the buck.

I was thinking more along the lines of assembled versions.  The 555 board should be easy to assemble pretty quickly, and reach a wider audience.

The two yellow caps on the 3-D rendering look to be non-polarized?  The 4.5uF cap on schematic is shown polarized.  4.5uF is a somewhat unusual value - I assume this was just a placeholder value?  Do you have the parts list - this would need to be a relatively tight tolerance part to keep the cycle time to 1.0 sec (+/-10%).

Hmm...  I thought that was changed to 4.7uf.  The real values are 4.7uf, 300K and 4.7k, that gives us a period of 1.008, probably close enough.  The cap will likely be 5%, and the resistors can easily be 1%.

Couldn't find the footprint for the electrolytic, been using too many SMD parts lately.   I fixed it the cheap way.

A #56 bit (smallest I could find) is slightly larger than the DCS-DC PCB holes. Found an old, cheap digital caliper (luckily the battery worked) which measured just over 0.040", very close to that looking through a microscope. The pin header I have is right at 0.025" and it fits with room to spare. Can even be rotated or pushed off in any direction at an angle while stuck in the hole, so there is some slop. 

GRJ, I only found 10% caps in thru-hole so I got some extra resistors in various sizes. I was using an electrolytic that was probably at least 20% on the breadboard. It was a cheapie from a grab bag of assorted parts. Hopefully we will be close with these. Ther resistors were cheap too, 70-80 cents or so for 10 of them. Probably odd sizes, but I have very few 1% resistors here, mostly 5%. I need one of those mass quantity packs of 1% resistors from ebay.

Here are the parts I have on order so far.

296-1857-5-ND - TLC555CP - IC OSC SINGLE TIMER 2.1MHZ 8-DIP
399-14082-ND - C317C100K2G5TA - CAP CER 10PF 10% 200V C0G RADIAL
445-173369-1-ND - FG24X5R1H475KRT06 - CAP CER 4.7UF 50V X5R RADIAL
RNV14FAL300KCT-ND - RNV14FAL300K - RES MF HV .25W 300K OHM 1% AXIAL
RNF14FTD301KCT-ND - RNF14FTD301K - RES 301K OHM 1/4W 1% AXIAL                                                                                                                  RNF14FTD309KCT-ND - RNF14FTD309K - RES 309K OHM 1/4W 1% AXIAL                                                                                                                   
RNF14FTD274KCT-ND - RNF14FTD274K - RES 274K OHM 1/4W 1% AXIAL                                                                                                                     
280KXBK-ND - MFR-25FBF52-280K - RES 280K OHM 1/4W 1% AXIAL
287KXBK-ND - MFR-25FBF52-287K - RES 287K OHM 1/4W 1% AXIAL
294KXBK-ND - MFR-25FBF52-294K - RES 294K OHM 1/4W 1% AXIAL
316KXBK-ND - MFR-25FBF52-316K - RES 316K OHM 1/4W 1% AXIAL
324KXBK-ND - MFR-25FBF52-324K - RES 324K OHM 1/4W 1% AXIAL

And yes, I would only send the 'gold plated' 3 pin header with an assembled PBW, but I was thinking of also offering kits (which would have the regular old solder type pin header). 

Hmmm...I don't see any PCB orders for these? Could have sworn someone had ordered those...Guess that someone better get on that.

The certain someone in question may have too many irons in his fire! 

What's the 10pf cap for?  10nf is nano-farads, 10,000pf.

I couldn't readily find any 5% 4.7uf caps either, I guess they're pretty scarce.

Change up the values and use a .47uf cap, Kemet R82DC3470AA60J, 33 cents.

R1 becomes 3.0 meg and R2 becomes 47K.  Both are available cheap in 1% resistors.  This should get us close enough in frequency.

Xicon 270-3.0M-RC 3.0m 1%, 3.0meg 1/8w

Vishay RN55D4702FB14 1%, 3.0meg 1/8w

Thanks everyone for the continued work on this device. I have two RCs waiting to be modified. Like Tom, I suspect there are many unused units from early RTR sets where people have upgraded to full DCS to run multiple engines.

I'll stand by as the new design gets proofed out.

The new one "should" be a no-cal affair, just press the 3-pin connector into the DCS-RC, and plug in the board.

The pf must have been an Oops?   I have 10nf on my schematic which is identical to yours above, that is what I was using for parts or trying to use anyway. 

I'll change things to the new parts you have listed above and give them the breadboard treatment when they arrive. 

How about this for the 1% R2 47k:  Xicon 270-47K-RC    I think it's the same as the 3meg only in the 47k flavor.  22 cents for one, but only .021 cents for 10 (and 53 cents for 25).   Didn't notice until I tried to order, but the other one is showing out of stock until Nov.  

Also just ordered some PCBs from OSHPark using the 0.8 thickness and 2 oz. copper you were telling me about.

It'll work fine, any 1% 1/8w resistor is fine for the task. 

I didn't notice they weren't stock, I normally work with Digikey, and they have an easier interface for searching than Mouser.

Ok, I have them in the cart ready to go. I am also more used to Digikey, but Mouser had some good prices on some of those resistors. Some had a massive price drop at Qty 10 instead of 100, like the 47k one.  

I just found something interesting at JLCPCB. I was ordering some PCBs and fiddling around with quantities. It's 5 (up to 100mm x 100mm) for $2 and $5.24 shipping. So I tried 10, it went to $5 for 10 and $5.24 shipping, then I tried 15 and it went to $4.40 for 15 PCBs and $5.24 for shipping.  Had to quit and didn't get to try 20, will have to do more checking on that later. 

I've noticed some PCB houses have a "fixed" quantity they want to have you order and if you order more, the price for each one shoots up.  So maybe there's a "knee" in the order curve somewhere where that will happen here.  Once they're processing one board, it costs them very little to do a larger quantity.

That was an interesting discovery, hope it wasn't just a glitch and they continue to do that. Used to be 10 PCBs for $2, but they have now cut that to 5.  I'll have to try that again, keep upping the quantities and see what happens. With my luck, I'll probably end up getting a great deal and ordering 50 with an error somewhere in the mix...

So what's a couple hundred bucks between friends?

A friend of mine keeps telling me "you can't take it with you when you go". 

What am I going to do with all those trains???  I'm going to have to have a separate coffin buried with me with my trains.

I was hoping they had a nice electronics lab at my final destination, maybe I should also look into taking a few things along just in case. 

And as a FYI, Mouser doesn't seem to ship as fast as Digikey. Ordered from both on Sun, got Digikey order yesterday, still waiting on Mouser order.

Yep, Digikey has the shipping skids well oiled, they ship in a flash.

Digikey is a finely tuned machine. I just got an email from USPS that the Mouser order has been delivered down at the mailbox. Guess it's time for an afternoon walk. 

I got some of the Mill-Max 801-43-003-61-001000 in my Digikey order. They seem to be a pretty snug 'press fit' from what I can tell so far. Nifty little devices they have here, they appear to be well made.

My problem is I have 2 DCS-RC receivers, both with soldered pin headers from the first run of the PBWs. I was testing the 'press fit' version on some of the unused holes, a bit awkward. I was a bit reluctant to fully seat it in that position for fear of causing other problems, not being able to remove it after fully seated. 

Also, I can't think of a good way to test the 'press fit' to be sure it's making good contact? Only being partially inserted in the holes is another problem with that. All I can think of is finding a generic perf board with identical holes for the test fit. I have some here somewhere, but it seems they had larger holes as I recall.

Any suggestions on how to proceed with testing these to be sure they will work correctly?    

I "presume" that the other unused holes on the DCS-RC you have are the same size, try pressing a couple into those.  If they press in with any kind of force, I'm sure they'll be making contact just fine.