Stc-1000+

[disney7]

The latest work code is running on my dorm fridge and on another unit on the desk (with different set points).

As for the 1.1 hardware. The chip has no markings (I hate it when they do that). If you can tell me anything else to check I will do it. On that header, three pins run directly to the daughter board and the other three stay on the main board, at least for the first jump. The IC is a 20 pin chip.

 

[alphaomega]

The latest work code is running on my dorm fridge and on another unit on the desk (with different set points).

As for the 1.1 hardware. The chip has no markings (I hate it when they do that). If you can tell me anything else to check I will do it. On that header, three pins run directly to the daughter board and the other three stay on the main board, at least for the first jump. The IC is a 20 pin chip.

Cool! Hope it works better now then.

The chip probably does have markings, just not very readable ones. Reverse engineering the 1.0 revision, I needed to desolder the entire board and had to use a 10x magnifier to be able to make out that it was a 16f1828.

Yeah, that is just it, looking at the first pictures you took, those look like they are just vias, not a header. Probably signals for relays and buzzer.

 

[truvr]

As for the 1.1 hardware. The chip has no markings (I hate it when they do that). If you can tell me anything else to check I will do it. On that header, three pins run directly to the daughter board and the other three stay on the main board, at least for the first jump. The IC is a 20 pin chip.

The IC on my 1.1 board looks the same. A long 20 pin chip. Mine does have markings.

N79A8211
1347XA11181

A search on the N79.... seems to indicate that this is an MCU made by a Chinese company. I couldn't turn up a datasheet or anything, at least nothing in English. The best info I could find was at the following link. I'm thinking the 1.1 is gonna be a no go, unless someone has some serious tricks up their sleeve.

http://zhangyan520797.en.b2b168.com/shop/supply/13635504.html

 

[truvr]

Here's the best photo I could get. Interesting that there is a five pin block just to the side of the processor. Wouldn't surprise me if this was a PIC knockoff, but who knows.

 

[truvr]

Hmmm, maybe not. The app doesn't want to let me attach the photo...

 

[truvr]

Ok, HERE is the picture.

 

[alphaomega]

Yeah...
Ok, so the link you provided says it is an 8052 compatible chip.
SDCC can compile for 8052. That is good. The hardware specs (at least the parts that we use) seem pretty similar to the 16F1828. That's good. It can be in circuit programmable. That's good,
Bad news is it is not a PIC, so even if it can be programmed from the Arduino, it is back to square one. A whole new sketch is needed.
It actually has half the RAM and EEPROM of the PIC. So even if it can be reprogrammed, current functionality would probably need to be reduced.
Also, the code would pretty much also need to be written from scratch.
And on top of that, finding decent datasheets and in english, might not be all that easy.
I say that it is prolly not something I plan on doing...

I just ordered two more STC:s in panic, hoping for rev 1.0's...

 

[disney7]

You're right, I missed that 5 pin header on mine. On the v1.1 I have they're all nice and clean. I think you could easily use some 90* pins and make a press and hold type connection for as long as it takes to program.

Assuming the chip is compatible with Alpha's code of course.

Oh, and the chip on mine has no markings at all. Checked with a good light and magnifying lens.

 

[disney7]

Just saw Alpha's last post. I'd say ditch the 1.1s. Not worth the trouble.

 

[uatuba]

I wonder if there is a new revision because it is cheaper for them to manufacture? The price on these things has been dropping and the sales have probably skyrocketed

They are going to run out of 1.0s eventually. Would emailing the US distributor do any good?

 

[andrewmalizia]

Thankfully I don't need this functionality on my serving fridge so I can swap out my old with a new once I get a freezer. What about some of the other brands selling? Agptek and the like? It would obviously be worth the extra $5 or so for the v1.0


Sent from my iPhone using Home Brew

 

[balrog]

I got my first one from eBay, user mixtea, and it was the stc-1000+ capable version. I ordered one almost immediately from Amazon, and it is the newer version, not PIC not STC+ capable. My point being that you *MIGHT* have better chance of getting alphaomega-flashable STC from eBay. Then again if there is a cheaper model flooding the pipelines, nothing is a safe bet.

EDIT: Prices at Amazon still dropping; and I just got another one (just in case) from "xseries" and it is flashable!

 

[atoughram]

I got my first one from eBay, user mixtea, and it was the stc-1000+ capable version. I ordered one almost immediately from Amazon, and it is the newer version, not PIC not STC+ capable. My point being that you *MIGHT* have better chance of getting alphaomega-flashable STC from eBay. Then again if there is a cheaper model flooding the pipelines, nothing is a safe bet.

I got a flashable unit from Amazon a week ago (3/21) from this guy

http://www.amazon.com/gp/product/B008KVCPH2/?tag=skimlinks_replacement-20

 

[uatuba]

I got a 1.0 from Elitetech two days ago (four really...it was two day shipping). I'm sure they have a mix of both at this point, but that's soon going to end it seems.

 

[Wilberforce]

The stickers on the outside of the units have slight differences. Perhaps we can determine a v1.0 board From the graphics?

 

[Wilberforce]

That looks like an interesting board. I've played some with the ethernet shield, but I have not much use for it, since I don't have access to wired ethernet where I would need to use the Arduino.
Anyway, it says it is 100% Arduino compatible, so my first guess would be that USB won't be able to provide enough current. If you have the UNO, then do try that.

I managed to flash a v1.0 with the etherten board. It was a wiring issue, I had the vcc going to the arduino vin, rather than vcc. Dooh.
How it managed to power up the stc-1000 from that is a wonder!

 

[Wilberforce]

Is there room on the setup menu to add ver menu?
Then you could show the firmware build number.
This would be useful for determining if a unit needs to be updated after it's been in the field for a while...

 

[disney7]

Is there room on the setup menu to add ver menu?
Then you could show the firmware build number.
This would be useful for determining if a unit needs to be updated after it's been in the field for a while...

Yes, that would be sweet.

Also, is it possible to not overwrite the config settings if they have already been defined or are legitimate values?

BTW, the ramp test is going well. It held the initial step for 8 hours and then properly ramped up 10 degrees over the next 8 hours and is holding there for a bit until it is set to ramp again. It will take several days for this to finish. I'm also running through shorter scenarios on another STC+ to see if I can confuse it. So far everything is working great.

I forgot to mention this a while back... the curve for the temp probe is so good that I have not had to change the calibration from 0.0 on any of the three units that I have flashed. This is verified at several points around 60*F with a ThermaPen.

 

[alphaomega]

I get mine from ebay, and my experience so far is that the images there means very little. You get what you get... Maybe amazon is better.
Cool though that you got it working with the etherten board and that you reported back.
Regarding version number, I think that is a pretty good idea. I might not want to add it to the 'Set' menu as I don't think it would be a good idea for the user to be able to actually set it, so I'm thinking along the lines of adding it after 'sp' is displayed when pressing up or even like a combo, so it is displayed when pressing both up and and down.

 

[alphaomega]

Yes, that would be sweet.

Also, is it possible to not overwrite the config settings if they have already been defined or are legitimate values?

BTW, the ramp test is going well. It held the initial step for 8 hours and then properly ramped up 10 degrees over the next 8 hours and is holding there for a bit until it is set to ramp again. It will take several days for this to finish. I'm also running through shorter scenarios on another STC+ to see if I can confuse it. So far everything is working great.

I forgot to mention this a while back... the curve for the temp probe is so good that I have not had to change the calibration from 0.0 on any of the three units that I have flashed. This is verified at several points around 60*F with a ThermaPen.

It should be possible to not overwrite the EEPROM values. I have just not had much time lately to work on this. I think I will keep it simple and just separate the hex files and add a one command (actually two, one for C and one for F) for uploading hex without EEPROM values. This should be simple enough and still be good for casual upgrading without loosing the settings.

I am very glad to hear that ramping seems good so far. Even though it is not true ramping and there may be slight rounding errors (i.e. result could be one tenth of a degree off) and in the general case (where duration is not multiple of eight) it might not be obvious at what time 'sp' is updated, I still do think this is better than not having any ramping, which is the alternative.

Also, awesome to hear that temperature readings are good

Have you tested any negative temperatures and/or setpoints? It wouldn't surprise me if you haven't as F scale would be only positive in the range of interest.
I might have to do some testing on this myself soon, as in C scale it would not be unreasonable for cold crashing. It should work, but you never know if the compiler honours sign bit when shifting and so on...

 

[disney7]

No, I haven't tried any negative values. I've been focusing on ranges that you would use for fermenting or cold crashing. The lowest I'd set anything for a cold crash would be around 0.5/1 C or 33/33.5 F.

 

[etrmedia]

A quick follow-up to the folks with a V1.1 board:

I grabbed a friend's two STC-1000s he's been running in his house for the past two years, and they're both V1.1 and both have the N79A8211 chips on them with the different header location. I don't think this is a new development in the STC-1000; I think it all depends on your source.

 

[balrog]

A quick follow-up to the folks with a V1.1 board:

I grabbed a friend's two STC-1000s he's been running in his house for the past two years, and they're both V1.1 and both have the N79A8211 chips on them with the different header location. I don't think this is a new development in the STC-1000; I think it all depends on your source.

Most. Curious.

 

[alphaomega]

With any luck, elitech has the 1.1 revision and the others still use the 1.0. I just hope production hasn't switched to 1.1 all the way around.

 

[disney7]

I succumbed to the panic buying mentality and ordered two more AGPTeck units from Amazon. The one I got from them a few weeks ago was 1.0. They should be delivered tomorrow, so I'll let everyone know what I get.

Recall that I had previously ordered two EliTechs from Amazon and one was 1.0 while the other was 1.1.

I wonder if there is any realistic chance of getting in touch with someone at the factory in China to find out what the story is.

 

[atoughram]

A quick follow-up to the folks with a V1.1 board:

I grabbed a friend's two STC-1000s he's been running in his house for the past two years, and they're both V1.1 and both have the N79A8211 chips on them with the different header location. I don't think this is a new development in the STC-1000; I think it all depends on your source.

So you had me curious... I grabbed one that I bought a year ago and opened it up (I replaced it on the ferm chamber with a flashed one..) It's a 1.1

 

[disney7]

I'll bet there is someone around here with knowledge of the workings of Chinese manufacture who can get to the bottom of this.

It may actually be that one design is being made in one plant and a different design in another plant. If something is working well, I doubt they would change it and offer no feature improvement unless the new materials cost so much less that changing the manufacturing process/tooling and recoding for the new processor would be offset quickly by the cheaper cost in materials. But, what do I know? :cross:

There are a lot of different controllers popping up using the same enclosure. There is a Fahrenheit version available now (see the other threads on HBT) for $20 and it is laid out completely differently. It has an alarm relay in addition to the heat and cool, but all three relays are not independent (they share one leg), which I do not like at all.

I e-mailed one of the manufacturers who claims they make the devices (Lerway Tech) and asked some questions. They're probably not really the manufacturer and I'll be surprised if I get an answer.

 

[balrog]

I'll bet there is someone around here with knowledge of the workings of Chinese manufacture who can get to the bottom of this.

It may actually be that one design is being made in one plant and a different design in another plant. If something is working well, I doubt they would change it and offer no feature improvement unless the new materials cost so much less that changing the manufacturing process/tooling and recoding for the new processor would be offset quickly by the cheaper cost in materials. But, what do I know? :cross:

There are a lot of different controllers popping up using the same enclosure. There is a Fahrenheit version available now (see the other threads on HBT) for $20 and it is laid out completely differently. It has an alarm relay in addition to the heat and cool, but all three relays are not independent (they share one leg), which I do not like at all.

I e-mailed one of the manufacturers who claims they make the devices (Lerway Tech) and asked some questions. They're probably not really the manufacturer and I'll be surprised if I get an answer.

You're probably right about getting a response, but it'd be fun to find out. If there are two basic styles out there I'm surprised one hasn't won out for cost over the other, either for parts or manufacture, but who knows? The demand for these devices has been quite long lived, which will always generate multiple people trying to get some of the market.

 

[alphaomega]

One thing that I started thinking about when reports came in that old devices were rev 1.1 was your original post Disney,

S1000-Power_V1.1 where all my other boards say A400_P Version:1.0

These are not the same.
We could be lucky in that the S1000-Power really is older, and that is the one being discontinued (or not discontinued perhaps, maybe just not as common).

Now for some updates. I have pulled ramping to master, as it at least seems to work + it is disabled by default.
And I have pushed new changes to work branch.
I have added two commands 'b' and 'g' to flash only program (not EEPROM) for celsius and fahrenheit respectively.
Also I have added version number. Pressing 'up' and 'down' at the same time will show it. I did not want to rush to v1.0, so I set version no to 0.10 but due to how display algorithm work (dropping leading zeros) it will show ".10", it will correctly show as "1.00" when we get there, so I won't bother to try to 'fix' that.

Also, I've done some testing with ramping and negative numbers and it seems ok, but I'll be running a longer test during the night.

 

[balrog]

OUTSTANDING!

Flash time!

 

[disney7]

Alpha, we should be able to flash without overwriting the config as long as you haven't introduced a new setting (such as Ramp 0/1) since the last flash, right?

 

[atoughram]

I succumbed to the panic buying mentality and ordered two more AGPTeck units from Amazon. The one I got from them a few weeks ago was 1.0. They should be delivered tomorrow, so I'll let everyone know what I get.

Recall that I had previously ordered two EliTechs from Amazon and one was 1.0 while the other was 1.1.

I wonder if there is any realistic chance of getting in touch with someone at the factory in China to find out what the story is.

I also ordered two more...

 

[Wilberforce]

Regarding version number, I think that is a pretty good idea. I might not want to add it to the 'Set' menu as I don't think it would be a good idea for the user to be able to actually set it, so I'm thinking along the lines of adding it after 'sp' is displayed when pressing up or even like a combo, so it is displayed when pressing both up and and down.

Sorry - I was not clear. If it was on the menu, it would be a read only item - the up/down would not have any effect - only exit up a level.
Personally I don't like the multi-key sequence things - I always found it tricky on the original unit.
The version would need to be stored in programme space. If you also had a version in the settings eeprom, you could then "check for compatibility" of the settings if there was a new feature added that invalidated the eeprom settings. however this does make things more complicated!

 

[alphaomega]

Alpha, we should be able to flash without overwriting the config as long as you haven't introduced a new setting (such as Ramp 0/1) since the last flash, right?

Yes, but even if a new config value is just added, you could opt not to overwrite data and just set that value manually. Really, only if flashing a brand new device or if there needs to be bigger changes (like how the data is laid out, or if we need to remove one profile to allow other data to be stored), then it could be convenient to know values are initialized to sane values.

Sorry - I was not clear. If it was on the menu, it would be a read only item - the up/down would not have any effect - only exit up a level.
Personally I don't like the multi-key sequence things - I always found it tricky on the original unit.
The version would need to be stored in programme space. If you also had a version in the settings eeprom, you could then "check for compatibility" of the settings if there was a new feature added that invalidated the eeprom settings. however this does make things more complicated!

No, you were perfectly clear. Let me explain my point of view. First, I think it would be strange to have a setting that is not a setting. Second, the state machine to set config values need to make assumptions to keep it simple, having a read only setting would add a special case and increase complexity and there are too many special cases already. Third, I can't imagine checking version is something normal user would be doing on a daily basis, so I think it is acceptable that it is slightly 'tricky'.
Program version really is not related to EEPROM data, so it is still very much possible to add an EEPROM version that is just not shown in menu, to keep track of if EEPROM data neess to be updated. However, if that functionality should be implemented, I need to do some thinking first. I'm not sold on the idea of uploading and not knowing if EEPROM data is retained or not. So, I actually don't think it is too bad to leave it up to the user to decide.

 

[Wilberforce]

No, you were perfectly clear. Let me explain my point of view. First, I think it would be strange to have a setting that is not a setting. Second, the state machine to set config values need to make assumptions to keep it simple, having a read only setting would add a special case and increase complexity and there are too many special cases already. Third, I can't imagine checking version is something normal user would be doing on a daily basis, so I think it is acceptable that it is slightly 'tricky'.
.

One option is to add to the arduino serial commands, 'v'. It could show the code version, and possibly EEPROM content version if you go that way.
Another would be a power on version number, however this would require a start up timer of sorts... So may not simple!

 

[disney7]

Eh, the way it works now is fine, especially if it is the most efficient way to implement it.

I'm not going to be checking the version number frequently, so the dual button press thing isn't a big deal for me and I'd rather not have another config option to have to cycle past.

 

[alphaomega]

One option is to add to the arduino serial commands, 'v'. It could show the code version, and possibly EEPROM content version if you go that way.
Another would be a power on version number, however this would require a start up timer of sorts... So may not simple!

Thinking about this, there are actually a few 'User ID' bytes near the config words, they could probably be used to store version and EEPROM version. It would be pretty easy to include that in the HEX... Hmm... Not a bad idea... It wont use up any EEPROM and can easy be checked at programming time.
I'll look in to that.
Still I think the decision to initialize EEPROM data should be left to the user, but with this it should be easy to determine if it is 'needed' or not.
I'll probably add it to the 'd' command though.

 

[ulnpiper]

Just received my two STC-1000s and both are v1.1 Interesting that the date on both my boards is 2009.6.20 and I noticed that disney7's in post #344 shows 2013.11.18 (assume that means Nov 18th, 2013 versus my June 20, 2009).

I'm going to spend a little time tracing the schematic and see if I can figure something out.

-gary

edited to add picture:

 

[disney7]

My two AGPTek's from BrainyDeals on Amazon were just delivered and they are both 1.0s. Though one is missing one of the orange mounting clips (think I'll keep it though).

 

[DrunkleJon]

Got my two elitecs friday. Both 1.0.