Youyue 858D+ some reverse engineering + custom firmware

[DGM]

Tnx for tips DGM.

I got the same problem when I tried to replace original MK1840 with modded one, the heater went ballistic on me, got glowing red while fan was on. It worked before that change so I tried to put the old mc back but still the same thing happens.

I'll measure it and see if neutral is like it should be.

Since I have no idea what all you did with the mod, I can't really refine what I'd recommend to troubleshoot.  That said, I don't see many ways a botched mod would directly cause the condition.  Unless there were latent faults that were unrelated (which is always a case with counterfeit semiconductors), I don't see how a misconfigured controller could damage the triac or opto alone. 

One way the controller could cause the stuck condition is if it caused the heater to be grossly overheated.  The only thing insulating the nichrome heater element from the grounded barrel/nozzle is a layer of mica paper.  The heater can certainly get hot enough to melt through the mica and short to ground.  As I've mentioned before, this condition may persist regardless of whether the triac is on or off, depending on the wiring.  I imagine there are any number of ways the controller might drive he heater to extreme temperatures: improper ADC reference voltage, missing or misplaced input from the TC amplifier, or wrong input scaling.  The default code (line 197 in youyue858d.ino) still assumes you're using the original board (not one of the newer MK1840/41-based ones).  Unless you rescaled the raw ADC value to match the response of the TC amp on your clone, it would have overheated. IDK what board you have, but I mentioned this in a blurb about the D06 board: https://www.eevblog.com/forum/reviews/youyue-858d-some-reverse-engineering-custom-firmware/msg1400946/#msg1400946, and I know there are other scaling factors for different clones posted here and there in this thread.

IDK if any of this conjecture is really accurate though.   Like I said, I don't know what you did in the mod or which clone species you have.  Really, at this point, this thread kind of needs a summary of all the different variants and their mappings/scalings.

[stj]

maybe if the triac is on too long, it cooks and shorts out.
overheating was the original problem on the pump-in-base models, the triac had no heatsink!

i'm getting ideas now, maybe a bigger heatsink or replacing the triac with an SSR bolted to the case!
that could solve the mains-isolation issue AND add zero-crossing switching.

[texaspyro]

The Pace desoldering stations over-drive the motor when you first press the vacuum switch and then decrease the pump voltage... they call it "Snap Vac".   It works really well to get a clean and thorough un-joint.

[DGM]

maybe if the triac is on too long, it cooks and shorts out.
overheating was the original problem on the pump-in-base models, the triac had no heatsink!

i'm getting ideas now, maybe a bigger heatsink or replacing the triac with an SSR bolted to the case!
that could solve the mains-isolation issue AND add zero-crossing switching.

That's kind of why I said I doubt that the triac could be caused to fail alone.  I assumed all of the clones had heat sinks on the triac.  Under that assumption, I'm pretty sure you'd end up with damage to the heater before the triac had been on for long enough to overheat.  If it doesn't have a heatsink, then I suppose that's a potential problem too.

If I were out to upgrade that, I might just use a bigger heat sink, or just switch to a non-insulated variant (so long as the heat sink has clearance and isn't connected to anything).  Since we're talking about a cheap ebay soldering tool, I'm assuming nobody would buy a Crydom for it.  Buying an ebay SSR to replace the triac/opto would just be inviting more problems, since they're the exact same counterfeit BTAxx insulated-tab triacs inside those SSR's, and you'd only be adding more thermal interfaces to the heat path.  Also, a bunch of them have input circuitry that's incapable of reliable triggering unless they're driven with >5v anyway.
http://epicbeardquest.blogspot.com/2014/10/how-to-destroy-fake-fotek-ssr.html

Changing to an inexpensive SSR wouldn't add any more isolation or ZC to the system.  All the board schematics I've seen use the MOC3041 opto, which implements the ZC functionality.  Similarly, all the schematics show the use of a BTAxx triac, which is an insulated-tab variety.  Considering that the inexpensive SSR's use the same parts, you wouldn't be getting any different isolation.  If you meant that it'd solve the isolation problem by allowing you to move all the mains-connected stuff from the PCB, I suppose that might be true.

At this point, we're basically guessing without more information.

[kulla]

Tbh, I didn't do anything else than replacing controller with modded one, no fan modification, no cutted traces or anything like that. I have tested 3 different mod boards, all with the same result, heater overheating and as it happens with the original controller I know it's not mod related.

You have right about heater and triac, I had spare heater as the first one melted burning mica paper in the process and blowing the fuse when it created short. All that in very short time, before I could even react. So now the heater inside is a new one, with new mica paper as isolation and it's the same thing.

I will have to trace whole board and see what is the issue, as triac is obviously geting logic gate open. I'll put the better picture of the board later.

Tnx for help, appreciate it.

[DGM]

Tbh, I didn't do anything else than replacing controller with modded one, no fan modification, no cutted traces or anything like that. I have tested 3 different mod boards, all with the same result, heater overheating and as it happens with the original controller I know it's not mod related.

You have right about heater and triac, I had spare heater as the first one melted burning mica paper in the process and blowing the fuse when it created short. All that in very short time, before I could even react. So now the heater inside is a new one, with new mica paper as isolation and it's the same thing.

I will have to trace whole board and see what is the issue, as triac is obviously geting logic gate open. I'll put the better picture of the board later.

Tnx for help, appreciate it.

Okay, so it's a D06 board.  I'm assuming the picture shows the original MCU reinstalled.  You say the triac is getting logic gate open; you mean the MCU is turning off the LV side of the optocoupler?  Did you verify that the triac is/isn't conducting?  I'd probably just start with an unpowered continuity test of the heater circuit to make sure everything is sane first (triac open, optocoupler output open, nothing shorted to earth). 

[DGM]

On a slightly different note, I wonder if anyone has considered replacing the handpiece cable on their 858D.  The cable on mine was only about 30" long, making it impossible for me to swap hands or use a stand.  The PVC jacketing was stiff and always retained the kinks and folds from packaging despite my attempts to straighten it with a heat gun.  The part that unsettled me is that the conductors appear to be 24AWG, even the heater and ground conductors.  I don't know if the other clones are similar (short cable, small wire), but I wouldn't be surprised if it were yet another 'feature' of my particularly crappy clone. 

Of course, trying to buy small quantities of multiconductor cable isn't going to be easy or cheap, especially if you want mixed conductors or a more flexible jacket.  I opted to make my own cable.  I kind of doubt that anyone is going to repeat this, but I figured I'd throw this in with all the other mod info. 

The materials:

• 3x 18AWG silicone-jacketed wire (ebay)
• 7x 24AWG TP jacketed wire (old VGA cables)
• 8mm x 6mm clear silicone tubing (ebay)

The method:

• Dye the tubing -- I used Jacquard iDye Poly (a fabric dye).  The dye won't take unless the bath temp is above about 90C.  Don't think you can use the more common RIT dye.  RIT has been reformulated and is now useless for dyeing plastics.
• Make a register plate -- The wire needs to be guided neatly into the tubing without crossing and overlapping.  The cable layout is essentially a circle packing problem; figure out the pattern you want and drill it oversize in a scrap of wood or something. 
• Prep the wires -- Feed the wires through the register plate, strip them and solder them together.  A leader must be attached to this.  A sufficiently strong wire could be soldered at this point, or a loop could be formed for attaching a strong string. 
• Pull the wires -- Fish the leader through the tubing.  Lubricate the wires and the inside of the tubing with a fluid of your choice.  I used window cleaner.  Incrementally push the tubing over the wire bundle while keeping tension on the leader and keeping the register plate close to the end of the tubing.  Remember that the tubing can't be pulled because it will constrict.  Work the tubing along its length, pushing it bit by bit.

The resulting cable is supple and lays flat.  I included an extra pair of smaller conductors in case I wanted to do something in the future with airflow measurement or something ridiculous.  In retrospect, I probably could have fit another extra pair in the tubing.  The new cable is thicker, but I was able to reuse the panel bushing.  I did use some adhesive to prevent the stretchy silicone from being able to worm its way out of its restraints at either end. 

This is the whole thing reassembled.  Included is the original cable to show the wire size (see green conductor).

[bitseeker]

Nice job, DGM. Yeah, a little larger but pretty close, and it lays nice and smooth.

[kelchm]

With the D06 board, are there any readily available PCBs I can order from OSHPark or elsewhere to at least give me an ISP header?

[DGM]

With the D06 board, are there any readily available PCBs I can order from OSHPark or elsewhere to at least give me an ISP header?

I don't know whether there are such boards already shared on OSH Park's site or not.  I know there's a lot of ambiguity in the conversations throughout this thread about what's compatible with different board versions.  I touch on that at the beginning of a post a while back: https://www.eevblog.com/forum/reviews/youyue-858d-some-reverse-engineering-custom-firmware/msg1400946/#msg1400946 and I gave a link to an even older post from floobydust in which he gives the design package for his adapter for the D06 board.  I imagine those files could be used to have the board printed where ever you wanted. 

As to what you actually want, the wording makes it sound like maybe you only want the header to access the MK184x.  I don't know if that's easily possible, but I'm pretty sure nobody has come forth with such a board.  I think someone did for the original AVR based boards, but not the newer ones.  If you just want to do the typical AVR mod, but you want to be sure to have the ISP header, floobydust's board has one.  I'm pretty sure that the adapter I posted for the D06 is the only one in the thread that has no ISP header.

[DGM]

I got sick of all the ambiguity and apparent contradictions, so I spent yesterday transcribing gerbers and going through this entire thread to compile a table of known boards and adapters.  I'm only interested in the non-AVR clones (20-pin Samsung/Zilog/whatever MCU).  I have identified 7 unique board layouts and 12 adapters. Everything is in a spreadsheet with an index of reference links to posts throughout the thread wherein board examples/info are shared. Links and info are provided for all shared boards on OSH Park's site. Here is a preview of the first sheet.  I may keep updating this spreadsheet in the future.  Don't assume that the preview picture stays up to date.

EDIT: I added a section for references to AVR-based boards as well.  (no mappings yet)



There are many points of uncertainty that come with the informality of a forum.  A lot of individual users never respond with mention of mod success or failure.  A lot of photos have board names or rev dates obscured.  There is at least one seemingly contradictory mapping given for the 858D04 board (poorman), and there is a unique mapping given for a Yihua board variant (YH858D V?), but the actual version number is never stated clearly. I've tried to make some notes to keep track of these issues.

Disregarding the one oddball D04 mapping, the 858D04 and 858D06 boards are not the same.  They differ in that the enable lines for digit 1 and digit 3 are swapped. Some D06 owners seem to be using (or were attempting to use) adapters for the D04, but nobody ever mentions this mismatch.  I'm going to assume people are just changing the pin assignments in the code.


After dredging through all this confusing ambiguity, there's one last thing I'm compelled to say. Maybe it's a bit late in this thread for a PSA, but I'm going to be a jerk and put it in giant text anyway because it drove me nuts:
STOP IDENTIFYING THE BOARD BY REFERRING TO THE MCU NUMBER
Asking "my board has a MK1840, will it work with the adapter for the MK1841?" is like asking "my car has a V6 engine.  will the windshield wipers from a V8 car fit?".  It's absurd.

If your board has any 20-pin non-AVR MCU on it, don't even bother reading the part number.  That number is of absolutely no use.  These are configurable parts; there's no justification to assume that identical parts have the same pin assignments in their firmware.  That's all moot anyway, since you're going to be replacing it. Unless you want to change a bunch of pin assignments in the new firmware, what matters is the actual PCB layout.

For example, these mutually incompatible boards were found with the same MCU marking:
YH858D V2 with S3F94C4EZZ
YH858D V7 with S3F94C4EZZ
youde 858DV4 with S3F94C4EZZ
AF858D with S3F94C4EZZ

and these compatible boards were found with different MCU markings:
858D06 with MK1840D3
858D06 with MK1841D3
858D06 with blank MCU

Refer to your board by whatever name and revision is printed on the silkscreen, or any other attribute that unambiguously identifies the board.  Even better would be to use multiple pieces of information and/or take a good photo. 

If you run across a board that has no prior examples known, the only way to determine the pin mapping is to transcribe the board. The MCU number will not tell you.

[bitseeker]

Cool reference, DGM. Thanks for putting that together.

[DGM]

Cool reference, DGM. Thanks for putting that together.

yw
just so you know; i might keep updating the file frequently for the next day or until i stop obsessing over minutia.

[cowana]

Thanks DGM, that's an excellent summary this thread has been lacking!

The important bit you're missing from me is the name of the 'unknown name board' - as shown in the attached picture, it is marked AF858D, and dated 20161212.

In case you want to add it, the details of the adapter board I developed for this PCB are below:

• Schematic: attached
• PCB: https://oshpark.com/shared_projects/NQGwEEH0
• Features:
• TL431 reference
• Vfan and current sense support
• ISP header
• MLF-32 ATmega package

The ADC scaling factor for this board when used with a 2.5v reference was 2.4 in my case.

[DGM]

Thanks DGM, that's an excellent summary this thread has been lacking!

The important bit you're missing from me is the name of the 'unknown name board' - as shown in the attached picture, it is marked AF858D, and dated 20161212.

In case you want to add it, the details of the adapter board I developed for this PCB are below:

• Schematic: attached
• PCB: https://oshpark.com/shared_projects/NQGwEEH0
• Features:
• TL431 reference
• Vfan and current sense support
• ISP header
• MLF-32 ATmega package

The ADC scaling factor for this board when used with a 2.5v reference was 2.4 in my case.

holy shit!  you knew just what i wanted!  your board was a big uncomfortable mystery, and i was contemplating PM'ing you about it.  as old as this thread is, i was reluctant to bother people about old posts.  thanks.
I'll probably get this into the next update later on tonight or tomorrow (or whenever i'm sober enough to see straight again). 

[bitseeker]

(or whenever i'm sober enough to see straight again)

Yes, please.

[DGM]

In trying to sort out conflicting information, I took a closer look at Jetiman's photos to try and transcribe what I could.
https://www.eevblog.com/forum/reviews/youyue-858d-some-reverse-engineering-custom-firmware/msg873591/#msg873591

I've updated the index with my tentative conclusions, but as an aside, I noticed one thing that I thought was curious.  There appears to be no resistor on the input side of the optotriac.  All the resistors over on that edge of the board are for the calibration network, the pushbuttons, and the TC amplifier input pullup.  The LED is being driven straight from 5v and being pulled to ground with a transistor.  At ~2x its maximum allowable input current, I have to wonder how long that'll last. 

Just thought I'd throw that out there.

[Miwer]

There are many points of uncertainty that come with the informality of a forum.  A lot of individual users never respond with mention of mod success or failure. 

Hi DGM,

That's a nice overview you created there, thank you for that.
I am in fact working on an adapter board for the 858D V1.0 board (or the "Hylko" branded 858D, as I tend to call it). I have even made the first attempt with a DIY single side prototype. However I'm having problems getting the temperatures and fan speed reading correct, and I'd like to have a working version without firmware modifications if possible, though I'm not sure I can avoid it. Also the spare time I can spend on this project is limited. But when I get something going, I'll be sure to post it here.

[DGM]

That's a nice overview you created there, thank you for that.
I am in fact working on an adapter board for the 858D V1.0 board (or the "Hylko" branded 858D, as I tend to call it). I have even made the first attempt with a DIY single side prototype. However I'm having problems getting the temperatures and fan speed reading correct, and I'd like to have a working version without firmware modifications if possible, though I'm not sure I can avoid it. Also the spare time I can spend on this project is limited. But when I get something going, I'll be sure to post it here.

Yeah, I'm thinking all the 20-pin clones are going to at least need ADC scalings and min/max limits changed in the firmware at the least.  Then again, I can't be sure of that.
Whenever you get something that works, I'll add it to the index. 

I'd decided to finish an aborted adapter for my "youde 858D4" board, and I kind of started throwing together adapter layouts for the other orphan boards.  Anything I put together for boards I don't have is liable to have bugs in it though.  I figured that editing a complete but unverified project file would be more convenient than starting from scratch.  Of course, it'd be better to have a known working layout, so I'd like to see whatever you come up with.

[DGM]

Okay, two things:  In all the disambiguation of the Yihua/WEP boards (major thanks to bitwelder), I'd started keeping notes regarding what board modifications certain boards would need.  I figured I might as well review what I have, since it seems fairly well nailed down at this point.  Also, this got me thinking that I should finish the half-baked "youde 858DV4" adapter I had started to lay out.  In the end, I came up with adapter layouts for the other orphan boards.  The board/adapter index has been updated accordingly.  Barring any errors in my work, this should tie up most of the loose ends in this thread.

REQUIRED BOARD MODIFICATIONS
Certain boards require modification to be adaptable without major firmware changes.  While the necessity for a given mod is described, the methods given are not the only possible way to remedy the problems.  These are simply the methods implied by the specified adapter concept.

General
The firmware expects some fan status input from all boards; these modifications are not board-specific.  Each method will require some configuration of the code to select the sensing method and define/tailor the expected min/max ADC limits. 

Fan voltage sensing can simply be done at the node connected to the positive pin on the fan connector (denoted by A).  Others have recommended to connect to the positive-most end of the pot (B), though I don't see the advantage.  Where ever the sensing is done, ensure that the selected divider resistors keep the ADC voltage within safe limits.  Current sensing is done by adding a shunt resistor in the ground path of the fan and measuring voltage across it (C).   Fan speed sensing requires extra circuitry to detect commutation noise, and is not covered here. 



YH858D V2, V6 (Yihua/WEP)
These boards use pins 17 & 18 for both button sensing and display element drive.  The mod required by the existing adapters (wguibas' blue adapter) is as follows:
- Traces between pins 17, 18 and the display should be cut.
- Wire should be routed between display pins 7, 10 and the adapter board.
User wguibas has posted instructions and images for adapting this board.

Additional mods: It may also be prudent to provide the TIP122 with a heat sink to keep Tj within allowable limits

YH858D V7,V8 (Yihua/WEP)
These boards use pins 12 & 13 for both button sensing and display element drive.  The mod required by the current adapter (DGM type 3) is as follows:
- Traces between pins 12, 13 and the switch resistors R24, R25 should be cut (or the resistors should be removed).
- Traces between pins 15, 16 and the unused EEPROM should be cut (or the eeprom removed).
- Wires should be routed between pins 15, 16 and the ungrounded switch nodes.
With internal pullups, switches should work as expected, and no extra flying wires to the adapter are needed.



Additional mods: The optotriac appears to have no current limiting resistor.  This would put its operating current over 100mA. These inputs are only rated to about 60mA.  Consider placing a resistor here.

youde 858DV4 (ZENY, etc)
This board uses a single ADC input to sense both buttons in a manner which prohibits the use of button chords as required by the new firmware. The original post describes the physical circuit layout and a potential mod to allow this ADC-sensing method to support chords. The current adapter (DGM type 2) uses a more conventional sensing method and requires the following modifications:
- Replace R21 with a jumper
- Remove EEPROM (or cut trace)
- Cut trace to isolate DN button from pin 11 node
- Add jumper from DN button to pin 18 or 19 (DGM adapter assumes 18)
This leaves 10k external pullup resistors in place for both buttons



Additional mods: TIP122 has no heat sink.  At ~2.5W dissipation, this needs a heat sink to keep Tj within allowable limits.

ADAPTERS FOR ORPHAN BOARDS
As I've stated before, these layouts come with caveats.  These are laid out with a mind toward hand etching/drilling (wider traces, no vias under parts, no attention paid to silkscreen, etc).  They may require changes to meet what a board house would want; at the very least, you'd want to clean up the silkscreen layers if you have someone else print it.  Also, any layout I throw together without being able to test it on the appropriate board is liable to have bugs in it.  I offer these on the assumption that editing a complete but unverified project file would be more convenient than starting from scratch. Just do your due diligence and verify that it matches your actual board.  If there's an error in the adapter or the board map from which I derived it, let me know. 

I went ahead and shoehorned an ISP header on them since I'm probably the only person that doesn't want one.
The YOUDE adapter has a TQFP version, but there's really no point to using it unless those are the only parts you have.  The extra vias would only make it more difficult to drill and assemble.
I tried to make sure the adapters clear obstacles on the board, but some things (transistors, small caps) might need to be laid down depending on how tall your headers are.
These are all Eagle project files.

YOUDE 858DV4 adapter


YH858D V7-V8 adapter


HD858D V1.0 (Hyiko) adapter

[anishkgt]

Any chances for the Atten 858D+. I just bought one from eBay and that seems to be a faulty one. Was hoping it would be fixed by the custom firmware mentioned in this thread but uploading it did not help, no display just the blower switches ON.

I have the PIC here in the repair thread (https://www.eevblog.com/forum/repair/new-atten-858d-received-with-faulty-display/msg1423943/#msg1423943) and the PCB looks almost the same.

[DGM]

Any chances for the Atten 858D+. I just bought one from eBay and that seems to be a faulty one. Was hoping it would be fixed by the custom firmware mentioned in this thread but uploading it did not help, no display just the blower switches ON.

If you think the display is faulty, pull the MCU and directly test the display in place. 
Just supply 5v to the CA pins (through the anode resistors!) and probe ground to each of the segment buses. 
Make sure it's the display before messing with trying to replace it. 
It's entirely possible to have damaged I/O sections on a MCU.
I've seen that more often than I've ever seen a dud display.

edit: if you end up having to replace it, look for something similar to the MCD4301Bx displays
10mm/0.4" char height, 12.7mm row spacing, common anode
there are probably a zillion chinese manufacturers with similar naming schemes
https://duckduckgo.com/?q=4031bs&ia=web
so long as the row spacing and pinout match
an ebay listing:
https://www.ebay.com/itm/0-28-0-36-0-4-0-56-0-8-Red-led-Display-7-Segment-Common-Cathode-Anode-1-4-Digit/253205048535?hash=item3af4326cd7:m:mkJWKwh1lzXl37aWlZDIHDw

edit again: if you're using that schematic for reference, the pin numbers are wrong there.
pin 6 is unused or clipped off, but it still counts:
1 e
2 d
3 DP
4 c
5 g
6 NC
7 b
8 CA3
9 CA2
10 f
11 a
12 CA1

[anishkgt]

What about the firmware. I will be sending it back if the seller would take it. Else I could repair the faulty display. Oh and if it’s MCU then I hope the seller takes it back and without the firmware the unit is just garbage unless I could figure out a way to upload a custom firmware on Atmega328P.

Is there a custom firmware for this model ?


Www.Georgehobby.wordpress.com

Equipments: DSO104z, Hakko FX888D, Brymen BM869s

[DGM]

i don't know about the atten units, but iif it's AVR-based, t's probably not too far from working with the firmware that's the topic of the thread. 
you'd just have to make sure the pin assignments and ADC scalings match. 
https://github.com/madworm/Youyue-858D-plus/blob/master/youyue858d.h

I've really only been paying attention to the non-AVR clones, so idk off the top of my head if there are any outstanding issues with the atten that would require code modification or otherwise make it a problem.
Maybe someone else knows.

[anishkgt]

I've downloaded the firmware by Madworm on this forum. (https://github.com/madworm/Youyue-858D-plus) and not sure if i should burn the fuses ? i had uploaded the sketch via arduino but that did nothing. For the display i had been looking at these (https://www.ebay.com/itm/Digital-Tube-LED-Display-Red-7-Segment-Module-Common-Anode-Cathode-Display/272590524676?ssPageName=STRK%3AMEBIDX%3AIT&var=571707668128&_trksid=p2055119.m1438.l2649)