Hi,

I’ve been following your thread up in dangerousprototypes, as I’ve also been doing research on solomon’s soldering irons (I want to buy one).

What you think about the following model?:

http://static1.tme.eu/katalog_pics/5/f/2/5f2f8951cf3e7465f47f0c654b71ae93/pensol%20sl30ce-i.jpg

The most I like from it is that it looks quite well designed, heating element included:

http://static2.tme.eu/katalog_pics/6/7/e/67e976e0cca81d3ba2da384130a7afa6/pensol%20sl20c-h.jpg

In adition to the one you bougth and the one shown above, there’s another scheme solomon also uses:

http://static2.tme.eu/katalog_pics/c/1/9/c195c5a9aff056034c9c843223ad6880/pensol-irone-n.jpg

It’s heating element (has a drill on the other end):

http://static1.tme.eu/katalog_pics/6/0/a/60a0cf80de49f23c4cff5741e5a4479d/pensol-sl10-ihn.jpg

And the tips this scheme uses (hence why the drill at the end on the heating element, the back side of the tip is in contact with the inner of such hole):

http://static4.tme.eu/katalog_pics/0/d/9/0d99805518d1479a9b2c700f660a8598/sr-623.jpg

So, what you think? which one would better to use in terms of capability of heat transfer and capability of temperature control?

I personally feel more confident with the former in terms of quality, though perhaps the heating element of the latter will be easier to repair at home with some nichrome-wire (the ceramic ones seems to use a plate of some other metallic material)

Also, have seen hakko duplicates as low as $7 plus shipment on ebay.

Which one to buy and why?

Regards,

The first picture states SL20, SL20 is not something I can recommend because SL20 uses the heater with hole and tip is pushed into the heater (just like SL10 that’s on the second picture). Anyhow that’s just the name of the picture, the iron on first picture looks like HQ type irons I’m talking about, they have tip that goes over the heater – as you see on my picture heater goes into metal tube and tip goes over that so heater is actually inside the tip. The contact area between tip and heater is much larger so works much better then SL10.

If your questions is “first or second” iron, go with first as second is not something I’d recommend.

If you ask solomon or hakko .. donno what to tell you about that. They both work perfectly, I personally prefer HAKKO, but I use original HAKKO irons. PRC knockoffs are also very good (especially if you use original tips), but original irons are not too expensive (under 20$ is original handle on ebay)

Thanks!

Well, the first I mentioned is not exactly like your SL-10. Tips are not as longer, but they go right over the ceramic heating element. A bit more like hakkos where no metal cover is between the tip and the heating element.

As I’m not an ebay expert and wouldn’t really know whether they’d be selling me an original hakko or just a fake, I’ll try luck with the solomon with ceramic part as all I’ve read about them from users is ok.

I also wanted to ask you if there is any formula or method you’re applying to find out how powerfull the power supply should be? I mean just as per just the soldering iron, extra ic-circuitry appart.

For example, these are 48W, should the supply be of just 48/24 = 2A, or better oversized as 1.5 times, 2, etc?

Common sense tells me it ougth to heat up
same as quick from 2A onwards, and that I should go for 2.5A (just to be sure that at least 2A are actually being delivered), but I might be completely wrong.

Solomon quotes the SL-30CMCESD, which is the soldering station which comes with the iron I’m interested in as 48W:

http://soldering.com.tw/soldering-station/SL-30CMCESD.html

Regards,

Ops,

Just noticed now the soldering iron I’m after is 30W not 48W, so power supply should be oversized by a factor of 1.6 it seems?

Regards,

The first pic is solomon HQ (sold in many other names in the World as solomon is one of the largest manufacturers of oem soldering equipment). It has tip that go over heater (like hakko only tip is bit longer). Looks like one you are talking about is similar to this one only with shorter tip.

The second pic show very old version of solomom SL and the tip is pushed into the heater (not over it like hakko or solomon hq). I don’t like these old ones.

You never know if they are selling original or fake, especially when it comes from PRC but you have to have some faith :). Hakko clones are very good quality devices too btw. just be sure you use original tips.

As for the transformer, I use what I have, I used 24VAC 2A, 24VAC 2.7A and 24VAC 3.33A toroidal transformers and 24VAC 2A EL transformer. Never had a problem. The electronics use very little, without lcd back light the controller I made use ~30mA. There is no need to oversize the transformer because it will never be 100% on for the long time. After a minute your iron will always be hot and you will after that only turn the transformer on in short bursts hence transformer will not be stressed … You can use stronger transformer (2.5A or more) but there’s really no need for it.. Anyhow, you can get a soldering iron with controller for very low amount of money now days .. ebay and prc online stores made that happen .. if you need to order every component you might go cheaper by purchasing ready made stuff (soldering station). Thing is that transformer could usually be salvaged from old equipment so find the 24V transformer you can salvage. If it can give 2A for 1 minute without heating up you can use it for soldering iron. You can find these in old soldering iron controllers, old modem/hub/switch transformers (most are 24VAC 1A but some are 2A).. some old transformers are over sized too so for e.g. 1A 3com 24V psu works like a charm for soldering iron. While initially heating up the iron producing 2A for a minute, transformer heat up to 44-50C and that’s more then acceptable. After the initial heat up transformer produces the 2A in bursts as requires and cools down to room temperature in few minutes… Now this transformer is “labeled 1A” I assume 3com added some security there .. but I assume any transformer that is capable of producing 1A 24/7/365 could give you 2A for 1 minute :) without any problems

Hi, thanks!

And sorry!, I can only access the internet from time to time, yet.

As you suggested, I finally bougth the hakko 907 replica found on ebay by $12.85, and yesterday I received it. Looks quite ok to me!.

I’ve also been given a bulky 24V 3A transformer, which actually ouputs +/-24.00V (it seems it was wound with that purpose).

So, I’m wondering If I can connect the iron temporally to it to test both and get some readings?. It seems to me the term AC is used in soldering irons’ slang as a shortcut for pulsed DC, though, so I feel a bit scared of trying.

Also, do you happen to have a table or something with your readings of sensor’s resistance across heating element or tip temperature? I couldn’t see any in your firmware, and would be nice to compare with the readings I’m willing to take.

I’d like to check from wattage this soldering irons consumes, to sensor resistance across temperature, as I’d be doing your very same circuit but with an arm (yes, overkill, but I’ve got a few samples and I’m learning this arch lately).

Also, There is an application note from AVR:

http://atmel.com/dyn/resources/prod_documents/doc8003.pdf

Where is explained how to oversample a 10-bit capable ADC to increase resolution to up to 16 bit, so probably with a DS18B20 or similar at the junction (if K-type support is wanted), and a 10-bit ADC, not even an amplifier might be needed.

Regards,

Eduardo.

AC is ALTERNATING CURRENT – it is no slang !!!
check out: http://en.wikipedia.org/wiki/Alternating_current

The output on the transformer is AC, if you put it trough rectifier and add a capacitor it will be DC. For heater it is same if you push AC or DC in so yes you can push AC 24V into heater.

I have no clue what you mean by +-24V, do you mean 24V AC or you say that you have 24V center tapped (so you have 2 secundar outputs of 12V) or you have 48V center tapped (so 2 secundar outputs of 24V).

Make sure when you connect the power to the heater that you push power into heater and not into sensor. For HAKKO the heater is cca 12R, you can see pinout and some resistance table on http://elco.crsndoo.com/wordpress/2011/09/hakko-907-i-kompatibilne-lemilice/

Also, HAKKO DO NOT HAVE THERMOCOUPLE, it has PTC.

And finally, no, you will not be able to read thermocouple (hakko don’t have it but some standard k-type) with atmel’s adc directly no matter how much oversampling you do. The input impedance of the ADC is too low for k-type to handle so you have to do some buffering at least, and when you put op-amp to buffer, then it’s easy to configure it as amplifier and condition that signal bit better.

Have fun :)

Hi, Thanks,

Ok, I get it!. I just meant AC refers to the full swing signal (from -V to +V), while DIY circuits usually feed power to the heater in the range (gnd – 24V DC, in this case), through a fet, triac, or whatever.

No, my transformer has just two output pins, across which, my multimeter exactly measures 24V AC (it normally measures more than 24 volts for a 24 volts rated transformer), that’s what I meant.

And you’re right again, I completely forgot about the high impedance of K-type thermocouples!. I know hakko does not have a k-type as such, as I’m also following your thread up in dangerous prototypes.

Then I’ll study what you say about using an op-amp as buffer, as you mention it can also be configured to behave as amplifier.

The first thing that came into my mind when I read the above datasheet, was that having as much of the soldering station’s functionality (and even accuracy) based on firmware, would be ideal, because it would allow one to experiment and improve everything without (or with minimal) changes to the hardware.

Sort of like an evaluation board. So as long as the ic can stand it all (in my case it’s an arm7 from ST, with a 12-bit ADC), it would be really nice.

Thanks also for the resistance table in your hakko’s article entry (I have even read it several times, but forgot was there). My clone measures about 57 ohms at room temperature, though.

But anyways, Thanks so much for all!

Regards,

> I just meant AC refers to

Now you know :)

> my multimeter exactly measures 24V AC

it really depends on the power and the type of the transformer. This will be adequate power source for the HAKKO heater (it is 24V 2A 48W heater, so your 3A transformer is perfect)

> Then I’ll study what you say about
> using an op-amp as buffer,
> as you mention it can also be configured
> to behave as amplifier.

Use the schematic that amplifies the signal 240 times. Then if you use 10bit ADC the read ADC value is exactly double the temperature in C. So if you read your ADC input (10 bit) and get value of 155 for example, the temperature difference between cold and hot joint is 155/2 = 77.5 degrees Celsius.

> My clone measures about 57 ohms
> at room temperature, though.

Looks like PTC in the HAKKO is not perfectly same across clones but 57R on room temp (assume 25C) is too big difference as that’s what my station assumes is 250C !!! if you heat up your hakko soldering station with another iron or hot air or whatever heat source to 100+C what resistance to you read? What type of connector you have on hakko iron? Male of female?

One important issue – do not measure any resistance on “cold iron”, heat the iron to min 100C and then do all the measurements :) (cold heater can show 2-3R, cold PTC can show any number of useles info as it is really set to work from 180-450C). It would be great if someone can do a test with a real hakko 937 station (I have a clone hakko station) with a single multiturn pot and a multimeter… but I know noone with real hakko 937 station :(

Hi!

My soldering iron is exactly same as your clone (the one with hakko printed on it). It has a male connector with that locked plastic freewheel holding the thread.

It seems all the hakko clones comes with polymeric grip, and only originals come with proper foam grip.

I don’t know of anybody with a 937 station either, though.

And yes, I tried at room temperature, then used a ligther at the tip of the heating element for a moment, and resistance raised.

I’ll buy tomorrow a female connector for the soldering iron, and will make a circuit with an atmega that allows begin taking measures.

For this, all I’ve got so far is a nordson temperature gauge from some machinery, like this sort:

http://www.brew-magic.com/images/temp_gauge.gif

It can handle from 10 to 290 degrees C, and it’s meant to have certain accuracy, so I’ll probably place gauge’s sensor and heating element one beside the other, and will keep them toghether with some bare copper wire.

Then, I think I’ll first try to study the heater itself. If I can keep a temperature as constant as possible by just measuring heater’s ressistance, then all should be a matter of stepping to some temperature, holding it long enough for the gague to also be at the same one, and then meassuring resistance across the ntc.

What you reckon? My idea is to try this at as many steps as possible such that I can have a profile as close as sensor’s behaviour as possible.

Perhaps it would be easier to just mount the solering iron, and do a small breakout board with a DS18XX sensor and a copper pad on which the tip of the iron rests.

At the end of the day we’re interested in tip’s temperature.

Regards,

Forget about the last “idea”.

It seems nobody does high-temperature sensor ic’s, and it’s obvious why. The silicon would melt, and the die would also go nuts at that temperature, hehe.

Regards,

Unfortunately it is not that simple. Contact temperature measurement is a science of it’s own. Measuring tip temperature with that device is going to be near to impossible as you will see yourself. For e.g. at the point when tip start to melt the solder wire your tip is hotter then 220C (as most solder wire melt between 180 and 230C depending on the chemistry of the wire) but measuring with that device just touching the tip will give you temp way lower then 180C :( because that “probe” is meant to measure “gas” or “liquid” and not to be a contact probe to measure one point temperature. I have for e.g. thermocouple and ntc, both show waaaaaay lower temperature when you touch them with the tip as the contact surface between them is just too low.

The only true way to measure tip’s temperature is to waste one tip – drill the hole on the tip, push thermocouple or a thermistor (with known characteristic) in the hole and close the hole with fire cement. This will give you (with a small delay) a true temperature of the tip.

And no, there’s no DS* that will go over 200C as the solder inside would melt.

Hi,

Yes, I noticed, hence my second post.

So how are you doing to get readings of the resistance at such high temperatures?

Perhaps before wasting a tip, it would be worth to try meassuring resistance in about 2 cms. of nichrome wire (tensored), while you let the soldering iron rest on it with a bit of solder to allow thermal propagation.

This is more or less how the hakko calibration equipment works.

So far for me, trying to measure resistance of the heating element is proving really difficult, but it would nice to have such subcircuit on the final design, so I’ll keep on reading on amplifiers to see if I can do it.

What I would to get the readings now, is to heat the ceramic element to, say 280 degrees, as 290 is the max I can read out right now, and then meassure resistance with the multimeter.

Then, will instruct firmware for never going beyond there, and take readings of sensor’s resistance from 280 to room temperature.

Regards,

> So how are you doing to get readings of the resistance at such high temperatures?

So far I didn’t, for solomon I just read the TC sensor and for Hakko I just made a table using existing clone driver and a pot. Now I will drill a hole in the tip, put the TC or NTC with known curve in and seal with the fire cement.. I have some destroyed tips (very fine tip that got burned out) so they can’t be used for soldering any more hence drilling a hole in them ain’t gonna be a problem :D

Making any measurements with a sensor that has just a contact with tip in one point and rest of the sensor to be in air is just “too slow”. Eventually it will reach the “stable” point but that’s too slow for pid tuning etc etc.. The only way to really measure tip temp is what I’m planning to do :)

Also, if you notice how new version of HAKKO, PACE and similar systems work, the sensor is no longer in the heater but is pushed out so that it contacts the tip.

Hi!,

I saw your pictures up in the DP thread, can’t wait to see your results!

I have a question about the schematics of the soldering station (current ones).

You connected a 47 Ohms resistor to the MOC3043, how did you get to find out that value?

The only similar part I’ve got at hand is a S21ME:

http://www.datasheetcatalog.org/datasheet/Sharp/mXtwsrs.pdf

But I can’t see which term did you use to set such resistors value within the MOC datasheet either.

As per the 1K resistor in the lower part,
I guess it’s there to force the triac open once it’s gate has been excited, even if no load is present.

Also, with this configuration, pulsed output is a must isn’t it? or can the MOC withstand continous current flow? (2A in my case)?

I’m not really understanding this datasheets :/

Regards,

The “results” came 3 years ago when I made the initial driver that I’m using since then … these are just improvements .. and I will probably be improving it for next few years too :D

47R resistor I “guessed”, because datasheet show 360R for 220V, 180R for 110V I assumed 47R will work ok for 24V and it does.

1k is IMO related to the TRIAC you are using. I tried 1k and it works great with triac’s I have.

wrt 2A, no real current goes trough MOC, your 2A will go trough TRIAC moc is driving.

Hi, and thanks for being always there for a comment!

Regarding my previous question, I found this:

http://www.simpleio.com/design/triacout/AppTriacOutGateResistor.asp

So I now understand. In fact, the article and the mentioned datasheets, makes me think that proper control of an AC soldering iron has more to do with phase control, than it has with PID as such.

Perhaps the easier would be having a dedicated MOC using cuadrants I and III of the AC wavefrom (I = zero -> v+, III = zero -> v-) triggering interrupts instead of firing the triac, and then driving the triac manually.

It should all be a matter of passing the MOC through the scope and studying if it’s zero-crossing detection + rise-time are within acceptable boundaries.

As per the IC, and being the MOC ok, I reckon that even a 4Mhz crystal should give it enough resolution for getting into the interrupt and turning on the triac quickly enough whether the load should be connected right at the begining of the cuadrant, or rigth at the end of it, which are the critical points.

Triggering it with offset respect to zero should be easier providing previous criteria can be met.

In my opninion, it would not only overcome the overshoot problems you’re having with your design, but will also allow a much faster response of the soldering iron.

For example, by constantly switching the triac near the end of the cuadrant, there must be a point where set temperature of the whole element (and hence, the tip) will be held constant when not in use.

When in use, and you notice the temperature tends to fall because it can’t be maintained with the given power, you just trigger the triac a bit before, until it does. Sort of a PID as well.

Same for the opossite. When the iron is cold, you just let the triac fully on, until it’s about 10 degrees from the set point or so, which when you start narrowing the swithing time towards the end of the wavefrom to achieve equilibrium.

What you think?

Regards,

> phase control

No, not really. The soldering iron is a “dummy heater” so phase, quadrant etc are completely irrelevant. You could use here a moc version without zero crossing detection and it will work as well as the one I’m using as the voltage is low and the load is pure resistive one.

> overcome the overshoot problems you’re having
> with your design, but will also allow a much
> faster response of the soldering iron

No, not really. It has nothing to do with “when” moc turns on the triac :D. We are talking about a very slow device here (heater). The PID control has only 16 steps (not needed to have any finer tuning really) and the frequency of the pwm is 2Hz. The shortest time iron is turned on is 32ms and the period of the ac signal is minimum 20ms (less in US) hence you at least have one full period pushed trough the triac, and then you have more depending on the pwm value.

You have to figure this is a very slow device (the soldering iron heater) and that doing the control too fast (inside a single period) is not important.

Wrt overshoot, there actually isn’t any. You forget that the temperature measurement probe in both irons I’m testing (hakko and clones and solomon) have heat sensor inside the heater and not inside the tip. There is a long delay in heat transfer from heater to the tip hence the overshoot you see on the graph that’s measuring value from the sensor in heater does not propagate to the tip at all. There will be a post with new measurements from the probe in tip in few days..

> When the iron is cold, you just let the triac
> fully on, until it’s about 10 degrees from the
> set point or so, which when you start narrowin

That is P control, the driver has full PID control so with only P control the results would be only worse.

[...] keramicki grejac snage 48W (24V) a HAKKO vrhovi su neverovatno kvalitetni. U poređenju sa Solomon lemilicom HAKKO ima manju termalnu masu (vrh je lakši) i kontaktna površina između grejača i vrha je [...]