My FX-801 Demo Experience (300w of Enlightenment)

[Hohn]

Hi Everyone-- first post here after lurking erratically for years.

Short bio-- Former USAF officer turned mechanical engineer. Have some background in electronics, just enough to sound like a dumb person's idea of a smart guy while looking dumb to the truly smart people. Basically, I know how to break things REALLY WELL. And my postings tend to be long. Sorry, I don't have time to make them shorter.

I took up soldering as a hobby to itself because so many of my other hobbies seem to require it. I'm using an FX-951 as a daily driver as my first "real" station upgrading from the cheap pencils that turn electricity into oxidation and frustration. I don't do a lot of PCB work and do mostly pro-audio, discrete type items. Things with lugs wired together. (vacuum tube sockets, turret boards, 1/4" and XLR cables, replacing switches, sockets, etc.). I've done PCB and I'm comfortable with it, but I'm not experienced with SMD/SMT.

The lack of soldering truly tiny things had me wondering if I couldn't make something like a Hakko FX-801 a one-station solution for me. Something that would do all that I do, but with the hot-rod performance of a 300W cartridge style handpiece. Since I'm a happy 951 user, I requested a Demo unit of the FX-801. It arrived a few weeks ago. I've been using it next to my FX-951 extensively, so what follows is a short review of the 801, contrasting with the 951, and some enlightening observations about using them side by side.

The 801 is impressive when you get it out and set it up. It has the same kind of microswitch stand as the 951. (pretty sure it's actually the same stand, not just the same kind). The housing of the 801 is real sheetmetal and has a flat, ventless top so it will stack. You flip it on and the not-quiet fan will remind you this is probably intended for industrial use and not lab use. It's pretty loud. It comes up to temperature about the same speed as the 951 does (at least with the slim series small chisels I requested). The handpiece is remarkably small and compact in hand given how much power is on tap. In terms of just generally accessibility, I think this could be an iron I could use for almost anything, because the slim series cartridges ("carts") are only 7.2mm diameter. vs the T12/15 diameter of 5.5mm. In practice, you don't much notice the larger diameter. Nor is the slightly longer distance from the grip much of an issue. It's probably a good thing to get the tip a little farther away from the handle when talking about this kind of heat potential. (settings up to 500C).

It has power to spare. Enough to where I basically could never load the tool enough to induce any meaningful tip temperature drop in terms of seeing the display dip below set point. Oh, I can hold it on a wet sponge and get it to drop 80C-100C, but in practical use, it was very difficult to make it drop temp.

But in the course of comparing it to my FX951, I realized I could also almost never get my 951 to show much of a temperature drop. On a wet sponge, absolutely, but in actual soldering, it seemed rock-steady.

So my engineer brain kicks and and starts asking: what kind of thermal load in soldering would actually pull down the temperatures and indicate that the station is completely overwhelmed? I happened to have a scrap of 3/4" copper pipe about a foot long. Knowing how oxidized this pipe was, I had generous amounts of both paste flux and liquid MG835 on ready-five. AHA! THIS is the mother of all heat sinks. All I need to do is try to solder a wire to this pipe and my 951 will come crawling for mercy when the temp drops 100C or more and I will show where the 300w Hakko reins supreme and conquers where the 951 fails. RIGHT?

WRONG. I tried my 951 at 330C. I managed to melt a rather large puddle of solder onto the pipe end, but it wouldn't wet or spread on the pipe on its own, I had to spread it with the iron tip. I had the tip of my 951's 3.2D completely submerged in solder and the station's feedback temperature never dropped below 310C from the 330C setting. Hmm. Clearly I'm not getting enough heat flow out of my tip and it's not because I don't have enough heat sinking occurring. So I crank up the temperature to 370C. The feedback temperature on the 951's display never got below 340C. This, with the tip well submerged in a massive puddle of solder on the end of a large solid copper pipe. (This tip is in excellent condition, it appears new on the tinned area, so it transfers heat as well as it ever could).

Unsurprisingly, the 300w Hakko did not much better at all at identical temperatures of 330C and 370C. I submerged its (larger) tip in a puddle of solder and it still wouldn't drop the feedback temperature very far at all-- less than 10C. It has less drop because it's a lot more powerful.


But why is there so little performance increase between the 951 and 801? This really bothered me. The 801 has much more mass in the tips. It has over 400% of the power of the 951. Yet when you start melting solder with each station, the real world difference between them is tiny at best, even with brand new tips in immaculate condition.

There's only one logical explanation I can think of, and it's this: the main limit in soldering heat flow is the tip interface and the solder itself. Think about it this way: you have a large section of copper pipe that's fairly massive and conductive. And because it has a large surface area relative to its mass, it will lose a lot of heat. You melt a substantial puddle of solder on one end and submerge the tip of an iron in it. But within the puddle you have temperatures of nearly 350C or so at the tip to barely over 180C at the edges of the puddle. There is a substantial heat gradient.

The gradient exists because heat is flowing in much faster that it's flowing out. But this means that eventually the heat will "pile up" and you'll have very little gradient at the tip. So heat stops flowing out of the tip. Which means the iron does a lot less work. Which means it doesn't need very much power because it's not drawing much power.

All that to say that I think there is some amount of actual station power such that having more power is totally moot. I am certain the FX-801 is beyond that point. You simply cannot flow that power out of the tip because whatever conduit you've got for the heat flow will not allow a more rapid rate of energy transfer. You get to a point where the iron and station simply aren't the weak link in your process. It's Pareto principle, and your soldering tool is no longer the far left column. Rather, tip geometry and cleanliness, and then ultimately the solder's conductivity and the components specific heat and thermal mass come into play. If a massive puddle of solder on a 12" section of copper pipe isn't enough thermal mass to pull the tip temperature down, then what is?

It's possible then that thermal recovery is a lot less important than we might first have thought. If you have a tiny tip, yes it stores a lot less heat, but it also stinks at transferring heat. So the recovery factor will come in sooner, but ultimately require less power because that small tip simply cannot conduct heat fast enough to tax the station. Likewise, a large tip will store a lot more heat. And it's possible that this stored heat will be sufficient to execute your joint before temperature drops appreciably and--again-- the power of the station ends up rather unimportant.

We're nearing the point that the thermal conductivity of copper itself is one of the main limiting factors to heat throughput. Or perhaps the iron plating on the tip is a big limiting factor. (hence the two different JBC tip series--hot rod thin plating vs industrial thick plating)

This explains why the T12 Clone vs 888 thread https://www.eevblog.com/forum/reviews/t12-clone-vs-888-practical-test-results/ showed somewhat "meh" results for the "superior" cartridge technology. Now I have some reasons to question that test based on extrapolating from a fake t12 tip to a real t12 tip and from a clone station to a real Hakko. But putting aside those objections, it perfectly plausible to me how it might be that a modern station-- even one with massive power and the latest technology-- could underwhelm.

I think it's just the case the soldering performance now is mostly limited by the solders and tip geometries we use and the iron plating we have to apply to keep our copper tips alive. Heat just wont flow out of the tip as rapidly as we want it to.


So the question of thermal performance of a station or iron really becomes one of sufficiency. Either it has enough performance to do what you want to do, or it will struggle. If only only need to get yourself and your pets somewhere, then a Honda civic is equally capable to a Ford SuperDuty. But only one of those will carry 10 tons. And it turns out that in most of our usage, we simply cannot load the vehicle with 10 tons-- it won't fit, the doors are too small, or it takes too long. There's nothing wrong with buying that huge truck just to use for commuting, but it comes at a cost and might not give much of any benefit in return if you can't exploit all of its capability.

[thm_w]

All that to say that I think there is some amount of actual station power such that having more power is totally moot. I am certain the FX-801 is beyond that point. You simply cannot flow that power out of the tip because whatever conduit you've got for the heat flow will not allow a more rapid rate of energy transfer. You get to a point where the iron and station simply aren't the weak link in your process. It's Pareto principle, and your soldering tool is no longer the far left column. Rather, tip geometry and cleanliness, and then ultimately the solder's conductivity and the components specific heat and thermal mass come into play. If a massive puddle of solder on a 12" section of copper pipe isn't enough thermal mass to pull the tip temperature down, then what is?

Yes. This is also how a 40W Metcal can outperform the FX951.
JBC as well, sensor and heater closer to the tip, thinner tip plating, etc.
Some photos of the tips cut apart have been posted here, which gives you an idea.

BTW the temperature shown on most soldering stations is not the actual temperature of the tip, they are often "optimized" to hide overshoot and not show the temperature jump up and down.

[floobydust]

Soldering-irons' limiting technology is the thermal resistance between tip and heater.
Hakko 936, FX-888 are extra poor performers because of the air-gap between tip and heater. It's several thou that shows up once you use up the tip's heat, you have to wait or crank up the temps, basically you are doomed at that point.
Weller stations have no air-gap and are much better. Hakko came up with the T12 technology to get around the air-gap. This is like the majority of soldering irons out there - cement, which has varying grades of thermal conductivity, is the thing between the heating element and the tip.

Having a tip with high thermal mass can cover up for the long delay to reheat the tip, but it gets really annoying once the tip goes cold from too much heat load. Then manufacturers are using brute force - super high heater power (temps) to overcome the losses which doesn't work I find. You still have the PID controller lag because the temp sensor is not representing tip temp, it's showing the heater's temp.
Cheap or crappy design soldering-irons have terrible heat transfer from heater to tip, i.e. older Pace products that use a side screw to mount the tip. That IMHO is Fred Flinstone tech.

I don't know what is better than cement, for the tech out there in soldering irons.

[tooki]

Soldering-irons' limiting technology is the thermal resistance between tip and heater.
Hakko 936, FX-888 are extra poor performers because of the air-gap between tip and heater. It's several thou that shows up once you use up the tip's heat, you have to wait or crank up the temps, basically you are doomed at that point.
Weller stations have no air-gap and are much better. Hakko came up with the T12 technology to get around the air-gap. This is like the majority of soldering irons out there - cement, which has varying grades of thermal conductivity, is the thing between the heating element and the tip.

I’ve struggled to find any record of who introduced the first cartridge heater soldering iron, but it certainly was not Hakko with the T12, which appears to have been released in 2005. I found some photos of a JBC AD2000 PCB with components with 1998 date codes, so at minimum we know Hakko wasn’t first. Pace had released its first cartridge heater system, HeatWise, by early 2003 at the very latest.

Cheap or crappy design soldering-irons have terrible heat transfer from heater to tip, i.e. older Pace products that use a side screw to mount the tip. That IMHO is Fred Flinstone tech.

Well of course it is. But everyone used that system (or something similar) originally. (And if done well, as in Pace, it still performs fairly well. Tight tolerances make a big difference.)

[floobydust]

Hakko T12 heater/tip patent US6054678A filed 1998, also related US6087631A.

Pace Inc. US6513697B1 filed 2001 granted 2003 specifically states "The construction and mounting of the heater 12 and Soldering tip 14, by themselves, form no part of this invention and can be of any type and arrangement known in the art, See, e.g., U.S. Pat. Nos. 4,839,501 and 6,054,678."

OP's soldering station patent US20170028497 Heater Sensor Complex With High Thermal Capacity, appears to be for the FX-801. It's target market is bigger through-hole in EV's, HEV's, solar etc.

[tooki]

As I said, earliest cartridge heater JBC station I could find evidence of is from 1998 (production unit in the wild), so I’m unconvinced of Hakko being first. But again, info is hard to come by. You’d think whoever invented it would prominently list that in their company history.

[Hohn]

I'd like to expand a bit on your point about the thermal resistance between tip and heater. I think that resistance is actually of secondary importance. The more consequential thermal resistance is between tip and SENSOR or thermocouple, at least for a temperature controlled station.

I've posted elsewhere that I've been using a scrap of 3/4" copper pipe for heat transfer experimentation and comparison of the three soldering tools I have on hand:
1) 25w radio shack "professional" pencil (4 tips, no control)
2) Hakko FX 951 (11 tips to test)
3) Hakko FX-801 (3 tips)

I have observed that unless either Hakko is showing some temperature drop on the display, it isn't providing much power to the iron. I have not yet tested this by measuring current draw (a plan is in the works), but it's clear enough by observation. When the tip is first applied to my test mound of solder on the pipe end, very little heat is flowing out of the tip. Which makes the performance of the heating element and the power of the station largely moot. There's simply too little thermal load.

If I heavily tin the tip to get a good glob for heat transfer, then I can start to pull down station temperature and force some power output. Remember, these stations use some kind of PID control and it's entirely reactive. So unless I can get heat out of the tip and induce tip drop, there's simply nothing for the controls to respond to.

The 951 testing shows this rather vividly. With a long thin bent chisel (JD14) , I can pull the tip temperature down so low that it will stick to the pipe. But the station still shows set temperature while this is occurring. The thermal gradient between tip and sensor is steep enough that the station can *think* temperature is fine when it isn't, so obviously no power is being drawn and thermal performance suffers. If the sensor doesn't think it needs to heat, then the thermal performance of the heating element and the bridge to the tip is entirely moot.

As an example of this, I will describe the experience with the 25w pencil with a 3.2mm chisel tip. (slip on T18 style like an FX-888d). While it takes a long time to do so, eventually the pencil can melt a mound of solder about 12-15mm in diameter (a puddle diameter) on the end of the copper pipe. The tip stays hot enough to melt solder wire if I apply more. This is because the pencil is *always* dissipating 25w and it needs no control signal to induce this thermal output. Given the inefficient thermal coupling, the tip is probably delivering <20w to the workpiece from a 25w input to the tip.

In other words, the poor coupling of the long bent tip on the 951 has the effect of reducing the 70w station output to something like 10w or whatever because of the gradient between tip and sensor. The sensor has no idea the tip is too cold to melt solder because at the location of the sensor, it thinks the station is nearly at set temperature (within the amount of damping and lying in what the station displays as tip temp vs what the internal controls of the station actually "see" as temperature.


If I replace the long bent tip of the 951 with the shortest tip I have (closest to the sensor), the thermal performance difference is incredible. The D3.2 stubby chisel has incredible performance because there's very little gradient between workpiece and sensor, so the station "sees" the heat loss and thus more power is forcibly dissipated. It will outperform the 3mm hoof (BC3) and the 4mm bevel (CF4) tips even though both of those are larger tips with more mass. Why do they have less performance? They are longer tips-- thus more temperature difference between workpiece and sensor.




You can see this dynamic in the station's displayed temperature. The stubby 3.2mm chisel actually shows significant temperature drop (20-25C) or more while the larger tips are difficult to get more than 10C drop. It's tempting to misinterpret this as "the bigger tips have more capacity and thus better thermal performance". In reality, they have poorer coupling to the sensor and thus cannot force the station to deliver more heat. On smaller workpieces where the tip mass is very consequential, the larger tips will of course outperform.

But when you're talking about a heat sink the size of half pound of copper and a foot long, the thermal mass of the tip is pretty much irrelevant because it's TINY compared to the thermal mass of the pipe. The pipe can drain all the heat from the tip and barely show a change of temperature. Instead, this test is almost exclusively a test of thermal coupling: of the tip to the workpiece, of the sensor to the tip, and of the heater to the tip.

Finally, my testing showed that there is an advantage to higher set temperatures if you are dealing with substantial thermal masses, the kind that can reduce a tip to being cold enough to where it won't melt solder. The maximum temperature gradient across any tip from sensor to workpiece is a fixed value at equilibrium, a property of the tip's thermal conductivity and geometry. In the case of my long thin tip, this value was well over 100C as the tip wouldn't melt solder (182C) when station set temp was 300C.  However, cranking up the station's set temperature to 400C showed substantial temperature drop on the station and therefore more heat demand trying to restore the set temperature against the large heat sinking of the pipe.

Higher set temperatures are effective at forcing more power dissipation from a station powering a tip with a significant gradient that mutes the temperature signal to the sensor.

The importance of tip "sensitivity" --in terms of ability to detect temperature drop--cannot be overstated. I think it is far more significant than the actual heating power from the heating element to the tip itself. (although in many designs they face similar thermal resistance and thus a tip with good sensitivity will also have more efficient heat throughput.)

[thm_w]

Good points. This is why most (?) iron vendors will sell "power" versions of their tips which are short and stubby, but they may not make it so clear when browsing their catalog. You should definitely have one of those tips, if you need to perform tough soldering/desoldering jobs.

However, sometimes its not convenient or possible to use such a stubby tip, and you want the reach of a longer thin hook. At that point you get much better results from a design that has the sensor as close as possible to the tip.

Reminds me I need to cut apart a metcal and see how far the core extends internally: https://www.eevblog.com/forum/reviews/metcal-sttc-soldering-cartridgetip-teardown/

[floobydust]

The thermocouple is right at the end of the heater coil, so it will easily read incorrectly and PID get fooled due to the high thermal resistance of the cement.
So far I've only seen big thermal mass sorta make up for this, but past that the tip cools and gets stuck to whatever you are working on even though the heater is on.
cheap chinese tips have little copper, mostly steel (use a magnet) and are the worst.

[gtm]

Quote from: Hohn on 01 April 2021, 05:40:28

...
But why is there so little performance increase between the 951 and 801
...

If you are using same sized tips, there shouldn't be any performance difference.
A soldering station/iron is a simple device, it just has to replenish the energy being transferred onto the joint, ideally at the same rate and in real-time. They don't really "pump" heat, heat simply travels from a hot place to a colder place. All the station/iron can do is ( try to ) maintain the selected tip temperature.
The only way to "pump" more and faster is by using a larger contact area and/or increasing the temperature.
The stations wattage is supposed to be enough for the worst-case scenario: ie. the largest tip, at the highest temperature, into the heaviest of thermal loads.
With a 3mm chisel at 400C onto a ground plane, that might be 40W, or 50W or whatever, both stations can provide that power.
With a 15mm, or 20mm chisel at 500C onto an ingot of copper the FX-951 would not be able to replenish the energy lost/transferred,  which is why there's probably no 15mm chisels on the T12 range, and almost certainly there is for the FX-801. (*)
That's the only real difference between high powered and lower powered stations: the size of tips that they can handle. (And a chunkier hand-piece for the high-powered one).

(*) I was wrong. I assumed that, being 300W, there were humongous chisels for the FX-801 ( like there are for the JBC HDE), but no, tip selection is rather small and the biggest chisel is 6mm or so.

[thm_w]

The thermocouple is right at the end of the heater coil, so it will easily read incorrectly and PID get fooled due to the high thermal resistance of the cement.
So far I've only seen big thermal mass sorta make up for this, but past that the tip cools and gets stuck to whatever you are working on even though the heater is on.
cheap chinese tips have little copper, mostly steel (use a magnet) and are the worst.

In the specific tip you show, its probably possible to drill a hole that extends almost to the end of the tip, and slide the thermocouple in there.

Of course this would cost more, and doesn't work on all tip geometries.
So for cost reasons, most likely every tip they make will have the same spacing you've shown.

[Hohn]

I packed up the demo unit FX-801 in preparation for return to Hakko, so I thought I'd post a final conclusion to this thread.

The FX-801 is surprisingly useful as a general purpose station for work involving mostly discrete components like turret board wiring and such. The handpiece is very comfortable, stays cool even at insanely hot tip temperatures and tips down to 3.2mm chisel are available.

Even though it can be pressed into service for work that would generally fall to a smaller station, most users will regret buying if it is used mostly for that purpose. It is a good second station to provide special capability in the unlikely case it is necessary, but would be a poor choice as an only station, even if you don't do "small" work.

I suggest that the only time this station (and its JBC competitor) can provide value anywhere near the substantial asking price is if you have exhausted the capability of your current "Regular" station with a high heat tip at or near its maximum tip temperature. Until and unless you are at the point of exhausting the capability of your smaller station and will have a performance advantage you cannot detect in normal use. And in return for having no perceptible benefit, you will have spent a lot of money, gained a loud exhaust fan, and gone to a larger and less versatile tip series with a longer reach handle.

I suggest you only consider a station like this if you are presently having problems with the performance of a less powerful professional station. If it's not solving a problem for you, it has no value proposition relative to a cheaper "pro" station with less power.

[Julienrl]

Thank you, great conclusion!