Need help with a tubehobby clock overheating

[Kiran Otter]

Hi folks, glad to find this group!

I've had a Tubehobby clock for several years, the NCV2.1 with the IN-18 tubes.  In the past Jonas has helped, and I even shipped him the main board for him to repair, but he hasn't responded to my last request for help, so I thought I would ask here.

Recently, I started to notice that other digits in the tubes were partially lighting up, and eventually the fuse blew.  My assumption was that the K155ID1 drivers had started to go, so I ordered six of them off eBay, and tried replacing them.. which isn't hard, everything is socketed.   Well it didn't help, so I contacted Jonas.  Jonas suggested replacing C6, which I did and it appeared to fix the problem.

Maybe a month later, I started to notice the left most digit was faintly showing numbers, and seemed to be influenced by the next to right digit.  So I thought perhaps the drivers I got from eBay weren't good, so I swapped them around, trying to see if it made any difference.  Unfortunately, I trashed the two original driver chips that came with the kit.  So far swapping the drivers around among the six I have, hasn't changed anything.. or if it has, the digits lighting that shouldn't be have moved from tube to tube.

Well I let the clock run like this for a week or so, and one day I just happened to feel around the voltage regulator U1 (L7805CV).. and it's blazing hot. I put a temp probe on it and it's running at 140F in open air, and when I built the clock, I epoxied a heatsink to it.  It never ever used to get this hot.  In fact the clock has run for years in a closed enclosure with very little ventilation.  It just never produced much heat at all.  I swapped both driver chips for two others, and it still gets just as hot.

When the clock shuts off the display at night, the temp drops to just above room temperature.

So my guess is has to be one of two things I replaced; C6, or the driver chips.  I think it's the drivers, and I'd like to get a pair from somewhere reputable so I can at least rule them out as the problem.  I've seen some that appear to be ceramic, instead of plastic cased.. claimed to be 'milspec' but I donno if that's BS or what.

Any help is appreciated!

Kiran

[Sture Nystrom]

Kiran,

Measure the input voltage to your clock. If voltage from your DC wall adapter have risen 7805 regulator will go hot. Also high voltage to Nixie tubes might also rise if input voltage to clock is too high.

Sture

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/a7625eb8-22e2-454e-91ea-8b7497882c22%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[blkadder]

I was just having a look at the manual for the clock, and was thinking that the adjustable trimpot at R26 should also be checked.  Being it is adjustable, could it be that it may have failed somehow?  Just a thought.

Ron

[Kiran Otter]

The voltage from the wallwart (12V, 1A) is 11.8V under load.  The high voltage to the tubes is 172.8V.  It's very difficult to get it right at 170V when adjusting R26.

Something else I wanted to mention; the separator tubes (separating hours from minutes, minutes from seconds,) one of them is mostly black, and neither of them light properly.  I'm wondering if they're the culprit.  I'm going to remove them and see if it makes any difference.

Also, I have the heatsink epoxied to both U1 and M1; maybe it's M1 that's getting hot, not U1?  I'll use a infrared temp gun and see if I can distinguish which is getting so hot.

Thank you for the replies!

Kiran

[JohnK]

Careful with epoxying a heatsink on. A heatconducting paste [dangerous chemical usually] OR a very thin layer of heatsink compound and a clip holding the heatsink is probably better. How much does the epoxy impede the heat flow?  [and note I said very thin re the compound?Just enough to fill the tiny voids that exist. The usual compounds are heat insulators, but are still better than air filling the voids.] 

PS. I know you said the overheating is recent, but I use the opportunity to mention this topic.

 

John k.

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/c3192283-c525-4403-bed6-a3979a06f065%40googlegroups.com.

[MichaelB]

I have had this same issue with 3 of Jonas' clocks and each time the fix has been a combination of replacing C6 and a dropping the voltage a bit to keep the temperature down on that VREG. I also ended up changing the design of my enclosed cases to allow better cooling internally and this seemed to fix the issue. These clocks all seem to do this when there is inadequate cooling in an enclosed case.

[Kiran Otter]

Ok, it's definitely U1 that's getting hot.  I removed the separator tubes, and lowered the voltage to 165V.

I still see some numbers/digits lighting faintly as the seconds tick by.. in the hour and minute digit.. 2 and 4 if you count the tubes left to right.

The temperature is now down to around 125F from 140F.  Still way hotter than it used to be.

I've attached another picture so you can see how I epoxied the heatsink to M1 and U1.. and in retrospect it was probably a bad idea because now the heat is being transferred from U1 to M1.

Any suggestions on good driver chips? Or is it just a luck of the draw?

Kiran

[Kiran Otter]

MichaelB, what did you lower the voltage to?  And I did replace C6.  Could I have a bad cap?  I had a hell of a time finding just one; got it shipped free from RS for for like a dollar.

[Kiran Otter]

By the way, I'm open to recommendations on a different clock kit that will use my IN-18s.  I haven't had much luck finding another make.

Kiran

[blkadder]

I have been using the kits from PV Electronics (www.pvelectronics.co.uk).  He has kits that will use pretty much every tube manufactured.  He does have a few that use the big Nixies like you have.  All the kits I have used are his QTC line, and they seem to have all the features that your current clock has.  If you see his online chat button active, he actually answers things right away.  Hope this helps.

Ron

[MichaelB]

 Around 165, can’t remember exactly, but enough so as not to sacrifice tube brightness and yes you could have a bad cap.

[Niek]

Do you have a scope? Is the output of the 7805 still a clean 5V? How about the output of the HV regulator? Can you measure the current going into the clock? (with your multimeter)

I'd be a bit suprised if the cause is the driver chips, because after replacing them it didn't seem to improve a lot (if any) apparently. 

I wouldn't be too worried about 140F for the 7805 btw, it should be able to go up quite a bit higher than that, though with the big heatsink, if attached well, it's a bit surprising.

You can try lowering the input voltage to 9V, and see if this helps. Keep an eye on M1 in this case, as it will probably have to work a bit harder to pump up the HV. 

[MichaelB]

Very nice kits and Pete is very quick to help

[Jeff Walton]

Kiran, 

PV Electronics makes a nice kit.   Their "Spectrum" model uses the IN-18 nixies and accepts a GPS and also has a motion sensor that turns off the high voltage when no one is around to extend the life of the tubes.  Really nice feature! 

They also have a couple different cases available for the Spectrum model.   

I recently bought one and would recommend them. It runs nice and cool. 

www.pvelectroniccs.com 

Jeff Walton 

-------- Original message --------

From: Kiran Otter <kiran...@gmail.com>

Date:03/28/2015 12:51 PM (GMT-06:00)

To: neoni...@googlegroups.com

Cc:

Subject: [neonixie-l] Re: Need help with a tubehobby clock overheating

By the way, I'm open to recommendations on a different clock kit that will use my IN-18s.  I haven't had much luck finding another make.

Kiran

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/bf9c36af-c129-4de0-a3fa-7b8423d521b5%40googlegroups.com.

[Nicholas Stock]

I cut a heat sink to fit under the 7805 like in this photo. Touch wood, out of the Tubehobby clocks I've made to date they're all still going strong. Hope everything is OK with Jonas, he's usually very responsive. I can also thoroughly recommend Pete's kits, lots of features and great customer support. Hope you get your problem sorted!

Sent from my iPhone

On Mar 28, 2015, at 10:51, Kiran Otter <kiran...@gmail.com> wrote:

By the way, I'm open to recommendations on a different clock kit that will use my IN-18s.  I haven't had much luck finding another make.

Kiran

--

[Niek]

You can of course get a new kit, but it will be much more fun and give you a better sense of accomplishment if you fix this one - and since I don't see anything inherently wrong with the design, you should be able to fix it fairly easily. Try lowering that input voltage to 9V, and could you also post a video of the digits issue, so it's a bit more clear what's going wrong?

Thanks!

[Niek]

One more thing: can you try completely removing that heatsink, and see if it makes any difference? I designed my own nixie clock with a 7805 (also 3x2 multiplex), and it hardly gets warm at all (you can easily touch it), without any heatsink. I'd just like to exclude this heatsink as a cause, since it's not in the original design, and i'm wondering if your "paste" is electrically conductive at all (remember, the metal tab of the 7805 is connected to ground, and the one for the IRF640 is connected to the drain).

[Jeff Walton]

 

Sorry, corrected link below.  Phones are fun...

 

Kiran, 

 

PV Electronics makes a nice kit.   Their "Spectrum" model uses the IN-18 nixies and accepts a GPS and also has a motion sensor that turns off the high voltage when no one is around to extend the life of the tubes.  Really nice feature! 

 

They also have a couple different cases available for the Spectrum model.   

 

I recently bought one and would recommend them. It runs nice and cool. 

 

http://www.pvelectronics.co.uk

 

 

Jeff Walton 

 

From: neoni...@googlegroups.com [mailto:neoni...@googlegroups.com] On Behalf Of Kiran Otter
Sent: Saturday, March 28, 2015 12:51 PM
To: neoni...@googlegroups.com
Subject: [neonixie-l] Re: Need help with a tubehobby clock overheating

 

By the way, I'm open to recommendations on a different clock kit that will use my IN-18s.  I haven't had much luck finding another make.

Kiran

--

[Jeff Walton]

Here is a direct link to PV Electronics "Spectrum" clock kit that uses the IN-18 tubes:

 

http://www.pvelectronics.co.uk/index.php?main_page=product_info&cPath=18&products_id=157

 

Jeff W

 

From: neoni...@googlegroups.com [mailto:neoni...@googlegroups.com] On Behalf Of Kiran Otter
Sent: Saturday, March 28, 2015 12:51 PM
To: neoni...@googlegroups.com
Subject: [neonixie-l] Re: Need help with a tubehobby clock overheating

 

By the way, I'm open to recommendations on a different clock kit that will use my IN-18s.  I haven't had much luck finding another make.

Kiran

--

[gregebert]

Be aware that higher temperatures will degrade electrolytic caps. I've seen some postings about this clock getting warm.

I have seen 30-year-old electrolytic caps basically short-out internally; I presume it was because they dried-out.

[Kiran Otter]

Niek,

I do have a scope, however I'd be lying if I said I knew how to use it.  I bought it used with the intention of learning how to use it, and it's sat since then.  You could probably say I know enough about electronics to be dangerous.  I know what resisters and capacitors and transistors do.. and I can somewhat read a schematic, but I wouldn't say I know much more beyond that.

That being said; I put my meter on the 7805 and there's a solid 5.042V coming from it.  I then removed the fuse and measured the current across the fuse holder, and at most it peaked at 0.5A.

I've now snapped the heatsink off; the epoxy never came in contact with the leads to U1 or M1.  And now I think I've discovered that it's M1 that's getting hot, not U1.  Going from the solder side of the board, with the heatsink, the board was much hotter behind U1 before.. but with the heatsink gone, I can hit the case of U1 and M1 with the temp laser.. and M1 is about 85F.. U1 is about 96F.

Now I've increased the voltage to the tubes back to 170V (from 165V) and U1 seems maybe a degree cooler.

I'm going to kick myself if all it was was this heatsink I put on.. plus I don't understand how it was causing such a problem.

Here's a video of the numbers faintly lighting. In this case it matches the seconds.
https://www.dropbox.com/s/3ex4b070e2r3s5p/2015-03-28%2019.52.58.mp4?dl=0

[Kiran Otter]

Now after running it a short while, U1 is about 98F and M1 is 112F.

[Kiran Otter]

Sorry, I think I was getting U1 and M1 mixed up in my earlier post. x.x

U1 is ok, it's M1 that's getting quite hot.  109F now from the PCB side.  It's hard to hit it on top because of the other PCB on top.  But with the heatsink gone nothing is anywhere near as hot as it was before.  Really makes no sense to me.

Kiran

[MichaelB]

Is it still exhibiting the same symptoms as in your video without the heatsink?

[Kiran Otter]

Yes, it is. I took that video after removing it.  I've also noticed the same effect in the minute digit.  It also follows the seconds.

I realize it's pretty faint, but this is the best it's been with the two driver chips I have in there now.  With another pair, the left most tube constantly displayed a digit based on the 10s of seconds. This is why I think the drivers are a bit wonky.

So far both M1 and U1 are around 105-108F.  I'll have to check it later when it's after 10pm and all 6 digits are in use.

Kiran

[MichaelB]

I would try replacing C6 again with a known good part and leave the clock out of it's case without the heat sink and see how it does.  This is the exact syndrome that mine exhibited twice on the one clock until I opened up the case and dropped the voltage and replaced C6. Then add a heat sink ,if you must, a la Nick's (Pramanicin) picture above and raise the voltage up a little.

Seriously doubt you have bad driver chips. Those things are pretty bullet proof

[Kiran Otter]

Well folks, I think I'm going to leave it like this for a while.  I've mounted a little 20mm fan to blow across U1 & M1.  The extraneous digits lighting isn't significant enough to be an issue; I just assumed it had to do with the over-heating.

I appreciate everyone's input. We'll see how long it goes like this.

Kiran

[Niek]

Glad removing the heatsink seemed to solved the issue.

It's interesting it has this behavior next to the blanked out tube: does it have the same behavior when using 24 hour mode? (if it has such a mode, or any other mode where it doesn't blank out any tubes). Blanking out a single tube (of a pair) is not really recommended by switching off only the cathode driver (the chip driver). It should be done by switching off the anode driver instead, but in this case you could only switch off a pair of two tubes that way (since this is a 3x2 multiplexed clock - each pair of tubes has a single, shared anode driver (the transistors)). The driver chip may be exposed to higher voltages this way, and this could lead to it failing sooner. 

However, in your case it seems to be showing the seconds in what I assume is the hours tube? So, this seems more like a timing issue to me, where the anode is switched to the next pair before or right at the same time as the cathode driver was switched to the next digits. To do this right, there should be some dead time between switching off the anode and switching to the next anode (see a recent (a month or so ago) thread in this group about a similar problem). Still - if this is a new issue (and you didn't change the firmware), that would also be strange. Are you sure it wasn't always like this?

[Kiran Otter]

Niek,

Yes, it's showing the seconds in the hour digit, and in the minutes digit.. though not as strongly.  If I force it to display the date or number of hours on the tubes, I can see whatever is in the most-right tube, faintly in the next to left tube.  And I swear I can see the 6 in the seconds tube coming on for like 4-5 seconds.  Again, it's all really faint, so I assume it's not really a problem.

Before I swapped the driver chips around, I was getting faint digits in the left most tube.  So it does seem to be driver related.  But I think it's OK for now.

Is there a known good source for the driver chips?  Someone on ebay?

Kiran

[Niek]

Hmm, well, to me it seems a firmware issue with the timing: if it were the drivers, then it would be random and probably more constant, not showing the seconds. The reason you see the seconds in the 2nd to left tube is that tube is driven by the same driver chip as the seconds (there are only two drivers, and three pairs driven by them in quick succession depending on the anode being driven - only one (of three) anode driver is active at any given time). So, you can imagine that if the anode driver is switching to the next tube set (the hours pair) while the driver chip hasn't been fully switched off yet or switched to the hours digits, then you will see the seconds on the hours pair. It may be that it works better for certain driver chips than others, if they switch off a teensy bit faster, but that's just bad design: it shouldn't be that critical.

It could almost certainly be fixed by a firmware fix quite easily. Maybe you can tell the designer to add some dead time. It would really tick me off having to look at this crappy faint seconds-in-hours where it should just all be solid. Every time you look at it, you get a little annoyed. That's not what you want in a clock. At some point it may be so bad that you smash it against the wall in frustration. You don't want to get to that point.

I'm not sure of any reliable source of driver chips, but you can try both the Russian equivalents of the 74141 as well as the 74141 itself. If you get them from two sources, surely one of them will be good (and most probably all will be good). Try disabling that leading 0 blanking in the future - it's not very good to do that.

[Nicholas Stock]

Could you have a faulty DS1307? Would that cause a timing issue?

Nick

Sent from my iPhone

--

You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/2a2ddbd2-87d6-4cc1-93fc-81dd50c2bc91%40googlegroups.com.

[Terry Kennedy]

On Saturday, March 28, 2015 at 7:55:48 AM UTC-4, Kiran Otter wrote:

I've had a Tubehobby clock for several years, the NCV2.1 with the IN-18 tubes.  In the past Jonas has helped, and I even shipped him the main board for him to repair, but he hasn't responded to my last request for help, so I thought I would ask here.

Recently, I started to notice that other digits in the tubes were partially lighting up, and eventually the fuse blew.  My assumption was that the K155ID1 drivers had started to go, so I ordered six of them off eBay, and tried replacing them.. which isn't hard, everything is socketed.   Well it didn't help, so I contacted Jonas.  Jonas suggested replacing C6, which I did and it appeared to fix the problem.

Yes, that kit tends to "eat" the HV filter cap. I've replaced it in multiple clocks multiple times. Eventually I suppose I'll just figure out how to get a Tayloredge HVPS module in there, and swap the supplies the next time the cap fails in each.

Jonas did have a bad batch of K155's way-back-when, which tended to have leakage between their output pins, but I'm pretty sure that would have manifested by now. It is possible your replacements have this or some other problem.

The IN-3 tubes used in the colons seem to be generally flakey. I've gone through a good chunk of a box of 100 to find "good" ones, which start acting up soon after being installed. I found some elongated NE-2 type tubes at Mendelson's years ago. They must have been very expensive as they were individually wrapped in foam with an inspection report and then packaged 1-per-box in cardboard boxes. Once I installed those, the colon indicators never gave me any more trouble.

[Nicholas Stock]

I'm a bit surprised at that to be honest. I've made well over 20 of these clocks for friends and family (both IN14 and IN18) and not a single one has developed a fault (I know, I'm tempting providence here....), some of them have been running for over 10 years. However, I've never put them in a sealed case, they've all been able to 'breath easily' as below....seeing as I'd had issues with 7805's in the past getting pretty hot, I always took the precaution of putting a heatsink in there as a precaution..perhaps I've just been extremely fortunate!

..

--

You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/54a9ef2a-7475-456e-9ac0-996743cf166c%40googlegroups.com.

[Dan Hollis]

On Sat, 28 Mar 2015, Terry Kennedy wrote:
> On Saturday, March 28, 2015 at 7:55:48 AM UTC-4, Kiran Otter wrote:
>> Recently, I started to notice that other digits in the tubes were
>> partially lighting up, and eventually the fuse blew. My assumption was
>> that the K155ID1 drivers had started to go, so I ordered six of them off
>> eBay, and tried replacing them.. which isn't hard, everything is socketed.
>> Well it didn't help, so I contacted Jonas. Jonas suggested replacing C6,
>> which I did and it appeared to fix the problem.
> Yes, that kit tends to "eat" the HV filter cap. I've replaced it in
> multiple clocks multiple times. Eventually I suppose I'll just figure out
> how to get a Tayloredge HVPS module in there, and swap the supplies the
> next time the cap fails in each.

Mine blew the fuse too, good to know i'm not the only one who had this
problem. Also good to know C6 is the likely culprit. My clock has been
sitting unused for a while.

-Dan

[JohnK]

I didn't look closely when I mentioned about how to properly use heatsink compound.

Was it hot before you added the heatsink?

Was it hotter after?

 

You pondered why the heatsink might make it hotter:-

 

Someone mentioned maybe glue conducting.

 

I mentioned the glue might insulate the heat.

Overlapping a couple of components, maybe one didn't like being heated by the other?

 

And, what if the added heatsink is acting as an UNWANTED capacitor between various pieces of the circuit?  You didn't join the heatsink to earth/gnd or a rail, it was floating?   Gnd-ing pretty well removes the coupling effect. [And the rail is effectively grounded.]

 

John K/

 

 

----- Original Message -----

From: Kiran Otter

To: neoni...@googlegroups.com

Sent: Sunday, March 29, 2015 8:53 AM

Subject: [neonixie-l] Re: Need help with a tubehobby clock overheating

--

You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/2a2ddbd2-87d6-4cc1-93fc-81dd50c2bc91%40googlegroups.com.

[Jon]

Have you thought about replacing the 7805 with a switch mode regulator? More expensive than cheap-as-chips 7805 but waaaayyy more efficient. That means they will run cold and you shouldn't need to worry about the ambient temperature in your case. Two to look at, both of which are pin-compatible with 7805 as I recall.

Murata OKI-78SR-5/1.5-W36-C

Traco TSR 1-2450

I've used the Traco one myself in a clock design, and it's fantastic - very small and straightforward to use. Likely around 90% efficient in your setting.

Cheers,

Jon.

[Dave]

I too was going to suggest the switchmode replacement approach.

Your heat will go away from the regulator, completely.

You can get one shipped to you for less than $9.
Click here:
 Traco regulator

[Nick]

I'm a touch surprised that no one has suggested that the 7805 might be oscillating in the MHz region...

I would 'scope the +5V rail making sure that all input filtering is off...

78xx regulators were notorious for singing if they coukd - hence the essential 100n input & output caps mounted right up against it....

Nick

[Mich...@aol.com]

Funny.

I bought many of his kits years ago.

I was always getting the spelling wrong.

 

So, I registered www.peteskits.com and simply redirected it to his site.  Much easier to remember.

 

Michail

 

In a message dated 3/28/2015 12:15:29 P.M. Pacific Daylight Time, jwalt...@gmail.com writes:

Sorry, corrected link below.  Phones are fun...

 

Kiran, 

 

PV Electronics makes a nice kit.   Their "Spectrum" model uses the IN-18 nixies and accepts a GPS and also has a motion sensor that turns off the high voltage when no one is around to extend the life of the tubes.  Really nice feature! 

 

They also have a couple different cases available for the Spectrum model.   

 

I recently bought one and would recommend them. It runs nice and cool. 

 

http://www.pvelectronics.co.uk

 

 

Jeff Walton 

 

From: neoni...@googlegroups.com [mailto:neoni...@googlegroups.com] On Behalf Of Kiran Otter

Sent: Saturday, March 28, 2015 12:51 PM
To: neoni...@googlegroups.com

Subject: [neonixie-l] Re: Need help with a tubehobby clock overheating

 

By the way, I'm open to recommendations on a different clock kit that will use my IN-18s.  I haven't had much luck finding another make.

Kiran

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/bf9c36af-c129-4de0-a3fa-7b8423d521b5%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/5516fdc9.5a876b0a.176f.ffff9e7f%40mx.google.com.

[Kiran Otter]

Something I want to note about how I used a heatsink; I epoxied it to the tops of the cases of M1 & U1; traditionally you'd bolt it to the back of the component but at the time I couldn't figure out a better way, and I had heatsink epoxy, so I just plastered it to the tops of the two components.  It's been like that ever since I built the clock, and never did it get so hot that you couldn't touch it.  Warm, yes.

I don't know how it could have become conductive across the cases of M1 & U1.. if it did, that's a neat trick.  But removing it certainly seems to have made some difference.

Using a laser temperature probe, M1 seems fine.. but U1 definitely seems cranky.. and it's heating the board up all around that entire end to ~110F.  Again, I never noticed it getting this hot.

To answer Johnk's question; the heatsink and epoxy weren't touching anything metallic, and I originally added it because I thought M1 & U1 were getting pretty warm from the start when I built it, and figured it wouldn't hurt to add it.  But maybe over time.. it started to conduct between them somehow.  I don't know.

Jon; thanks for that info. I'll try one of the 7805 alternatives you mentioned.

Nick; I'll drag out my scope and see if it even still works, and see if I can't check the 7805 with it.  I might have a question or two on using the scope. :D

So far, right now with a little fan blowing over the M1/U1 area.. the whole thing is keeping cool.  Barely running above 90F anywhere I check.  If I turn off the fan, it shoots up past 120F. I turned the fan back on at that point.

Niek; I re-enabled the leading zero, and still see the same thing.  In fact when going through the settings, I was able to catch the attached pic of the right most tube showing what was in the next-to-left tube.  So it seems to go both ways.

Kiran

[Jeff Walton]

Kiran,

 

If you get your scope running, check the output of U1 (7805 regulator) and look for a very flat DC.  If you see any type of sawtooth or superimposed oscillation, replace C1.  You can use a any higher value than 10uF as long as it physically fits and has a high enough voltage rating.  Any visible oscillation superimposed on the output pin could cause the overheating and also could cause the phantom digit illumination as well as other issues.  Just as was mentioned in a previous post regarding C6, C1 could also have dried out and a heated board in the same area will accelerate the dryout of any nearby electrolytic caps.  C2 is also filtering the input to the 7805 and in the direct vicinity of your hot board.  If it is dried out, it will also cause unwanted behavior.

 

Jeff Walton

 

From: neoni...@googlegroups.com [mailto:neoni...@googlegroups.com] On Behalf Of Kiran Otter
Sent: Sunday, March 29, 2015 12:56 PM
To: neoni...@googlegroups.com
Subject: [neonixie-l] Re: Need help with a tubehobby clock overheating

 

Something I want to note about how I used a heatsink; I epoxied it to the tops of the cases of M1 & U1; traditionally you'd bolt it to the back of the component but at the time I couldn't figure out a better way, and I had heatsink epoxy, so I just plastered it to the tops of the two components.  It's been like that ever since I built the clock, and never did it get so hot that you couldn't touch it.  Warm, yes.

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/508e5b80-f513-4cfb-9d71-8a46ab09022e%40googlegroups.com.

[Nick]

Just looked at the schematic. There are no suppression capacitors on the 7805. That is, IMHO, a major flaw.

I would, as soon as possible, add those two 100n caps on the legs of the 7805.

Nick

[Nick]

Just to re-iterate - solder a small 100nF capacitor between pins 1 & 2, and another between pins 3 & 2, preferably on the underside of the PCB directly to the 7805's solder pads.

The symptoms you are seeing are almost certainly nothing to do with the heatsinks etc - I'd put good money on this being a very common & classic issue with 78xx & 79xx linear regulators whereby they break into VHF oscillation and over-heat. This is extremely well documented and the manufacturer's datasheets all say that the two suppression caps are not a "nice to have" but are essential. Most of the datasheets actually specify a 330nF input capacitor (between pins 1 & 2) but a 100nF works well too.

Missing out these two capacitors is a basic error.

Nick

[Kiran Otter]

Nick,

Would replacing the 7805 with a 'non-isolated dc/dc converter' like Jon suggested solve the problem too?  I ask because I was going to try the Traco TSR 1-2450, or the Murata part (though I'm not sure it will fit.)  I ordered both.

I'll get my scope out today and see what I can see.

Kiran

[Nick de Smith]

I suspect they will "solve the problem" but at considerably more cost than two small capacitors which should have been there in the first place. ...

Nick

--
Sent from my Android device with K-9 Mail. Please excuse my brevity.

[GastonP]

The one of suppression capacitors in the 7805's is a very interesting topic in itself... that values of input and output capacitors have been changing in the datasheets for years.
The LM340T series of regulators (of which the 7805 is a direct descendant) used to need one 10uF tantalum capacitor in the input and another in the output. Being tantalum caps  known for its flashy temperament, it drove some designs to very interesting failure modes. I have seen datasheets that put the capacitors as unnecessary and other as must-have. I feel much better with suppression as a must-have and with the small size of SMD components, it seems nonsense not to install them.

[taylorjpt]

If you send me your shipping address, I'll send you one of the 7805 drop in's that I built.  I used it on my TI App note clock and it was so efficient (>95%) that I even hooked the HVPS to the 5V rail.

It is the same foot print as the 7805 and uses a 4-40 socket head cap screw for mounting.  It does not require any external capacitors.

5.25 to 25V in, 5V out at 1A, heat loss is 250mW at 5W out/5.25-25V in.  Compare to 7W dissipated by a 7805 at 12V in.

j...@tayloredge.com

[Kiran Otter]

Well I finally got my scope out and gave my self a sort of crash course on how to use it.  It's a 1975 Tektronix 475.  Last calibrated in 1986. :D  I found it on eBay about 10 years ago and it was cheap enough, I figured I could at least play with it and learn something.  Bought probes and all, but never got around to using it.

At any rate, I'm not sure this scope will show me what I want to know;  if the 7805 is oscillating.  I can see it's putting out 5V by the scope, but I can't adjust it enough (or don't know how) to get it to show if it's oscillating.  I can barely see a wave form but I have no idea how to read the frequency.  If anyone wants to tell me what to set all the switches and knobs to, I'll give it a try.

Meanwhile JT was nice enough to send me one of his little 5V regulators.  I'm waiting on a better desoldering iron to arrive before I butcher this clock any further.

Thanks for everyone's help so far!

Kiran

[Kiran Otter]

I should have said if I see oscillation from the 7805 that Jeff mentioned, pointing to a bad C1.  I may just replace it regardless.

Kiran

On Saturday, March 28, 2015 at 7:55:48 AM UTC-4, Kiran Otter wrote:

Hi folks, glad to find this group!

I've had a Tubehobby clock for several years, the NCV2.1 with the IN-18 tubes.  In the past Jonas has helped, and I even shipped him the main board for him to repair, but he hasn't responded to my last request for help, so I thought I would ask here.

Recently, I started to notice that other digits in the tubes were partially lighting up, and eventually the fuse blew.  My assumption was that the K155ID1 drivers had started to go, so I ordered six of them off eBay, and tried replacing them.. which isn't hard, everything is socketed.   Well it didn't help, so I contacted Jonas.  Jonas suggested replacing C6, which I did and it appeared to fix the problem.

Maybe a month later, I started to notice the left most digit was faintly showing numbers, and seemed to be influenced by the next to right digit.  So I thought perhaps the drivers I got from eBay weren't good, so I swapped them around, trying to see if it made any difference.  Unfortunately, I trashed the two original driver chips that came with the kit.  So far swapping the drivers around among the six I have, hasn't changed anything.. or if it has, the digits lighting that shouldn't be have moved from tube to tube.

Well I let the clock run like this for a week or so, and one day I just happened to feel around the voltage regulator U1 (L7805CV).. and it's blazing hot. I put a temp probe on it and it's running at 140F in open air, and when I built the clock, I epoxied a heatsink to it.  It never ever used to get this hot.  In fact the clock has run for years in a closed enclosure with very little ventilation.  It just never produced much heat at all.  I swapped both driver chips for two others, and it still gets just as hot.

When the clock shuts off the display at night, the temp drops to just above room temperature.

So my guess is has to be one of two things I replaced; C6, or the driver chips.  I think it's the drivers, and I'd like to get a pair from somewhere reputable so I can at least rule them out as the problem.  I've seen some that appear to be ceramic, instead of plastic cased.. claimed to be 'milspec' but I donno if that's BS or what.

Any help is appreciated!

Kiran

[gregebert]

Easiest way to hunt for oscillations is to use AC coupling, and start with 100mV/division, and 100usec/division. The exact frequency of the oscillation isn't important; what you want is less than 100mV peak-to-peak of noise. If you see 60Hz noise, either your 'large' electrolytic cap (usually > 1000uF)  is too small or it dried-out. To get rid of high-frequency noise, start with small bypass-quality caps, around 0.01uF as close as possible to the regulator inputs & outputs. Large caps have parasitic resistance and inductance that limits their ability to attenuate higher frequencies compared to smaller-size (hence smaller-value) bypass caps.

BTW, the 475 is a very nice analog scope; when I was a tech in the 1980s, I debugged a lot of computer boards with a 475.

[Jeff Walton]

Kiran,

 

A sawtooth waveform would point to a bad C1.  A waveform of a much higher frequency would be from an oscillation and more superimposed on the DC. 

 

As long as you are going to do some soldering anyway, you should just replace the 7805 with the drop-in switcher that you received from Tayloredge.  It is MUCH more efficient.  You could also replace C1 at the same time since you are already working on it.  Saves you the work later if it is deteriorating.  Bypass caps are also a good practice for circuits with high frequencies.  Many modern board layouts that work with much higher frequencies (probably not in your clock) build-in circuit traces that are engineered to act in the same manner as a bypass but it won't hurt anything in a clock design to add the bypass caps to a power rail.

 

Many years ago, during my time with Texas Instruments, I sold hundreds of thousands of the little 78xx and 79xx series regulators into a lot of different applications.  They were a new and novel solution for cheap and dirty regulation and efficiency was not a big deal back them.  Today, the small drop-in replacements for the 78xx series are a much more modern and elegant solution when there is a small power budget.   

 

Jeff Walton

 

From: neoni...@googlegroups.com [mailto:neoni...@googlegroups.com] On Behalf Of Kiran Otter
Sent: Saturday, April 04, 2015 1:52 PM
To: neoni...@googlegroups.com
Subject: [neonixie-l] Re: Need help with a tubehobby clock overheating

 

I should have said if I see oscillation from the 7805 that Jeff mentioned, pointing to a bad C1.  I may just replace it regardless.

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/db144a82-9537-4bb7-9a3b-6f1536af8c65%40googlegroups.com.

[Kiran Otter]

Well.. I replaced the voltage regulator with JT's part, it appears to be working perfectly.. a solid 5V right on the mark.  I also replaced C1.

The ghosting numbers are still there but very faint.. but the MOSFET (M1, an IRF640) is still getting extremely hot.  You know when a component gets so hot it has that hot-electronics smell?  It's that hot.  The voltage to the tubes is right on 170V.  I didn't have to adjust it after replacing the two components.

Should I replace the MOSFET?  Any other suggestions? :)

Kiran

[Niek]

Nice that the SMPS made it run cooler. I didn't expect it to fix the ghosting issue - that's probably a timing issue in the software, as I explained earlier. Would have been very strange if that had been related to the 7805. 

As for the hot mosfet: can you check Q4? It's the small transistor nearest to the mosfet. Is it connected correctly? Any issues with it? Can you check the solder joints, and touch it to see if it's hot at all? If that transistor isn't working (or if it wasn't installed correctly), then I would expect the mosfet to get hot. Check the part number on it (make sure it's the correct one), and also cross check with the datasheet of that component and the schematics of the clock, to see if it's correctly installed.

Let us know how it progresses! These posts are becoming like a good detective series ;)

[Sture Nystrom]

A faulty inductor with some windings short cuted will make switching transistor go hot.

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/a71bc0e6-d7a7-429c-95b4-416e7fa1a182%40googlegroups.com.

[Kiran Otter]

Niek, I don't see any problem with Q4, and it's been in there for 7+ years.  If the recommendation is to replace it, I can do that.. the bummer is every time I need a cap or some other part, it costs a dollar plus another $6 to ship it.

Sture, let me know if there's a way to test the inductor (270UH) somehow.  Or of course, I can just replace it too.

This is kind of reaching the point where I might ditch this board and buy a new kit.  I'd love to fix it, don't get me wrong, but I'm starting to have doubts of it being resolved.

Kiran

[Niek]

If you want to be sure that the high voltage power part is as good as it can be, then I would replace the most relevant parts, if you haven't already replaced them recently:

- L1 (inductor)

- M1 (mosfet)

- Q4 (transistor)

- D2 (diode)

- C6 (capacitor, but you already replaced it, so I wouldn't do it again).

- C2 and C7, just in case.

I would especially do this if you are sure that the mosfet used to be clearly much cooler before (which, from your description of it being so hot that it smells, sounds like it would be the case). Just get some flux, some desoldering wick, and you'll have it fixed in a fun hour or two.

I agree that it's annoying to have to pay $6 in shipping each time: but you can order these parts together - usually they're practically for free anyway (excl. shipping). Alternatively, try aliexpress.com, it's a great source of cheap components, often with free shipping (sort by price and select free shipping). But since it's shipped from China it can take a week or two to arrive. (that said, i've ordered countless of components through it, at least 50 separate shipments, and they always arrived). Sometimes on that site, you cannot order just 1, but you can order 10 or so for next to nothing. I wouldn't recommend it for all components though - capacitors especially can fail soon when they are low quality brands, but as far as transistors and resistors, there's not too much that can go wrong with those.

It's true that you can also get a new kit, but then your nixies won't easily fit to it - i.e. the board connector will be different, so you'll have some issues there. If you're willing to take the tubes out, or get new ones, then you can of course use any kit (most nixies are compatible with any kit).

Good luck!

[MichaelB]

I thought I remembered you saying somewhere in this thread that replacing C6 fixed the problem, but then it reoccurred? Yes? And did you not have your large heat sink on the 7805 at that time that seemed to be causing its own issues? Now that you've remove it and if you've got another electrolytic this value you might give it a shot rather than throw in the towel. These clocks are problematic and C6 has always been the problem with the 3 I have. my $.02

[Kiran Otter]

Thanks Niek.. I'll try replacing those parts you mentioned.  The IN-18s are socketed.. I wouldn't expect to have much trouble moving them to a new clock. I'd hope not, anyway.

MichaelB.. I replaced C6 first, when I was having a problem with all of the tubes lighting up all sorts of numbers like crazy.. and then the fuse finally blew.  It was Jonas' suggestion it was C6, and it fixed that particular problem.

With the heatsink epoxied to both the voltage regulator and the mosfet, I wrongly assumed it was the voltage regulator getting so hot. (Hard to tell even with a infrared temp probe.)  But now that I've removed the heatsink and installed JT's voltage regulator board.. it's obvious it's the mosfet that's cooking.

I'll get these parts ordered and we'll see what happens!

Kiran

[Kiran Otter]

Ok, I need some help to know which parts to order..

The inductor; I know it's 'inductance' is 270uh.. but what current rating?  There's about 330 choices. :D  I wish I could tell these things by looking at the schematic.

Is http://www.digikey.com/product-detail/en/RLB0914-271KL/RLB0914-271KL-ND/2352778 correct?

On the mosfet, there's only 2 choices.. one with a max power of 125W, the other 150W.  I'm going on the assumption that the 150W would be better..?

I can't find the 2SA1266 transistor.  Tried a cross-ref site and it gave me 213 choices. x.x

On the diode, there's two; 5ua@1000V or 10ua@1000V?

The capacitors I can figure out, at least.

Kiran

[Nicholas Stock]

Kiran, PM me...I have all the parts you need and can send them to you for the price of postage only. 

Cheers,

Nick

Sent from my iPhone

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/1fcacb70-ae75-4a8c-96d4-e0fdd8a36730%40googlegroups.com.

[Charles MacDonald]

On 15-04-12 10:53 AM, Kiran Otter wrote:

> I can't find the 2SA1266 transistor. Tried a cross-ref site and it gave
> me 213 choices. x.x

http://www.futurlec.com/Transistors/2SA1266pr.shtml

It is a Japanese type. but their are probably hundreds that would work...

use the specs in the link above to screen on digikey for a similar one,
or order some from futurlec, although they tend to take a while to ship.

--
Charles MacDonald Stittsville Ontario
cm...@zeusprune.ca Just Beyond the Fringe
http://Charles.MacDonald.org/tubes
No Microsoft Products were used in sending this e-mail.

[Kiran Otter]

Just an update for those morbidly curious...

I replaced several parts per Nick's suggestion.. (in fact Nick was nice enough to send me the parts!) but in the end, the MOSFET is still getting super hot.

So today I'm sending the board to Nick for him to poke at it.

Thanks Nick!

Kiran

[Dan Hollis]

So what's the clock kit to get these days that doesn't have these problems?

1) overheating
2) blue spots
3) singing due to multiplex frequency being in audible range

-Dan

> --
> You received this message because you are subscribed to the Google Groups "neonixie-l" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.

> To post to this group, send an email to neoni...@googlegroups.com.
> To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/496e8fb9-31b2-4ff4-9aba-63afc0418305%40googlegroups.com.

[Nicholas Stock]

PV electronics sells a direct drive IN18 kit (the Spectrum). Dieter still sells the direct drive 'Blue Dream' for IN18's also. They're both very nice.

Nick

[gregebert]

Since I'm picky & stubborn, I do my own designs. That way I can only blame myself. Snarkiness aside, making a production-worthy design is not trivial and I'm sure that the few kits that are available are designed by engineers who have day-jobs as well. On top of that, kits need to be affordable, so that means the design will have tradeoffs. Lastly, many kits might not include 100% of the parts (again, for cost reasons), so substitutions are inevitable. All of these things combined will result in no 2 boards being 100% identical, and that's an opportunity for things not to work as expected.

=========================================================

Fuzzy blue spots are a tube issue; spectra suggest it's mercury. But you definitely want tubes that contain mercury because they have a longer lifetime. I have yet to find any information that explains the low-level details why this is the case, such as chemical-reaction equations, etc. I've been watching a blue dot in one of my tubes for more than a year, and it has not changed in size, location, or brightness. I suspect there is an impurity on the cathode that attracts mercury.

Singing doesn't exist with direct-drive, but I suspect most designers avoid direct-drive for cost reasons.

Overheating can be caused by many factors, such as the case, and may have nothing at all to do with the circuit design.

[MichaelB]

Here are the two Pramanicin mentions with my scent on them:

http://www.badnixie.com/PV_Electronics_the_Spectrum.html
http://www.badnixie.com/The_Tangerine_Dream_%28Nocrotec_Blue_Dream%29.html

Anymore, Direct Drive is a must for me, because of the increased brightness and both of these are. The tube spacing is a little odd with the PV, but for my money, this kit is feature packed and the better value with a menu structure that makes a lot of horse sense.  This clock is definitely for those of us that prefer to build our own enclosure.

[Jeff Walton]

Dan,

 

For IN-18 tubes, the Nocrotec Blue Dream Clock or the PV Electronics Spectrum 18 clock are both very nice.  Both are direct drive and run nice and cool.  The Nocrotec clock has a stainless case available and the PV Electronics clock has three different cases with choices of clear or smoke grey acrylic.  Among the differences - the Nocrotec clock is blue backlit, whereas the PV Electronics clock has programmable colors on the tube backlighting and also has a very nice motion sensor option to save the tubes when no one is around.  Both have similar fading effects on the digits but the Nocrotec clock also fades the colons to match.  The cathode protection option on the PV Electronics clock is more interesting as the digits cycle independently of each other rather than exactly the same.  Both have good GPS support.  If you have daylight saving time, the PV Electronics clock has a one button DST control for easy changes.  Both have robust menus. 

 

I've built both and like both clocks very much!  

 

Jeff Walton

[Niek]

Hi Kiran,

Did Nick already have a chance to look at your clock? Any idea what was causing the issues?

Curious! :)

[Ian Sparkes]

I don't know if this will help, but I had something like this on a clock I was prototyping. The error I had was that the MOSFET was sometimes being left in the ON state for longer than I expected sometimes.

Once the inductor has "saturated", you effectively have a short from VIN to GND, and of course the MOSFET gets hot when that happens, and the consumption goes right up. This can happen if the frequency of the drive is too low, or the duty cycle is wrong (too much "on" time).

Have you investigated the frequency and duty cycle of the output of the TL494?

From the schematic, it looks like the frequency is fixed (at around 18kHz), and the voltage output is controlled by the duty cycle. I have no idea what the time constant of the inductor circuit will be, but after 5 of them, you are in dead short.

Just a thought...

[Nicholas Stock]

Hi Folks. I did have a poke at the board and the ghosting was apparently caused by a flaky K155ID1 as replacement of it made the problem disappear. However, when operating with my IN18 display board I didn't notice any heating issues, but Kiran still did when I returned the board to him and was using his display board (for those not familiar with the Tubehobby design, the board with the tubes can be removed from the driver board below...it's used with either IN14's or IN18's depending on what anode resistors you put in the driver board). He blew a couple of fuses and noted the Mosfet getting pretty hot again. Pretty much the only thing I didn't swap out was the TL494 as per Ian's comment....so now I'm curious...

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.

To post to this group, send email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/8e5b2d8d-7254-4549-b8db-11f2abeaef60%40googlegroups.com.

[gregebert]

Power supply voltage ? If the supply voltage is higher, it will likely cause more inductor/transformer current, and perhaps enough to cause saturation, which leads to heating.

I've noticed a similar problem with my wristwatch. Even though I have voltage-feedback from the HVDC supply, the current in my transformer increases at higher battery voltages, and it definitely heats-up faster at higher supply voltage. My MOSFET does *not* heat up because it's a very low Rds on device and it's only switching for a few seconds while the time is displayed. 

-----------------------------------------------

Load current ? As mentioned elsewhere the display board is separate, so if it's possible to swap boards with a 'known-good' clock you can quickly narrow-down the problem.

Again, I've noticed with my watch that when running fewer segments on the display, the reduced load current effectively reduces the duty-cycle on my HVDC converter, so the transformer heats less. There's no heating at all when the display is dark.

It may be that when the board tests-out OK, it's with a lower load-current due to smaller nixies, and/or lower supply voltage.

..................................................................................................................................................................

BTW, I still call it a nixie watch even though it uses a 3-1/2 digit panaplex display....who's heard of a panaplex watch?

I'll post another picture as soon as I get a decent watchband attached to the case. Although I could fix the transformer-heating problem with a larger transformer, I'm stuck with the pea-sized transformer I have because of size constraints. The wasted energy (heat) apparently isn't a major problem because my demo-board has been running almost 3 months since the last time it was charged.

[Kiran Otter]

Well after MUCH trials and tribulations.. I gave up. :)   I lack the abilities to figure out what the problem is beyond what I've already done.

I'm sending the boards to Nick tomorrow.. maybe he'll be able to figure out.

My ultimate solution: I spent $300 on a new clock, and used my IN-18s.  Took me 4 hours to build Pete's Spectrum 18 clock, and it's working perfectly.  The PRISM case is just wonderful.  I was really unsure about the multi-color LEDs at first, but now I really like them.

I want to thank everyone for their input and time.  Especially JT for sending the 7805 replacement, and Nick, for sending even more parts and putting up with dozens of my emails!

Kiran

[Jeff Walton]

Did you get the motion sensor?  It's a good feature! I was pleased with it. 

Jeff Walton 

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/b25d754a-9dee-4c70-821e-a2d6b657b6d1%40googlegroups.com.

[Kiran Otter]

No.. the clock is in view all day.. but I do have it shut down at night.

I did however get the GPS.

Kiran