Perculiar soldering iron problem.

[Dave Plowman (News)]

I've got a home made solder station. Comprises of Antex soldering irons and a Pace desolder one. I built it because of the cost of such things commercially at the time, and because none provided exactly what I wanted. It's basically 24v AC with temperature control - and the boards based on a Maplin design.

The problem is the element in the Pace blew recently, after several years service. Fitted my spare and that went too. So I then checked the board. It's basically just a Triac controlled by the temperature sensor in the hand piece. The same transformer feeds 24 volts to all the elements. No fault found.

So I bought another element at a high price, and was careful to set the temperature low - and actually to measure it with my Fluke. Before it got to about 250C tip temp that burnt out too. Any ideas?

[Andrew Gabriel]

Is the Triac blown (shorted) so the element is stuck on?

I'm still using a Weller TCP-1 I bought at University (at great expense for a student), when I found how much better they were than my dad's 25W uncontrolled soldering iron that I had used up until then.

[Frank Erskine]

It would be interesting to put a 'scope on the output of the control board and look for, for example, spikes, or anything else that might knack the temperature sensor.

[Adrian]

Don't suppose you could be getting some kind of voltage spikes superimposed on your 24vdc, could you - something to do with the triac switching, maybe ?

This would not show up on a meter - but would be visible on a 'scope...

I've tended to use the older type Antex & Weller irons - where the temperature control is a magentic control built into the iron itself - so I've no experience of triac-controlled irons.

To save money during the debug stage - maybe a couple of car headlamps in series would be a cheap way of providing a load, without despatching another (expensive) iron element...

Hope this helps Adrian

[Dave Plowman (News)]

The temperature sensor inside all the failed elements is fine, and I replaced the triac after the second element failure. I didn't look at the output on a scope - it didn't last long enough to hook it up. ;-)

But I'm not sure how the triac could produce spikes of such current that would burn out a heating element so quickly? The cold resistance of the (good) element is approx 10 ohms.

[Adrian]

No - neither am I ...... - grasping at straws a bit there

Is there any evidence that the triac is actually switching 'off' ?

Could you load the output of your controller with a light bulb or two, and then apply some heat to the sensor to see if it actually causes switching ? - or even replace the sensor with a pot that you can tweak to prove the switching circuit ?

Does sound as if the switching circuit is locked 'on' somehow.... does it 'fail safe' (= failure of the temperature sensor turns the triac 'off') ??

Have you a copy of the circuit somewhere on the web ?

Regards Adrian

[Dave Plowman (News)]

Yes. On testing it afterwards with one of the Antex irons it was switching.

The temp control does just that.

Trouble is it didn't work for long enough to get up to working temperature - ie only a matter of about 1/2 minute and I was measuring the temperature of the bit at the time. It did smoke, though. ;-(

I'm not sure how to do this as I don't have a website. Or rather I think I may have the space but have never used it.

But it's very basic - an instrument grade opamp amplifies the output from the temp sensor and feeds a comparator, the other input of which is fed from a variable supply. The output of the comparator goes via a 'status' LED to a zero crossing opto isolator which turns on the triac. When the amplified sensor voltage exceeds that of the control, it turns off. The control voltage source is stabilised and the gain of the first opamp set so that the maximum temperature allowed will produce an equal voltage to the control voltage set to max.

[Clint Sharp]

In message , "Dave Plowman (News)" writes

Can you be certain that the elements you have are exactly the same as the original and Pace don't make, for instance, a 12v version? I don't know the tool you are talking about but I'm guessing the bit has enough thermal mass to cause the element to be powered continuously until it nears the set temperature, if this is the case is the element fitted correctly and making proper thermal contact with the bit, is the temp sensor making proper thermal contact? Either scenario could possibly cause an element to burn out in short order, how long did it take the bit to reach operating temp when it was working?

[Dave Plowman (News)]

I have checked and there's just the one available.

Approx 45 secs. Rather slower than the Antex which is nominally the same power - although that's to be expected since the de-solder tool has a greater thermal mass.

On doing some more checks, I'm wondering if sharing the same 24 volt transformer is allowing a rather higher voltage when only one is heating - it's a 160 VA toroidal I had lying around whereas each unit is only fitted with a 50 VA individually. Looking at the volts with only one heating I'm getting over 27 rather than 24. However, that doesn't explain the original element having a reasonable life - unless they've change the design. The part numbers are on the elements and they are the same.

[Ron Lowe]

Is it a mechanical heat transfer problem between the element and the rest of the iron / temp sensor? Like the heat is not being sunk away to the tip and temp sensor correctly, and the most of the energy is over-heating the element only, as it tries to raise the temperature of the tip?

Ron ( Also with Weller TCP from student days. )

[Dave Plowman (News)]

The element and sensor are an assembly, and I fitted a new bit at the same time. There's nothing else on the hand piece that would effect things as it's all plastic.

[meow2222]

yes, it cant be that. I'd vote for significantly too much power to the element as the prime suspect. How & why is another matter, but I'd try running one at reduced power and see if it all behaved itself properly. There is also the poss of a bad batch of elements. Weller might be able to advise, though I'd start by sorting out the overvoltage.

Its a while since I used a tcp so I dont remember if theres any chance of cracking it open to self-rewind the element.

NT

[Dave Plowman (News)]

I doubt Weller could help as it's made by Pace. ;-)

I might rip one apart to see how it's made - I now have a few duff spares.

I've been doing some more research and it seems it's stated as being 19 volts 50 watts. I wonder if this is just because of the insertion loss of the triac? But I'm surprised a slight overvoltage would kill it if it's an ordinary ni-chrome winding.

[Clint Sharp]

In message , "Dave Plowman (News)" writes

I'd suspect that the 150VA toroidal is rather better at keeping itself at the specified voltage than an equivalent 'standard' transformer and is almost definitely better than a lower rated transformer of any type so you may be barking up the wrong tree here. I'd now be considering loading the thing up with a pair of 12v automotive bulbs in series (or one truck bulb) of the same and attaching a variable resistance where the temp sensor should be, adjust the resistance and observe bulb behaviour.

Or the transformer has changed/gone faulty and this is what killed the original element.

[meow2222]

Oh :)

27/19 is quite a big power increase, double the power. A series diode should fix it then, if its now getting full wave from the controller..

NT

[Tony Williams]

The temperature control circuit should be duty- -cycling the triac so that the raw 24V (27V?) ends up as a 19V average at the element.

24V is 80W instead of 50W. That could overheat the element. It does sound as if the problem is in the temperature control circuit, Dave.

Is the element 7.2 ohms resistance when cold?

[Dave Plowman (News)]

Do you mean some form of pulsing during a call for heat?

No - it's 8.2 - 9.5 according to the manual.

I do have the correct PS for the Pace handset, but it's microprocessor controlled and has all sorts of facilities I simply don't need, and some missing that I do. The circuit I've used has proved to be more than adequate for my needs and ultra reliable in use - as well as easily repaired if needed. And I'm slightly wary of over complication of these things after hearing of problems that Weller had with their 'clever' PS - at least in the early days ;-)

I'm going to replace the single transformer with two smaller ones and use a 18 volt one for the Pace. I don't mind if it's a bit slower to heat or doesn't reach the maximum temperature it states in the manual (some 450C)

[Tony Williams]

Sort of, yes.

A smooth servo would be using the triac to duty- -cycle each half-sine. This gives the smoothest temperature control.

A cheaper way would be bang-bang control, where the triac is on/off for many cycles, switched by a comparator, and doing a sawtooth around the avarage set temperature.

8.9 ohms average then. 19V*19V/8.9 is only 40W, not the quoted 50W.

8.9ohms will need a 21V supply to reach 50W.

If you think about it, a 23.5V transformer with 10% regulation drops to 21V at full load.

[meow2222]

Other way round, rated voltage is at full load. A series diode is much easier than replacing the transformer or playing with the waveform. If

2 irons, one diode in each direction so the tf only sees at most half rated current rectified load.

NT

[Tony Williams]

Oh yes thanks. Brain Phart.