Youyue 858D+ some reverse engineering + custom firmware

[x2net]

Sorry Bro, We are using SI units only. It's easy!

[tablatronix]

Word of advice, I have a yihua 858D, I had to add my own rubber feet, because the ones on it ate the finish off my desk, those feet will probably gum up over time also if i had to guess, crappy rubber.

[bitseeker]

Eek! I guess I should check the feet on mine.

[DGM]

Word of advice, I have a yihua 858D, I had to add my own rubber feet, because the ones on it ate the finish off my desk, those feet will probably gum up over time also if i had to guess, crappy rubber.

i didn't look at the feet, but the blower bushing in mine is that nasty black rubber that sweats acrid brown mystery oil. 
i've seriously been considering making a copy with some silicone to cut down on the smell.

[kPATm]

Just took delivery of mine today.

https://www.ebay.co.uk/itm/858D-220V-SMD-Soldering-Desoldering-Station-Hot-Air-Rework-Gun-Tool-3-Nozzles/301966004104?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

Had a quick look inside and it seems pretty well made. I took a pic of the board

[ncx94]

Just took delivery of mine today.

https://www.ebay.co.uk/itm/858D-220V-SMD-Soldering-Desoldering-Station-Hot-Air-Rework-Gun-Tool-3-Nozzles/301966004104?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

Had a quick look inside and it seems pretty well made. I took a pic of the board

Looks like the same one I have.

nc94

[kPATm]

Did you mod it mate. I havnt used it yet as its a birthday present from the Mrs. Need to wait :-( How does it function as standard? AND IS IT SAFE

[ncx94]

Did you mod it mate. I havnt used it yet as its a birthday present from the Mrs. Need to wait :-( How does it function as standard? AND IS IT SAFE

No because nothing is available yet for that one. Works great though, and mine was grounded correctly even though I wouldn't say such a pcb design could ever be safe

nc94

[kPATm]

Dont mind about being able to mod it if it works OK. How accurate is it? I wouldnt mind a few pointers in checking its safety?

[DGM]

After condemning my YOUDE 858DV4 board in a previous post, I ordered a different board from the stated aliexpress source.  Indeed, the board I received was an 858D06 board.  To fit in the ZENY enclosure, I had to remove two of the welded pcb standoffs.  I also flipped the 3-pin header for the hook switch and TC so that it matched what I had.  To avoid having to make new buttons for the different sized switches, I removed them and surface mounted the switches from the YOUDE board. 

After running an initial calibration cycle and letting it cool back down, I tested the overshoot to compare it to the old board.  With the large nozzle and 100% air, or with the small nozzle and 50% air, overshoot for a 300C setpoint was 40-60C.  I think everyone would complain about that, but considering the original YOUDE board's firmware was causing a 150-200C overshoot, I'm rather pleased.  While it does to lock the display and simply show the SP once it's been reached, it does appear to employ some form of rate compensation behind the scenes.  All in all, it's a nice, clean double sided board.  The darlington has a heat sink, and the firmware isn't complete garbage like the YOUDE was.

While I've drawn up an adapter layout for the 858D06 in eagle, I'll probably use it as is for now.  Like I said, 50C overshoot is a far cry from 150-200C overshoot.

[floobydust]

Did you mod it mate. I havnt used it yet as its a birthday present from the Mrs. Need to wait :-( How does it function as standard? AND IS IT SAFE

No because nothing is available yet for that one. Works great though, and mine was grounded correctly even though I wouldn't say such a pcb design could ever be safe

nc94

858D06 and 04 both have scary creepage/clearance fails between mains and the low voltage side. Not a safe design.

[ncx94]

Did you mod it mate. I havnt used it yet as its a birthday present from the Mrs. Need to wait :-( How does it function as standard? AND IS IT SAFE

No because nothing is available yet for that one. Works great though, and mine was grounded correctly even though I wouldn't say such a pcb design could ever be safe

nc94

858D06 and 04 both have scary creepage/clearance fails between mains and the low voltage side. Not a safe design.

Yes I know that's why I said such a design could never be safe ^^


nc94

[DGM]

considering the scope of this whole discussion, i think there's a realm of design defects that are just going to be accepted unless we go back to talking about complete board replacement.
i don't think i've seen any 858 clone layouts that were "good" in anything but a relative sense.

i'm pretty sure the worst spacing on the 06 board is in the corner.  creepage distance to the front panel standoff.
i'd be less worried about the LV system getting destroyed than having the (poorly grounded/ungrounded) enclosure energized.

having reduced those issues as much as I can with rewiring, bushings, copper removal and conformal coating
i think i'll take poor creepage distance to a low-voltage system with no exposed conductors. 
in my mind, that's safer than firmware that tries to melt the handpiece and start a fire every time you pick it up.
on a dirty board with a fan circuit that's just waiting to die -- and melt the handpiece and start a fire.

i felt i had  to point out that the 06 i got was "clean", which it was in comparison to the YOUDE 858V4
i don't think anyone ever mentions the spacing violated by having solder balls and dross stuck all over the board.

[ncx94]

I can only agree with you on that.. I wish I had the skills to do the high power part of the pcb... The other part is easy but I don't have the knowledge to do the HV part...



nc94

[floobydust]

Designing a safe replacement main PCB, with a decent MCU and better display, the issue is the rest of the station would remain unsafe.

The GX connector, the wand ground wiring and the wand cable- would never pass regulatory for mains use.

I run my Youyue 858D06station from a GFIC and that lowers risk enough for me.

[technogeeky]

I just wanted to check in. I got the "HFS" 959D model from the jolly old fat man we all know and love.

This one has the three memory buttons (CH1 CH2 CH3).

I took a peek inside and I found a green PCB with a socketed MCU. I used a NIST traceable and recently calibrated (in November) Thermapen held (with my hands, not a stable setup) about 1 inch away from the nozzle, with no attachment nozzle. I did not test at the pre-programmed 300 °C because the thermapen sharply cuts off there.

I found:

• 100 °C measured 105 °C
• 200 °C measured 198 °C
• 290 °C measured 289 °C

I will do a more detailed review later, with pictures and identification of the board. Some early thoughts:

• The auto shut-off feature is surprisingly nice!
• The nozzles that came with this are the bolt-nut tightened versions, contrary to the review on Amazon.
• There is a flashing dot at the bottom right of the digital display. This seems to indicate when the control feedback has stabilized. It flashes quickly and then pauses for a moment.
• The temperature control is fast and active: If you blow into the nozzle with varying strength, the temperature drops significantly and can be seen to recover quickly.

Some notes about the cosmetic appearance of the unit I got:

• The faceplate does not say YIHUA. In fact, it doesn't even say 959 or 959D.
• The buttons are red, not blue.
• The knob is black, plastic, bell-shaped, and attached with two hex nuts. (as opposed to faux aluminum, cylinder shaped, and presumably attached by clamping pressure alone)

Other than the fact that the wand unit is permanently attached, this unit may be the best starting points for these type of hot air guns -- and therefore may be the best and cheapest starting hot air unit.

Please let me know any questions or specific things you want checked out from the teardown.

[DGM]

IDK what everyone else wants, but my curiosity is to merely see if this variant breaks the trend(s) of known flaws.  It seems to some degree it's already looking above average, but I suppose I can throw down a few points of curiosity for any freshly discovered clone.  I have no idea if you even intend on bothering with figuring out pinouts or schematics, but consider it a general statement of my own general curiosity regarding this entire thread.

is the chassis/handpiece grounded?
is the switch/fuse in the phase or neutral?
is the MCU a Zilog/Samsung or can we hope for an AVR again? (not holding my breath)
does the MCU share any known pinouts (may be of interest to a prospective buyer looking to mod, but in this case, i'd assume it doesn't)
are there significant changes to the fan control and TC amp circuits? (this might influence how any mods are adapted to the board)
are the power semiconductors provided with a heat sink? (the darlington in my clone wasn't)
are there any terrible layout decisions made regarding isolation between mains & low voltage systems or mains & chassis?
what about the dynamic response; how bad is the overshoot from cold?

maybe other people expect more perfection out of a dodgy chinese appliance that smells like shredded tires, but i'm just out to see if it can perform repeatably without incinerating itself or the work.   that's enough of a challenge to start with.

[biohazrd]

Just got mine in.

Looks like a 858D06 PCB (date 2017.03.04) with a MK1840D3 micro.



From what I can tell, it looks like the big changes between the D06 an the D04 PCB is the use of surface mount components for most resistors and other passives.

I saw some posts back in the thread that the MK1841D3 might be pin compatible with the MK1840D3 but no confirmation. And that this OSH park board might be the one I need, or this board that was posted in the thread.

Does anyone have any experience with this particular micro before I go forward?

[DGM]

I don't see how differences between MK1841 and MK1840 really matter here.  Pin-compatibility of the MCU doesn't mean an adapter will work correctly with a given board.  What matters is the layout of the board. 

This is the schematic and pinout for this the 06 board:
https://www.eevblog.com/forum/reviews/youyue-858d-some-reverse-engineering-custom-firmware/msg1349350/#msg1349350

floobydust also posted details of an adapter specifically for the D06
https://www.eevblog.com/forum/reviews/youyue-858d-some-reverse-engineering-custom-firmware/msg1206460/#msg1206460
Keep in mind that the pin assignments on the AVR correspond to the TQFP package, not the PDIP package.

I don't remember the specifics of what those other adapters went to.  The OSH park link is incomplete,
You'll basically have to look at the adapter layout and figure out what nets connect to the AVR pins. 
EDIT:  The board made by wguibas in your second link won't work for the D06 board.  Assignment of pins 2 and 7 are swapped between the D06 and whatever board he was using.  You may be able to make it work if you change pin assignments in the firmware.
EDIT AGAIN: If you want a summary of board and adapter compatibility, check here: https://www.eevblog.com/forum/reviews/youyue-858d-some-reverse-engineering-custom-firmware/msg1418224/#msg1418224


These are my own notes for this board. Might want to double check your physical board, but it's easier than trying to sort through this entire thread and all the myriad incompatible bits of information:

Code: [Select]

map for MCU adaptation (D06 BOARD)
NAME ATMEGA168 MK1840D3
PB0 digit 1 14 2
PB1 triac drive 15 6
PB2 down switch 16 9
PB3 NC 17
PB4 reed switch 18 4
PB5 up switch 19 8
PB6 digit 3 9 7
PB7 digit 2 10 3
PC0 ADC temp sense 23 11
PC1 NC 24
PC2 NC 25
PC3 fan on/off 26 10
PC4 NC 27
PC5 fan sense 28 FAN SENSE CKT
PD0 A 2 15
PD1 E 3 16
PD2 D 4 17
PD3 F 5 14
PD4 DP 6 5
PD5 C 11 18
PD6 G 12 19
PD7 B 13 13
VCC 7 20
AVCC 20 20
GND 8 1
AGND 22 1
AREF external vref 21 AREF CKT
RST 1 terminate
AREF needs to be provided:
- VCC/2 divider (no supply rejection)
- LPF+TL431 or similar (e.g. LM336) (see schematic of original board (below) for example)
https://github.com/madworm/Youyue-858D-plus/blob/master/Docs/Datasheets/858D_RevEng_Schematic.pdf

ADAPTING THE CODE:
using TC amplifier output range for D06 board (from schematic pdf)
round([0 1.217 2.44] * 1024/2.5) = [0 498 999]
this demonstrates conversion of the full range of the pin voltage to the full range of register values
use this to adjust lines 104,105 in youyue858d.ino and the default value in the .h file (and possibly other values i missed)
by calculating the relevant intermediate register values for min/max/default, etc

the original board (and default code) assumes raw adc values are scaled to be equivalent to degrees C!
the required rescaling factor is based on amp output slope (4.88mV/K) (see schematic pdf)
4.88E-3 * 1024/2.5 = 1.9988
this scaling factor needs to be used for adjusting the calculation of temp_inst on line 197 in youyue858d.ino

Code: [Select]

temp_inst = (adc_raw / 1.9988) + temp_offset_corr.value; // approx. temp in °C
similarly, FAN SENSE needs to be provided based on whether you're measuring fan voltage or current.
the min/max register values have to be set accordingly

if measuring the fan voltage using a voltage divider:
given original version voltage divider (10k and 330R) for [9 32] volt range on Vfan (again, from the schematic pdf)
adc register value range is:  round([9 32]*330/10330 * 1024/2.5) = [118 419]

if using the current sense mod:
fan current sense mod is a 1R resistor in the ground connection for the fan
default limiting adc register values are [30 71], coresponding to [74 174] mA  (lines 155,156 in the .h file)
if the range of fan currents extends beyond these min/max values, or if a different resistor value is used, the min/max values have to be recalculated to correspond to your hardware

i don't really see that by default FAN SENSE is used for anything other than startup testing and PV snooping via the display
so i don't really see the point unless you're out to make some more complicated calculations based on the fan state.

also, all these calculations are based on an assumed 2.5v reference voltage
everything should  be adjusted based on actual AREF provided by the adapter board.

[DGM]

For the sake of completeness, I'm going to go ahead and offer my Eagle files for the 858D06 board. 
I kind of tailored it to my own use with no intention of anyone else seeing it, so idk if anyone would want it.

Caveats:
I have not had a chance to print this, so it has not yet been physically proven. 
I did not include a programming header, because I didn't want one.  I tend to just use a socket.
The layout is based on the assumption that I would print/etch/assemble it myself; it does not meet the requirements of a board house (and the silkscreen layers are garbage).
If printing/drilling this yourself, keep in mind that there are some topside pin connections that need to be made under the IC socket (if used)

Unique bits:
This adapter uses a LPF and voltage reference (LM336M) for the AREF source
There is provision for an end-of-cycle beeper (common 13mm magnetic buzzer)
The routing minimizes vias and topside connections (convenient when plated through-holes are unavailable)

[kelchm]

Just received my Youyue 858D+ that I ordered from Amazon. Seems to work well and be of a decent build quality. I've also received the 858D06 board that several other users reported.

One oddity that I'll note is that the grounding wires are soldered to crimp connectors. At least they didn't tin the wire and then crimp them.

[kulla]

Any of you had issue that heater is aways on? 

[stj]

dont think so,
that was a problem with the older style with a pump in the base though.

[DGM]

Any of you had issue that heater is aways on?

Given how frustrating it is to do searches within threads on this forum, I'm going to just assume you're not trying to troubleshoot a modded 858 build you've mentioned prior. 

Assuming the controller isn't actively turning the heater on constantly, it's certainly possible to have a triac stuck on when it's failed.  That much is easy enough to diagnose with a multimeter while off. 

Otherwise, like my experience with the ZENY clone with the YOUDE 858DV4 board, it could simply be horrific overshoot and faked user feedback making it seem like the heater gets stuck on.    I know it sounds dumb, but it might be a good idea to actually measure and make sure the heater is drawing current when you think it is.

Finally, the worst thing I can imagine is that the unit is wired with the switch, triac, and fuse in the neutral (like they all seem to be), and for some reason the heating element shorted to the nozzle (earth).  If there is no RCD/GFCI in the branch circuit, that'd result in the heater being stuck on.

[kulla]

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.