advertisement


building +5V ALWSR and doubts

geoturbo said:
Auch no luck today.. better sleep now..
Click to expand...

OK, so we know the issue lies with the super-reg element.

I think a poke around the reference section, since the 2.5V there is needed to maintain a stable output voltage - does the reference voltage drop as soon as the output drops? If yes, that is an area to look at, if no, then the output dropping is related to the regulator element. What voltage reference are you using?
 
Testing Super reg alone it has been working fine for more than 20 minutes, so decided to wire prereg and superreg back up again, but it failed to Keep the 4,9 or so output Volts for more than few minutes. So I Am back at the starting point.

ZD1 voltage reference is by a LM336Z-2.5, I will Keep An eye what happens there when voltage drops, Thank you Andy.
 
I've taken a look at the AD817 datasheet and might have identified the problem. The worst case minimum output voltage of the AD817 is 1.5 volts. Add the 3.3v + 5% of the zener and assume you get the same Vbe across the reg's output transistor and the pnp at the output of the opamp and you get a possible minimum output voltage of 4.965v for the regulator to work. That's too close to your measured 4.9 volts to be reliable IMHO. Adding the pre-reg makes the opamp sink more current (as the 317 provides some straight to the load) which further raises this minimum voltage for operation and which prob explains why it then fails. The Vbe of the output transistor also reduces as it heats up under load which also reduces the voltage margin over time, hence the delay before failure.

An easy way to check if this is indeed the faults cause, without removing the 3.3 volt zener, is to tack solder a green or red LED across the zener. The zener should not conduct and the LED should light and conduct instead. Alternatively you could tweak the output voltage up to say 5.1 v by putting a resistor across the lower arm 1k resistor setting the reg's gain. The ideal value of parallel resistor for 5.1 volts is 12.5 k. Try something between this and say 20k to raise the output voltage to just under 5.1 volts. Personally I prefer the LED solution as it completely removes any risk of this issue altogether.

John
 
Thank you guys.
Timh, I doublecked and the LM336Z-2.5 is the correct way 'round.
Andy, measured and the drop occurs at the zd1 LM336Z-2.5, from 2.49V when the alwsr is outputting 4,9V to about 1.22/1.28V oscillating when output voltage drops.

John you're gracious (great would not be enough), giving me hope for the nights to come.. yes, It seems like when some component is heating up than It starts to give trouble.. I will duly check your reply, not sure about resistor values around.. but I should have somewhere some green leds indeed!
Planned to build some more ALWSRs btw..
 
The green led in parallel with the 3,3V 1N5333B zener (D2 in part list) does not light up, and instability reigns, after a while. I soldered the led up with the same polarity as the 3,3V zener (As in first picture of this thread), if this is correct.

If It is correct I reckon I should try next to take out the 3,3V zener and use the green led only in place of D1. Is this a good plan?
 
I am confused, which may not help you much, but D2 on the 5V ALWSR build in the online manual parts list is a 1N5333B, but the schematic has a LED, and the component overlay has a LED. I think the online manual 5V build schematic and component overlay are a bit out of date?

I believe a zener needs to be the opposite way round to a LED.

From the online manual component overlay the LED is shown anode at the top, cathode at the bottom.
From your photo, although the PCB is not the same, it is shown in the same orientation, and has the zener cathode at the top, anode at the bottom.

So you would need to add the LED in the opposite orientation to the zener, as shown in the online manual I believe.
 
Quite, in that zener diodes and LEDs are drawn differently (for good reason - the zener is reverse-biased, in that it is designed to 'break-down' above a certain applied forward voltage; but the LED is forward-biased: when both are a PN junction) So the manual shows both correctly.

For the average 5mm round LED, the 'flat' of the D shape case should be next to R6 / link wire in that position, as shown on the ALWSR schematics
 
Thank you, gents.
I did not verify but very likely install the led on the zener the wrong way round, hoping I did not damage anything.. let's see..
I put this one on hold to take a deep breath and as my "new" weller station started misbehaving too..

Will check back in due time
 
Looking at your photos R7 is 1K, I don't say it is right or wrong but might just be worth noting that there is some inconsistency between the ALWSR online manual parts list R7 1K and online manual schematic and component overlay 499R and the google docs parts list 499R.

Similar inconsistencies in the online manual parts list for R2 & R3 499R with the schematic and component overlay showing 1K and the google docs parts list showing 1K.

In all cases I have gone with the schematic and component overlay and google docs parts list, and not the user manual parts list. Dunno which is correct!
 
R2 and R3 need to be equal values to get 2.5v out from the pre-reg so 499r or 1 k doesn’t matter (probably)
 
Agreed for R2/R3, unlikely to make much difference to the setting of pre-reg as long as they are not crazy values and are the same.

But what about R7? The user manual suggest the R8/R9 in parrallel and R7 should be about the same? So if R8/R9 are 1K then the parrallel resistance is 500R so R7 should be 500R?
 
Just looked at one of my ALWSRs and r7 is 1k
I’ve used a green led rather than a zener (flat pointing away from the DH44) and an LM431 as vref. Happy to post a picture if it would help
 
More pictures the better from my point of view! It isn't that I think R7 is the issue, it is just that there is confusion in the user manual and parts lists and schematics and component overlay plus stuff in the user manual that suggests there is a purpose to impedance matching R7 with R8/R9 in parallel.

TLDR: I notice stuff but don't know it is relevant
 
nobeone said:
Agreed for R2/R3, unlikely to make much difference to the setting of pre-reg as long as they are not crazy values and are the same.

But what about R7? The user manual suggest the R8/R9 in parrallel and R7 should be about the same? So if R8/R9 are 1K then the parrallel resistance is 500R so R7 should be 500R?
Click to expand...

It's a counsel of perfection, looking for absolute lowest noise; and for the differences you've picked-up, utterly not worth worrying-about. [ back of envelope says << 0.7nV/RT(Hz) - when all other possible contributions on the board are decade or more higher. ]
 
FB1-D9-CBD-91-A4-4-BAB-B775-BEC681-D02-DFD.jpg


Ignore the red wire. This board is for 3.3v output which isn’t enough for the opamp so I had to cut a track and feed it directly from the prereg output
 
martin clark said:
It's a counsel of perfection, looking for absolute lowest noise; and for the differences you've picked-up, utterly not worth worrying-about. [ back of envelope says << 0.7nV/RT(Hz) - when all other possible contributions on the board are decade or more higher. ]
Click to expand...

Thanks Martin, much appreciated, the problem for me is, as a noob to ALWSR it is hard to know what is important, and what is not, but the state of the documentation is really not helping. So I see contradictions and point them out. Do they matter? No idea. Do they hinder a noob? Absolutely. I appreciate this is a lot of work shared for free, i point out the issues in the hope they can be fixed and so others will feel it is easier to follow. I really appreciate the manual and the parts list, and have a my first tested ALWSR, but is has been a bit of a slog :)
 
You must log in or register to reply here.


advertisement


Back
Top