HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron driver

A place to document your own projects.

Re: HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron dri

Postby ian » Fri Nov 18, 2011 2:06 am

I always prefer an integrated board, daughter-cards need connectors and make a project less professional. IMHO only though :)
User avatar
ian
Crew
Crew
 
Posts: 10803
Joined: Mon Jul 06, 2009 6:14 am

Re: HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron dri

Postby arhi » Fri Nov 18, 2011 3:12 am

Yes, that is true. The stacked pcb's can look very professional but I'm not sure if this would be the case ... I'll see what I can push on a single board, but I fear smd will be required :D, or at least a bigger board
User avatar
arhi
Hero Member
Hero Member
 
Posts: 2160
Joined: Thu Jun 24, 2010 11:41 am
Location: Belgrade, Serbia

Re: HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron dri

Postby ian » Fri Nov 18, 2011 3:24 am

It is tempting to push this project to be through hole because beginners can DIY a soldering iron, but maybe the 28pin and 16F versions are the beginner 'lite' version, and it is ok to have an SMD advanced version. I know I would certainly be happier doing this project with a 44pin TQFP PIC and SMD small passives :) Making room for all the PTH stuff is a pain, especially a 40 pin .6 DIP chip.
User avatar
ian
Crew
Crew
 
Posts: 10803
Joined: Mon Jul 06, 2009 6:14 am

Re: HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron dri

Postby arhi » Fri Nov 18, 2011 3:27 am

well, I can make a small changes on the 16F version so that both PTC and TC work properly (PTC is too noisy with current design) and then do a 18F version in smd to support fancy options and dual irons :D
User avatar
arhi
Hero Member
Hero Member
 
Posts: 2160
Joined: Thu Jun 24, 2010 11:41 am
Location: Belgrade, Serbia

Re: HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron dri

Postby arhi » Mon Nov 21, 2011 12:38 pm

Here's the upgrade of the 16F version. It uses different (proper) path to read HAKKO (and any other RTD/PTC sensor) and uses a jumper to select what signal is going into pic (PTC or TC). The DS is replaced with NTC. I used a quad op-amp instead of dual one and I had one free so I used it to buffer NTC output.

It should work perfectly with HAKKO too now (the older version worked perfectly with solomon and "just worked" for hakko). I'm doing some tests tonight so tomorrow I'll post up some graphs with this changed schematic :)
User avatar
arhi
Hero Member
Hero Member
 
Posts: 2160
Joined: Thu Jun 24, 2010 11:41 am
Location: Belgrade, Serbia

Re: HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron dri

Postby arhi » Tue Nov 22, 2011 12:19 am

Just few more tests but looks like this is a final schematic for the 16F690 version.

I didn't make PCB for it. I'm still testing it on stripboard+pcb+some dead bug design :D but I should have final results in few hours. Then maybe Arakis can try to make one sided board out of it (I suck big time in making one sided boards so I will let him practice :D ) who knows :D .. anyhow, I don't suggest any board making before I do the final tests :D

Some hints (should be obvious from schematic but)

  • I removed schmidt inverter, the simple RC is there now, should be easier to fit on one sided board (and tad cheaper)
  • Board can be powered by AC or DC.
    For AC operation
    • connect AC source to POWER1 and POWER2
    • Close JP_AC (JP2) and open JP_DC (JP3)
    • You do not need to mount U6, R38, D9 and Q2
    For DC operation
    • connect DC source to POWER1 and POWER2 (POWER2 is PLUS, POWER1 is MINUS)
    • Open JP_AC (JP2) and close JP_DC (JP3)
    • You do not need to mount U3, U4, R32, R33, R34, R35, C14
  • Board can read PTC/RTD and K-type thermocouple as a main temperature probels
    For PTC/RTD operation (HAKKO, WELLER etc)
    • Switch JP1 so that PTCOUT is connected to TEMPERATURE (close 1-2)
    • You do not require R11-R22, C6-C8, C12, C13
    • The R27 and R28 define gain in this stage. Gain for HAKKO is 40, for another PTC sensor you need to change this gain to get best results.
    For TC operation (Solomon)
    • Switch JP1 so that TK_OUT is connected to TEMPERATURE (close 2-3)
    • You do not require R23-R28, C9-C11, D1, Q1
  • U4 is any Triac that can handle power of your soldering iron, this is in most cases 24VAC and 2A so almost any triac will do
  • U6 is some npn optocoupler, depending on the current required trough the opto's input adjust the R37
  • Q1 is general use PNP transistor, taking into account that max voltage here is 5V and collector current is 1mA any pnp should do
  • Q2 is any Ntype FET that can handle 30V and 2A, most can handle that without a problem
  • There is NTC input on the board too. NTC can be used to detect ambient temperature required for TC or it can be used to detect a secondary temperature source (for e.g. I use ntc to read temperature in the TIP itself so I can do a better calibration). The resistors in the U2:A are defining the gain of the NTC, in most cases you need gain of 1 so R30 should just be short and R31 should be left open, but if a NTC that requires gain is required you can define gain with those two resistors
  • E1 and E2 are inputs from rotary encoder, EB is a "menu button", if you have rotary encoder with button - use it, if not mount button separately
  • B1 and B2 are secondary buttons, have no function attm but could be used as "pedal boost" (turn on heater and ignore pid) that you can use when you need extra heat (soldering to220 back to pcb for e.g.), and the other button can be used for e.g. to turn heater on/off without turning the whole device off - or as some fancy devices do, it can be a switch mounted in the iron carriage so when it detects iron in carriage it lowers the temp of the iron to 50-60% so that your tip lasts longer...
  • J6+J7 are for debugging purposes only exposing UART
  • Calibration, D1 is voltage reference diode, depending on diode you are using chose the CAL1 pot (basically you need to have resistor there that is similar to reference voltage of LM385 so if you use LM385 with 1.24V reference use 2k pot so you can easily find 1k24 position, or if you use 2.5V reference then 5k pot will be easiest). To calibrate device initially, disconnect sensor and instead sensor connect milliampermeter. Turn the CAL1 until you have 1mA flowing trough PTC1/PTC2

Now I'm off to do that final test (so far this works perfectly with original Hakko iron, problem is that some clones are very low quality so I'm about to test few of them too)
Attachments
sh1.png
sh2.png
sh3.png
SolderingIronControllerGen2.PDF
Soldering Iron Controller GEN2, V4.0
(185.24 KiB) Downloaded 2013 times
User avatar
arhi
Hero Member
Hero Member
 
Posts: 2160
Joined: Thu Jun 24, 2010 11:41 am
Location: Belgrade, Serbia

Re: HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron dri

Postby arhi » Tue Nov 22, 2011 3:52 am

Here it is, RAW ADC - no averaging, no LF filter in software - pure adc data - and it's smooth as a baby's butt :D

It looks like the gain of 40 is too much as I see some clipping when 270C marker was hit (my NTC goes only up to 270C) but everything looks perfect. The only change is that value of R28 need to be reduced so that we can measure up to 500C. Since other PTC sensors (for e.g. Weller use PT20, or to be more correct, 5 PT100 sensors wired in parallel) require different signal amplification maybe a 500K trimmer could be added instead of R28.

All in all, both sensors work perfectly now, Ktype and PTC on Hakko iron.

What could be noted is that I'm using only half of the ADC swing (2.5-5V) as values below 2.5V are below room temp, most of the region we are interested in is between 4V and 5V. In order to be able to use DC power also I can't add a negative 2.5V to the signal in order to pull it down and stretch it so that the interesting temperature region spreads trough most of the ADC range, but taking into account precision required for soldering and that we have a sensor in heater and not tip, it really makes no real difference...
Attachments
screenshot18.png
User avatar
arhi
Hero Member
Hero Member
 
Posts: 2160
Joined: Thu Jun 24, 2010 11:41 am
Location: Belgrade, Serbia

Re: HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron dri

Postby arhi » Tue Nov 22, 2011 6:31 am

More graphs. Here's the graph with (finally) real temperatures.

Note that red line is the temperature of the tip measured with NTC. Since the 320C is the absolute max of the NTC (it is accurate up to 270C) you can notice the clipping on 400000ms and around 500000ms (the flat tips).

As you can imagine, the heater was 100% on, then fully off, then on, then off, on, off and then it cooled to room temperature (I turned logging off bit before room temp but irrelevant).

Important facts from the graph
1. heater is always hotter then tip (yeah right :D )
2. heater heats up faster then tip (as you can see the heating curves for the heater are steeper then for the tip). This is caused by the heat transfer delay between heater and the tip.
3. heater and tip cool at the roughly the same pace (the cooling curves are almost identical)
4. peaks are very interesting, the second and third peaks don't show much as the tip temperature is clipped but first peak show both delay and temperature difference between the tip and the heater.

Way more informations will be available when I run the same setup only with PID control and not manual on/off so that we can see how fast the tip will catch up to the heater and how tip temperature drops when the soldering is involved :)

One not so obvious thing that can be seen on the graph is, if you look at the end of cooling period between second and third heating cycle you see on the TIP chard that it cooled "faster" then usual, then started to heat up and then it started cooling again. I will upload also the zoom of this region so it can be more obvious. What happened here is that during cooling period (no power on heater) I pushed the dump sponge on to the tip of the iron, as you can see it started to cool rapidly, and no noticeable effect on the heater temperature can be seen. Then, when I removed the sponge, even without any power, tip starts to heat up because heater is still way hotter then the tip and then as they go into equilibrium they continue to cool together until I initiate the heating process again.

I'm off now to try to assemble the firmware again so that I can have both pid and serial output and NTC in the same firmware but I am afraid that it will not be possible since the flash in 16F690 is very tight :(
Attachments
screenshot19.png
screenshot19.png (7.88 KiB) Viewed 14126 times
Screenshot.png
User avatar
arhi
Hero Member
Hero Member
 
Posts: 2160
Joined: Thu Jun 24, 2010 11:41 am
Location: Belgrade, Serbia

Re: HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron dri

Postby arhi » Wed Nov 23, 2011 2:21 am

[rambling]
I can't push everything I want into 4k without moving to assembler or exchanging 16x2 to 7seg's ... and I don't want to do either so I will now move immediately to GEN3 and 18F ... I spent some time deciding what 18F to go with and I just decided that I do not need more pins, this as is works great, and I have already 2 extra buttons, one could be used for sleep and other for foot pedal. So attm I'm in lookout for 18F with 20-28 pins. I know Ian wants USB but I really have no need for usb on my soldering iron. Serial port that I can access via bus pirate is more then enough. On the other hand, 18F2550 is right size, 32k, and I have one in my drawer... it has usb so if Ian likes he can add usb support to the firmware. My first choice was actually 18F2685, same 28pin package but 96k .. 3 times what 18F2550 has ... donno .. as I have 2550 already in my drawer I'll start writing GEN3 firmware for 18F2550 but if I get short with flash, I'll switch to 18F2685
[/rambling]
User avatar
arhi
Hero Member
Hero Member
 
Posts: 2160
Joined: Thu Jun 24, 2010 11:41 am
Location: Belgrade, Serbia

Re: HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron dri

Postby arhi » Wed Nov 23, 2011 3:32 am

Here's the 18F2550 version of the schematic I'm writing firmware for ...

The only untested path is the DC heater driver (page 3, driving fet trough opto). I don't have any fet's lying around and I don't anyhow plan to drive the iron with DC but if someone can check out the page 3, I think it is solid, the npn opto is driving fet, zener is keeping the gate voltage at 10V (since the DC input will probably be 20V or more) ...

Actually I haven't tested the lcd backlight brightness circuit also but don't see what could go wrong with a general purpose npn :)

Everything else is tested
- moc + triac works
- dcdc works
- thermocouple signal conditioning works (tested with Solomon HQ30 soldering iron)
- PTC signal conditioning works (tested with HAKKO and QUICK soldering irons)
- NTC signal conditioning works (tested with 204GT 200k NTC embedded in HAKKO tip)

If anyone has comments - you know where to find me :D
I will start working on firmware now. Hopefully 32k is enough for what I want. I will be making pcb only after I finish firmware (as if 32k is not enough I will be moving to 18F2685 and that is bit different pinout so pcb would have to be redone)
Attachments
SolderingIronControllerGen3.PDF
Soldering Iron Controller GEN3 v1.1
(229.07 KiB) Downloaded 1643 times
User avatar
arhi
Hero Member
Hero Member
 
Posts: 2160
Joined: Thu Jun 24, 2010 11:41 am
Location: Belgrade, Serbia

Re: HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron dri

Postby magius » Wed Nov 23, 2011 5:35 am

Thnks for the design!

If it helps, i'm tryng to mount it for Hakko pencil using a 18F2550. I'm compiling with Pinguino (a platform similar to Arduino but less mature). It's useful because supports 18F2550 (28 pins), 18F4550 (40 pins) and a 32bit PIC. With this MPUs you can cover from small to large projects.

One idea: the 18F2550 has 2 op.amp. inside, could be possible to save one op.amp. (and to save space) using the internal one (but it needs 3 pins of the 18F)...I'll try soon, when I have a bit of time.
magius
Newbie
Newbie
 
Posts: 3
Joined: Mon Nov 07, 2011 6:16 am

Re: HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron dri

Postby arhi » Wed Nov 23, 2011 5:40 am

I need pins on the mcu. And iirc 2550 only has comparators not op-amps. Also this all already works on my table only with 16F690 .. I just need bit more space in flash for lcd messages :D debug info etc etc .. so that's I moved up to 18f2550 .. without debug info and with only temp display the 16F690 works like a charm.

Also all 4 op-amps are in the same package (mcp619 is a quad op-amp) so no need to save any :D
User avatar
arhi
Hero Member
Hero Member
 
Posts: 2160
Joined: Thu Jun 24, 2010 11:41 am
Location: Belgrade, Serbia

Re: HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron dri

Postby arhi » Wed Nov 23, 2011 7:17 am

Forgot to add, yes the GEN3 with 18F is still single iron driver, not a dual one. For dual driver pic with more pins is required (for e.g. 18F4550) and bigger PCB, and since I purchased a PACE ST-50E with TD-100 iron for myself I really don't need more drivers on the table :D. When PACE arrives I'll try to integrate driver for TD-100 into SID GEN4 :D but I can't guarantee too much work on that area as TD-100 is too expensive to break and I'm not sure how interesting would it be for hobbyists since you can get 10 hakko pencils for a single pace pencil.
User avatar
arhi
Hero Member
Hero Member
 
Posts: 2160
Joined: Thu Jun 24, 2010 11:41 am
Location: Belgrade, Serbia

Re: HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron dri

Postby arhi » Thu Nov 24, 2011 5:50 am

Few final tweaks (added oscillator, replaced gain for PTC resistor with a pot ..)
Attachments
SolderingIronControllerGen3.PDF
Soldering Iron Driver Gen3 V1.3
(236.59 KiB) Downloaded 1514 times
User avatar
arhi
Hero Member
Hero Member
 
Posts: 2160
Joined: Thu Jun 24, 2010 11:41 am
Location: Belgrade, Serbia

Re: HAKKO (907ESD) and SOLOMON (SL-10/30) soldering iron dri

Postby arakis » Thu Nov 24, 2011 12:05 pm

you better build a final version soon, I ordered a Weller PES51 soldering pencil with a k type sensor that screams for a SID :D
BTW the original WES51 soldering station costs 100+$ and it is analog and doesn't even come close to the specs the SID can/does/will have
best regards FIlip.
arakis
Crew
Crew
 
Posts: 1044
Joined: Wed May 25, 2011 11:15 am
Location: Belgrade, Serbia

PreviousNext

Return to Project logs