Pace ADS200 soldering station

[KL27x]

^+1
I have done production soldering for several years, now.

Why do they use a 17 bit ADC despite 5 degree resolution of temp setting? Despite the 1 degree resolution of the display? Those things don't matter.

In production soldering you don't look at a display. And you don't need to adjust the temp to a tenth of a degree or even 1 degree. The human operator makes the final adjustment to his own technique after the best setting is found for the task at hand. What matters is consistency and repeatability, so that once the groove has been found, it doesn't change, and the iron behaves as expected throughout the other 1000 of the same operation. When soldering stations do the soldering by themselves, then they can bother adjusting temp down to a tenth of a degree... without turning knobs or pressing buttons... and they won't need a display to see where they're at. Until then, the best way an iron can behave is simply to be as consistent as possible, down to as fine a resolution as possible, day in day out, without drifting with time or temp or humidity... and it doesn't matter if that shows up on a display or not. As a production solderer, there are only 2 reasons to have a display at all. 1. To aid in adjusting the temp and/or to quickly return to previous known settings and 2. To debug in the event of problems with the station.

That Pace uses 17 bit ADC and doesn't have a USB port just shows their priorities are pretty much correct for their traditional market. They are working at improving the part that matters. Whether they succeeded or not, only time will tell. But I find it impressive that Dave took two random tips and they both measured spot on. If you can replace a worn tip with a new one, and it behaves exactly the same as before, that would be pretty nice.

[exe]

9-bits sufficient? Maybe not, there's 500 steps alone between tip extremes - min(350F) and max(850F). Someone else may want to comment on higher resolution and/or over-sampling to establish delta, but more data is preferable than less.

1) Discussion was about MCUs not having 17bit ADCs. I question this is needed. 10 or 12 bit ADC would do the job. Jelly bean MCUs have it.

2) It doesn't display to that resolution, so, yes, 9bit is fine. In fact, if it rounds to 5deg C, then even 8bit ADC would do the job (assuming ADC is only for display and control loop is analog).

3) No point in having more data if it is thrown away. Also, ADCs with a lot of resolution are harder to make them properly working. They easily pick up noise from environment, power rails, etc. Even 12bits is not easy to utilize (as stm32 demo boards show, I've seen noise up to 3 LSB).

Just in case, I'm not saying they are doing it wrong or something. I just like to discuss the design and learn something from it. There can be non-design reasons to make the station this way. May be they just had tons of these 17bit ADCs for free and decided to use them. Or they just wanted reduce design costs/time and put parts with overkill specs, who knows.

BTW how does AccuDrive works? Everyone says "wow, it is spot on", but I'd want to know details. Because may be Dave received a special calibrated unit with pre-selected tips.

PS any hi-res photos of the PCB available? Want to see how active rectification works... I have two ideas how it can work (e.g., using comparators to switch mosfets, or a self-driving circuit if load is resistive).

[KL27x]

3) No point in having more data if it is thrown away.

A soldering station isn't a temperature measuring device. None of the resolution is necessarily discarded. It doesn't need to be displayed, because the display doesn't do any soldering. Presumably, that high resolution ADC is used to improve response and temp stability. If it is only there to put a number on a screen, and the temp control loop doesn't use it, well that would a pretty big joke. Let's hope for Pace that this isn't the case.

Some of the Metcal stations don't even have a temp display or adjustment capability. The user input and display have a resolution of 1 bit. It's on or off. But every bit of resolution that the Curie RF system produces (which is infinite, since it's analog) is put to work. In the Metcal, the further the tip drops in temp, the more it increases in skin resistance, and the more power gets put into the tip. something like that, anyhow. If you wanted to do that digitally, you have to differentiate to some level of resolution. How much will matter? Well, enough is enough, but more can't hurt. 

[exe]

3) No point in having more data if it is thrown away.

You don't seem to understand the post I made just previous to yours. A soldering station isn't a temperature measuring device. None of the resolution is discarded. It doesn't need to be displayed, because the display doesn't do any soldering.

I'm questioning 1) how many ENOB they can get from the circuit 2) if there is a digital control loop or analog (you claim it is digital, are you sure?).

Just in case, a 17bit ADC supplied from 5V has 0.038mV step. Good luck getting this resolution and precision in a circuit supplied from a 7805 regulator (or whatever there, I don't have access to fast Internet to check what was there) with no shielding and layout considerations. It's a resolution of a 5.5 digit DMM. I don't remember seeing lm399 on the board. Also, ADC with high resolution are slow. So, definitely resolution was not a reason to put a 17bit ADC.

[KL27x]

You make good points.

If the temp control loop doesn't use those bits, well that would be curious. Maybe it IS just for show..

[tooki]

which suitable micros have a 17 bit ADC?

Not needed at all. 9+ bits is enough. Also, looks like the display can only display numbers that are power of 5 (was my impression from Dave's video, don't have access to any good internet to check).

9-bits sufficient? Maybe not, there's 500 steps alone between tip extremes - min(350F) and max(850F). Someone else may want to comment on higher resolution and/or over-sampling to establish delta, but more data is preferable than less.

I'll mention that too. The thought is that it can be annoying to have all those single digits flashing. But some of our older equipment had single digit resolution, so why not here? You let it go single digit for 10 digits, then switch to 10 degree jumps ...

I think the issue isn’t the granularity (1 degree or 5 or 10), but rather the display update rate and the hysteresis. A value that’s switching between 280 and 285 every second is every bit as annoying as one flapping between 283 and 284. It depends on the display type, but on an LED display, an update rate faster than about 8-10 updates per second becomes impossible to read. Somewhere between 3-7 is my personal preference for most devices. (Non-TFT) LCDs are slower, so 3-5 is about the max for them to remain legible. On a soldering station, I’d lean towards a gentler rate like 3/sec.

But hysteresis is the critical thing here: Let’s assume 5 updates per second. If the setback logic is maintaining 280F ±2F (actual tip temp), then you might set the hysteresis to 2 degrees, so that it won’t update the display until the temp actually leaves the 278-282F range, and then you update in 1 degree steps.

There is a balance of too much information vs. just the right amount.  Too much can be distracting.  Too little can not be informative enough.  I don't mind a granularity of 5 degrees at all.

In actual practice, 5 degrees granularity is more than enough for soldering. But as a nerd, somehow, I like seeing the individual degrees!





Since we’re on the topic of display annoyance: I think I’d find the blinking setback to be very distracting. I wonder whether there’s enough CPU cycles left to do fancy effects like dimming the display (or gently “throbbing” it) when it’s in setback — or doing something with the Ready LED, like turning it yellow or throbbing it.

[exe]

BTW, consider making a series of "performance tips" with better heat transfer (and shorter tip life). It's a win-win: 1) better performance 2) people will have to buy more tips from you if they want performance 3) less coating material used

Another idea: sell the station as a kit. I'd love to assemble one by myself.

[Cliff Matthews]

Some observations: TC500ACOE is not so costly, it lists as a $2.50 part (est. 1000's price). Date sheet header says designer can trade resolution for speed, but no more than 10sps. Marco Reps investigation shows unbuffered 120hz DC to the tip..

*edit - Opps forgot, here's the index to the point he made: https://youtu.be/erKCA71q7cg?t=843

[exe]

TC500ACOE

Wait, but that's not ADC. It's an ADC _frontend_. Now it's getting really weird...

Marco Reps investigation shows unbuffered 120hz DC to the tip..

You mean 100Hz? (He's in Germany, afaik, so should be 50Hz mains) Uhm, so, I guess, there is no control to fets at all, are they connected like this? https://electronics.stackexchange.com/questions/203315/storing-the-the-charge-from-a-mosfet-bridge-rectifier . But then I wonder why rectifying at all? Why not drive iron from AC?

[Cliff Matthews]

There's also an ICL7660A to run +- 5V to that front-end..

[alank2]

I don't know too much about these things, but when it comes to temperature regulation and PID loops, wouldn't more resolution/precision offer superior control, irregardless of the display?  Could the ADC resolution be primary there to drive control?  Maybe this is obvious..

[labjr]

BTW, consider making a series of "performance tips" with better heat transfer (and shorter tip life). It's a win-win: 1) better performance 2) people will have to buy more tips from you if they want performance 3) less coating material used

Another idea: sell the station as a kit. I'd love to assemble one by myself.

I wonder if performance could be improved with a more aggressive ramp-up curve with a bit of overshoot. Or the ability to select different performance curves required for the task.

Kits would probably need too much support.

[nanofrog]

BTW, consider making a series of "performance tips" with better heat transfer (and shorter tip life). It's a win-win: 1) better performance 2) people will have to buy more tips from you if they want performance 3) less coating material used

Another idea: sell the station as a kit. I'd love to assemble one by myself.

I wonder if performance could be improved with a more aggressive ramp-up curve with a bit of overshoot. Or the ability to select different performance curves required for the task.

Kits would probably need too much support.

In regard to selectable curves, Ersa does that on their i-Con 1/2/4 models.

Kits would be a horrible idea for Pace. Just way too much aggravation, potential damage to their reputation, and increased costs (they're not Heathkit, which is long gone).

[tooki]

I wonder if performance could be improved with a more aggressive ramp-up curve with a bit of overshoot. Or the ability to select different performance curves required for the task.

In regard to selectable curves, Ersa does that on their i-Con 1/2/4 models.

All the i-Con series have that, even the nano and pico. (It’s configured via SD card, using the desktop app.)

[exe]

Just watched Marco Reps's video. Afaik, there is no "active" rectification, just mosfets connected in "ideal diode" configuration. Which is fine.

The station switches heater at zero crossing. This means decisions are made at the rate twice the line frequency. So, in the worst case there is a delay of 10ms. Why I'm saying this? Just to say that even a slow SMPS would work just fine (I'm not sure they are really slow as some users report here). But, again, doesn't really matter.

Concerning display and other fancy features. It is true they are not essential for soldering. But noobs like me like them. So, it's just a matter of target market. Also would make possible to change profiles, preset temperatures, etc easier (ersa with configuration on sdcard is a joke).

I personally like rotating encoders with push button. Minimalistic and convenient.

[Cliff Matthews]

Just watched Marco Reps's video. Afaik, there is no "active" rectification, just mosfets connected in "ideal diode" configuration.

Not so fast, take a look at 15:25 (https://youtu.be/erKCA71q7cg?t=843 ) bottom left is right under the mosfets. I couldn't get a good enough look at the tiny part number, but Z1 and Z4 could be dual zener's who's tab could be under uC control.

*edit - those are ZXMN4A06G n-ch mosfets, if that makes any sense (it might have been easier with p-ch?).

[exe]

There are two N-fets and two P-fets (yeah, they are different, hard to see on video, but part numbers slightly different, they are ZXMP, not ZXMN). I wasn't able to trace all the traces, but I think it's this bridge:

.

Looking at the video, it seems the station has more part than I expected. Still curious how they use adc frontend.

[nanofrog]

I wonder if performance could be improved with a more aggressive ramp-up curve with a bit of overshoot. Or the ability to select different performance curves required for the task.

In regard to selectable curves, Ersa does that on their i-Con 1/2/4 models.

All the i-Con series have that, even the nano and pico. (It’s configured via SD card, using the desktop app.)

I wasn't aware the SD card was usable for that.

Seems a bit of a PITA though vs. the rotary encoder with button operated menu system in the bigger siblings.

[Cliff Matthews]

There are two N-fets and two P-fets (yeah, they are different, hard to see on video, but part numbers slightly different, they are ZXMP, not ZXMN). I wasn't able to trace all the traces, but I think it's this bridge:

.

Looking at the video, it seems the station has more part than I expected. Still curious how they use adc frontend.

You have good eyes! I'm still curious where Vin for the 7805 comes from (it won't boot without it..) The body diodes?

*Edit - In future model updates etc..
If Pace considers a single transformer with dual winding's, they might consider internal wiring only, since a station incorrectly strapped for 120v on 240v would require 80v mosfets and good crowbar to protect (yet again another reason for SMPS as some have suggested).

[Alex Eisenhut]

I don't like how small the sponge cubby is. Solder blobs fly out all the time. The sponge is half the size of the Hakko 936's.

Really don't like this.

[Bud]

JBC sponge is also tiny.

[Alex Eisenhut]

JBC sponge is also tiny.

I probably wouldn't like it either. 

[Shock]

No brass wool holder on the Hakko 936 stand, so 90s

[nanofrog]

If you'd rather use a sponge only, then why not just remove the brass wool box so you can install a larger one?

Or is there some technical reason preventing this that I'm not seeing in the photos?

[gabinetex]

Not to mention,  JBC sponge-life is notoriously short. They get corroded, I heard somewhere.