Hi folks!
Today I've a bit of a special problem: due to size limits I have to think about using a 16V Panasonic TS-UP 4700uF cap in a 18.5V circuit.
I don't know anything about Panasonic cap tolerances, so if somebody has some experience on this it would be greatly appreciated!!
What do you think about that? Just stupid or already idiotic?
All the best and thanks a lot! Hannes
Today I've a bit of a special problem: due to size limits I have to think about using a 16V Panasonic TS-UP 4700uF cap in a 18.5V circuit.
I don't know anything about Panasonic cap tolerances, so if somebody has some experience on this it would be greatly appreciated!!
What do you think about that? Just stupid or already idiotic?
All the best and thanks a lot! Hannes
Hi,
when is the supply @ 18.5Vdc?
What ripple is on that supply?
What happens to this supply voltage if the mains voltage changes?
How high can mains voltage go?
What happens to the supply voltage if the load is removed?
Is the load fused?
Add all the worst factors together and estimate the highest supply voltage.
Go buy a cap that meets or exceeds this worst case voltage.
I suspect you may need a 25Vcap.
when is the supply @ 18.5Vdc?
What ripple is on that supply?
What happens to this supply voltage if the mains voltage changes?
How high can mains voltage go?
What happens to the supply voltage if the load is removed?
Is the load fused?
Add all the worst factors together and estimate the highest supply voltage.
Go buy a cap that meets or exceeds this worst case voltage.
I suspect you may need a 25Vcap.
Hi Andrew!
This is the idle voltage, under load I expect it to decrease.
Dunno, but 18.5 is the voltage after the diodes, so ripple should not increase this value.
It changes accordingly. However since it's already running at 230V mains, no higher values should occur.
No, does not seem to be the case.
18.5 VDC ;-))
And where? I need a cap with max height 20mm, diameter max 25mm. The only that fits the bill is a 2200uF Rubycon 25V. 4700uF would have been prefered.
Well, I had hoped somebody would just say, Panasonic caps have 20% tolerance, so not a real problem.
Guess, it was just day dreaming ;-))
Thank you very much! All the best, Hannes
This is the idle voltage, under load I expect it to decrease.
Dunno, but 18.5 is the voltage after the diodes, so ripple should not increase this value.
It changes accordingly. However since it's already running at 230V mains, no higher values should occur.
No, does not seem to be the case.
18.5 VDC ;-))
And where? I need a cap with max height 20mm, diameter max 25mm. The only that fits the bill is a 2200uF Rubycon 25V. 4700uF would have been prefered.
Well, I had hoped somebody would just say, Panasonic caps have 20% tolerance, so not a real problem.
Guess, it was just day dreaming ;-))
Thank you very much! All the best, Hannes
Hi,
you're on 230Vac. I bet you have a tolerance on that.
In the UK we used to have +-6% but the Europeans forced us to change our specification to allow 216Vac to 254Vac as the normal range of supply voltage.
I expect your supply company to adopt some range, find out!!
The transformer has regulation. as load decreases the voltage rises. That's why I asked if the load is fused. If the load becomes opencircuit, for example you disconnect the load during some testing procedure, the voltage on the caps rises to near peak voltage coming from the secondary AND the voltage on the secondary rises because the load has been disconnected either accidentally or deliberately.
Now look at ripple on the DC supply.
When you measure the DC voltage you are measuring the average voltage. The ripple rises above and below the ripple voltage. The maximum voltage on the smoothing cap is roughly average voltage + half the peak to peak ripple voltage. However some capacitor manufacturers state that the worst case voltage to be used for selecting smoothing capacitors is worst case average voltage + Vpp of the ripple voltage (not +half Vpp as you might expect).
If you add all these factors together you will find that your 18Vdc is exceeded on many occasions (like one hundred times every second) and on a few occasions exceeded by a large factor.
Some builders suggest running smoothing caps at just 60 to 70% of rated maximum working voltage rather than doing the sums. I prefer to calculate the worst case using real numbers in my models. You choose what you want to do, but don't explode a smoothing cap in your face!
If you cannot understand the safety message in what I describe, I suggest you should not entertain DIY electronics as a hobby.
you're on 230Vac. I bet you have a tolerance on that.
In the UK we used to have +-6% but the Europeans forced us to change our specification to allow 216Vac to 254Vac as the normal range of supply voltage.
I expect your supply company to adopt some range, find out!!
The transformer has regulation. as load decreases the voltage rises. That's why I asked if the load is fused. If the load becomes opencircuit, for example you disconnect the load during some testing procedure, the voltage on the caps rises to near peak voltage coming from the secondary AND the voltage on the secondary rises because the load has been disconnected either accidentally or deliberately.
Now look at ripple on the DC supply.
When you measure the DC voltage you are measuring the average voltage. The ripple rises above and below the ripple voltage. The maximum voltage on the smoothing cap is roughly average voltage + half the peak to peak ripple voltage. However some capacitor manufacturers state that the worst case voltage to be used for selecting smoothing capacitors is worst case average voltage + Vpp of the ripple voltage (not +half Vpp as you might expect).
If you add all these factors together you will find that your 18Vdc is exceeded on many occasions (like one hundred times every second) and on a few occasions exceeded by a large factor.
Some builders suggest running smoothing caps at just 60 to 70% of rated maximum working voltage rather than doing the sums. I prefer to calculate the worst case using real numbers in my models. You choose what you want to do, but don't explode a smoothing cap in your face!
If you cannot understand the safety message in what I describe, I suggest you should not entertain DIY electronics as a hobby.
Thanks a lot for you replies and especially AndrewT for taking the time to reply in detail!
Don't worry, I'm really not interested in building a time bomb. And honestly, I already decided to go along the only professional and reasonable route, that is go for the 2200uF 25V cap.
For a couple of reasons I would prefer the 4700uF, but hey, I will not build something that *may* work for some time. DIY is after all for pursuing his very own high quality standards, so I will definitely not produce something that is not perfectly professional.
But you know, giving up his favourite solution is sometimes not soo easy, that's why I posted initially.
No cheating on quality.
Thank you very much! All the best, Hannes
Don't worry, I'm really not interested in building a time bomb. And honestly, I already decided to go along the only professional and reasonable route, that is go for the 2200uF 25V cap.
For a couple of reasons I would prefer the 4700uF, but hey, I will not build something that *may* work for some time. DIY is after all for pursuing his very own high quality standards, so I will definitely not produce something that is not perfectly professional.
But you know, giving up his favourite solution is sometimes not soo easy, that's why I posted initially.
No cheating on quality.
Thank you very much! All the best, Hannes
h_a said:
Hannes,
an experiment for you :
multiply the rated capacity with the voltage rating of all regular different type/size electrolytics you have which are larger than 1000uF, divide that by the cubic volume of the specific cap.
The closest you can find for 25 Diameter x 20 Height is 3300uF/25V
Oh, you have to spoil all the fun, haven't you Jacco ;-))
I really hoped that nobody will see that thread, it's stupid and unprofessional all the way.
And then you step up and state obvious things.
It's not my fault that no manufacturer - at least none that I've heard of - builds magic capacitors with double capacity per m^3 than all the others.
Or maybe I just say your mentioned statement was just a rhetorical question. Therefore the 16V cap has to do its duty. Makes me appear in a better light, ahem.
Err, forget about that.
All the best, Hannes (maybe somebody could delete that thread, he?)
I really hoped that nobody will see that thread, it's stupid and unprofessional all the way.
And then you step up and state obvious things.
It's not my fault that no manufacturer - at least none that I've heard of - builds magic capacitors with double capacity per m^3 than all the others.
Or maybe I just say your mentioned statement was just a rhetorical question. Therefore the 16V cap has to do its duty. Makes me appear in a better light, ahem.
Err, forget about that.
All the best, Hannes (maybe somebody could delete that thread, he?)
Why stupid ?
There's a choice out of several brands if you had room for a 20 mm Diameter by 25 mm Height cap.
This obsolete Elna silk capacitor almost equals the volume of your preferred capacitor size, and 3300*35/4700*25 ~ 1
Only thing is that they're able to squeeze more capacity/voltage into a thinner cap, thickies have lower (C*V)/Volume efficiency.
Both capacitance and sustained voltage of a capacitor can vary, but you have no guarantee.
Using a high surge value capacitor would be a safer bet, unfortunately those are not even near to found in the category you need.
There's a choice out of several brands if you had room for a 20 mm Diameter by 25 mm Height cap.
This obsolete Elna silk capacitor almost equals the volume of your preferred capacitor size, and 3300*35/4700*25 ~ 1
Only thing is that they're able to squeeze more capacity/voltage into a thinner cap, thickies have lower (C*V)/Volume efficiency.
Both capacitance and sustained voltage of a capacitor can vary, but you have no guarantee.
Using a high surge value capacitor would be a safer bet, unfortunately those are not even near to found in the category you need.
Attachments
DID I SAY STUPID?
C'mon.
And no, 25mm height is a definitely no-no, since the cap would press against the enclosure bottom which could sorta break the pcb.
Only solution I've found so far, is a Rubycon ZLH 2200uF 25V cap. Ultra-low impedance. Just the contrary of what I would need (at least standard impedance).
Grr. It's annoying that the perfect part has too low volts.
Well.
Jacco, no nice weather up there? Or already've been out with your Duc? I would also love to ride today, if the carburator wasn't in bad mood.
Seems in the moment the parts have their fun with me.
All the best, Hannes
C'mon.
And no, 25mm height is a definitely no-no, since the cap would press against the enclosure bottom which could sorta break the pcb.
Only solution I've found so far, is a Rubycon ZLH 2200uF 25V cap. Ultra-low impedance. Just the contrary of what I would need (at least standard impedance).
Grr. It's annoying that the perfect part has too low volts.
Well.
Jacco, no nice weather up there? Or already've been out with your Duc? I would also love to ride today, if the carburator wasn't in bad mood.
Seems in the moment the parts have their fun with me.
All the best, Hannes
17C and sunny, doing woodwork/painting on my diy roof construction, restaurant hopping in a couple of hours when Mrs Nut returns.
Started on this thing 3 years ago, 10m x3.5m x 2.5m for 3 rooms at the toplevel. All epoxy and wood trim construction, hope to get it finished before it gets to cold, epoxy resin/glue requires 15C minimum.
Parts can be a pita.
Started on this thing 3 years ago, 10m x3.5m x 2.5m for 3 rooms at the toplevel. All epoxy and wood trim construction, hope to get it finished before it gets to cold, epoxy resin/glue requires 15C minimum.
Parts can be a pita.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.